Rational Choice Theory: Resisting Colonisation (Critical Realism: Interventions)

  • 70 792 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up

Rational Choice Theory: Resisting Colonisation (Critical Realism: Interventions)

Critical realism: inte"entions Edited by Margaret S. Archer, Roy Bhaskar, Andrew Collier, Tony Lawson and Alan Norrie C

1,660 810 2MB

Pages 270 Page size 432 x 647.28 pts Year 2011

Report DMCA / Copyright


Recommend Papers

File loading please wait...
Citation preview

Critical realism: inte"entions Edited by Margaret S. Archer, Roy Bhaskar, Andrew Collier, Tony Lawson and Alan Norrie

Critical Realism Essential n:adings Edited by Margaret Archer; Roy Bhas/uu; AndrewCollier, Tony Lawson and A/on Norrie

The Possibility of Naturalism A philosopbic:al critique of the: contemporary human scicnc:cs Roy Bhaskar

Being and Worth AndrewCollier

Qaantum Theory and the Flight from Realism Christopher Non-is

From East to West Odyssesy ofa soul Roy Bhaskar

ReaUsm aDd Racism Concepts ofrace in sociological research BobCorler

ReaUst Perspectives oD Maugemmt and OrganisatiODs Edited by Stephen Ackroyd and Stew Fleetwood

Rational Choice Theory Resisting colonization

Edited by Margaret S. Archer and Jonathan Q. Tritter

London and New York

First publisbed 2000 by Routledge 11 New Fetter laDe, London EC4P 4EE SimullaDeOusly published in the USA and Canada by Routledge 29 West 3Sth Streec, New York. NY 1000 I

Routkdge Is an Imprint ofthe Taylor & Francis Group This edition published in the Taylor & Francis c-Library, 2001. C 2000 Margaret S. Archer and Jonathan Q. Tritter, selection and editorial mauer; individual chapters, the. contributors

All rights reserved. No pm of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, oow known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the pubUshers.

British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library ofCongnus Cataloging in Publication Data Ratiooal choice theory: a critique I edited by Margaret S. An:hcr and Jonathan Q. Trittcr. Includes bibliographical references and index. I. Rational choice theory. l. Archer, Margaret Scotford. II. Triuer, Jonathan Q., 196S­ HM49S .R37 200 I 301 '.OI-dc21 ISBN 0-41S-24271-1 (hbk) ISBN 0-41S-24272-X (pbk) ISBN 0-203-13389-7 Masterc-book ISBN ISBN 0-203-18424-6 (Olassbook Format)


Rational Choice Theory

Rational choice theory is flourishing in sociology and is increasingly influential in other disciplines. Contributors to this volume are convinced that it provides an inadequate conceptualization of all aspects of decision-making: of the individuals who make the decisions, of the process by which decisions get made and of the context within which decision-making takes place. The critique focuses on the three assumptions which are the bedrock of rational choice:

Rationality: the theory’s definition of rationality is defective because it is restricted to ‘instrumental rationality’. RCT cannot deal with ‘substantive rationality’, which is not a means to an end, but a commitment to those ‘ultimate concerns’ which define who we are. What we care about most involves our emotionality and normativity, yet neither affect nor norms can be satisfactorily incorporated into the RCT model of the cool, individualist bargain-hunter, seeking to maximize a ‘utility’ whose constitution remains mysterious.

Individualism: the methodological individualism of RCT is unsatisfactory for conceptualising structure or agency, but especially the relations between them. On the one hand, society’s structure and culture enter into the personal goal-formation, thus undermining the atomistic character of Homo economicus. On the other hand, rational choice theorists can only view structure and culture as aggregates and cannot incorporate their influences as emergent properties, which effect decision-making by shaping the situations that people confront.

Temporality: RCT presupposes the possibility of identifying people’s preferences atemporally, because of their supposed stability. On the contrary, preferences have a history and a trajectory. Where agents are concerned, it is only by undergoing certain experiences that they can reflect upon their past preferences. Yet they become different people through this learning process, which then makes it more difficult to rectify and realize their preferences in the future. Similarly, the history of structural change conditions preference transformation over the same time span. Instead, RCT stereotypes a stable preference order, thus denying the creative reflexivity of agents, as they respond to changes in their circumstances over time.

The critique is grounded in discussion of a wide range of social issues, including race, marriage, education, homeworking, the underclass and crimes against humanity.

Margaret S. Archer and Jonathan Q. Tritter are both members of the Department of Sociology at the University of Warwick.


List of illustrations Notes on contributors 1 Introduction

vii ix 1



Rationality 2 The bird in hand: rational choice – the default mode of social theorizing




3 Homo economicus, Homo sociologicus and Homo sentiens



4 Is rational choice theory ‘unreasonable’? The neglected emotions




Individualism 5 Social theory and the underclass: social realism or rational choice individualism? JUSTIN CRUICKSHANK



vi Contents 6 (Ir)rational choice: a multidimensional approach to choice and constraint in decisions about marriage, divorce and remarriage



7 Switching allegiances: decisions by schools to ‘opt out’ to self-management



8 Rational choice or ‘Hobson’s choice’? Intention and constraint in UK higher education




Temporality 9 ‘I do’: a theoretical critique of Becker’s rational choice approach to marriage decisions




10 Decision-making as a process over time: the careers of home-located cultural workers



11 The decision to commit a crime against humanity



12 ‘Race’, ethnicity and housing decisions: rational choice theory and the choice–constraints debate





13 ‘When the battle’s lost and won’



Bibliography Name Index Subject Index

234 249 253


Tables 6.1 9.1 9.2

Characteristics of the interviewees Calculating a utility balance Freda’s matrix of possibilities

99 148 155

Figures 9.1 9.2

Freda’s three scenarios Freda’s preference order and choice

154 154


Margaret S. Archer completed her postgraduate research at the London School of Economics and the École Pratique des Hautes Etudes, Paris. She produced four books on education and the social structure of Europe before moving to the University of Warwick in 1973. There she completed her Social Origins of Educational Systems (1979), which opened up her main interest in social theory. This has been pursued in a trilogy, beginning with Culture and Agency: The place of culture in social theory (1989 and 1996), and followed by Realist Social Theory: The morphogenetic approach (1995). The final book, completing the trilogy, Being Human: The problem of agency was published in 2000. At the 12th World Congress of Sociology (1986), she was elected as the first woman President of the International Sociological Association. She is also a founder member of the Pontifical Academy of Social Sciences (1994) and a Trustee of the newly founded Centre for Critical Realism (1997). James A. Beckford has focused in his research on the theoretical and empirical aspects of religious organisations, new religious movements, church-state problems, civic religion and religious controversies in several different countries. His publications on these topics include The Trumpet of Prophecy: A sociological analysis of Jehovah’s Witnesses (1975), Cult Controversies: The societal responses to new religious movements (1985), Religion and Advanced Industrial Society (1989) and (with Sophie Gilliat) Religion in Prison (1998). He has also edited New Religious Movements and Rapid Social Change (1986), The Changing Face of Religion (1989) and Secularization, Rationalism and Sectarianism (1993). His teaching and research have taken place at the Universities of Reading, Durham, California at Berkeley and Loyola at Chicago. His current research deals with the relationship between religion and politics, with civil religion and with the treatment of religious minorities in prison. Justin Cruickshank studied at the University of Warwick and gained the MA in ‘Philosophy and Social Theory’. He became a General Teaching Assistant for four years, during which time he completed his PhD thesis, ‘Anti-foundationalism and social ontology: Towards a realist sociology’. He was awarded his doctorate in 2000 and his thesis is now in

x Notes on Contributors production as a book with Routledge. He is now a Lecturer in Philosophy at the Nottingham Trent University Graduate School for Social Research. He is currently working on a book about the underclass debates. Robert Fine joined the University of Warwick Department of Sociology in 1973, where he is currently Reader and Director of the Centre for Social Theory. His books include an edited collection (with Edward Mortimer) on People, Nation and State (1999); Being Stalked (1997); an edited collection (with Shirin Rai) on Civil Society (1997); a study of the relation between labour and nationalism, Beyond Apartheid (1991); a study of the relation between liberalism and Marxism, Democracy and the Rule of Law (1986); and an edited collection on critical theories of law, Capitalism and the Rule of Law (1989). His two most recent books are an edited collection with Charles Turner, Social Theory after the Holocaust (2000), and a monograph on the legacy of Hegel’s political thought for sociology, Elements of Political Modernity (2001). He is also doing research on crimes against humanity. David Hirsh is currently a doctoral research student at the University of Warwick. His thesis addresses the prosecution of individuals for crimes against humanity at the International Criminal Tribunal for Yugoslavia and in nationally organised criminal proceedings. His particular concern is with the significance of these trials for nationalistic and cosmopolitan narratives. Richard Lampard joined the Department of Sociology at the University of Warwick in 1990 and focuses primarily on marriage/marital histories and on social mobility/class-related inequalities in his research. He is a research methods specialist with a background in social statistics and a particular interest in the statistical modelling of cross-tabulated data and of life/work histories. His research on the formerly married and remarriage has involved both quantitative and qualitative approaches and has led to a paper (co-authored with Kay Peggs) in the British Journal of Sociology (1999). His other recent publications include a study of parents’ occupations and their children’s occupational attainment, in Sociology (1995), and a study of party political homogamy in Britain, in European Sociological Review (1997). Andrew Parker is Lecturer in Sociology and Research Fellow in the Centre for Educational Development, Appraisal and Research (CEDAR), at the University of Warwick. His research interests include the sociology of education and the sociology of sport and centre specifically on issues of gender relations and sexuality. Published articles reflecting these research interests have appeared in journals such as Gender and Education, Sociology Review and the Journal of Education and Work. Alongside his involvement in numerous research projects in CEDAR, he is currently carrying out research into the personal and professional life experiences of newly qualified primary school teachers.

Notes on Contributors xi Kay Peggs is Senior Lecturer in Sociology in the School of Social and Historical Studies at the University of Portsmouth. Her research interests include couple relationships, gender, ageing and consumption and the sociology of risk. She has published in journals such as the British Journal of Sociology and the Sociological Review, and in edited collections on women and pensions and on women and health. Ian Procter gained his BA from Durham in 1969 and joined the University of Warwick Department of Sociology in 1971. He has a long-standing involvement with widening access to higher education, including working with local Access courses and collaborating with the Workers’ Educational Association. Previous publications include Service Sector Workers in a Manufacturing City (1988) and a series of papers on women’s work history. His book (with Maureen Padfield) on Young Adult Women: Work and Family was published in 1998. This grows out of research on women, work and family at a key point of the life course. The latter and the general issue of the age-structuring of society is a developing interest. Peter Ratcliffe joined the University of Warwick’s Department of Sociology in 1971 and also works closely with the Centre for Research in Ethnic Relations. His books include (with J. Rex, S. Tomlinson and D. Hearnden) Colonial Immigrants in a British City: A class analysis (1979), Racism and Reaction: A profile of Handsworth (1981), Ethnic Discrimination: Comparative perspectives (ed. with J.A. Diaz and A. Bacal, 1992), ‘Race’, Ethnicity and Nation: International perspectives on social conflict (ed., 1994), Social Geography and Ethnicity in Britain: Geographical spread, spatial concentration and internal migration (ed., 1996) and ‘Race’ and Housing in Bradford (1996). Jonathan Q. Tritter graduated from the University of Chicago and then completed a DPhil at Nuffield College, Oxford University. Before joining the Department of Sociology at the University of Warwick he was a Research Fellow at the Social Sciences Research Centre at South Bank University. His doctoral research examined the impact of socialization and the reproduction of moral values in religious schools. This interest in education continues in a recently completed ESRC-sponsored study of grant-maintained status and church schools. He also has interests in medical sociology and particularly the nature of satisfaction, dissatisfaction and complaining. Current research on the experiences of cancer patients builds on previous studies, Patient-Centred Cancer Services? What patients say (1996) and Meeting the Needs of People with Cancer for Support and Self-Management (1999). Peter Wagner is Professor of Sociology at the University of Warwick and Professor of Social and Political Theory at the European University Institute, Florence. His main research interests are in the social and political theorising of modernity and in the comparativehistorical sociology of institutions and discourses. Currently he is working in the area of a social theory and political philosophy of Europe. His book publications include: A

xii Notes on Contributors Sociology of Modernity (1994), Der Raum des Gelehrten (with Heidrun Friese, 1993), Discourses on Society (co-editor with Wittrock Bjorn and Richard Whitley, 1991) and Sozialwissenschaften und Staat (1990). His most recent books are Not All That is Solid Melts into Air. A history and theory of the social sciences (2001), Theorizing Modernity. Inescapability and attainability in social theory (2000) and Le Travail et la nation. Histoire croisée de la France et de l’Allemagne (co-editor with C. Didry and B. Zimmermann, 1999, German edn 2000). Simon J. Williams completed his postgraduate studies at the University of London with a doctorate on the sociological consequences of chronic respiratory illness. Before joining the Department of Sociology at the University of Warwick, he was a Research Fellow in the Centre for Health Services Studies at the University of Kent. In 1995 he took up his current post as a Warwick Research Fellow. He has published articles on chronic illness, pain and disability, lay perspectives on health and medicine, and health promotion, including recent contributions to Social Science and Medicine, Sociology, Sociology of Health and Illness, Health and Body and Society. His books include Chronic Respiratory Illness (1993), The Lived Body (1998) (with Gillian Bendelow) and two co-edited volumes: Modern Medicine: Lay perspectives and experience (1996) (with Michael Calnan) and Emotions in Social Life (1998) (with Gillian Bendelow). His book Emotions and Social Theory and a further co-edited volume (with Jonathan Gabe and Michael Calnan), Health, Medicine and Society, were both published in 2000. His present research interests lie in the areas of the sociology of embodiment and the sociology of health, with particular reference to pain, emotions and sleep. Carol Wolkowitz is a Lecturer in the Department of Sociology at the University of Warwick. Her research has involved a number of different areas of gender studies. She has a longstanding interest in gender in Indian history and politics, stemming from her doctoral research on women politicians’ careers in South India. Since then much of her work has focused on gender and employment. She is co-author of two books on homeworking and home-located work, Homeworking Women: Gender, class and racism at work (1995) and Homeworking: Myths and realities (1987), and she is currently working on a book exploring various aspects of the relation between embodiment and the labour process. Her most recent publications are A Glossary of Feminist Theory (1997), with Terry Lovell and Sonya Andermahr, and several articles on women’s roles in the American communities established by the Manhattan Project during the Second World War. She was also co-editor of Of Marriage and the Market: Women’s subordination in international perspective (1981 and 1985).


Introduction Margaret S. Archer and Jonathan Q. Tritter

Rational choice could plausibly lay claim to being the grand theory of high modernity. Its metanarrative is fundamentally about the progressive rationalization of the West and then the rest of the world. Only insofar as global forces for disenchantment are massing can this theory take such giant imperialistic steps forward. Yet immediately we have to pause over cause and effect. Is it the case that the world is changing independently, in such a way that rational choice theory (henceforth RCT) is becoming increasingly appropriate for capturing its workings? Alternatively, is it rather the case that the growing influence of this approach in international economic and financial institutions, such as the OECD, the World Bank and the IMF, is now shaping larger tracts of the world in line with the theory? This volume brings together a range of scholars who address different aspects of RCT. It has been spurred by a sense of the increasing influence of RCT upon explanations about society in a range of academic disciplines, and growth in its proclaimed relevance to local, national and global policy-making. RCT has underpinned the neoliberal reforms of the public sector in much of the Western industrialized world. With these reforms have come the rollback of the traditional social welfare state and withdrawal of its services. The justification for applying RCT to policy-making continues to be the more efficient provision of services. This has resulted in the intended decline of expenditure by the public purse and instead cost-shifting to the grey sector. Thus the welfare services which we have come to expect as part of what is provided by the public sector are now increasingly dependent on altruistic acts of charity. The attraction of neoliberal reforms is readily apparent, but so are its problematic outcomes, and this is both the achievement and the threat of unquestioned application of rational choice theory. The spread of neoliberalism raises the important question as to whether the progressive colonization of public policy by RCT is not in practice dependent upon it being supplemented by something antithetical to it, namely altruism and its practical outworkings – the voluntary creation of collective goods. Thus, for example, there has been an increase in the extent of services provided by charities and voluntary organizations that have become embedded in

2 Margaret S. Archer and Jonathan Q. Tritter what was the public sector. In 1997–8 Health Authorities in England spent £500 million on services provided by the voluntary sector, 70 per cent of which was for the provision of residential care (CAF 2000). In Britain, Macmillan Cancer Relief provides a classic example of this phenomenon. By 1998 the charity had funded 1700 Macmillan nurses and 200 Macmillan doctors; these personnel, together with the charity’s information service, had contact with over 200,000 new patients and their families. The charity provides funding for Macmillan nurses and their training and has financed the modernization and improvement of cancer wards and the building of hospices (Macmillan 2000). Thus the public provision of cancer care in Britain is dependent on healthcare professionals and clinical settings paid for by charitable donations and overseen by a voluntary organization. This shift in the nature of the funding and delivery of public services is increasingly apparent in Britain and around the world. But is it a consequence of changes in society which make RCT more readily applicable, or is it due to the colonizing of policy-making by economic theory? If colonization is indeed the answer to reductions in public service provision, then a contradiction emerges. Such policies are ultimately reliant upon supplementation by the altruism of voluntary organizations. The ‘success’ of neoliberalism is thus parasitic upon forms of motivation which it cannot comprehend. In other words, the undergirding notion of instrumental rationality can neither embrace nor explain the altruism upon which it depends in practice; that is the expansion of the indispensable grey sector. This is because the attempts made by rational choice theorists to account for altruism all fundamentally result in defining it out of existence. Charitable impulses are simply how some individuals gain their satisfaction. Hence the pursuit of altruism becomes as selfish as that of any other preference. Do not ask for its wellsprings, for these are a matter of taste, like a preference for raspberries over strawberries, which is taken as given on the basic principle that de gustibus non est disputandum. Yet we would question whether altruism does not indeed have its sources in human commitments to moral values which override selfish concerns. In this case, altruism remains authentic and the Macmillan Nurse or his/her sponsor cannot be clawed back into those with a simple personal taste for the seemingly altruistic. Yet, if it is the case that neoliberal colonization only ‘succeeds’ insofar as charitable impulses plug the gaps in the public provisions made, then we must seriously question the sheer technical efficiency by reference to which an RCT-based approach recommends itself to public policy-makers. This question is particularly pertinent because rational choice is a maverick amongst grand theories. It has not attained its influence by projecting a vision of world history which wins ideological adherents through the power of its totalizing image. On the contrary, the impetus towards totalization advanced dramatically from exactly that moment when RCT shed its overt ideological baggage. Certainly, at its origins during the British and French Enlightenments, it was not lacking in ideologues of post-traditionalism – namely social

Introduction 3 contract theorists in politics, liberals in political economy and Utilitarians in public policy. Admittedly too, commentators have not ceased to note that there is rather more than an ‘elective affinity’ between philosophical individualism and the methodological individualism which forms the explanatory charter of rational choice theory, and that together these undergird and promote neoliberalism. Nevertheless, the totalizing impulse gained ground, in Habermas’s terms, precisely when RCT, as an empirical-analytical theory, gave up explicit ideological posturing and won out by naturalizing its framework of analysis. Its influence then expanded through serving the seemingly neutral interests of technical control. Over the last half century, rational choice theory has forged ahead on ‘purely’ instrumental grounds, by demonstrating its empirical successes in an increasing variety of fields. It has done this through the simple analytical manoeuvre of taking ‘instrumental rationality’ to be the dominant and universal mode of human decision-making. Because the theory is totally permissive about what substantive ends we may seek to pursue, and only offers to model how individuals maximize their net utilities within constraints of cost, time, information, effort, and so on, it recommends itself in wholly technical terms as proffering a powerful helping hand. Paradoxically, this technical neutrality has accounted for its spread. The growing imperialism of RCT has rested upon its pragmatic appeal and not on its neoliberal allure. Yet there is no doubt that expansionism is prominent on the RCT agenda: the ‘economic approach’ only regards its pre-eminence in microeconomics as a showcase for successful techniques which can be generalized to other disciplines. Thus Gary Becker (1976: 8) sets out his colonial mission: ‘I have come to the position that the economic approach is a comprehensive one that is applicable to all human behaviour.’ Undeniably, evangelization is going well, since we now encounter rational choice theories of religion itself. Obviously, ‘all of human behaviour’ embraces whatever chunks of it have been addressed by sociologists of different persuasions. The point therefore is that the whole field of sociology is seen as ripe for a take-over by RCT. The proffered framework and its techniques are not taken up merely by those who tackle micro, meso and macro social problems which respectively approximate most closely to consumer behaviour, the theory of the firm or market mechanisms. Now it is very doubtful indeed whether sociology has ever had a dominant, let alone a single, paradigm, which is the first necessary premise for it fitting Kuhn’s portrayal of paradigmatic shifts, held to be constitutive of revolutionary science. RCT, at least to those who share Becker’s aspirations, bids to become such a paradigm. If so, then in Kuhnian terms it must demonstrate how it can cut a swathe through the accumulation of anomalies which dog current social theorizing. For those less than convinced by the Kuhnian model of theoretical transformations, then on Lakatos’s version, a progressive rather than a degenerative research programme is one in which the former can explain all that the latter could, plus some more,

4 Margaret S. Archer and Jonathan Q. Tritter which means that it can eliminate certain old anomalies by something better than ad hoc means. Now there is no doubt that current sociology has its central problems, and there is a surprising degree of consensus about what these are, as distinct from how to solve them. Thus contenders as different as social constructionists, structuration theorists and critical realists show considerable agreement over the ‘big three’: how to link, transcend or otherwise cope with the dichotomies represented by ‘subjectivism and objectivism’, ‘agency and structure’ and ‘synchrony and diachrony’. If rational choice theorists are making a take-over bid, then a theory which claims to be applicable to ‘all of human behaviour’ cannot avoid taking on these major anomalies. The litmus test of its promise in sociology has to be its demonstrable superiority in confronting these central problems of social theory. Thus the book is divided into three parts which correspond to the three problems listed above. What we examine and query in each part is the adequacy of RCT in relation to the ‘big three’ problems. Since we are a diverse group of theorists with very varied substantive preoccupations, we are not defending or advancing any particular theoretical viewpoint, its derivative explanatory programme or associated practical social analyses. Instead, from our different affinities and interests, we simply set ourselves the common task of exploring the decision-making process. We are thus playing on rational choice terrain because we are drawing upon those parts of our work where in principle that theory should have most to contribute. The ball then is in the RCT court; what we want to know is what returns its defenders make. In particular we will have our eyes firmly fixed on their strategies for keeping the ball in play rather than ending at the bottom of the net – three times over.

Rationality Succinctly, the problem of ‘subjectivity and objectivity’ arises because of the ontological differences between the two main constituents of the social world. Many of the ‘agents” properties and powers are subjective in nature and entail their capacity to entertain meanings and to act in relation to them, whilst the ‘parts’, which constitute their structured environment, have objective properties and powers to constrain and enable action. Both therefore and necessarily have to feature in any account of action, such as decision-making. The problem consists in how to combine them. Human meanings need to be subjectively apprehended and hermeneutically understood, whereas environmental influences have to be objectively ascertained and causally explained. Attempts to by-pass the problem by construing all as meaningful (ethnomethodology) or everything as causal (behaviourism) have merely served to reinforce the indispensable contribution of the element which they suppress. Therefore the rest of us, whatever our theoretical persuasions, confront the fact that there is an irreducible hermeneutic element, which means that our research cannot dispense with

Introduction 5 the meaning that actions have for the agent in question. Also there is an equally irreducible contextual element, whose consequence is that neither can any research dispense with objective features of the situations in which action takes place. There are various solutions on offer to this problem within sociology, but these will not detain us here, for our concern is to know what solution rational choice theorists bring to sociology and then to assess its adequacy. In a nutshell, their solution consists in acknowledging agents’ meaningful values and goals (aka ‘utilities’ and ‘preferences’), which they seek to maximize in the outside world, whose constitution attaches various ‘costs’ to their realization. The bridging concept which enables hermeneutics and causality to be linked is rationality. Thus, in Weberian terms, interpretative understanding consists in identifying the rational element in human action, which lies in agents’ own reasons for what they do, and then considering the rational way to maximize their subjectively defined ends in the light of the objective constraints and enablements which they confront in the environment. Thus, as Hollis (1987: 7) summarizes the matter for RCT, the ‘category which lets us make most objective yet interpretative sense of social life is Rationality’. The basic difficulty is that rationality immediately boils down to ‘instrumental rationality’. The reason for this is straightforward: ‘substantive rationality’ (Wertrationalität in Weber) is not a matter of cost–benefit analysis at all, for the pursuit of a value is oblivious to the costs involved or of any prospect of external success. It is in this sense that Weber (1964 [1947]: 11) wrote that ‘absolute values are always irrational’.1 Thus the adequacy of the RCT solution turns exclusively upon the validity of the proposition that all human agents, all of the time, are beings who work solely in terms of ‘instrumental rationality’. Supposedly they are eternal bargain-hunters in their environment, even though the specific bargains sought are subjectively defined and can consist of any goal which an individual seeks to maximize. Be these ends self-regarding or philanthropic, cost–benefit maximization means that individuals will never pay more than they need nor settle for less than they must. There are three major objections to this model, which are explicated in Part I. As a bridging notion, the proposed solution to the problem of ‘subjectivity and objectivity’, namely ‘instrumental rationality’, is presumed to be a universal characterization of human action in its context. The force of Peter Wagner’s contribution is to deny its universality and to point to the specificity of the socio-historical configurations whose objective properties underwrite the local and temporal salience of ‘instrumental rationality’. Only when profound social disruption occurs, especially when accompanied by extensive physical relocation, are thick collective registers of evaluation fragmented; then, by default, individuals with their selfdefined instrumental rationality stand forth as the principal political actors, raising the new problem of how their activities sum in relation to the well-being of the body politic. America is presented as the exemplar, and consequently it is no surprise that this is the home-base of

6 Margaret S. Archer and Jonathan Q. Tritter RCT. Yet necessarily this represents a radical challenge to the supposed universality of ‘instrumental rationality’. What Wagner is stressing here is the existence of other social configurations in which ‘traditional action’ shades over into wertrational action, as Weber (1964 [1947]: 117) admits is the case, thus denying hegemony to the Zweckrationalität. Therefore, pure ‘instrumental rationality’ will rarely be found (unaccompanied by value commitments and emotionality), and even if it were, we would be justified in questioning its claim to be the universal exemplar of rational human action. This Archer does by querying whether all of our actions are means to some indefinitely deferred point at which we become ‘better off, or rather whether we agents do not have ‘ultimate concerns’ (wertrational) which are simply expressive of who we are – a terminus rather than a stepping stone. Where ‘ultimate concerns’ are involved, cost– benefit analysis is inappropriate. In actions like supporting one’s friends or loving one’s partner and children, we are not seeking to maximize anything. These actions are about who we are and what we care about most. Williams’ chapter is complementary with this argument, because could we be said to care deeply about anything without having some emotional attachment to it? We could not properly say that anyone was devoted to his or her art, wife/ husband or career without the assumption that considerable affect is involved. Conversely, if ‘instrumental rationality’ is devoid of emotionality, then what supplies the ‘shoving power’ which motivates the maximizer? It would be a very odd kind of ‘preference’ whose maximization supposedly leaves us ‘better off, but whose attainment leaves us feeling we ‘couldn’t care less’. Not even the miser contemplating his pile, with no intention of spending it, approximates to this state, since archetypically he is held to gloat over it. Putting these contributions together, then, the objective environment cannot be reduced to the costs incurred in maximizing our independent and individual preferences, since the sociohistorical context shapes preference formation itself (Wagner). Nor can our subjectivity be reduced to cost–benefit analysis, since as moral and affective beings we pursue other goals in other ways (Archer and Williams). In short, ‘instrumental rationality’ is not a bridge which can fully span the subjective and the objective domains, nor is it sufficiently broad to accommodate all the trafficking between these two domains. We conclude that the RCT concept of rationality is defective in itself and deficient as a solution to this first central problem of sociology.

Individuality Even more succinctly, the problem of ‘structure and agency’ can be expressed as ‘the vexatious fact of society’; namely that agents are ultimately responsible for the formation and transformation of society, but they are themselves partly formed and transformed in the process of shaping society. How do we disentangle this intertwining? On the one hand, the

Introduction 7 properties and powers of people are very different from those of structure (thus a person can be cowardly, which an institution cannot be, but an organization may be centralized, whilst an individual cannot be). This difference in properties and powers could well be the basis for distinguishing between structure and agency. However, what makes for the problem is that structures and agents are interdependent, yet their properties are irreducible to one another, so some means has to be found for theorizing their mutual influence and interplay. To maintain that reductionism is not a solution is to reflect upon the long-drawn-out and ultimately sterile debate between those who held that individuals were the ‘ultimate constituents’ of the social world and those who conversely championed that ontological status for ‘social facts’. Generically, this debate between methodological individualism and methodological collectivism has been left behind in sociology, in favour of a new debate about non-reductionist solutions, specifically between those who seek to transcend the dichotomy between structure and agency, by conceptualizing them as mutually constitutive, as in structuration theory, versus those who treat the two as distinct and emergent strata and then examine the interplay of their causal powers, as in critical realism. However, the point here is not about the relative merits of the participants in the new debate. Instead, it is to note that RCT still holds to methodological individualism and thus encounters all the difficulties traditionally associated with it. Specifically, RCT is based upon atomistic, individual decision-makers, who are not only regarded as their own ‘sovereign artificers’ but are also credited, along with others like them, with making their society too – through the aggregate effects of their decisions. The question thus becomes: does RCT have anything new to add which buttresses the defence of methodological individualism? In particular, what can RCT contribute to the two conundrums, the perennial rocks upon which reduction to the individual has come to grief? Firstly, how individual is the methodological individualists’ individual; for is it not the case that structural and cultural properties have to be incorporated into his or her constitution? Secondly, can the social structure be reduced to the aggregate doings of ‘other people’, that is, ‘without remainder’, in Mandelbaum’s (1973) phrase? Tritter’s chapter on the decision of schools whether or not to opt for self-management is a classic example of the continued relevance of the methodological collectivist critique, and makes two very pertinent points about the ‘perennial rocks’. On the one hand, he asks why, when ‘opting out’ was legally designed to be a collective decision of the school, did it effectively turn out to be taken by the headteacher? Instead of seeing head-teachers as atomistic individuals, Tritter finds it impossible to understand their decision-making powers without making recourse to the collective concept of power, and its use of the ‘manipulation of bias’ (Lukes 1974) to organize parents out of the decision-making arena. On the other hand, when examining the five reasons volunteered by headteachers to account for their preference for opting out of local authority control, he finds that not one of these attitudes can even be

8 Margaret S. Archer and Jonathan Q. Tritter identified without reference to collective entities, such as funding agencies and central or local authorities. These cannot be construed as ‘other people’, as is indubitable when he finds that one of the entities towards which this decision-making is oriented is its future impact on school league tables. This nicely illustrates Gellner’s (1968: 259) contention that one good reason why we cannot reduce collective entities away is precisely because people think and act in terms of them. That the interplay between structure and agency can be reduced to the RCT formula of individual goals as modified by contextual costs is now in doubt. However, Parker gives this RCT strategy a fair innings when he considers students’ shifting orientations to their studies over the three-year degree period. The first year’s ‘push’ to do well enough to pass, even when reinforced by the debts that most students have contracted, is shown to be overridden by peer group norms of socializing, enjoyment and overspending. Group variables have intervened and normativity again undermines the atomistic character of the individualistic Homo economicus. Equally, can the finalists’ greater commitment to hard work simply be attributed to the putative resurgence of personal desires for a good degree, plus the more stringent costs of failing to do so? After all, a new set of normative expectations have arrived on the scene or have gained in salience: bank repayments, parental hopes, employers’ requirements and staff encouragement. Nothing has shown that these group variables are inefficacious in galvanizing the ‘pull’ towards academic success, a preference which has already been shown to have considerable plasticity for the majority of students. In relation to the ineliminability of ‘group variables’, Peggs and Lampard’s discussion of the reasons their interviewees supplied for their decisions about marriage, divorce and repartnering is of particular interest. This is because if any example should conform to the prototype of decision-making whose parameters are confined to ‘other people’, then these interpersonal relations should surely constitute the prime case. Yet, on the contrary, many of their respondents were found volunteering that their actions had been consciously influenced by social norms and values, or that these gendered relationships were incomprehensible without introducing power and domination, deriving from wider society and shaping the situations in which couples found themselves, rather than these factors being generated by the actors involved. Potentially such findings predispose in favour of Zafirovski’s (1999a) reformulation of a distinctively ‘sociological rational choice’ model. Now, undoubtedly, this multivariate approach, which incorporates macro-variables, such as norms and values, and also political variables, like power and domination, does indeed break with methodological individualism. Nevertheless, the question remains as to how successful Zafirovski is in coping with the ‘problem of structure and agency’ from within his radical revision of the rational choice framework. Basically, both micro- (individual) and macro- (structural) features now appear additively within his new formula (Zafirovski 1999a: 59). This is still a formula for rational

Introduction 9 choice as these are all goals subject to maximization, and thus one can still ‘speak of strict rationality, optimality or optimizing behavior’ (Zafirovski 1999a: 104). However, this additive procedure ducks the whole question of ‘structure and agency’. After all, norms and power cannot seek to maximize themselves (reification apart): they are only operative through people, so the entire problem resurfaces. At least ‘classical’ RCT tried to show how, by analysing structural factors (defined aggregatively) as variable ‘costs’, actors then had to take them into account when optimizing or satisficing their individual preferences. In Zafirovski’s model, these structural factors are bundled in with people as part of the predictive matrix for decision-making; as if ‘structure + agency’ could substitute simple summation for the complex interaction between the two. Zafirovski ends up, basically, with a multi-factoral theory, still grounded in the positivist-predictive tradition, which has forsaken the respectable enterprise of RCT, namely to account for how agents maximize their goals in the face of structural constraints. The latter at least recognized that the interface between structure and agency required examination, even if the most severe deficiency of the approach in relation to the problem of ‘structure and agency’ was that RCT held the former ultimately to be reducible to the latter. It is the intransigence of this problem within RCT which leads Cruickshank to pass over Zafirovski’s (1999a: 47) ‘creative destruction’ of RCT and to recommend its abandonment in favour of a realist approach. He maintains that in RCT the agent remains incurably essentialist: his or her preferences are what they are, de gustibus non est disputandum, and there is nothing more to be said – least of all about preference formation and reformulation and how structure enters into this. Such essentialism also entails determinism at the level of individual agents: knowledge of their stable preferences tells us what they will try to do. Because structure cannot be conceived of as possessing emergent causal properties and powers, due to RCT’s commitment to methodological individualism, then it does not interact with agency, at the level either of preference formation or of realization. Consequently RCT leaves us with an incorrigible model of passive agents, programmed for action by their preferences, who merely encounter their environment as ‘carrots and sticks’, which facilitate or impede the maximization of these stable preferences. Thus, holding to the RCT linkage of ‘structure and agency’ basically degenerates into latter-day behaviourism. Only if we reject the individualists’ individual can we restore active agents who are reflective, rather than simply calculative, and allow, as Peggs and Lampard, Tritter and Parker show, that one of the things they are reflexive about is their social structure. Only if we reject structure as the aggregate effect of individual doings and grant it sui generis emergent properties can we examine the dynamic interplay of two sets of causal powers (those of ‘agency’ and those of ‘structure’) over time. This would serve to explain not only how structure pre-groups collectivities of agents involuntarily, in terms of society’s distribution of scarce resources, and thus conditions both their opportunities and aspirations,

10 Margaret S. Archer and Jonathan Q. Tritter but also how agents, both individually but especially collectively, can subsequently act back to transform their social context for themselves, at later ages, and for the next ‘generation’ of agents.

Temporality Thus, the last ‘central problem’, that of temporality, has already issued itself in. Basically, the sociological problem is less about time than tense. Almost no-one today would advocate an atemporal sociology because the detection of any pattern of stability or change entails the passage of time. Nevertheless, time can often be seen as a passive medium through which events take place, rather like water for swimming, whereas in fact time is a variable intrinsic to any theoretical framework. Whether implicitly or explicitly, theories are ‘tensed’ by their variable incorporation of past, present and future. The basic problem is how many tenses the discipline needs to use, given that time goes by in all of them. Because of the methodological individualism upon which it is based, RCT is committed to the ‘autonomy of the present tense’ (a generous form of synchronicity) when dealing both with the context of decision-making and with decision-makers. Thus, as far as the context is concerned, since society is made up of people, there is nothing in the social environment which people cannot change. To Watkins (1968: 271), ‘the central assumption of the individualist position ... is that no social tendency exists which could not be altered if the individuals concerned both wanted to alter it and possessed the necessary information’. Here, structural properties have become the effects of contemporary action, for whatever the origins of the structural tendencies observed, their present existence is due in some way to the people there present. In other words, the present is severed from the past, for there is no allowable sense in which historically emergent structures (such as resource distributions) can be granted causal efficacy without, as it were, the sanction of contemporary ‘other people’. The present is also cut off from the past for the agent. The bed-rock of explaining their decision-making is their ‘preferences’. Yet this presupposes the possibility of identifying preferences prior to their manifestation in a social context. Such a presumption entails the peculiarity that ‘it seems to preclude a priori the possibility of human dispositions [such as their preferences] being the dependent variable in an historical explanation – when in fact they often or always are’ (Gellner 1968: 260). The five contributors in Part III will be found agreeing and thus seeking a more broadly tensed theory to account for preference formation. Equally, since all five authors deal with preference transformation too, they cannot work in terms of a theoretical framework which also severs the present from the future. Thus, if preferences can never be the dependent variable, capable of transformation through accumulated experience and changing circumstances, then an agent lives in an eternal present, governed by preferences which are set in concrete. De gustibus numquam est

Introduction 11 disputandum. Thus, ‘one does not argue over tastes for the same reason that one does not argue over the Rocky Mountains – both are there, will be next year, too’ (Becker 1996: 24). Even if preferences were to remain so resilient to change, which all our contributors contest, then the fact that structural transformations are never allowed a future tense either, such that they could feature as the new initial conditions of a subsequent causal sequence of action, should not be immaterial to RCT. Were preferences indeed to remain constant, nevertheless ‘costs’ undergo change over time, due to the consequences of present-tense decisions, and will thus necessarily change outcomes in the future. Yet, adherence to methodological individualism prevents rational choice theorists from countenancing the development of emergent properties from interaction in the present which will then exert their new causal powers upon agents in the future – be they the same people at a later date or a new generation of agents. Although unintended consequences can be traced back to their individualistic origins (which is the aim of reduction), to concede their influence in future time would be to accord them sui generis properties and powers, ones which are not ‘recoverable’ by determined and knowledgeable individuals in the present tense. The chapters by Fine and Hirsh and by Ratcliffe show how indispensable the contrary assumption proves to be in order to explain how, over time, contextual changes profoundly alter the decisions made in the ‘present’, compared with those made in the ‘future’, wherever these are situated in historical time. The first two chapters in this Part (Procter and Wolkowitz) deal with preference formation and transformation from the point of view of the agent. What they are basically challenging is RCT’s ability to incorporate time into the initial specification of preferences and their revision. This implies that preferences have a history and a trajectory, rather than being as stable as the Rocky Mountains. As far as the past is concerned, Procter and Wolkowitz respectively question the two elements in the term ‘individual preference’. Procter shows how Becker’s latest attempt to build our past experiences and early socialization into an ‘extended utility function’ is inconsistent with RCT, in one of two ways. Either the incorporation of references to socialization is not consistent because the resulting preferences cease to be the agents’ own calculations of utility. Or inconsistency arises because the introduction of choice of preferences, rather than the usual calculation and comparison of utilities deriving from stable preferences, cannot take the form of rational choice. This is because such decisions entail judgements about ‘overall lifestyles’, about whom and what we want to be – thus echoing Archer’s main point. Such judgements necessarily invoke substantive rather than instrumental considerations about what it is morally worthwhile to become as a person. But then these are not rational choices, according to the canon. Wolkowitz, on the other hand, contests the individual nature of preference formation. She shows that the needs of others are not just externalities which are taken into account, but rather that the family unit engages in collaborative decision-making. Where the decision to work at

12 Margaret S. Archer and Jonathan Q. Tritter home is concerned, she demonstrates that those whose biographies have left them childless do indeed conform more closely to cost–benefit analysis. This therefore makes conformity to the RCT model contingent upon being able to act as a lone individual. Secondly, both authors raise a serious question about the role played by the acquisition of information over time, and thus about agents’ approximation to having ‘perfect information’. According to RCT, agents cease searching for more knowledge at the point where the utility to be gained from a known prospect outweighs the marginal cost of further searching. Instead, both Procter and Wolkowitz stress that certain information can only be acquired by experience: the only way to know is to do. However, the information generated through our experiences can transform our preferences and thence our outlook on the future. No-one can know what a marital relationship is like without undertaking one (Procter), and no-one can know that homeworking may result in loss of confidence and deep feelings of isolation (Wolkowitz) unless they have tried it. Had they known that exercising their past preferences would later result in divorce or depression, they would presumably have (rationally) renounced them, but they could not know these things. Yet they have become different people in the process of learning them, which also makes it more difficult to rectify their preferences and to realize them in the future. The divorcee is a less attractive partnering prospect because of his or her past relational failure; the homeworker pays for his or her relational decision to step off the career ladder by the loss of his or her network, which makes it more difficult to step back on. Thus, past, present and future constitute a trajectory of preference transformations which make the Rocky Mountains seem decidedly plastic. Turning to the context of decision-making, Fine and Hirsh show that Nazi soldiers were not those with prior preferences for antisemitic activities, who merely ‘awaited’ circumstances in which their hatred of Jews could be maximized at tolerable costs to themselves. Instead, the rise of the Führer principle represented a new context which transformed people themselves, and with them their preferences. ‘Ordinary men’ temporarily became extraordinarily hardened mass murderers. This Nazi ‘present’ was discontinuous from the soldiers’ civilian past, precisely because a stable preference schedule could not be presumed. Similarly, the future was discontinuous with the present for the same reason. In the political structure of the post-war period, former mass murderers were not found seeking further ‘outlets’ for their acquired anti-Semitism, but, in one notorious case, settled down as a routine clerk for British Rail. Here, preferences show a malleability which can only be understood as produced by the reflexive mediation of the influence of circumstances, rather than as permanent dispositions whose expression became feasible in a period of circumstantial change. Exactly the same point is made by Ratcliffe. He begins, like Cruickshank, by challenging RCT’s essentialist assumptions. Here, this concerns the assumption that different ethnic groups were each the bearers of some homogeneous ‘traditional culture’, which underwrote

Introduction 13 the stability of their housing preferences over time. Instead, Ratcliffe highlights that these preferences are transformable, given that contextual change allows new goals to be formulated and pursued, following modifications in legal or public policy. Thus new strategic decisions about housing emerged from those ethnic agents who were not stringently constrained; and, in turn, these began to transform the housing market. This then became the new context for preference transformation, or for preference formation amongst the next generation of (modified) agents, now living within a modified structure. The whole process is therefore ‘tensed’ through and through. Static, essentialist snapshots only serve to perpetuate stereotypes of stable preference orders which deny the creative reflexivity of agents, as they respond to changes in their circumstances over time.

Rational choice theory in the real word We began this Introduction by outlining the ways in which RCT has increasingly come to influence thinking in academia, public policy and business. We went on to suggest that the presentation of RCT has provided a framework for summarizing and explaining human action that is insidiously attractive but also highly problematic. One classic critique of RCT that has not been a key part of the work collected in this volume is that provided by altruistic acts, given that such actions cannot properly be presented as the maximization of philanthropic preferences. Yet altruism plays a key part in both everyday and life-chance decision-making. Such decisions are problematic for RCT because they draw either upon social normativity or upon sources of morality, both of which should play no part in the constitution of Homo economicus. These may be better understood by examining the voluntary, sector, which now employs 485,000 people in Britain (NCVO 2000). Similarly, charitable donation, at least in the United Kingdom, has increased dramatically in the last decade, despite the work of the National Lottery (CAF 2000). The grey sector, that part of the economy which does not seek a profit yet still provides key services and employment for thousands, is thus growing in importance. But why would an individual choose to seek a career in this sector instead of in more traditional areas? Employment within the grey sector is dominated by low salaries, often less security of employment (as employers in this sector are typically dependent on charitable donations or ‘project’ funding) and only limited opportunities for promotion. The individual who moves into the grey sector cannot hope to maximize his or her earnings or ‘career’ opportunities. Instead he or she must be motivated, at least in part, by altruism, which cannot be reduced to an individualistic yen to ‘do good’ since it implies reference to some collective belief system which prioritizes the well-being of others over self-promotion. Similar arguments could be put forward about working within the public sector, as a school teacher or tax collector. There has long been recognition that pay in the public sector lags

14 Margaret S. Archer and Jonathan Q. Tritter behind that of the private sector. For instance in higher education in the United Kingdom, since 1981, compensation for staff has only increased by 1 per cent in real terms, while average earnings have increased by up to 40 per cent (AUT 2000). Public sector reform has also removed the job security that was often seen as the compensatory motivating factor in the public sector. The introduction in the United Kingdom of quasi-markets and Executive Agencies, with their separate employment bargaining arrangements, has destroyed what was once seen as a united ‘civil service’. Yet people still choose to make their careers in the public sector; they are still motivated by a public service ethos. The difficulty of rational choice theorists in explaining altruism has long been seen as a weakness. What we are suggesting is that altruism plays a much larger role not only in informal terms but also in structural ways in contemporary society. We have suggested that the grey economy based on the work of voluntary organizations, charities and nongovernmental organizations makes a substantial contribution to the economy as both a service provider and employer. We suggest that research examining the grey sector can provide important evidence of the inadequacy of RCT in explaining people’s motivation either in their allocation of monies or in their decisions about jobs. Discussions of altruism in terms of voluntary organizations give rise to debates around public goods, a conception that is central to the economic theory of government (Speth 1999), yet difficult to incorporate into the framework of RCT, for, as Bourdieu (1998: 105) puts it, the public interest, whether one likes it or not, will never, even by juggling the figures, be produced by the accountant’s view of the world (once one would have said ‘grocer’s’) which the new belief [neoliberalism] presents as the supreme form of human achievement. The provision of public goods is based on the needs of the collective, traditionally nation states, and provides the basis for the functioning of civil society. Further, public goods cannot be produced ‘Without a mechanism of collective action’ (Kaul et al. 1999: xx). The production of public goods does not maximize any one individual’s needs, but without such goods there is little chance of their maximization. Recently, discussions of public goods have generated a conception of global public goods (Kaul et al., 1999; Reinicke 1998). These become more difficult to attain as the globalization of business increasingly leads to conflict between the interests of multinational corporations and the priorities of national governments. One central challenge is the fact that what is and what is not in the collective or public interest often differs across countries. Hence for global public policy to be successful, not only would participating societies have to show a willingness to cooperate in the

Introduction 15 establishment of internal sovereignty; in practice they would have to recognize and accept others’ notion of the collective interest and others’ definition of the public good. (Reinicke 1998: 86) Rational choice theory and neoliberalism, when applied to the public or private sphere, have stressed the need for the definition and measurement of individualized outcome indicators. Such an approach promotes models and predictions of action that increasingly become defined by what they can measure. As we have already suggested, such an approach, which prioritizes instrumental rationality above all else, individualizes, isolates and insulates decisions and actions from a social and historical context. More importantly, it must miss those key aspects of individual and collective action that cannot be measured. The worth of education provided by a teacher becomes distilled into the measurable outcome of pupil exam performance, with little reference to the nature of the pupil–teacher interaction that is at the heart of learning. Absent from this conception of teaching and learning is an understanding of the ways in which values and meanings are embedded in education and define individual and communal relationships and personal well-being. These aspects of the social have always been fundamental objects of sociological inquiry. Rational choice theory presents a blinkered model of the world and urges decision-makers to take account of only those issues that can be measured, thus undermining the collective nature of society. As Bourdieu (1998: 98) suggests, This comes through the setting of individual objectives; individual appraisal interviews; personal increments or bonuses based on individual competence or merit; individualized career paths; strategies of ‘responsibilization’ ... all methods of rational control which, while imposing over-investment in work ... combine to weaken or destroy collective references and solidarity. The complexity of society and the global system leads us to seek a theoretical framework that can make the world comprehensible. Neoliberalism and RCT present themselves as such a framework. While potentially providing a useful tool for conceptualizing aspects of social action, however, they fail to incorporate the complexity of the measurable and immeasurable aspects of society. Altruistic acts, at an individual, collective or global level, provide a clear illustration of such limitations. We have argued that contemporary society is evolving in such a way that altruism and voluntarism are becoming more central and enmeshed in the interstices of the public and private sphere. Simultaneously, globalization is altering our traditional models of the public sphere and the relations between a state and its citizens. Such developments raise issues of central importance and interest to social science, but understanding their nature requires us to reach beyond a particular economic theory of action.

16 Margaret S. Archer and Jonathan Q. Tritter This volume is the product of two years of meetings, discussions and seminars between members of the Department of Sociology at the University of Warwick. As a group we came together with research experience in different substantive and theoretical areas but with a common desire to contest the ways in which RCT explained our areas of expertise. Our critical perspective is not limited to academic and disciplinary concerns but also encompasses the growing colonization of political and public policy thinking and action. Rational choice theory, as a key explanatory model for human action, has developed and spread rapidly both within academia and more broadly in the public and private sectors, as Beckford rightly stresses in his concluding chapter. The contributors to this volume have wrestled with the ways in which such an explanatory model provides an attractive framework for thinking about problems but at the same time is both inadequate and dismissive of many of the key questions sociologists ask. We are left pondering whether the growth of RCT in academia led or followed its colonization of the political sphere, but such concerns will do little to help develop more appropriate theoretical explanations of action, nor to resist the process of colonization in the real world. Note 1 The quotation continues, Indeed, the more the value to which action is oriented is elevated to the status of an absolute value, the more ‘irrational’ in this sense the corresponding action is. For, the more unconditionally the actor devotes himself to this value for its own sake, to pure sentiment or beauty, to absolute goodness or devotion to duty, the less he is influenced by considerations of the consequences of his actions. (Weber 1964 [1947]: 11)



The bird in hand Rational choice – the default mode of social theorizing Peter Wagner

Entangled in words, as a bird in lime-twigs. (Thomas Hobbes, Leviathan)

Of sparrows and doves There is a saying in the English language that holds one bird in the hand to be worth two in the bush. The recommendation presupposes an order of preferences in which, firstly, a ‘bird’ is considered to be a good worth having and, secondly, the possession of this good in increasing quantities has a positive utility attached to it. There is nothing extraordinary about this assertion thus far; economic theorizing has long formulated such propositions and elaborated them further. The interesting part of the recommendation comes with the comparison between actually having a certain quantity of the good, on the one hand, and having a greater quantity in view, on the other. In the metaphorical language characteristic of everyday wisdom, ‘hand’ stands here for actual possession, whereas ‘bush’ refers to present unavailability and potential but uncertain future possession. From the point of view of the preference-holder, the difference in the spatial location of the goods is one of accessibility. For reasons that will become clearer in a moment, let me underline that a difference in quantity of the good is here related to a difference in accessibility, and that the question of accessibility itself is confined to ‘having’ and ‘not now (or yet) having’. On the face of it, there are no further significant connotations that come with ‘hand’ or ‘bush’ in this context.1 Before moving on, an observation on the status of this proposition is in order. The recommendation is clearly one of prudence, but this does not necessarily imply that the social world in which it is uttered and accepted is one in which prudence reigns. In general, everyday wisdom is known to be internally contradictory. A counter-recommendation, for instance, is contained in the proverb ‘Nothing ventured, nothing gained’. The existence of such

20 Peter Wagner contradictions can be interpreted in a variety of ways. Often, it is assumed to be inappropriate to expect coherence and consistency from a body of knowledge which has been accumulated naturally, such as everyday wisdom. This is precisely what is considered to distinguish scientific-philosophical reasoning from common sense. Alternatively and more interestingly, however, one can see everyday wisdom as working with a plurality of repertoires. In that case, its single maxims would not be meant to be applicable across all kinds of situations; rather, their application is situation-dependent, or, more precisely, dependent upon the interpretation of the situation. In the German language, there is a saying that at first sight seems to make precisely the same recommendation as the English one just cited. Its literal translation is: ‘The sparrow in the hand is better than the dove on the roof.’ In this proverb, having a bird is also better than only hoping to obtain one. But this saying moves beyond triviality, not by increasing the quantity of the birds-not-possessed, but by introducing a qualitative distinction – between a bird of lower utility and one of higher utility. At the same time, the metaphor chosen to mark the distinction between that which is accessible and that which is inaccessible is also cast in different terms. While ‘hand’ (in lay Heideggerian terms) stands here, too, for what is available, conversely that which is not accessible is moved, we may be justified in saying, to a higher, a superior, plane – using the philosophically time-honoured and linguistically almost unavoidable association of altitude with value and importance. If, just for a moment and for the sake of developing the argument, we worked with a standard cultural-linguistic theory of social life and assumed that linguistic forms express cultural values, would there be an important difference between an ‘English’ society and a ‘German’ one, using these names as short-hand for societies whose members understand and accept, in principle, the one or the other maxim? Moving beyond the shared emphasis on prudence contained in these particular sayings, rather different registers of evaluation can be identified which stand behind the choice- and decision-making situation. The ‘German’ choice is between goods of different quality, whereas the ‘English’ one is between different quantities of the same good. Some may want to say that this aspect should not be exaggerated – after all, a bird is just a bird. And the fact that the one is larger than the other and that the one is widely considered edible and the other is not do not make much of a difference – or only a difference that makes the preferences assumed in these proverbs plausible, as is necessary for their proverbial status. If those points exhausted the differences between a sparrow and a dove, this objection would be valid. Either the difference is located on a common scale of size, that is, large/small (or more/less) and the size of a ‘good’ equals its value, or this difference is translated into the ability of the good to satisfy a universal human need, that is, the need for food.2

The bird in hand 21 Such a reading, however, would miss some significant interpretative possibilities. Firstly, it ignores the irony that often occurs in everyday wisdom. For all intents and purposes, a sparrow is no good at all, and this observation means that the proverb should possibly not be read as advice in a situation of choice and decision. Instead, the German saying may indeed be telling us that there are moments in which one has to resign oneself to quite a miserable life in spite of high-flying ideas and ambitions. Among the latter are peace, and also love and freedom, for all of which the dove can stand as a symbol. Because of that, we may also find explanations for why the German dove is on the roof, while the two English birds are in the bush. That which is inaccessible in the English saying is just in a different place, a site different from that of the subject and one which he or she could only reach by making an effort. In the German saying, by contrast, inaccessibility refers to a higher plane, where ordinary life does not take place, and from which one would look at life from a different perspective.3 Thirdly, the suspicion that there may actually be no choice or decision to be made in this situation can be read in the opposite way. If the saying is taken to mean that a sparrow in the hand is obviously better than a dove on the roof, then it may call for a halt to further inquiries. By digging further into what one’s preferences actually are and how they can best be satisfied, one may destroy any possibility of satisfying them. Ultimately, there may be no way to avoid the connotations of intimacy, since it is here that this aspect is most evident. If one keeps questioning oneself about whether the person you are with is the one you love, then one precondition of love and friendship, namely their existing unquestioningly, will cease to hold. Theorists as diverse as Maurice Merleau-Ponty, Hannah Arendt and Jon Elster (1986b: 13) have discovered this insight. I will call a halt here to further inquiries in comparative cultural meta-physics, since this is not the purpose of this chapter – worthwhile as such an effort could be, if it were pursued on more secure grounds than one constituted by two proverbs. This little comparative exercise was meant to direct attention to some features of ‘decision-making’ that are often neglected, and which I will now explore in more technical language.

Of rags and riches Two observations have been made about the relation between the two proverbs (and it will have become clear that they are indeed two proverbs and not, as they appeared to be and as dictionaries treat them, two versions of the same proverb). Firstly, I have suggested that they vary in their construal of the goods at stake. A qualitative difference in the one case turns into a quantitative one in the other. This raises the issue of the commensurability of these goods. If there is a situation of simple increases in marginal utility, then inhabitants of the ‘English’ world could probably indicate for which number of birds in the bush they would give up the

22 Peter Wagner one bird in the hand. But could ‘Germans’ conceivably give the sparrow up for, say, an eagle – that is, for (collective) power and glory instead of for peace and love? Possibly, but there would be no formal method, at a level comparable to marginal utility theory, of approaching the question. Eagles (as well as doves)4 may belong to a quite different order than sparrows, and the ones could not be traded against the others. This observation leads into a discussion about what Charles Taylor calls ‘hyper-goods’ as well as ‘irreducibly social goods’ (e.g. Taylor 1989, 1995), both of which are difficult to conceptualize from a rationalistindividualist point of view. Secondly, pursuing that kind of observation further, one could argue that a world in which there are sparrows, doves and eagles is a world quite different from one in which there are only birds existing in different quantities. In other words, the ‘Germans’ and the ‘English’ use quite different registers of evaluation when interpreting and judging the world in which they find themselves. The theoretical conclusion that follows here is that there may be differences between registers of evaluation such that they cannot be reduced to differences of preferences, at least not without further implications. Such a conclusion can be related to a Wittgensteinian emphasis on languages rather than on words or concepts, and to the idea of the ‘form of life’ that emerges from there. More recently, Luc Boltanski and Laurent Thévenot (1991) have attempted to show that disputes in contemporary Western societies are indeed approached from a number of different registers of evaluation, many of which have their own internal consistency and coherence, but which are incommensurable and thus mutually irreducible.5 At this point, we may ask what happens if one tries to make goods commensurable or to translate from one language or register to another, say from ‘German’ to ‘English’. This would be an effort at redescription, which, in Richard Rorty’s (1989) understanding, changes the world. In other words, when moving from ‘German’ to ‘English’, one may indeed make different goods commensurable and reinterpret differences in registers of evaluation as differences in preference-orders, but at the same time the situation is altered for those human beings who face the choices. Arguably, this is precisely what the application of rational choice theorizing does, and what such theorists normatively recommend human beings to do. To broaden the view on rational choice theorizing as a mode of social theory still further, I shall take up the two other observations made briefly above. I have claimed that with some kinds of ‘preferences’ it is essential for satisfying them that they should not be revealed in the situation of ‘decision-making’ – without this being a question of deception of others or even of oneself. Reasoning against the background of an ideal state of complete information, rational choice theorizing holds information to be a good, of which it is always preferable to have more. This good is weighed only against the cost of obtaining yet more information, which is conceptualized as a form of transaction cost. In contrast, the example developed

The bird in hand 23 above held that acquisition of further information may make a certain good unobtainable and a given preference unsatisfiable. Thus, it undermines the general validity of the concept of grounding rationality in information. As in the earlier two cases, following the advice from rational choice theory would considerably alter the situation of ‘decision-making’. This alteration, however, may be highly undesirable; and the degree of such undesirability might be untheorizable from the point of view of rational choice theory. A final observation relates to the question of the ‘situation’ itself. Building on the insight that singular human beings may avail themselves of more than one register of evaluation, the interpretation of a situation gains crucial significance. To elaborate further on one example, the relation to another person may be one of justice – for example, between colleagues at work – or it may be one of love (see, e.g., Boltanski 1990). The observably different registers of evaluation which human beings apply would have to be interpreted as a violation of the need for consistent preferences as presumed by a rational choice theorist. The only alternative, in order to avoid strongly counter-intuitive outcomes, is to claim that situations can unequivocally be defined as ‘given’ – that is, as one of justice or as one of love – and can then be treated unproblematically as the ‘context’ of decisionmaking. The rational pursuit of preferences ‘under constraint’ could thus again be made consistent. Situations, however, are often not unequivocally defined, and the interpretation of a situation may be exactly what is contested or in doubt. In this case, then, rational choice theorists would apply the maxim of scrutinizing both one’s preferences and the situation, since this is seen as the only maxim that is applicable to any situation. But it would invariably favour the calculating individual. This, however, means preferring a particular register of evaluation, and transforms the situation to make it amenable to this register. Four problems of decision-making have now been raised: incommensurability of goods; a variety of registers of evaluation; inconsistent preferences across a variety of situations; and situation-dependent rules of application. Already at first sight, they pose enormous difficulties for theories of the rational chooser. Rational choice theorists have pursued two different strategies to deal with these complications. Accordingly we can group those theorists into two categories, the doves and the hawks. Doves and hawks are equally convinced of the superiority of rational choice theorizing. Thus, a major spokesperson for the doves, Jon Elster, claims unequivocally that this approach is unrivalled as a normative theory. But let us see what price, to use hawkish language, he has to pay for that assertion. He and his fellow-doves allow their individuals all kinds of preferences – they may follow social norms, engage in collective action, accept institutions as bounds of their rationality, or may forgo the best outcome for an inferior one because they are satisfied with it. In short, they appear like quite ordinary and reasonable human beings. But what is their claim to rationality and in what sense is this normatively superior behaviour?

24 Peter Wagner Technically speaking, all particularities of human lives are assimilated into either the preferences of the actors or the context of the decision-making situation, both of which are outside the realm of the rationality of the choice. The claim to normative superiority ultimately boils down to an understanding of rationality in which to be rational, pace Hegel, means to do what you think is good for you. While this is quite agreeable, it leaves all the tasks of social research and social theorizing to others – namely to find out how preferences emerge and change and what the social contexts are in which human beings happen to find themselves. Hawks, in contrast, keep their eyes firmly on the calculating individual. In their view, reason-endowed atoms are what the social world consists of, and the more this fact is recognized, the better everything will be. Rather than accepting the particularities of human social life and trying to adapt their theorizing to it, as the doves do, they recommend ways of acting and behaving. Summarized briefly, in the light of the above observations, their advice runs as follows: Consider yourself an individual who knows what you want before you enter into any situation of decision-making. When in such a situation, aim at establishing cognitive control over it; that is, provide yourself with as complete information as is available (or as much as you can afford). Systematically relate the information gathered to your preferences. Design strategies to see which preferences you can satisfy to what degree. Weigh your preferences, that is, make them comparable, so that you can establish a hierarchy of strategies. Decide. Looking back at the decision-making problems discovered by interpreting the proverbs, we see that the following happens here: if commensurability of goods was not given, it is now established. Any diversity of registers is reduced to one. Preferences are being ordered in a consistent way throughout. The situation is now unequivocally defined. The two presuppositions that are needed for this procedure to work are the assumption of an individual with fully established preferences and the assumption of the possibility and desirability of absolutely context-independent judgement. Rather than always/already being in some world full of strange birds and other animals, the human being is seen here as initially standing outside the world and only moving towards it with some intention and purpose. Or, more precisely, since it is a theoretical operation we are talking about, the distancing from a context is seen as the precondition for rational action of the individual. However, this specifically entails an alteration of the situation, in the sense in which I referred to such an occurrence above. It is inappropriate to say that a rational choice approach allows a different (read: superior) way of dealing with a given situation. Rather, a rational choice approach changes the way in which a human being is situated in the world. It is an intervention in the world rather than an analysis of it.

The bird in hand 25 A number of rational choice theorists will not be at all unhappy with such a diagnosis. After all, some may see themselves as engaged not only in knowing the world, but also in improving it. And there can be no doubt that a rationalist-individualist approach, in a broad sense, has enabled human beings to gain insights and possibilities, many of which they would not want to miss. (Though space is too short to go into detail, I will implicitly provide some illustrations below.) However, the action that is performed is one of reducing that which is not reducible. Translations introducing commensurability and unequivocality lead to an impoverishment of the available repertoires of evaluation. The messy richness of the world is exchanged for some clean rags. While hardly any reasonable being would prefer rags to riches as a matter of principle, there may be situations in which one has reason to believe that one cannot do otherwise.

Of wars and revolutions As a mode of social theorizing, rational choice theory is – or, better, relies on – a theory of modernity. It works with a postulate of autonomy; human beings have wills and are, in principle, able to act according to them. And on this basis, it takes the pursuit of their strivings, with a view to accomplishing their objectives, as that which human beings will want to do, that is, as expressions of rationality. In terms of Enlightenment philosophy, rational choice theory takes a version of the combination of freedom and reason as its basic philosophy. It thus relates closely to what I like to call, in somewhat broader terms, and following Cornelius Castoriadis (e.g. 1990: 17–19), a commitment to autonomy and mastery as the double signification of modernity. However, rather than merely situating itself within modernity, rational choice theory provides a very particular interpretation of modernity. As I have tried to argue at some length elsewhere (Wagner 1994, 2000), both of those meaning-providing terms (autonomy and rationality) are underspecified and ambivalent on their own, and tension-ridden as a doublebarrelled concept. The social signification of modernity is widely open to interpretations. Autonomy, for instance, can be understood predominantly on individual terms, but it can also be read as collective self-determination. Rationality or mastery can be conceived of in purposive, instrumental and then procedural terms, but it can also be related to substantive concerns. And these formulae for ambivalence – individual/collective, instrumental/ substantive – do not yet capture anything approaching the richness of possible interpretations, nor are they even necessarily the best way of framing the issue. Rational choice theory, in contrast, proceeds from an unequivocal starting-point. Its social entities are individuals, and they behave according to instrumental rationality.

26 Peter Wagner If rationalist individualism (from now on I will use this broader term because moving on to historical considerations) is a theory of modernity, but a very particular one, the question arises as to the grounds on which it was preferred to others. To approach an answer to that question, a rapid survey of the history of this thinking is useful.6 Rationalist individualism initially emerged in social contract theories, as first spelt out unequivocally in Hobbes’s Leviathan, and was emphatically developed as a basic theoretical approach by thinkers such as Condorcet in the context of the French Revolution. In parallel to the latter, Smith’s moral philosophy provided a space for rationalist individualism within the confines of the economy, an approach which also inspired Marx. Abolishing the elements of a separate moral-political philosophy, this thinking was radicalized during the marginalist revolution in economic thought, which led to what is now the dominant thinking in this field, namely neoclassical economics. During the inter-war years of the twentieth century, European cultural critics identified a degraded version of rationalist individualism, in the form of a conjunction of atomism and conformism, as the prevailing attitude in mass society, in particular the North American one. Weber, though not a typical proponent of this thinking, laid many of its foundations. After the Second World War, rational choice theorizing gained the form in which we now know it and spread from economics to the other social sciences. With the help of these few historical reference-points, it is possible to provide elements of a contextual understanding of the way in which the particular, rationalist-individualist interpretation of modernity imposed itself (the following draws on Wagner 2000). I will anticipate the general argument and will then situate it in various contexts. Rationalist individualism may emerge and find acceptance, or even impose itself, as a social theory by default. By default I refer to a move that is made in situations in which other interpretations of the modern condition, while they may be available in principle, cannot be utilized. The default situation may arise as a consequence of the exigency that all other such interpretations have to make stronger social presuppositions or, in other terms, would need to assume more substantive social prerequisites than the individualist-rationalist one. To relate this idea to the signification of modernity: not to interpret autonomy as purely individual autonomy requires, if not a coherent and stable collectivity, then at least socially rich ways of relating to others, that is, both singular others and networks of others. Not to interpret mastery or rationality in instrumental terms requires other substantive value orientations, which again need to be, if not shared with, then at least communicable to and acceptable to others. If this consideration is generally acceptable, the next task is to identify the conditions under which such other interpretations become difficult or impossible so that the default situation arises. Very abstractly speaking, such conditions prevail in times of destruction of a social configuration, in particular rapid and forceful destruction, and in times of the founding or refounding of social configurations, especially if this occurs under pressure or comes from

The bird in hand 27 a great diversity of sources. Destruction may be caused by imposed social change, such as was the case historically due to the capitalist revolution and through the building of bureaucratic state apparatuses. Here we recognize Marx’s theory of alienation and Weber’s theory of rationalization as responses to such experiences; and now we would also need to include the experience of totalitarianism. But destruction of a social configuration also occurs through warfare and revolutions. And in this context it is significant that individualist-rationalist modes of theorizing made early breakthroughs in the context of the seventeenth-century religious wars and the eighteenth-century revolutions. In such situations, no other way seemed available to conceptualize the return to peace and order. To the present day, Hobbes’s Leviathan (1962 [1651]) is one of the key references when discussing the attempt at grounding political order on an abstract conception of the individual human being and ‘his’ rationality. The abstraction from any concrete situation, later to be called the hypothesis of the ‘original position’ (John Rawls (1970, 1993)), would then provide the required distance and make rational-scientific knowledge possible. However, rather than reading Hobbes’s Leviathan as the inauguration of scientific method in political thought, it can also be interpreted as a contribution to solving an urgent socio-political problem, namely how to bring an end to violent religious strife. In such a contextual perspective, Hobbes’s attempt shows a clear awareness of an inescapable dilemma. On the one hand, there is a deep intellectual consciousness of the lack of stability and certainty. Radical doubt is indeed the starting-point of the reasoning. On the other hand, Hobbes was also driven by the conviction that humankind could not live well without some categories of social and natural life which impose themselves on everybody, and such categories would only impose themselves if and because they were considered as undeniably valid. The situation of uncertainty that was experientially self-evident, where devastating religious and political strife had to be overcome, was to be transcended by appeal to an instance that, in his view, could only be found outside such experience. The coexistence of these two, apparently incompatible, convictions gives his work a character which one might label as dogmatically modernist: radical in the rejection of unfounded assumptions, but inflexible in its insistence upon some definable minimum conditions of cognitive and political order that could, and would have to, be universally established. After the religious grounding of unity had been irretrievably lost, it could, in his view, only be replaced by individuals and their rationality. All other categories could be contested, and strife would be renewed. Rather than giving intellectual support to the idea of a strong state, Hobbesian thinking opened the way for an increasingly consistent derivation of political power from the combined will of the individuals in a given collective. The American and the French Revolutions marked attempts to implement such a conception. Thus, they revived the

28 Peter Wagner theoretical problem and gave it a new urgency. Edmund Burke (1993 [1790]: 8–9), sceptical observer from across the Channel, issued an early warning in his reflections on the French Revolution: ‘The effect of liberty to individuals is, that they may do what they please: We ought to see what it will please them to do, before we risque congratulations.’ In this situation, the rationalist-individualist approach reaffirms itself, alongside the other emerging social sciences, as one way of finding out ‘what it will please them to do’. The hope and aspiration was that the moral and political sciences should and could now achieve ‘the same certainty’ as the physical sciences (Baker 1975: 197). Certainty was a requirement of some urgency, since the new political order needed assurances of its viability. But it was also regarded as a new historical possibility, since political action was liberated from the arbitrariness of decisions made by rulers of doubtful legitimacy and given into the hands of the multitude of reason-endowed human beings. The ‘blend of liberalism and rationalism’, which Keith Baker (1975: 385) observed in Condorcet’s convictions, can thus be explained as stemming from the Enlightenment linkage of freedom and reason. In such a view, the rights-endowed individual became the only conceivable ontological as well as the methodological foundation of a science of political matters after the revolutions. Once the rights of man had been generally accepted as self-evident and inalienable, it seemed obvious, to Turgot and Condorcet for instance, that they were also ‘the logical foundation of the science of society’ (Baker 1975: 218). In rights-based liberalism, the individual is the only category that need not, and often in fact cannot, be contested. The individual is simply there, whereas everything else – for instance, what criteria of justification are to be applied when determining the collective good – is subject to argument. Substantive aspects of human interaction are dependent upon communication and consensus. And it is even uncertain, to make the issue yet more complicated, with whom should one enter into communication, because the boundaries of the community are not themselves given, but subject to agreement. Once this assumption was accepted, basically two avenues for constructing a science of the political had opened up. One possibility was to try to identify by theoretical reasoning the basic features of this unit of analysis, individual human beings, and their actions. Since this unit was conceived of as an ontological starting-point, devoid of all specific, historical and concrete ties to the world, its characterization had to proceed from certain inherent features. From earlier debates, those features had often been conceived of as twofold, namely as passions and as interests. In the late Enlightenment context, the rational side of this dichotomy was regarded as the one amenable to systematic elaboration. It thus allowed the building of a scientific approach to the study of at least one aspect of human interaction with the world, namely the production and distribution of material wealth. This approach inaugurated the tradition of political economy, later to be transformed into neoclassical economics and, still later, into rational choice theory.

The bird in hand 29 While political economy was based on a highly abstract, but for the same reasons an extremely powerful assumption of human rationality, the other conclusion from the individualist foundational principle was possibly even more reductionist but much more cautious. Avoiding any substantive assumptions whatsoever about the driving forces in human beings, the statistical approach, often under the label of political arithmetic, resorted to the collection of numerical information about human behaviour. The space for substantive presuppositions was radically evacuated in this thinking, but the methodological confidence in mathematics seemed to have increased in inverse proportion (Brian 1994; Desrosières 1998). This approach would later lead into quantitative-empirical research and behaviourism. Up to this point, I have interpreted the seventeenth-century religious wars and the eighteenth-century revolutionary upheavals as events of social disruption that made observers inclined to adopt a rationalist-individualist position. The objective of this withdrawal from richer social theories was to identify the point from which order could be rethought, in the first instance, and then reconstructed. At the same time, the establishment of the European state system, in the Treaty of Westphalia, and the creation of the French Republic, in the course of the Revolution, were obviously also moments of the founding or refounding of social configurations. In the case of the French Revolution, a certain rationalist individualism indeed informed republican political language. Overall, however, European consciousness saw other substantive resources with which to refound social order as still being sufficiently available after these events, and thus rationalist individualism did not achieve an overall breakthrough in social and political thought. However, from a later European point of view, this was much less the case for the creation of the American republic, the master-case for the founding of a polity in the West.

Of Europe and America During the early twentieth century, in particular between the two wars, a European image of America developed in which an individualist-rationalist modernity was seen to have been established in America and, owing to some of its features, had expanded from there to transform Europe in a similar way. I shall briefly summarize the main features of this view: ‘America’ is what we may call presentist, that is, without history and tradition. It was a country without tradition, ‘where no medieval ruins bar the way, where history begins with the elements of modern bourgeois society’, as Friedrich Engels (1958 [1887]: 354) already pointed out in 1887. It was even ‘the specific country of ahistory’, as Alfred Weber, Max Weber’s brother, remarked in 1925, in stronger tones (Weber 1925: 80). America is also individualist, that is, there are no ties between human beings except for those that they

30 Peter Wagner themselves create. And it is rationalist, that is, it knows no norms and values except the increase of instrumental mastery, the striving to use efficiently whatever is at hand in order to achieve one’s purposes. Again Tönnies (1922: 357), here using Max Weber’s concept of rationality, succinctly expressed his view on American public opinion as ‘the essential expression of the spirit of a nation: it is “rationalistic” ... in the sense of a reason which prefers to be occupied with the means for external purposes’. Or, in D. H. Lawrence’s (1962: 28) words, the American ‘is free to be always deliberate, always calculated, rapid, swift, and single in practical execution as a machine’. And, finally, America is what we may call immanentist, that is, it rejects the notion of any common higher purpose, anything that transcends individual lives to give them orientation and direction. In other words and in sum, this view saw the ideal society of rational choice theorists as already established on earth. There are very strong reasons to doubt whether this image provides a proper representation of American social life, but this is not the question that I want to raise here. It is for two other reasons that this image of America is important in the context of a discussion of rational choice theorizing. Firstly, it provides the major, and historically early, example of a use of such theorizing to describe an existing social configuration; and it does so in comparison to another one, namely by way of the – largely implicit, but sometimes explicit – comparison of Europe and America that informs these writings. Since in doing so it also suggests that the ‘other’ society provides Europeans with a view of their own future, a hypothesis about the direction of history can be derived from it. More precisely, it suggests that the history of humanity led to the emergence of a social order inhabited and sustained by rational individuals. To be sure, most of the European traffickers in this view also criticized and rejected this perspective. Secondly, we can nevertheless infer a normative perspective, namely a metaphysics of the rational individual, by inverse reasoning, that is, by identifying precisely what the Europeans were rejecting. I shall briefly discuss these two features before drawing conclusions about the particular force of their combination. The European image of America resonates strongly with a specific mode of theorizing, and specifically of theorizing modernity. This theorizing is modernist, rather than modern, because it builds on the double notion of autonomy and rationality, which are key characteristics of modernity, but it also turns this notion into an unquestioned and unquestionable assumption for theorizing the social world. Such thinking pervades many areas of intellectual life, and possibly its strongest version is rational choice theory in social thought. At its core, such theorizing proceeds by a double intellectual move, as discussed earlier. It first withdraws from the treacherous wealth of sensations that come from the sociohistorical world to establish what it holds to be those very few indubitable assumptions from which theorizing can safely proceed. Subsequently, it reconstructs an entire world from those very few assumptions. Its proponents tend to think that the first move decontaminates

The bird in hand 31 understanding, any arbitrary and contingent aspects being removed. The second move is held to create a pure image of the world, with scientific and/or philosophical validity, from which yet further conclusions, including practical ones, can be drawn.7 This description of theoretical steps fits a certain view of American history, namely as one of self-foundation ex nihilo. The European writings that convey this image of American rationalist individualism certainly contain no thorough sociological interpretation of the specificity of America. However, there are fragments of such an interpretation, and they point to the effects of large-scale migration and of the creation of a society by emigrants and exiles. The experience of emigration and exile is one of distancing from that which was known and familiar, from all that which appeared as given and natural. Such distancing – even when it was imposed – can liberate the mind to think of the world in different terms. Emigration and exile, as forms of distancing, therefore lend themselves to a formal discourse in social theory and political philosophy, one that aims at avoiding or eliminating anything contextual or particularistic. In the light of such distancing, the experience of going away is sometimes seen to change the character of the social bond with other people, at least at the level of the entire society or polity. Raymond Williams (1961: 107) once suggested that the exile will usually ‘remain an exile, unable to go back to the society that he has rejected or that has rejected him, yet equally unable to form important relationships with the society to which he has gone’. Williams suggests that there is something irrecoverable, once one leaves one’s place of origin. The social bond cannot be re-created in the same way in which it existed before; the same density of social relations and density of meaning in the world around oneself can no longer be achieved. Sometimes emigration has even been related to a radical loss of the ability to give meaning. The liberating effect that distancing may have turns into the inevitability of negation and, normatively speaking, into cynicism. One of the protagonists in The Plumed Serpent (Lawrence 1987 [1926]: 77–8) holds that America, having no roots, no history, no tradition of its own, is incapable of any creation. Its very emergence in an extended process of migration is indicative of some degree of exhaustion of the creative power of humankind. In such characterizations, we re-encounter the two steps of rationalist theorizing, albeit in a different form. In the process of migration, it is alleged, all substantive ties in social life are shattered and human beings are left without any important moral orientations. This is the first step. Then, from the loss of all these fetters, grows the hubris of wanting to re-erect a world without any such substantive bases. Rather than referring to a territorially locatable society, ‘America’ stands here for a specific – a ‘modern’ – way of living, which is traced to something like a generalized migratory experience. The – liberating – willingness to throw off the burden of history goes hand in hand with the impossibility of relating positively to a society with ‘thick’ social bonds. The experience of migration and exile that sociologically is taken to

32 Peter Wagner account for the specificity of ‘America’ can then be related to the tendencies towards rationalization and of individualization, as postulated in social theory from Marx and Weber onwards. The historical hypothesis contained in such views thus focuses on a decreasing historical depth of social life due to the recomposition of societies, and on the overburdening of singular human beings with the task not only of reconstructing their lives but also of reconstituting the guiding frameworks for the social world of which they have become a part. As a result, those beings are seen to be left on their own, without substantive ties to others, and with reason as the only resource they could reliably draw upon, given that other resources presuppose that they are to some degree shared or recognized by others. In the course of ‘modern’ human history, it is seen as inevitable that a ‘flattening’ of the temporal depths of social life and a ‘weakening’ of the social bonds occur. In the most general terms, these are considered to be effects of the dynamics of modernity, although precise interpretations vary considerably in terms of their socio-historical explanations. Looking at the history of social theory, we find here a theme which unites its critical tradition from the eighteenth century onwards, in particular that strand which runs from German idealism to the Frankfurt School. Obviously, in critical theorizing, these tendencies are both diagnosed and opposed. It is precisely the task of critical theory to identify the conditions under which they came about or the forces that brought them about, with a view to challenging them. Rational choice theory basically provides for an inversion of that perspective. If it had a historical dimension (which most often it has not), it could broadly accept a Marxian–Weberian narrative of individualization and rationalization. But rather than deploring this development, it would celebrate this course of human history as the progress of reason.8 At the same time, it takes licence – again implicitly – to derive a general theory of human action and social life from this supposed historical tendency. However, this brief analysis of images of America has shown that such reasoning starts out from a quite exceptional situation, namely the founding of a society by – for a variety of reasons – ‘disembedded’ people.

Of words and worlds The diagnosis of an increasing rationalization and individualization of the social world is a workable hypothesis for sociological inquiry, even though a complex and difficult one. It would have to be tested by research in comparative-historical sociology.9 In contrast, the presupposition that it is possible to analyse human social life as if its basic units were rational individuals is a statement of basic social ontology, or of ‘social metaphysics’ (Boltanski and Thévenot 1991). It would have to be judged in terms of its adequacy, necessity and consequences.

The bird in hand 33 Starting out from the latter, the observations at the beginning of this chapter were meant to demonstrate that the application of rational choice reasoning leads to an impoverishment of social theory. A wide variety of registers of social and political philosophy are reduced to the application of instrumental rationality by individual human beings. Sometimes this step is justified in terms of an increase in the consistency of theoretical elaboration. However, the differentiation within the field of rational choice theorizing between doves and hawks shows that such consistency has indeed not been achieved. Crucial questions have always remained open to a variety of different answers. If the consequences cannot be considered as unequivocally desirable, then the case for rational choice theory would have to be made in terms of its empirical adequacy. If the world were populated by rational choosers, it would certainly have to be described and analysed in terms of rational choice theory. An analysis of existing discourses about moral-political evaluation, along the lines of the comparative cultural metaphysics hinted at above, would show that this is far from being the common way in which human beings justify their actions. In addition, comparative analyses of disputes and controversies in social and political life also reveal the variety of forms of justification that are actually applied (Lamont and Thévenot 2000). This leads us to the case for the necessity of the rational choice approach. The brief historical analysis of ‘contexts of discovery’ resulted in the observation that rational choice theory is a state of emergency thinking. It applies either in situations of violent strife within a given social order, or in situations of founding an order without common ground. The two kinds of situations are indeed not entirely distinct. Hobbes transformed an analysis of conflict into a hypothetical rationale for the founding of an order. And the French Revolution was the attempt at founding an order that turned into violent strife. The problem being addressed is the lack of common cultural resources – or, in the terminology used throughout this chapter, of a common register of moral-political evaluation – to deal with a socio-political situation. Individualist rationality is then proposed as some kind of bottom-line on which everybody can agree – or at least would be willing to agree to end a dispute.10 It is in this sense that rational choice theory provides the default mode of social theorizing – and it has a value and a place among the modes of social theorizing on this very ground. However, the social world is neither in a state of permanent crisis and strife nor in a condition of continuous change of such dimensions that everything is always uncertain and in question. In most situations most of the time, much more than such a very limited repertoire of evaluation is available to human beings in order for them successfully to ‘go on’. Only a social theory that remains sensitive to the richness of those repertoires will be able to understand social life.

34 Peter Wagner There is even more at stake than ‘mere’ understanding. The adoption of a rational choice perspective in situations in which richer repertoires are indeed available is not just a matter of description or analysis. It is an alteration of the situation, in the sense described above. By means of redescription, it aims at turning that situation into one in which individual human beings make means–ends decisions on the basis of a conscious preference ordering. In other words, it suggests reading a situation as one of deep crisis, distrust and lack of common resources, in which a default mode then would need to be mobilized, when in fact there may be no such crisis at all and common resources may be abundantly available. If successful in persuading the actors themselves of such an interpretation, the rational choice approach would provide the ground for its own application. But the price to be paid would be the loss of those common resources. The rational choice approach is the bird which social theorists will always have in hand. But out there in the social world there are many other animals, and social theory should not stop striving to grasp them, unless forced by necessity.

Notes Thanks are due to Heidrun Friese, David Lambourn and Falk Reckling for helpful comments and suggestions.

1 Or at least not for the purposes of my reasoning. Arguably, there are sexual connotations in this proverb, as in the other ones in different languages that I will quote below. A consideration of those connotations would make this introductory argument much more complex. Since it seems safe to assert, however, that my line of reasoning would only be further strengthened through their inclusion, I will largely leave them out for the sake of brevity. 2 If this were the case, however, one could also ask why the ‘Germans’ did not develop or appropriate the much clearer form of ‘English’ wisdom. As we know, even folk wisdom travels. Those who may be inclined to think that the German form typifies the common obscurity of Continental thinking should be aware that there is a version of this wisdom in French which is closer to the English than to the German, at least with regard to the quantitative aspect. In ‘Un “tiens” vaut mieux que deux “tu l’auras”’ (‘Something you have is worth more than two things you will have’), the temporal dimension is explicitly introduced in addition as an aspect of human interaction (bringing in issues of trust). Rational choice theorizing is notorious for having difficulties in dealing with future time, since the preferred strategy, namely discounting the future, is open to a number of objections. 3 May it not be the case that these proverbs refer to birds, among the many goods one may want to have, because they are always inclined to fly away, because of the difficulty of durable possession? 4 Eagles as well as doves can symbolize freedom, but possibly the eagle – as a state symbol – stands rather for collective freedom and collective self-determination and the dove for individual freedom.

The bird in hand 35 5 This is, however, far from saying that communication and compromise are impossible, as is sometimes alleged. 6 In the view of its own proponents, the real history of rational choice theory only starts in the middle of the twentieth century, but an insight into its deeper roots or ‘predecessors’ can occasionally be found. Similarly and significantly, rational choice thinking also lacks a long historical view on the development of Western societies, although it could and should have one, a matter to which I return below. 7 Whatever dissonance there may be between sensations and this image will then be treated as the secondary problem of the relation between theory and empirical observation. 8 But then it may be the cunning of reason rather than its progress of which we find evidence here. From the middle of the twentieth century onwards, we find the earlier European view of America partially confirmed when, even if no overall individualist-rationalist way of life emerges in America, then at least its intellectual foundations proliferate at – predominantly – US universities. 9 This issue can obviously not be pursued here. Let it just be noted that approaches that see modernity strongly in terms of the destruction of ‘traditions’, that is, of the withering away of common registers of moral-political evaluation, tend to underestimate the human ability to re-create richer forms of social life, even after crises. This is a theme insistently put forward by Hans Joas (1996a, 1996b), for example. 10 As doves of rationalism have recently argued: ‘In the absence of strong environmental constraints, we believe that rational choice is a weak theory, with limited predictive power. ... The theory of rational choice is most powerful in contexts where choice is limited’ (Satz and Ferejohn 1994: 72). The authors, however, move from that insight to arguing for the compatibility of rationalism with ‘structuralism’, without considering the criticism the latter approach has encountered over the past twenty years.


Homo economicus, Homo sociologicus and Homo sentiens Margaret S. Archer

Rational choice theory requires rational actors: insofar as they deviate by behaving as normative or expressive agents, they vitiate the theory. Development of the model of ‘rational man’1 has leant heavily upon his instrumental rationality. Whatever the goals he cherishes, egotistic or altruistic, and whatever constraints he encounters in realizing them, his means– ends calculus is that of ‘economic man’. The trouble is that he lacks substantive rationality: like the shopper writing out his list, he can be allowed, within his budget, to put on it whatever he likes. The provisos are all adjectival: preferences must be fixed, transitive, complete and informed, which says nothing about the substance of what is preferred. This is the central problem: does ‘economic man’ choose his ends by rationality alone? Put a different way, can his substantive preferences be construed as those whose realization would objectively be ‘the best for him’? This problem has exercised rational choice theory from its Enlightenment origins onwards, for initially our choice of ends was not held to be a matter of rational choice. On the contrary, ‘economic man’ was born fully Humean, and thus only weakly rational: his passions dictated his priorities and reason was confined to the role of their ingenious slave.2 This substantive irrationality dogged him throughout modernity. The Utilitarians sought to tame us into rational actors by transmuting our ‘passions’ into enlightened ‘pleasures’, but this left the (substantive) conundrum as to why someone should continue to prefer pushpin to poetry. Hence, subsequent generations of rational choice theorists have worked at making us more strongly rational by attempting to take the passion out of our preferences. This chapter questions the outcome of that endeavour, which has supposedly established the rationality of ‘economic man’. The first section shows that he is a product of rational choice theory rather than its premise. The second section examines the product and finds that this rugged individual, ‘economic man’, cannot be detached from his social context, and especially from the influences of normativity upon his decision-making. However, it is

Homo economicus, Homo sociologicus and Homo sentiens 37 maintained that no compromise can be negotiated between Homo economicus and Homo sociologicus because, as such, this would still fail to account for the substantive ends sought by human agents. The third section concludes that Homo sentiens, the moral agent whose decisions are expressive of who he is, rather than instrumental means to further ends, remains central to any acceptable model of human agency. The argument thus comes full circle back to the beginning, by contesting the cornerstone of rational choice theory, namely its construct of ‘economic man’.

Taking the passions out of preferences: taming substantive rationality This is a long-drawn-out process, which can be summarized in four stages that are roughly chronological. Overlaps occur because the practitioners of several disciplines – philosophy, economics, politics and sociology – all worked to the same end – that of excising emotion from desire, such that the remaining preferences could be drawn into the ambit of rational choice.

From ‘pleasure’ to ‘preference’ The first radical shift came at the hands of the economists and consisted in the reconceptualization of ‘pleasures’ as ‘preferences’. In empiricist terms, this had the advantage of turning an unobservable human quality into observable and measurable behaviour. ‘Preference has a positivistic respectability that pleasure lacks: we can tell what a person prefers by offering him a choice and seeing what he does’ (Gibbard 1986: 167–8). Yet positivism is never innocent. Here it entailed substituting for a generative causal mechanism which is internal to the person (the pursuit of pleasure) his external behaviour (choosing), which could be readily ascertained on the perceptual criterion. Now, the satisfaction of preferences could in principle be synonymous with the successful pursuit of happiness, or even emotional feelings of well-being, but this was not the case. Instead the ‘standard here is the satisfaction of preferences, not the satisfaction of people’ (Gibbard 1986: 167-8). The move to take preferences as the standard of human welfare, rather than the degree to which humans report themselves to be satisfied and happy, simply does not settle what is best for people. It assumes that whatever course of action a person prefers is shown by the fact of his choice to be better for him. It reached its apogée in Samuelson’s theory of ‘revealed preference’. If a basket of goods x could have been bought within the budget of an individual, yet he was observed to purchase another basket y, then it is presumed that he has revealed a preference for y over x, with preferences being inferred from the choices observed. This, Samuelson (1938: 71) argued, enabled him to ‘develop the theory of consumer’s behaviour

38 Margaret S. Archer freed from any vestigial traces of the utility concept’. Presumably, by this he means that his overt behaviourism has obviated any need to peer into people’s heads to find out what they think they are doing, or into their hearts to discover what they feel they are doing. This shopper has certainly been divested of strong feelings, but we immediately need them back in order to explain why his choosing y over x on one occasion and x over y on another is not inconsistent. Most of us do shop differently for birthdays compared with weekdays, and we often ‘let ourselves be extravagant’ with a treat for ourselves or others. Secondly, revealed preference leads directly into the Prisoner’s Dilemma. This is because to follow their self-interest results in solitary prisoners making confessions which produce longer sentences than would have resulted from collaborative or altruistic non-confession. Is this ‘choice’ of more years in prison a reliable guide to their underlying preferences (Sen 1973)?

From subjective preferences to ‘ideal preferences’ Among those who said ‘no’ to this question were neo-Utilitarians, whose next move was to clean up our preferences and to purify them from their defective subjectivity by shifting over and talking instead about our ‘ideal preferences’. In other words, ‘one reason for rejecting the subjectivist conception, or taking special care with its formulation, is that a person’s preferences can be defective in ways which discredit them as a measure of his welfare’ (Schwartz 1982: 195). Hence, enter ‘ill-informed preferences’, ‘irrational preferences’, ‘poorly cultivated preferences’ and ‘malconditioned preferences’, as contrasted with ‘rectified preferences’, namely ones which are better for a person and which he or she would prefer were he or she fully and vividly aware of everything involved. What is significant here is that among the things to be cleaned up are the emotions and emotional interference, which now appear on the ‘belief side of the equation (Desires + Beliefs = Preferences + Action). Situating them there makes it easier for reason to get a grip on them, by the innocent means of being supplied with better information. This is Harsanyi’s strategy. For it to be credible we need to be convinced that we do have passionately irrational beliefs (rectifiable), but that our emotionality does not extend to our desires. More generally, we also would need to accept that emotion is a thoroughly bad thing in preference formation. Now it is easier to accept this as formulated for a defective belief, which is demonstrable, than for a defective desire, a notion that seems always to remain contentious. Thus Harsanyi writes that it is well known that a person’s preferences may be distorted by factual errors, ignorance, careless thinking, rash judgements, or strong emotions hindering rational choice etc. Therefore, we may distinguish between a person’s explicit preferences, i.e. his preferences as they actually are, possibly distorted by factual and logical errors – and his

Homo economicus, Homo sociologicus and Homo sentiens 39 true preferences, i.e. his preferences as they would be under ‘ideal conditions’ and, in particular after careful reflection and in possession of all the relevant information. In order to exclude the influences of irrational preferences, all we have to do is define social utility in terms of individuals’ ‘true’ preferences, rather than in terms of their explicit preferences. (Harsanyi 1977: 41, my italics) This is probably the Utilitarians’ best move, but it rests on questionable assumptions which threaten its workability. It seeks to retain the (rectified) link between subjectivity and welfare and can account for the fact that it is possible that what is good for people nevertheless violates their preferences. The difficulties are two-fold: it is asserted that ‘non-defective preferences are more felicifically efficient (other things being equal) than defective ones: their satisfaction is more satisfying. Not only is this not obviously true; it is not obvious how to tell if it is true’ (Schwartz 1982: 198). It might seem that to encourage someone to clean up his or her irrational beliefs is something no rational agent should resist, because the injunction must come with a demonstration that the belief in question is misinformed. My aim is not to champion ‘happy fools’, but is to stress (i) that there is no necessary correlation between the passion with which a belief is held and its erroneousness, and (ii) that many of our deepest held beliefs (religious, political, ethical) simply cannot be demonstrated to rest upon defective information. The corrigible beliefs of the discerning shopper are not the appropriate model for the inherently contestable beliefs upon which our lives are run.

Eliminating ‘strong feelings’ from rational desires If working upon our beliefs to make them more rational was difficult, then to do the equivalent with our desires might appear to be a non-starter. Yet this has been the main contribution of political science. It is significant that Jon Elster (1999b: 179) has sometimes called his work ‘studies in the subversion of rationality’, and he agrees that it is a ‘nonstandard move’ to consider what induces ‘irrational desires’. As these studies have progressed, it has become clearer that emotions and addictions (it being no accident that they are examined together) are regarded as the main sources subverting the rationality of desires. Hence, ‘by virtue of the high levels of arousal and valence they induce, emotions and cravings are among the most powerful sources of denial, self-deception, and rationalization in human life’ (Elster 1999b: 200). It is these ‘strong feelings’ which need cleaning up in the interests of rational autonomy, roughly meaning that people have the inner freedom to plan their desires in relation to the possible.

40 Margaret S. Archer ‘Strong feelings’ result in discontent, an unhappiness which is more diabolical than divine. Rational desires do not result in happiness – that link has been broken for good – but they do issue in that sober Protestant contentment associated with living within one’s means and making virtues out of necessities. Thus to Elster (1983: 119), ‘there is a respectable and I believe valid doctrine which explains freedom in terms of the ability to accept and embrace the inevitable’. This is deeply embedded in the way he defines ‘rational desires’, and the main question it raises is why agents should treat social conditions with such inevitability and why those who refuse to do so are branded as irrational. To Elster (1986b: 15), the definition of a rational desire would include, first, an optimality property: that of leading to choices which maximise utility. If people very strongly desire what they cannot get, they will be unhappy; such desires, therefore, are irrational. A rational desire is one which is optimally adjusted to the feasible set. Maximum utility is thus defined and delimited within the politics of the possible, for the ‘feasible set’ is a complete list of actions which do not contravene constraints, such as wealth, opportunity or discrimination. It is that which is really (ontologically) possible to which rational actors have to accommodate their desires, not their perception of possibilities (epistemological) to which they have to adapt their utilities. All desires have causal origins, but some have ‘the wrong sort of causal history and hence are irrational’ (Elster 1983: 16); they are under the sway of strong feelings rather than being coolly and deliberately chosen. Thus, over-adaptive preference formation (sour grapes), or under-adaptive preferences (Ulysses unbound) (Elster 1979),3 result from ‘a “hot” rather than a “cold” process’ (Elster 1983: 26), from ‘unconscious drives’, from ‘non-conscious psychic forces’ which are incapable of that cool deferred gratification which is cognizant of the reality principle.4 Hence, ‘rational desires’ are ones which display an optimal adjustment to what is feasible, and thus generate the maximum utility which is realistically possible. Yet there seems to be something both inherently and unacceptably conservative about defining the rationality of desires in terms of the social status quo at any given time. Basically, why is it more rational to entertain only those desires which are attainable within existing social constraints, rather than to envisage the transformation of such constraints in a better society? Nevertheless, this seems to be what is entailed by Elster’s notion of rational character-planning. Thus he comments that, ‘whereas adaptive preferences [sour grapes] typically take the form of downgrading the accessible options, deliberate character planning would tend to upgrade the accessible ones’ (Elster 1983: 119). The message is that rationality in our desires is a matter of getting reconciled: if you do not get the promotion sought, then change your lifestyle to benefit from the greater leisure opportunities of a less prestigious position (Elster 1983: 123). Yet suppose you are convinced that the promotion was refused

Homo economicus, Homo sociologicus and Homo sentiens 41 because you were female or Black, or spoke with a working-class accent, why is it rational to become reconciled to discrimination? Why is the anger inappropriate, or the passionate commitment to working for more egalitarian processes an irrational response? Why are our most reasonable preferences those which are ‘causally shaped by the situation’ (Elster 1983: 121)? After all, the choice is not limited to either acquiescing to external constraints or succumbing to irrational, internal, hedonic ‘psychic forces’ – unless freedom-fighting is illegitimately assimilated to the latter! Elster’s insistence upon defining the rationality of desires in terms only of that which one can entertain within social constraints has the perverse effect of welcoming closed doors. If, at a particular time, no-one happened to exercise freedom of speech or association, must we conclude that they do not value these freedoms? However, if they do value them passionately in times and circumstances when they do not exist, must their attempts to kick the doors open be written off as instances of irrational desires? One of the conclusions to be drawn in the next section is that Elster’s rational choice, and rational choice theory in general, cannot cope with the phenomenon of social movements.5 In part this is due to defining rationality in relation to the status quo, and in part it is due to the methodological individualism which can only see society changed through aggregate actions rather than by collective action. In part too, it derives from that purging of the emotions which condemns them as sources of irrational ‘wish-fulfilment’, rather than allowing that they can foster the vision and commitment which rationally support social movements pursuing the fulfilment of wishes.

Choosing our emotions rationally In the fourth stage, the attempt is made to rid the theory of these tempestuous emotions by fully incorporating them into rational choice as items considered alongside others when people maximize their utility. Thus in Accounting for Tastes, Becker (1996: 131) introduces the proposition that ‘Forward-looking individuals consider how their choices affect the probability of developing rewarding emotional relations.’ Such relations turn out to be those which save us money if we avoid them in ourselves and bring us monetary reward if we inculcate them in other people.6 What follows from Becker is some good mercantile advice about monitoring one’s movements in order to minimize the pains of guilt or love. Encounters with beggars make (most of) us feel guiltily uncomfortable, feelings which we would rather avoid; and the stronger the beggar’s appeal, the larger the contribution elicited from us. Hence, evasive action is recommended, so that by ‘anticipating the negative effects of begging on their utility, guilty contributors may be able to avoid harmful shifts in their preferences’ (Becker 1996: 233). Similarly, Becker’s marriage guidance counselling advises that rather than placing oneself in a position where generous love finds one willingly making financial transfers to a

42 Margaret S. Archer poorer spouse, it is better to stick to those neighbourhoods and clubs which will supply partners from the same class, who do not occasion such expensive sentiments (Becker 1996: 236). Similarly, in family life Becker advocates ‘investment in guilt’, by which parents financially promote the acquisition of ‘merit goods’ in their offspring with the intention of inducing sufficient guilt in them to ensure that they themselves are cared for, in return, in their later years. Yet this assumes that parents are actuated by investment considerations, rather than genuinely caring for their children, and that the latter are guiltily responsive later on, rather than simply angry at having been manipulated. In all of this, emotions like love and caring within the family have been subordinated to cost–benefit calculations which will later be cashed in. More generally, in society, the development of norms and the formation of preferences entail emotional involvement, so the simple advice here is to buy the appropriate values through indoctrination. Thus the provision of churches for the lower classes is recommended, where the poor are compensated for lending themselves to the inculcation of beliefs because the rich subsidize expensive buildings and the clergy. Compensation is limited to start-up costs, because once the norm of churchgoing is established, the poor will police one another in these socially useful religious practices (Becker 1996: 229–30). A major difficulty here is that indoctrination must not be seen as such, hence ‘the rich’ cannot do it directly but needs must work through clergy who actually are religious believers. Yet even if the clergy are not averse to financial contributions, what ensures that the donors’ ‘message’ is prominently preached or prevails in the conflict with religious values of compassionate caring? Nothing does, unless the assumption is made that religious commitments, like all others, have their price-tag. This fourth strategy, which subjects emotions to the process of rational choice itself, necessarily commodifies them. When they were treated as externalities of the market, as in Becker’s earlier work, they could impinge, positively or negatively, upon people’s welfare, but in their own terms. Now that they are chosen (or deselected for), alongside other commodities, then they must be presumed to be commensurable. Becker does not hesitate to make this assumption and to make cash the common denominator between emotions and anything else which is rationally chosen. It means he is presuming that emotional ‘returns’, like the love and regard of others, can be bought. I would simply deny that these are commodities or that they are for sale. Instead, they are some of our deepest human concerns, ones which are constitutive of who we are, rather than being instrumentally rational means to yet further ends – to some indefinitely deferred point at which we ‘forward-looking individuals’ become ‘better off’. Yet, if considered as ‘ultimate concerns’, as I will try to show in the final section, then these concerns which express who we are, and thus entail affectivity, are not commodities which can be costed. Rather they should be considered as commitments which are quite literally priceless. Conversely, emotions can only be tamed and finally

Homo economicus, Homo sociologicus and Homo sentiens 43 brought within the ambit of rational choice if they are forced to take up their abode in the marketplace.

The return of the exile So far I have represented the upbringers of modernity’s man as four generations of carers, all of whom were determined that he came to maturity a good deal more rational than he started off life. They did this progressively by taming his passions into preferences, then working on the beliefs with which these operated in order to purge him of ‘defective preferences’, and finally by attempting to purify his utilities until these became matters of ‘rational desires’. What were exiled to the nursery were the temper tantrums associated with his ‘hot’ emotional ‘I want’s’, so that his coming of age was the emergence of a ‘cool’ reasoner capable of dispassionate ‘character planning’. Then, modernity’s man would have become a fully rational chooser and thus the ideal subject for rational choice theory. Yet the trouble is that this theory aspires not only to give us a ‘model of man’ as the rational actor, but also to derive complex social structures directly from him, that is, from some property pertaining to this (idealized) human being. In other words, as Hollis (1987) put it, this second Adam was intended to be a ‘sovereign artificer’ not only of himself, but, together with those like him, of his society. Since Adam is the topic here, I will only make a brief detour to comment on his abilities in making his social environment before returning to the assessment of his character. The point of this excursion is to show that rational choice theorists themselves came to regret the exiling of the emotions and sought the repatriation of ‘emotional man’ as a necessary partner in the building of society. The first contender as a society-builder was the ‘rational man’ of classical economics, whose calculus, consistency and selfishness resulted in choices which summed to produce social reality. The trouble was that this model of ‘rational man’ could not cope with manifest social phenomena like voluntary collective behaviour or the voluntary creation of public goods. Only the offer of selective incentives or the imposition of coercive measures could persuade the ‘free-rider’ of rational grounds for co-operating in collective action (Olsen 1965: 61–6, 106–8). Nevertheless it was admitted that ‘irrational’ forces such as ideological inspiration, altruism or personal commitment could do the trick of reforming the free-rider. This led some theorists, who conceded ‘rational man’’s inability to deal rationally with the Prisoner’s Dilemma or free-riding, to complement him with an inner running mate, rather than abandoning him to his paradoxical fate. Enter ‘normative man’, who shifts to a different logic of action under circumstances in which he realizes that he is dependent upon others for his welfare. This is someone who will act in terms of normative expectations and social duty and thus becomes subject to a sense of interdependence when public non-divisible goods are desired. However, inexplicable macro-

44 Margaret S. Archer level effects still remained, and ‘emotional man’ joined the team to mop up structural and cultural properties which were based upon expressive solidarity or a willingness to share. Above all this third man was needed to account for social stability deriving from solidarity and for instability resulting from collective protest. Thus as ‘a “loving” participant in collective action, emotional “man” contributes to the creation of public goods, and as a creator of corporate actors, to the stabilisation and regulation of otherwise intermittent and fleeting feelings, but also to a general emotional restraint’ (Flam 1990a: 42). On the other hand, emotional action ‘represents the volte-face potential in, or outer limits of, everyday normative or calculating behaviour. As an angered participant in collective protest, emotional “man” is a source of social change and political instability’ (Flam 1990a: 42). Thus those emotional qualities on whose expunging so much effort had been devoted were now brought back out of exile as necessary components in explaining what held societies together or precipitated them towards transformation. What was being done, however, was not to revise the various ‘models of men’, but to multiply the incommensurable situational logics to which human beings were responsive. The technique was to multiply the denizens of every human being with these different ‘men’ all harmoniously cohabiting the same body – ‘rational’, ‘normative’ and ‘emotional’ – without the previous fear that one undermined the other, because they were now confined to separate spheres of action and were supposed to know their social place.7 There are problems with this multiplication of complements, all inhabiting the same being. The first concerns Adam’s own mental health. What is it that ensures that the complements do know their proper place? Once they have been housed in the same body, what prevents one from usurping another? Supposedly they await the appropriate external social cues for activation, but why should we not suppose instead that there is some internal effort after personal integration? If on occasions one can be either dutiful or loving, why should one not strive to be lovingly dutiful? That does not sound too bad; but when instead we try being angrily rational, all the old problems of impetuousness resurface which supposedly dogged modernity’s man. The second problem with this notion of complements making up the same human being is that it eventually comes full circle, ending up with the ‘multiple self’ (Elster 1988) and the suggestion that we treat ‘man’ like an organization. Yet this is a completely vicious circle: some sort of ‘man’ was wanted to explain that which was problematic, namely social organization, but now we are enjoined to use the explanandum in order to conceptualize the explanans, the nature of man! Matters are now getting out of hand: Homo sentiens can neither be left in exile nor can he be repatriated on rational choice soil without doing untold mischief. Hence my suggestion that the upbringing of modernity’s man might fare better if we left him with his emotions intact and allowed them to develop with him. Certainly, I am going to

Homo economicus, Homo sociologicus and Homo sentiens 45 maintain that he will not then much resemble either Homo economicus or Homo sociologicus, but he does stand more chance of presenting a human face.

Homo economicus and the circumstances not of his choosing Fundamentally what accounted for the gradual development of this ‘multiple self’ with its plurality of inner personas – rational, normative and emotional – is what always goes wrong with methodological individualism’s individual. He begins as ruggedly autonomous, radically atomistic, a man whose individual dispositions plot his actions and life course.8 Yet methodological individualists immediately break with these requirements of their position, since the facts about people which are allowed to figure in rock-bottom explanations are neither solely individual nor solely dispositional.9 This is exactly what is going on with the multiplication of inner complements inhabiting rational choice theory’s ‘man’. Because economic choices are obviously constrained by our spending power, the social distribution of wealth is disaggregated into what you or I have laid our hands on, and is then incorporated into us as our ‘budget’. Thus preferences, a dispositional term, can now be expressed individualistically in the market-place ‘within our budget’. Similarly, because our actions are often constrained by normative expectations, these social duties are also disaggregated into a personal sense of co-operative interdependence when individual welfare depends upon co-operation. The creation of public goods can now be dispositionally explained in voluntaristic fashion because there is no longer any need to introduce external normative constraints, since the work they do has been taken over by the individual’s recognition of his or her own good. Finally, because social solidarity frequently seemed to exert a Durkheimian binding force over us, this too is disaggregated into a personal willingness to share or to collaborate, which depends upon individual emotionality. Thus generous emotional dispositions account for the voluntary creation of public goods, and angry ‘men’ make for collective protest. In all three cases, those aspects of the social context which are indispensable for explanation – the social distribution of resources, the social pattern of normative expectations and the social conditions of solidarity – are themselves incorporated into individual terms. As Lukes (1968: 125) puts it, ‘the relevant features of the social context are, so to speak, built into the individual’. There are two serious ontological objections to this procedure. On the one hand, in what recognizable sense are we still talking about ‘the individual’, when he or she has now been burdened with so many inalienable features of social reality? On the other hand, can the social context really be disaggregated in this way, such that solidarity and protest are purely interpersonal matters, normative beliefs are only what certain people hold in their own interests, and resource distributions are just what each of us has on personal deposit? There is

46 Margaret S. Archer a further methodological problem due to this ‘desperate incorporation’ of social factors into the individual, which Gellner (1968: 267–8) expressed as follows. ‘Algy met a bear, the bear was bulgy, the bulge was Algy’; the individual may consume what Durkheim and others have called social facts, but he will bulge most uncomfortably, and Algy will still be there. I suspect that actual investigators will often ... prefer to have Algy outside the bear. I agree, and want to show good reasons why it is distinctly preferable to keep Homo economicus and his ‘circumstances’ clearly differentiated from one another. The first reason concerns the relationship between ‘preferences’ and ‘circumstances’, or rather its absence in rational economic ‘man’. Preferences are assumed to be given, current, complete, consistent and determining. Here I want to argue that the societal distribution of resources actually has an independent effect upon each of these elements, and one which is badly obscured by the process of disaggregation. Economists typically deal with given preferences: they freeze the frame in the present tense and generically take no interest in their source or formation. In Becker’s (1996: 24) extreme version of this, as was noted in the Introduction, tastes neither change capriciously nor differ importantly between people. On this interpretation one does not argue over tastes for the same reason that one does not argue over the Rocky Mountains – both are there, will be there next year, too, and are the same to all men. Other rational choice theorists of course allow for changes over time and for inter-personal variability. Nevertheless, a normal given preference (of the shopper rather than the addict) is supposed to show price elasticity. Thus those on low pay will stock up with the supermarket’s cheapest offer on baked beans, but if their incomes rise they will head for the luxury lines, their basic preference being for the best food they can afford. Now nutritionists may have their doubts about whether sudden wealth does produce good eating, but the bizarre diet of Elvis Presley is just where the economist would invoke de gustibus non est disputandum. However, sociologists remain rightly worried when they suspect the converse, namely that it is circumstances which are forming people’s preferences. Working-class parents were not being good bargain-hunters when they used to declare a greater preference for the secondary modern school over a free grammar school place, on the grounds that the latter was ‘not for the likes of us’; and Bourdieu showed middle-class parents to be exceedingly good at winkling-out educational safety nets which protected against their children’s social demotion (Bourdieu and Passeron 1977). Is this difference in economic manhood merely a nondisputatious question of tastes, in which case why is there a class-patterning, or is it a highly contentious effect of generations of educational discrimination?

Homo economicus, Homo sociologicus and Homo sentiens 47 Similar difficulties attach to the requirement that a preference order should be both complete and consistent, meaning that for any given x and any given y, the individual prefers one to the other, or rates them equally, but must have an overall ranking of all alternatives. Necessarily this means that agents must know their preferences for experiences which they have not yet tried and which might alter them considerably if they did so; that is, for relationships which they have not yet undergone, but which, like motherhood, there is no going back upon; and for ways of living which they have not sampled but which, like early school leaving, will alter their life-chances if they do. Here, information cannot be first-hand, yet the distribution of second hand information is as socially skewed as are other distributions of resources. Appeals to family and friends, the availability of role models and the advice given by teachers will be filtered through their own cultural capital and perhaps reflectively adjusted to the cultural assets of the inquirer. Again, the differential social accessibility of information plays a role in shaping preferences, and not vice versa, as is presumed. Consistent preferences are ones whose ranking shows ‘transitivity’, namely a ranking such that if A is preferred to B, and B to C, then A must also be preferred to C. It is immaterial what coinage is used to order these incommensurables, hence the handy word ‘utility’, but it is needed in order to know what is being rationally maximized or for rational action tout court. Imperfect information now makes a second appearance, affecting our preference ordering as distinct from its formation. Think of the thousands who agonize annually over their choice of university. Suppose they have come down to three criteria: ‘league position’, ‘location’ and ‘course content’, all of which are considered equally important. Then an applicant grades University A as 3, 2, 1 (meaning ‘excellent’, ‘good’ and ‘acceptable’ on these criteria), whilst University B scores 1, 3, 2 and University C gets the scores 2, 1, 3. Were the applicant to consider only a difference of two points in any category to be significant enough to dismiss an institution, but that one point difference is unreliable, then transitivity is broken (Black 1990: 148–9). Thus A would be preferred to B (for decisive superiority on the league table), B to C (for decisive superiority in terms of location), and C to A (for decisive superiority in course content). Our applicant has now violated the Bayesian choice axioms and yet is not indifferent to these universities, as if they were regarded as equal. Rather than behave like Buridan’s ass, our applicants could do many things. They could behave impeccably and introduce a tie-breaker like ‘accommodation’. Yet why should they, for their main concern is to pick a university, not to establish a transitive preference order? So they are just as likely to pour over the brochures, meet some congenial people at one particular open day, or ask friends for their opinions. None of these are irrational reactions, but their circumstantial nature means that neither are they ‘rational choices’. In fact this is where we remain deplorably ignorant, for though we can invoke ‘luck’, or ‘contingency’, which sounds better, what we do not know is where they turn, what they take into account, and how they weigh it in their own personal balances.

48 Margaret S. Archer This fundamental problem becomes even more acute when we move on to the determining force which preferences are supposed to exert. On our common understanding, the idea of ‘choice’ is an active process of choosing in which agents weigh the pros and cons, in terms of whatever ‘weights and measures’ they employ, and then come up with their decisions. Of course, some regularities are to be expected amongst those similarly placed, precisely because the circumstances of those differently placed attach different costs to the execution of the same project. All of this is familiar in realist social theory, which sees the constraints and enablements emanating from the societal distribution of resources as being mediated to agents through the situations which they confront and the different opportunity costs associated with different courses of action for those who are differently placed. However, choice itself requires an active agent who reflectively weighs his or her current circumstances against the attainment of his or her goals, and who alone determines whether the price can be afforded. This is very different from the passive ‘man’ of rational choice theory, for there, ‘If we recall that preferences are given, then it will be seen that the agent is simply a throughput’ (Hollis 1987: 68). Determination of a course of action is a matter of the hydraulic pushes and pulls which prices exert on preferences. The difference between the active chooser and the passive throughput can be seen if we consider bright, working-class Sharon, who left school early, took her vocational qualification and became a hair stylist. For a time she is buoyed up by her pay packet and progress up the salon hierarchy, but she becomes increasingly frustrated with her job. Now this initial choice is corrigible, but further costs are attached, some necessary and some contingent, for life goes on as corrective preferences are contemplated. The trouble is that whilst on day release she met and later married Darren, who is unenthusiastic about the sundry costs which would fall on him if Sharon pursues the university access course which a client has just mentioned to her. His concerns are allayed by the birth of Warren, but Sharon’s costs of extricating herself have now redoubled. She has acquired new vested interests in her family and home, which constitute new obstacles to enrolment on the access course. It can still be done, but, presented with these situational constraints, not too many will be undeterred and decide that they have good reason to do so at that time and under those circumstances. Suppose that Sharon does not abandon the project entirely but postpones it until Warren is at school: she finally enters the university as a mature student and, in an uncertain job market, has collected another penalty for being older. Moreover, not only does Sharon now have to juggle childcare and study requirements, but in realizing her abilities she is also acquiring a new set of friends and a new outlook, which bode ill for her marital stability. Divorce is another common price to accrue, often leaving those in Sharon’s position as single-parents, with courses to complete and uncertain occupational futures. The moral of this story is not meant to deter our mature students. On the contrary, they have triumphed over their circumstances, but as extremely active agents they leave us with the

Homo economicus, Homo sociologicus and Homo sentiens 49 question of what made them battle uphill and what exactly they were weighing in their balances which made these sacrifices worthwhile to them. What will not do is to render Sharon passive and to write off her determination as being due to the effects of ‘socially random influences’, in the words of one rational choice theorist (Gambetta 1996: 186). Neither, on the other hand, is reproduction (which in this case replicates the well-known socio-economic differences in educational attainment) merely a matter of routinization. For let us now take another story, that of Karen, coming from much the same background as Sharon and sharing the same circumstances, but who capitulated and gave up her educational aspirations. Far from being an account of a hydraulic process, this entailed reluctant resignation, strenuous exertion against the odds and a bitter failure to meet the costs of overcoming situational constraints. To have tried and failed does not mean that she had a different preference schedule from Sharon. (Routinization might appear as a more appropriate characterization of the perpetuation of structural privilege, but this would be deceptive because it also requires an active and reflective co-operation with enablements. Habitual action may be highly inappropriate for cashing them in. Enablements are advantageous for allowing people to stay ahead, not to stay where they are, and the former means being ready and able to innovate with new courses and openings, which is beyond the wits of passive ‘man’.) Since, as has been seen, the rational choice theorist aspires to explain both the choices we make and the social context in which we make them, it is a fair question to ask for an account of Homo economicus’s circumstances which are not of his choosing. Rational choice’s diachronic account construes enablements as the ‘winnings from previous games’ and constraints as their losses, as if the privileged and underprivileged come from the lines of good and bad card players respectively. This has a certain plausibility, until it is recalled that the ‘games’ of game theory have pretty stringent regulative and constitutive rules attached (there would be no Prisoner’s Dilemma without assuming prison sentencing, criminal justice or solitary confinement). What are the rules of the ‘Life-Chances’ game? The social distribution of resources was not brought about by our ancestors playing ‘Modernist Monopoly’, for new players could enter and change the rules (of the Constitution, of enfranchisement, habeas corpus or combination) and new strategies could not be ruled out (of professionalization or unionization). Nor, for the same reasons, can the present generation of the privileged and underprivileged be construed as playing some ‘Us and Them’ game. The alternative synchronic solution to explaining and incorporating constraints is Becker’s invocation of ‘personal’ and ‘social’ capital. The former is roughly the accumulation of our childhood sins and virtues. ‘For example, investment may depend on smoking, attending church, or playing tennis because these types of consumption build up stocks of habitual capital’ (Becker 1996: 7). Conversely, ‘an individual’s stock of social capital depends not primarily on his own choices, but on the choices of peers in the relevant

50 Margaret S. Archer network of interaction’ (Becker 1996: 12), plus his ethnicity, family history, nationality and religion, because culture changes so slowly and with such a low depreciation rate that it is effectively ‘given’ to the individual for life. This then enabled Becker to deal in the following manner with what he agrees are ‘neglected constraints’: ‘Preferences and constraints no longer have independent influences on behaviour since personal and social capital are constraints that operate through preferences’ (Becker 1996: 21, my italics). What is happening here is the typical methodological individualists’ manoeuvre of ‘desperate incorporation’ of group variables (class, status and power) into the individual person, as was discussed at the start of this section. We thus do not confront constraints like Marx’s ‘alien powers’; rather they constitute us and thus we involuntarily ‘prefer’ to reproduce them. Presumably we should neither congratulate Sharon nor commiserate with Karen, since both were only expressing their preferences. At most we should look for an educated role model or encouraging teacher in Sharon’s biography and expect to find that all Karen’s family and friends spent their school days smoking behind the bicycle sheds. Well, this may or may not be a way of putting flesh on those ‘socially random influences’, but it is a long way from regarding their struggles as real and their outcomes as dependent upon their own, and still opaque, reasons for choosing. Finally Homo economicus is passive in another important sense. Because he and others like him are passionless, atomistic individuals, the formation of a social movement is beyond them. It was of course to enable him to engage in collective action that Flam had complemented him with ‘emotional man’. This was the final way of incorporating into the individual those properties which were needed to make his behaviour approximate to the manifest abilities of social agents. Although it is true that social agents need to have their passions involved before participating in social movements, otherwise all would have reason to be free-riders, this makes collaborative action depend upon ‘collective effervescence’ alone. Instead, we need to introduce the objective conditions for collective action and the objective goals which ‘corporate agents’10 pursue. After all, every social movement is for something, even if its ends are imperfectly articulated and the organization is strategically uncertain about the most effective means to employ. The missing elements, which are necessarily lacking in a methodologically individualistic account, are interests and identity. As social agents, collectivities of people are involuntarily pre-grouped at birth in relation to the social distribution of scarce resources. For simplicity these are called the ‘privileged’ and the ‘non-privileged’, and the latter confront plenty of daily exigencies, given their poor life-chances. They thus have vested interests in common in ameliorating their position and hence the best of reasons for struggling to develop collective organizations (unionization, franchise, civil rights movements and feminism), to achieve together what they cannot bring about alone. Their promotive activities are too innovative to be construed as ‘games’, for there was no nineteenth-century game called ‘winning the

Homo economicus, Homo sociologicus and Homo sentiens 51 franchise’, no regulative rules governing political conflict (repressive strategies changed all the time), and any constitutive rules were ex post since they entailed constitutional changes as a product of this conflict itself. What the introduction of the collective social movement allows is something which rational choice theory cannot supply, namely an account of how agents are not condemned to a static distribution of life-chances and an equally static array of accessible positions. In other words, the picture of individual agents ‘living within their budgets’ is transformed when we introduce processes through which they collectively alter the structural principles of budgetary allocation. However, we cannot do this without also allowing that the inegalitarian distribution of life-chances furnishes interests in their transformation and that collective passions fuel politics. Neither can be accommodated in a model of Homo economicus as an individualistic passive agent whose circumstances push and pull him around but who lacks the capacity to combine in order to change circumstances himself.

Homo sociologicus and normative constraints I would agree with Greenwood (1994) that the mere mobilization of collectivities to obtain civil rights or women’s rights does not solve the problem of acquiring a social identity. It cannot, because the mere acquisition of new rights does not secure strict numerical identity, whose criteria cannot be satisfied by more than one candidate, and thus never by being members of a social movement with others sharing the same interests. What the successful movement does do is to open doors to new parts of society’s role array which were formerly inaccessible to them, thus making new social positions available in which people can invest enough of themselves to feel at home with what they have become. However, here is the rub for rational choice theory. Those who take on social roles are subject to normative expectations. Insofar as their conduct is subject to social norms, they act out of a sense of duty rather than according to a ‘rational man’’s self-interest. Attempts to deal with norms within this theory have sometimes taken the form that norms are merely grist to strategic manipulation, such that people invoke them as a cloak to legitimate their selfinterest, whilst not really subscribing to them. This seems impossible because if no-one believed in the norms, they could hardly be manipulated through them. Alternatively, it is the avoidance of the sanctions associated with norm-breaking which, on a cost–benefit analysis, makes ‘rational man’ obedient to them. Against this we can note that some norms are followed in the absence of any observers who could sanction violations. Thus many people vote even when nobody would notice that they do not,11 and many do voluntarily use our bottle banks. Because the efficacy of social norms cannot be denied, but neither can they be reduced to the machinations of individuals’ rationality, there have been attempts to explore ‘a negotiated

52 Margaret S. Archer settlement between the standard homo economicus and the standard homo sociologicus’ (Hollis 1987: 147). There is much that is very attractive in Hollis’s (1987) exploratory account, especially the arguments that a role is both a source of reasons for action and a vehicle which gives a person control over his or her own social life. The chain of rationality is not broken by the subsumption of action under normative expectations, because cultural dopery is avoided by asserting that the reasons for action associated with a role only move actors when they are adopted as their own good reasons. The great attraction is the maintenance of an ‘active’ agent whose role-taking is a matter of ‘intelligent stewardship’, since role requirements entail something both more involving and more demanding than a mindless reading of the small print. In short, since roles are not fully prescribed, it is better to think of individuals as personifying roles, that is, monitoring themselves in order to bring the character to life, rather than extinguishing themselves by simply impersonating characters. This makes them into loyal servants rather than subservient dopes, and it has the supreme advantage of conceptualizing social selves for Adam and Eve which, whilst dependent upon society, also enable them to meet the strict criteria of identity as particular persons. Thus through choosing (and being chosen) to be Professor of Sociology at Warwick University, I personify the role in my own fashion, which makes me distinct from all others who hold the same job contract. My obligations I accept as mine, but their execution, far from swamping me, are used to define me. Basically the settlement between Homo economicus and Homo sociologicus unravels because it requires a moral agent to make it work, one who has to be introduced from outside the settlement. No hybrid of the two can provide this moral character. On the one hand, ‘rational man’ is only induced to contract into social norms because they are in his enlightened self-interest: he does not accept them as binding but endorses them calculatively. Thus our rogues playing a series of Prisoner’s Dilemmas will both calculate that acting according to the ‘no-grassing’ norm and keeping silent leaves them both better off, until they both later calculate that cheating on this rule leaves each best-off. ‘No-grassing’ is normatively unstable and deteriorates iteratively into the regular, although sub-optimal, ‘tit for tat’ (Rapoport and Chammah 1965). Hence there can be no contractarian ethics because these rogues’ codes do not bind morally and the rogue will revert to free-riding or side-dealing, provided he calculates that he can get away with it. In other words, authentic ethical behaviour can only work if persons are ultimately motivated not by personal gain, but by something other than the satisfaction of their own preferences. On the other hand, over-binding them so that they become normatively dopey does not do the moral trick either, since it produces socially conventional behaviour rather than that moral sensitivity which bridges the gaps in conventions or copes when they clash.

Homo economicus, Homo sociologicus and Homo sentiens 53 ‘Rational man’, the bargain-hunter, looks more active than he is because of his remorseless pursuit of self-interest. However, if he is pre-programmed by a fixed preference schedule, he is deprived of the ability to reflect morally upon his preference-set. Thus he could not do any of those things which Elster (1979) advises in order to modify our undesirable or addictive preferences by ‘pre-committing’ ourselves in ways which outsmart our weakness of will. Ulysses needs to be a morally active agent to ask his sailors to bind him to the mast so that he can hear the Sirens without being drawn to his doom by them. However, if he is such an agent, then there is more to him than a programmed robot. Yet that which is ‘more’ than his appetites, because it can control them, cannot be filled in by social norms. For one thing, the Classical Sea Captains’ manual did not contain the instruction ‘Immobilize yourself when in the vicinity of beguiling Sirens’, because roles cannot be scripted for every contingency. Therefore the successful role-player cannot sink into the passive follower of normative instructions whose only reasons for action are located in the role itself, or because satisfactory performance gives access to further roles. Even role theory’s explanation of social change, which is rooted in role clash, requires an active agent with the capacity to determine what to do under these circumstances because the ‘cultural dope’ will simply be paralysed. Thus, if Homo economicus is crossed with Homo sociologicus, although ‘each of the two models explains how actions are programmed, ... they cannot be combined to explain why action is not programmed’ (Hollis 1987: 172). Yet if we truly personify roles, then we bring to them something other than the normative stuff out of which roles are made. What we bring is judgement of a kind ‘more like moral judgement than like either ready reckoning or rifle drill. It consists in acting intelligently for reasons which a role-player can justify’ (Hollis 1987: 172). From there on we have to try to understand the ‘man’ both who is more than a pursuer of his preferences and who animates his roles above and beyond their normative obligations. Who is he? It will be argued that Homo sentiens is required in order to give real animation to the ‘man of judgement’ who has more than an eye to the main chance and is something better than word-perfect in the parts he plays.

Homo sentiens and his ultimate concerns Homo sentiens is a character who is capable of making moral commitments and who has a reason for keeping them which derives from his involvement in society (which thus means more to him than a public means to his private ends). Hollis works towards this notion of social commitment, which stems from an active agent who does the committing (passive agents being puppets of socialized committal), through asking what people are doing in trying to become ‘better-off’. In other words, what are their ‘pay-offs’ and in what currency are they dealing? We need to know the human ‘weights and measures’ applied in relation to the

54 Margaret S. Archer ultimate ends that they seek. Here, in judging actions, ‘it is not enough to know that the agent prefers its expected consequences. We also need the measure of value being applied, including the end allegedly better served’ (Hollis 1989: 174). When we seek to be loved, regarded and respected, not only are these things not for sale, but also they are something like a terminus in that they do not lead on to further ends which could be achieved by an additional dose of instrumental rationality. Ends like these to which we are ultimately committed are those things which we care about most. As such, they are not only extensions and expressions of ourselves, but also ones which can be irreducibly social. In other words, those social relationships to which we are committed as our deepest concerns (marriage, family, career, church, community) are not for any agent the ‘means to his flourishing but its constituents’ (Hollis 1989: 174). Here there is no sense in asking why it pays someone to give their partner a birthday present or to help their friends out, for these actions are expressive of their relationship, not matters of investment and quid pro quo. Moral commitment of this kind is neither calculative nor socialized, yet it is both reasoned and social, for our relations to these significant others are the expression of who we are and where we belong. This serves to cut through both the calculation and the individualism of ‘rational man’ and to keep him social without his becoming over-socialized. Thus ‘a rational agent’s ultimate reference group cannot be himself alone. He needs some group to identify with in relationships whose flourishing is a measure of his flourishing’ (Hollis 1989: 179). By necessity this has to be authentic because when another’s interests are part of one’s own, short-cuts which simply give the appearance of belonging are self-defeating to a person whose real need is really to belong. What this implies is that Weber’s Wertrationalität, far from being expelled from a disenchanted world, remains part of our lifeworld, which cannot be reduced to the bargain-hunter’s bazaar. Therefore what is crucial to our decision-making are our ‘ultimate concerns’, namely that to which we are emotionally drawn, as modified by our knowledge about them.12 This is the source of what people commit themselves to in society: it is the key to their social identity and one which involves considerable affect. As Frankfurt (1988: 83) puts it, a person who cares about something is, as it were, invested in it. He identifies himself with what he cares about in the sense that he makes himself vulnerable to losses and susceptible to benefits depending upon whether what he cares about is diminished or enhanced. What we are committed to thus shapes our lives and determines our pay-offs. When explaining particular decisions, it is these commitments which supply their own ‘weights and

Homo economicus, Homo sociologicus and Homo sentiens 55 measures’. Without this knowledge about what emotionally moves people, we simply do not know what counts to them as a cost or a benefit, or how strongly it counts.

Notes 1 ‘Rational man’ was the term current in Enlightenment thinking. Because it is awkward to impose inclusive language retrospectively and distracting to insert inverted commas, I reluctantly abide with the term ‘man’, as standing for humanity, when dealing with this tradition, its heirs, successors and adversaries. 2 ‘Reason alone can never be a motive to any action of the will ... reason is and ought only to be the slave of the passions and can never pretend to any other office than to serve and obey them’ (Hume 1969 [1739/40]: Book II, Part III, Section 3). 3 Elster’s Ulysses and the Sirens (1979) deals with akrasia, that is, weakness of the will as a source of irrationality, and praises Ulysses’ ‘pre-commitment’ in having himself bound to the mast to overcome this condition. 4 These unconscious drives are ‘conceived of as non-conscious psychic forces that are geared to the search for short-term pleasure, as opposed to conscious desires that may forgo short-term pleasure to achieve some longer-term gain’ (Elster 1983: 25). 5 Typical of this is the following methodological-individualist statement. The real bottom line is that there are individual actions, that there are outcomes of these actions, and that individuals choose actions in terms of their outcomes, using some decision rule or another. This is the heart of rational choice theory, and it does not admit of the possibility that groups of individuals choose actions in some way that is fundamentally more than an interactive product, however complex, of the choices of individual members of the group. (Laver 1997: 28) 6 One difficulty is the very limited range of emotions Becker considers, which gives pause to wonder whether the same principle is held to operate over the four hundred or so emotions discriminated in the English language. 7 Thus Helena Flam (1990a: 39) wrote that it was her intention to propose a model of the emotional ‘man’ as a complement to the models of rational and normative man [and] not to argue that the mode of emotional action should replace either of the other two. Rather it is to advocate model pluralism in lieu of the present model duopoly. The model of emotional ‘man’ is useful because it helps to explain some aspects of collective action which the rational and normative man models cannot handle.

56 Margaret S. Archer 8 As J.W.N. Watkins (1968: 270) stated the matter, ‘According to this principle [methodological individualism], the ultimate constituents of the social world are individual people who act more or less appropriately in the light of their disposition and understanding of their situation.’ 9 Instead, Watkins (1968: 270–1) allows that acceptable predicates can include ‘statements about the dispositions, beliefs, resources and inter-relations of individuals’ as well as their ‘situations ... physical resources and environment’. Some of these, of course, are not individual and some are not dispositional, and it is arguable that none are both. 10 These are agents who are organized as social movements and have articulated their goals. See Archer (1995: Ch. 8). 11 For ingenious but unconvincing attempts to explain ‘the paradox of voter turnout’, see Green and Shapiro (1994: Ch. 4). 12 For a more detailed discussion of the relationship between our emotions and our ultimate concerns, see Archer (2000: Chs 6 and 7).


Is rational choice theory ‘unreasonable’? The neglected emotions Simon J. Williams

My starting-point, contentious perhaps, is that reason itself, qua instrumental rationality, is indeed ‘unreasonable’. The ‘irrational passion for dispassionate rationality’, an abiding theme throughout the course of Western history, has proved something of a mixed blessing: both enlightening and blinding, liberating and binding all at once. Rational choice theory (RCT), the modern-day inheritor of this orderly, predictable vision of the world, and the fear of the ‘irrational’ it contains, provides a paradigmatic example of these legacies and the need to move beyond them. The target of any such critique, to be sure, is a moving one. Rationality itself has multiple meanings, just as there is no one rational choice theory, only rational choice perspectives (with numerous qualifications and disclaimers to boot) (Zey 1992a). To this we may add the more or less pervasive influence of rational actor/choice assumptions, implicit or otherwise, in much past and present sociological theorizing: from Weber’s privileging of Zwekrationalität to Giddens’ (1992) own somewhat calculative approach to the ‘pure relationship’. It is not my intention, however, to provide a detailed balance sheet of the pros and cons of rational choice theory. Instead I seek to focus on the neglected issue of the emotions, including their role in (rational) decision-making itself, as well as the crucial part they play in the explication of micro–macro sociological links. The chapter takes off with a brief sketch of the so-called ‘orthodox’ view-point on highminded reason and the outlaw emotions it denounces. Whereas this was conceived in largely negative terms as subversive of reason, the second section, in contrast, provides a somewhat more balanced assessment of what precisely emotions are, including their multi-faceted nature and irreducible qualities. This in turn paves the way, in the following section, for the main thrust of the chapter: the question, that is, of whether or not emotions are themselves ‘reasonable’. The former orthodox viewpoint is contrasted, in this respect, with two alternative stances in which this relationship becomes progressively more intimate, so to speak. The issue here, as Zey rightly notes, is not so much rationality per se, as its ‘constituents’. Whether rationality is the only basis for action, however, is equally open to question (Zey 1992b: 15). Implicit in these considerations is the search for a more satisfactory

58 Simon J. Williams sociological resolution of the micro–macro, structure–agency problematic than rational choice theory can supply. Here as elsewhere, I suggest, emotions provide a crucial ‘missing link’, exposing once again the limits of rational choice theory. The final section, in drawing these arguments together, suggests some possible ways forward, including the need for more sensitive methodologies with which to capture these complex multidimensional decisionmaking processes, alongside a call for a more ‘passionate sociology’ (Game and Metcalfe 1996) in general. Rational choice theory, I conclude, in keeping with this volume as a whole, is not in fact a ‘rational choice’, but is indeed ‘unreasonable’.

What is ‘reason’? The orthodox viewpoint The orthodox viewpoint, dominant throughout the course of Western history, is one in which a wedge is firmly driven between reason, on the one hand, and emotion, on the other: the former a peculiarly disembodied enterprise, the latter banished to the ‘irrational’ margins of Western thought and practice.1 Emotion, from this perspective, perverts the course of ‘reason’ – the latter conceived as an impartial, dispassionate, disinterested pursuit – and the search for truth and wisdom. To be rational, Plato declares, is truly to be ‘master’ of oneself, a mastery based on calm and collected self-possession. The realm of passion and desire, in contrast, is deemed ‘chaos’, our souls ‘torn apart’ through a ‘perpetual state of conflict’ (Taylor 1989: 116), or the ‘miasma of the indeterminate’, in Bauman’s (1991) terms. The Cartesian cogito, likewise, turns reason into a somewhat disembodied enterprise, a form of mental disengagement far removed from the corporeally based ‘Passions of the Soul’, or the embodied forms of knowing and being-in-the world so central to phenomenologists such as Merleau-Ponty (1962). Emotions, for Descartes, cannot be entirely ‘controlled’ by thinking. They can, however, be regulated by thoughts, especially those that are true. Reason, Kant proclaimed, should be sovereign, an independent faculty set in opposition to emotions qua nature. To speak of ‘reason’ and ‘rationality’, however, is to invoke terms which are far from settled. Elster, for example, in his discussion of the French moralists – from Montaigne to La Rochefoucauld, Pascal to La Bruyère – notes how ‘Reason’ must be sharply distinguished from ‘Rationality’: the latter conceived as the ‘instrumentally efficient pursuit of given ends’, the former ‘any kind of impartial motivation or concern for the common good’ (Elster 1999a: 102). We may, for instance, from this viewpoint, be ‘rational in the pursuit of some private interest, but then we are not reasonable’. Similarly, we may be ‘reasonable when arguing in terms of the categorical imperative (“what if everyone did that”) but then we are not rational’, at least in Elster’s terms (Elster 1999a: 102). The key sociological treatment of these issues, of course, comes from Weber (1991) and his ‘incapacitating fear’ of the ‘irrational’. In highlighting the existence of dual rationalities

Is rational choice theory ‘unreasonable’? 59 or two value spheres, Weber drew a sharp distinction between four different types of action. Two of these, traditional (habit) and affectual (powerful emotions), were deemed ‘nonrational’ (not, that is to say, the result of ‘deliberate choice’). The other two, Wertrationalität and Zweckrationalität, were seen as rational. Wertrational action, Weber emphasized, was undertaken as a result of belief in the ultimate or intrinsic value of acting in a certain way (i.e. substantive rationality). Zweckrational action, in contrast, a dominant force in contemporary Western society, involves the calculation of appropriate action (means) to be taken to achieve desired goals (ends) – that is, instrumental rationality. Officially enshrined in modern-day bureaucracy, this is a world, in Weber’s terms, in which calculable rules, ‘without regard for persons’, rule the day: When fully developed, bureaucracy stands, in a specific sense, under the principle of sine ira ac studio. Its specific nature, which is understood by capitalism, develops the more perfectly the more bureaucracy is ‘dehumanized’; the more completely it succeeds [sic] in eliminating from official business love, hatred and all purely personal, irrational and emotional elements which escape calculation. This is the specific nature of bureaucracy and it is appraised as its special virtue. (Weber 1991:215–16) Weber, without doubt, was ambivalent about the various effects of rationality in the modern, disenchanted world – as his arresting ‘iron cage’ metaphor suggests. The truly human person, none the less, was one ‘guided by reason, who transforms impulses and desires into a systematic lifeplan, exercises choices, and can improve the world’ (Hillier 1987: 196). This, he thought, distinguished human life from natural events. What is ‘rational’ from one viewpoint, however, as Weber himself stressed, may well be ‘irrational’ from another. Value conflicts were therefore inevitable, competing rationalities emerging, paradoxically, as part and parcel of the overall process of rationalization, without being reducible to one another (Hillier 1987: 197; see also Brubaker 1984). From these traditional rationalist legacies, it is only a short step to rational choice theory itself. A progressive ‘taming’ of the passions, in Archer’s terms (this volume), from the Utilitarians onwards, into ‘enlightened pleasures’, ‘ideal preferences’, ‘rational’ as opposed to ‘irrational’ desires, and finally as items to be considered alongside others when people maximize their utility. To the extent that emotions are addressed here, the economic/ Utilitarian assumptions underpinning these analyses casts them, in large part, as mere ‘givens’ or objects of rational choice itself (see Becker’s [1976, 1991, 1996] work on ‘marriage markets’, altruism and child-rearing). Alternatively, as in much of Elster’s (1983, 1989) work, they are treated, in an explicit fashion, as ‘irrational’ and therefore subversive or undermining of rationality, particularly so-called ‘strong feelings’ (Elster 1999b), which are

60 Simon J. Williams likened to addictions – see, however, his somewhat more balanced assessment in Alchemies of the Mind (1999a). A central tenet of rational choice theory, as this suggests, is a view of Homo economicus as the bearer of given sets of discrete, non-ambiguous, fixed, hierarchical and transitive preferences, providing the criteria to be optimized. The postulate here is that the actor will supposedly choose the action with the best (i.e. optimal) outcome, which maximizes the difference between benefits and costs (Coleman and Fararo 1992b: xi). To be ‘rational’, in this sense, is to act in a way which is consistent with one’s (stable) preference rankings: establishing the optimal relation between the goals and beliefs of the agent (Elster 1989: 30) and his or her situation. Micro–macro linkages are then specified in a way which shows how actions which are ‘reasonable’ or ‘rational’ for actors at the individual level, consciously or otherwise, can combine to produce a variety of systemic social outcomes. These, in turn, are ‘sometimes intended by actors, sometimes unintended, sometimes socially optimal, sometimes non-optimal’ (Coleman and Fararo 1992b: xii). These contentions have generated considerable debate – see, for example, Coleman and Fararo (1992a), Zey (1992a), Holton (1995), Zafirovski (1999a, 1999b). Suffice it to say, in the present context, that the central issue here concerns the extremely narrow definition of rationality which rational choice perspectives embrace. Not only do rational choice theories, in many cases, appear more utilitarian, economistic and individualistic than contemporary neoclassical economics itself, but rationality, as noted above, is ‘multidimensional rather than unidimensional’, oriented not only to profit and other economic variables, but also to ‘power, prestige and related non-economic variables’ (Zafirovski 1999a: 103–5). Rational choice theories, in this respect, commit a ‘double fallacy’ of misplaced reduction and disqualification, by dissolving the non-economic forms of rationality into the economic, or by dismissing these forms as sheer irrationality, ignorance or prejudice. These two treatments of nonutilitarian rationality – its reduction to the utilitarian and its disqualification as irrationality – are logically contradictory to each other. (Zafirovski 1999a: 105) Again we confront here the constituents of rationality itself, including the neglected role of habit, emotion, moral and ethical values as alternative, if not dominant, motives behind the selection of means and ends in economic, political and social decision-making processes. This in turn raises the related question of whether rationality itself, however defined, is the only basis for action (Zey 1992b: 16). Rational choice theory in sum, as this preliminary excursion into the realm of the so-called ‘rational’ suggests, often involves an extremely narrow (i.e. utilitarian) interpretation of

Is rational choice theory ‘unreasonable’? 61 rationality, one which does not simply deny or conflate its multiple forms, but neglects the multidimensional (i.e. emotional, habitual and moral) elements in both ‘rational’ decisionmaking processes and other so-called ‘non-rational’ forms of action. What, then, are emotions, and how might they be conceptualized in other, less negative terms? It is to this particular issue that I now turn, before considering the more specific question as to whether or not, contra this orthodox viewpoint, emotions are themselves ‘reasonable’.

What are emotions? The nature and status of emotions is indeed a hotly contested issue. Throughout much of the discussion and debate, however, emotions have remained poorly conceptualized and inadequately understood. To the extent that they have been theorized, contrasting either/or viewpoints have tended to cancel one another out in self-defeating ways. The sociology of emotions, in this respect, given a variety of competing perspectives, may have muddied the water still further: a victim of its own success perhaps? (See Wouters 1992.) It is possible, none the less, cutting a swathe through these debates – views which range from ‘organismic’ (i.e. biological) accounts at one end of the spectrum, to ‘“strong/weak” constructionist’ accounts at the other – to define emotions in the following multi-faceted terms. As thinking, moving, feeling complexes, emotions are irreducible to any one domain or discourse. As emergent human compounds, located at the intersection of biology and society, they involve a complex mixture of physiological substrates, lived embodied feelings, experiences and expressions, as well as cognitive appraisals/antecedents. These factors, in turn, are linked to ongoing relational practices of a socio-cultural and historical kind (including ‘emotion talk’ and the ‘public language games’ within which it is embedded). Whilst so-called ‘primary’ emotions are doubtless involved here – rooted as they are in our evolutionary biological make-up and shared amongst all human beings as universal features – they are endlessly elaborated, like colours on a painter’s palette, across time and place, history and culture, including fundamental social processes of differentiation and socialization, management and change (Gordon 1990). Emotions, from this perspective, display a ‘deep sociality’ (Wentworth and Yardley 1994), forged within a social habitus, which is at once both communicative and embodied. The cognitive component, however, is complicated and contested: an issue dating back at least as far as Aristotle. Cognition, for example, may ‘trigger’ an emotion. Emotion, in contrast, may ‘influence’ a cognition (Elster 1999a). A cognition may also, as rational choice theorists themselves are happy to endorse, have an emotion as its intentional or propositional ‘object’ (Elster 1999a; see also the Lazarus [1984] and Zajonc [1984] debate). Others, in contrast, stress how this relationship is not merely ‘external’. Instead the ‘cognitive, evaluative-situational aspect is inherent in the very feeling itself (Heller 1990: 80; see also

62 Simon J. Williams Damasio 1994). Consciousness and intentionality, moreover, as Sartre’s (1971 [1939]) and Merleau-Ponty’s (1962) commitment to the pre-reflective structures of being-in-the world reminds us, need not necessarily entail self awareness as such. Nor, of course, should the unconscious elements of human motivation and desire be overlooked: forces which display a continual resilience in the face of rational attempts at ‘emotion management’ or ‘cognitive control’ (Craib 1995). Viewing emotions in this way enables us to grapple with the seemingly problematic fact that whilst, in some sense, they always have a ‘social’ or ‘worldly’ referent – even, as Freud’s classic case-studies reveal, when repressed or unconscious – they also appear, at times, to possess a peculiar ‘life of their own’. This includes ‘feelings we cannot express to our satisfaction’; feelings we can ‘express but that others find difficult to understand’; and, perhaps most importantly of all, the ‘regular experiencing of contradictions between our thoughts and our feelings’ (Craib 1995: 153). Acknowledgement of these seemingly ‘troublesome’ qualities, it should be emphasized, does not, tout court, rule emotions out of the province of reason. To do so is to conflate these recalcitrant, if not ‘irrational’, tendencies with emotions in general, thereby missing the significant role which emotions play in reason and decision-making of a utilitarian or nonutilitarian kind. Emotions, in other words, may be Janus-faced, but one should not confuse the part with the whole. To this we may add the fact, as will be discussed more fully later, that emotions provide a crucial missing link in the interplay between structure and agency. Emotions, from this viewpoint, qua potent social forces and emergent human compounds, are socially constituted rather than socially constructed (Greenwood 1994): a view which refuses to reduce the world to either biology or language, or to conflate what we know with what there is to know. The ‘figurational’ approach of writers such as Norbert Elias (1978b [1939], 1982 [1939]), including his subtle and sophisticated approach to emotions themselves (Elias 1991), consolidates and extends this multidimensional viewpoint: an approach in which biological and social, learnt and unlearnt components interlock in complex ways across the long historical curve of the ‘civilizing process’ (see also Newton 1998; Scheff 1992; Williams and Bendelow 1998). Conceptualizing emotions in this way, however, is merely one preparatory step, given the broader task at hand. The need, in short, is to rethink the relationship between reason and emotion as a challenge to rational choice theories and the ‘unreasonable’ assumptions upon which they rest. It is to this more general issue that I now turn.

Are emotions ‘reasonable’? The question as to whether or not emotions are ‘reasonable’ or ‘rational’, as we have seen, remains far from settled. Within the rationalist tradition itself, for example, the division between reason and emotion has never in fact been absolute or watertight. Plato, for instance, compared reason to a ‘charioteer’ and emotions to ‘galloping horses’, the skill of the former

Is rational choice theory ‘unreasonable’? 63 being acknowledged as ‘worthless’ without the existence of the latter (Jaggar 1989). Aristotle and Spinoza, likewise, accorded emotions more than a peripheral role in their respective philosophies: emotions helping clarify and understand, both morally and spiritually, our place in the world and relation to others. ‘The potentialities of human emotions that are in us’, Aristotle declared in the Poetics, become more violent if they are hemmed in on every side. But if they are briefly put into activity, and brought to the point of due proportion, they give delight in moderation, are satisfied and purified by this means, are stopped by persuasion and not by force. (Aristotle 1987: 59, my emphasis) Aristotle, moreover, was one of the first to highlight the cognitive antecedents and consequences of emotions, and the importance of Katharsis – a much used and abused term since his time (Nussbaum 1986) – in the process of clarification. For Weber too, the rational and irrational were dynamically related, something, he maintained, which could result in one ‘intensifying the other’. A number of issues are at stake here, including: (i) the ‘impact of emotions on the rationality of decision-making and belief formation’; (ii) whether or not ‘emotions themselves are more or less rational, independently of their impact on choice and belief formation’; and (iii) whether or not emotions, as mentioned earlier, can become the ‘object of rational choice’ (i.e. whether people ‘can and do engage in rational deliberation about which emotions to induce in self and others’) (Elster 1999a: 283–4). At least three further models, in this respect, suggest themselves (Barbalet 1998): approaches, opposed to the former orthodox viewpoint, in which emotions are no longer outlawed or banished as the scandal of (male) disembodied reason.

Emotion in the ‘driving seat’: the affective deployment of reason? Contrary to traditional scientific opinion, feelings are just as cognitive as other percepts. They are the result of a most curious physiological arrangement that has turned the brain into the body’s captive audience. (Damasio 1994: xvii, my emphasis) If emotions are not simply ‘irrational’; if they may be channelled in certain ‘appropriate’ socio-cultural ways (as some within the orthodox tradition are prepared to concede); and if they include some internally or externally related cognitive component – contentions, as we have seen, which I myself am happy to endorse – then the way is open for a view in which, far from being the mere saboteur of reason, emotion actually drives or guides, supports or directs its age-old adversary.

64 Simon J. Williams Damasio (1994), for instance, as the above quotation suggests, is a key contemporary exponent of this viewpoint. Reflective thought, he convincingly demonstrates, requires the ‘tagging’ of cognition with emotions. Without this capacity, decision-making becomes difficult, if not impossible, as there is no criterion with which to ‘drive cognition’ in a given direction. Emotions, from this viewpoint, are central to the ‘a/effective deployment’ of reason itself, providing it with salience, direction and purpose (i.e. goal formation). This is not, Damasio stresses, to deny that emotions can ‘wreak havoc’ with the processes of logical thought and rational decision-making; they can and do. The absence of emotion, however, as a variety of studies now convincingly show, is no less damaging or devastating (Damasio 1994). A classic example of this is provided by the case of Phineas Gage, a nineteenth-century railroad worker whose frontal lobes were damaged when an iron bar shot through them by an accidental explosion. Along with the ‘emotional deficits’ which followed this damage, Gage had great difficulty in planning his ordinary life, making disastrous social decisions, whilst dithering endlessly over inconsequential issues. A number of patients with similar damage have now been studied, the general conclusion being that it is this ‘socio-emotional guidance system’ which was affected in the brain of the original Phineas Gage and his modern-day counterparts (Damasio 1994). A key factor here, for Damasio, concerns the notion of ‘somatic markers’. Somatic markers, he argues – ‘somatic’, because the feeling generated is bodily, and ‘marker’, because it marks an image – direct attention to the negative outcome to which a certain situation may lead, functioning as an ‘automated alarm signal which says: Beware of the danger ahead if you choose the option which leads to this outcome’ (Damasio 1994: 173). Feelings, in this sense, let us ‘mind the body “live” when they give us perceptual images of the body, or “by rebroadcast” when they give us recalled images of the body state appropriate to certain circumstances, in “as if” feelings’ (Damasio 1994: 159). Somatic markers in this sense – many of which are based on ‘secondary’ (i.e. learnt) emotions – actually protect against future loss, allowing in the process for choice amongst fewer alternatives. There is still room here, Damasio stresses, for ‘cost/benefit’ analysis and ‘proper deductive competence’, but only after these somatic markers have done their work, so to speak, ‘drastically reducing’ along the way the number of options available for serious consideration. From this it follows that, far from irritating variables, somatic markers actually ‘increase the accuracy and efficiency’ of decision-making processes’ (Damasio 1994: 173). The same, moreover, applies when these somatic markers have a positive rather than negative valence. Somatic markers, in short, are a special instance of feeling, generated from secondary emotions. Those emotions and feelings have been connected, by learning, to predicted future outcomes of certain scenarios. When a negative somatic marker is juxtaposed to a particular future outcome

Is rational choice theory ‘unreasonable’? 65 the combination functions as an alarm bell. When a positive somatic marker is juxtaposed instead, it becomes a beacon of incentive. (Damasio 1994: 174) Emotions, from this perspective, ‘guide’ reason, furnishing it with salience and direction, purpose and priorities, amongst multiple goals and options (Damasio 1994: 174). What makes this a less than ‘optimal’ position, however, as Barbalet (1998: 44) notes, is that it remains somewhat apprehensive about the possibility of emotion undermining reason, particularly in its technical or instrumental guise. Again we return here, as with Aristotle, Spinoza and others, to a view in which emotion can only contribute ‘so much’: a more balanced assessment, without doubt, but one in which emotion continues to play a largely supportive, if not subservient, role. Taking a step back in time, Hume redresses, if not reverses, this imbalance. ‘Passions’, he proclaimed, ‘direct the will, and reason serves the passions’ (Hume 1969 [1739/40]: 462). Reason, on this reading, is well and truly the ‘slave’ of the passions, pretending to no other office than to ‘serve’ and ‘obey’ them – that is, emotion, qua passion, is in the driving seat and reason lost without it. Hume, moreover, raises a further possibility, one which feeds directly into the next synonymous/radical viewpoint (Barbalet 1998). In differentiating between socalled ‘calm’ and ‘violent’ passions – reason being associated more with the former than the latter through the ‘tranquil actions of the mind’ (Hume 1969 [1739/40]: 484) – he points to a view in which passion does not simply drive or direct reason, but, in an important and ‘properly understood’ sense, actually constitutes it: What we commonly understand by passion is a violent and sensible emotion of mind, when any good or evil is presented, or any object, which, by the original function of our faculties, is fitted to excite an appetite. By reason we mean affections of the very same kind with the former; but such as operate more calmly, and cause no disorder in the temper: Which tranquillity leads us into the mistake concerning them, and causes us to regard them as conclusions only of our intellectual faculties. (Hume 1969 [1739/40]: 484, my emphasis) Control, in short, as this suggests, is exercised not strictly by reason, as ‘improperly’ understood, but by a ‘calm and reflective form of passion – an enlightened self-interest’.2

One and the same: a ‘passion’ for reason? The juxtaposition of reason and emotion is a heuristic device forged by rationalist philosophers. If something went wrong with this device, it will not be rectified by

66 Simon J. Williams transferring emotions into the bracket of ‘the cognitive’ while leaving feelings and the body stranded on the wrong side of the great dividing line. (Heller 1990: 79, my emphasis) Here we arrive at a third, in Barbalet’s (1998) terms, more radical viewpoint in which the boundaries between reason and emotions themselves become wholly blurred. Instrumental rationality, from this perspective, like any other form of reasoning, is itself founded on particular emotions (i.e. a passionately held belief and a cherished ideal). Barbalet, for example, discusses James’ (1956 [1897]) insightful essay ‘The sentiment of rationality’ in this connection. Within this essay, the title of which provides the clue, James emphasized the human passion for clarity and order: a passion, that is to say, for ‘generalizing, simplifying and subordinating’ (James, quoted in Barbalet 1998: 54; see also Barbalet 1999). To divorce reason from emotions, in this respect, is like trying to separate two sides of the same coin: a philosophical sleight of hand to be sure. From this it follows that rationality itself is fundamentally embodied. We human beings, as Johnson (1987: xix) states, have bodies. We are ‘rational animals’, but also ‘rational animals’, which means that our rationality is embodied. The centrality of human embodiment directly influences what and how things can be meaningful for us, the ways we are able to comprehend and reason about our experience and the action we take. Our reality is shaped by the pattern of our bodily movements, the contours of our spatial and temporal orientations, and the forms of our interactions with objects. It is never a matter of abstract conceptualization and prepositional judgements. Emotion, on this reading, is no mere adjunct to cognitive processes or the means to practical knowledge. Instead it is woven into the very fabric of our reasoning, from scientific observation and the generation of hypotheses, to ‘moral understanding’ and the ‘communicative rationality’ of the lifeworld (i.e. emotions as ‘accountable, contestable and defeasible’ [Crossley 1998: 36]). ‘Theoretical investigation’, echoing a fundamental Popperian insight, ‘is always purposeful and observation is always selective’, thereby dispelling the myth of dispassionate scientific inquiry once and for all (Jaggar 1989: 161). Just as certain emotions may contribute to the development of knowledge, so too the growth of knowledge may contribute to the development of certain emotions (Jaggar 1989: 161). Rather than repressing emotion in Western epistemology, therefore, it is necessary fundamentally to rethink the relation between knowledge and emotion and construct conceptual models that demonstrate the mutually constitutive rather than oppositional relation between reason and emotion. Far from precluding the possibility of reliable knowledge, emotion

Is rational choice theory ‘unreasonable’? 67 as well as value must be shown as necessary to such knowledge ... the ideal of dispassionate enquiry is an impossible dream but a dream nonetheless or perhaps a myth that has exerted enormous influence on Western epistemology. Like all myths, it is a form of ideology that fulfils certain social and political functions. (Jaggar 1989: 157) Underpinning these issues, Barbalet (1998) stresses, is an attempt to bring emotions of a ‘foreground’ and ‘background’ nature into fuller focus, and to highlight the manner in which conventional understandings of both reason and emotion serve to mask as much as they reveal, not simply about these terms themselves, but also their close relationship to one another. The conventional opposition between reason and emotion, in this sense – following Heller’s (1990) thesis of ‘emotional impoverishment’ – may in large part be due to the ‘cultural discounting of certain “background emotions”: emotions which underpin instrumental rationality but are seldom acknowledged or seen as belonging to some other category (e.g. attitudes, customs) which fundamentally obscure their emotional nature’ (Barbalet 1998: 29–30). Context is equally crucial; differing emotions, and the same emotion in different contexts, having different relations with reason itself, however conceived (Barbalet 1998: 32). To this we may add the observation, alluded to above, that emotion itself, grounded as it is in the action-oriented life-world, is central to the validity claims upon which Habermasian communicative rationality itself rests. Emotions, that is to say, are ‘accountable’, ‘contestable’ and ‘defeasible’ (i.e. reasonable). Habermas, to be sure, is neglectful of this fact, yet the claims he makes, as Crossley astutely observes, are in no way antithetical to it (see also Scambler 1987). What this amounts to is that emotions can strike us as either appropriate or inappropriate, rational or irrational, that we find them perfectly intelligible when we encounter them in others, that we explain this in terms of reasons rather than causes, and that we hold people responsible for them, just as we do for any other of their actions ... even apparently irrational emotional responses are not completely divorced from the rational-intersubjective world. Emotional outbursts do not preclude communicative reasoning, even if they sometimes distort it. (Crossley 1998: 30) Emotion, as this suggests, cannot be factored out of decision-making, any more than it can be driven out of the intersubjective lifeworld and the communicative rationality/validity claims upon which it rests. Habermas, of course, is part of a long tradition of critical theorists, including Adorno, Horkheimer and Marcuse, each of whom took issue with the ‘false’ or ‘distorted’ rationality

68 Simon J. Williams of modernity.3 To this we may add a variety of other critical, if not radical, viewpoints, including feminist epistemological critiques, touched on above (e.g. Jaggar 1989), and the postmodern liquidation of the Cartesian rational subject: critiques, both separately and in conjunction, which point to alternative ways of being and knowing as a challenge to dominant Western traditions of masculine thought and the disembodied claims and unreasonable practices they entail (Seidler 1994). It is not, however, the time or place to enter into a detailed discussion of these latter viewpoints. Enough has already been said, I hope, to cast a long shadow over the first, so-called ‘orthodox’, viewpoint, of which rational choice theory is a key contemporary example. If rational choice theory is defective on this first count (i.e. an ‘unreasonable’ form of reason), then it also suffers, through this self-same neglect of the emotions, from an inadequate conception of micro–macro linkages. It is to this further problem therefore, that I now turn.

Micro–macro relations: emotions as a ‘missing link’? Rational choice theory, we are told, is a strategy which ‘blinds itself to deviations from rationality [sic], with the aim of getting on with a different task – the task of moving between micro and macro levels’ (Coleman and Fararo 1992b: xi, my emphasis). The key point of interest here, it seems, is the way in which institutional structures produce systematic behaviour and systemic outcomes, intended or otherwise (Coleman and Fararo 1992b: xi). Rational choice theory, in this sense, however flawed, constitutes a seemingly rigorous (quasi-scientific) approach to the modelling, prediction and explanation of human behaviour, rooted not in the murky depths of individual psychology or intra-psychic processes, but in socio-economic structures of incentives which lead any ‘reasonable’ or ‘rational’ person to act in a certain predictable way in a given set of circumstances, including the decision not to join in collective action (the so-called ‘free-rider problem’). Committed as it is to these (narrow) ‘domain assumptions’ (Bohman 1992), rational choice theory cannot therefore make the transition from the micro to the macro without severe conceptual limitations (Mouzelis 1995). Whilst its logico-deductive strategy may prove useful when linking ideal-typical micro decision-making models with macro phenomena of an aggregate nature, it is much less useful, Mouzelis (1995) argues, when applied to phenomena of a figurational character (see Elias 1978b [1939]; Newton 1998; Scheff 1992) which entail emergent properties that cannot be reduced to their constituent parts. Rational choice theorists therefore, in Mouzelis’s terms, are faced with the following dilemma: the maintenance of their ‘formalist elegance’ at the expense of ‘reductive generalizations’; or the consideration of emergence and context, including issues of power and spatio-temporal context, with a corresponding ‘loss of logico-deductive rigour’ (Mouzelis 1995: 40).

Is rational choice theory ‘unreasonable’? 69 Again we confront here the limits of an approach, like much of sociology to date, which fails to take emotions seriously. Fortunately, however, a variety of recent work has begun to redress this imbalance (Collins 1981, 1990; Hochschild 1983; Scheff 1990a, 1990b), pointing instead to the crucial contribution which emotions, alongside other emergent properties, make in the explication of micro–macro relations and a variety of other sociological issues besides. It is indeed surprising, on reading this literature, how sociology, as the science of society and geometry of social forms, could ever have done without an explicit emotional analysis in the past. These issues are not, however, as Archer (this volume) points out, resolved by some compartmentalized appeal to ‘emotional man’ (Flam 1990a, 1990b, 1993) to ‘mop up’ those macro-structural effects left unexplained by ‘rational’ or ‘normative’ man: a circular line of argument in which man, qua multiple self, is treated like an organization, and the explanandum (social organization) is conflated with the explanans (the nature of man). Nor should these emotional claims be overstretched or overstated. Emotions, to be sure, play a key contributory role, yet structural and cultural emergent properties which are neither upwardly or downwardly reducible to emotions (e.g. inflation) ensure that this is never the end of the story.4 Barbalet (1998) again provides a promising lead here, augmenting the work of writers such as Scheff (1990a, 1990b), Kemper (1990) and Collins (1981, 1990) in important new ways. Emotion, Barbalet stresses, inheres simultaneously in social structures and social actions. Patterns of ‘structured relations’, in this respect, form the basis of action, which in turn ‘consolidates’ or ‘modifies’ social structures at some later point in time. With respect to class and resentment, for example, it is the structure of class relations themselves, Barbalet (1998: 68) stresses, which tends to ‘determine the emotions which individual class members feel’. Although in no way a simple or direct relationship, the ‘emotional tone’ of class members can none the less be explained in terms of these patterns of class relations. This, in turn, ‘disposes social collectivities to particular types of action, or the absence of action, which feeds back into the pattern of class relations’ (Barbalet 1998: 68). Emotion, from this viewpoint, is both ‘socially responsive’ (in relation to existing social structures) and ‘social efficacious’ (in relation to their action-oriented reproduction or transformation over time) (Barbalet 1998: 81). This, I venture, resonates with the morphogenetic account of structure– agency links advanced by Archer (1995); one which places a premium on analytical duality and temporal principles of interplay and emergence. Emotion, in short, functions in a ‘macrosociological framework in which it is given a social-relational form and structural effect’ (Barbalet 1998: 81). This, in turn, given these socio-relational forms, raises the issue of whether or not the individual (decision-maker) can ever really be the sufficient unit of analysis. Where, then, do we go from here?

70 Simon J. Williams

Discussion and concluding remarks In drawing the diverse themes and issues of this chapter together, at least five points/possible ways forward, themselves mutually reinforcing, emerge. First and foremost is the need fundamentally to rethink reason itself, in other less ‘unreasonable’ terms, which no longer exclude the emotions. Not only does this open the way to an appreciation of the multiple forms which reason and rationality may take, including their communicative and non-utilitarian forms, it also forces us to confront, via the radical or synonymous viewpoint, the role of emotions in the very constitution of (instrumental) rationality itself. To focus on the subversive potential of emotions as the ‘saboteur’ of reason, in this respect, is fallaciously to substitute part for whole, and in doing so to rule out the many different ways in which the former contribute to, if not converge with or constitute, the latter. Emotions, from this viewpoint, may indeed display certain troublesome qualities, but in no way should this blind us, as so often in the past, to their equally ‘reasonable’ and indispensable character. This, in turn, suggests the need for more complex, multidimensional, models of decisionmaking in which emotions, alongside habit, customs and moral values, play a part. Even if social action does result from an ‘individual calculus of interests’, the emotional basis of these interests remains crucial (Barbalet 1998: 65). Emotions, as we have seen, provide reason with ‘salience, direction and purpose’ (i.e. goal formation) (Damasio 1994). They are also bases for the selection of both means and ends in ‘economic, political and social decision-making processes’ (Zey 1992b: 16). Humans however, and this should equally be borne in mind, ‘do not always make rational choices’ (Zey 1992b: 16). Nor are all our behaviours and choices – as notions of habit, practical logic (Bourdieu 1977, 1984, 1990), the pre-reflective (MerleauPonty 1962; Sartre 1971 [1939]) and the ‘unconscious’ (Freud 1984 [1915]) suggest – necessarily the result of cognitive reflection or reflexive self-awareness. Co-opting emotions, like ‘habit’ itself, within existing explanatory frames of reference, as objects of rational choice (see Becker 1976, 1991, 1996),5 takes us little further forward in this respect; reducing them to mere ‘things’ rather than ongoing relational processes with further emotional correlates/consequences (Barbalet 1998: 181). Emotions, preferences and decisions, in this sense, are dynamic rather than static or transitive. To this, as Archer (this volume) notes, we may add that our relations and moral commitments to significant others, including Becker’s (1996) examples of marriage and child-rearing, are the expression of who and what we are, not matters of calculative investment and quid pro quo. This criticism, it seems, applies equally well to Giddens’ (1992) analysis of the ‘pure relationship’ as an ideological ideal of late modernity (Craib 1995): something individuals supposedly weigh up and exit from when costs exceed rewards. The place of emotion in the explication of human behaviour, both rational and non-rational, utilitarian and non-utilitarian alike (Zafirovski 1999a, 1999b), as this suggests, is extremely complex. It includes behaviour that follows

Is rational choice theory ‘unreasonable’? 71 from the ‘action tendencies’ of an emotion (e.g. striking someone in anger); behaviour which is ‘inseparable from the emotion itself’ (e.g. sadistic behaviour and the pleasure of hurting others); behaviour generated on the basis of ‘emotionally induced beliefs’ (e.g. A’s love for B induces A to believe B’s story that C hurt him or her, causing A to hurt C); and behaviour ‘against one’s own better judgement’ under the ‘influence’ of so-called ‘powerful emotions’ (e.g. marital infidelity) (Elster 1999a: 328–31). If these first two points are mutually reinforcing in relation to (non)decision-making processes, then bringing emotions back in also enables us, as we have seen, to contribute to the resolution of a series of other sociological conundrums, including the much debated issue of micro–macro linkages. Emotions, I have argued, echoing Barbalet (1998), inhere simultaneously in both individual experience and social structure: a sequence in which ‘situated relations’ give rise to ‘emotion-based actions’ which, in turn, consolidate or modify social structures at some latter point in time. Emotions, in this sense, central as they are to identity and affiliation, social interests and social action, are socially responsive and socially efficacious (Barbalet 1998): principles in keeping with the morphogenetic emphasis on temporality, interplay and emergence (Archer 1995). Time, as this suggests, is more sensitively and adequately handled, compared with rational choice theory, within this latter morphogenetic framework, built as it is into its very foundations. This, in turn, reminds us that the socially constituted world, with its irreducible properties, cannot be conflated with its socially constructed characteristics (Greenwood 1994). Embedded in these first three points is the need for more sensitive methodologies and multi-method approaches with which to capture these complex, multidimensional (non)decision-making processes, including the structural contexts, emotion-based actions and practical logic of everyday life. A sociological commitment to the emotions, in this and other respects, necessitates a whole new way of ‘seeing the world’ (Hochschild 1983; 1998), including an innovative array of techniques and methods with which to grasp it. Here we arrive at a fifth and final conclusion. The need, that is, for a more emotionally founded, if not ‘passionate’, sociology itself (Game and Metcalfe 1996): one, contra rationalist ‘myths’ of dispassionate or disembodied inquiry, which takes seriously the embodiment of its practitioners as well as those it seeks to study. Sociology, like science itself, is indeed a passionate vocation, one involving imagination and insight, creativity and flair. As a reflexive social discipline/discipline of the social, it celebrates an immersion in life, a compassionate involvement with the world and with others. ... An engaged or passionate sociology involves a sensual and full-bodied approach to knowing and to practices of knowledge such as reading, writing, teaching ... passion, social life and sociology only exist in the in-between, in specific, moving social relations. (Game and Metcalfe 1996: 5, my emphasis, except final word)

72 Simon J. Williams Not only would this serve to re-animate the sociological imagination (see Mills 1959), breathing new emotional life into its classically rationalist bones, it may also help to dispel, once and for all, the ‘myth of dispassionate rationality’. Rejecting emotions, in short, is not a rational choice!

Notes 1

Although ‘dominant’, this view has never in fact gone unchallenged. See, for example, Taylor’s (1989) treatment of the ‘expressivist turn’ as one of the ‘constituent streams’ of modern culture and ‘sources of the self’. See also Bauman (1991), Maffesoli (1996), Mellor and Shilling (1997) and Williams (1998), amongst others, on the Janus-faced nature of modernity in relation to the emotions.


Although not required by Hume’s philosophical theory, the gendered implications of this view of reason, as Lloyd (1995: 55–6) points out, none the less take on traditional associations with ‘maleness’.


See, however, Mestrovic’s (1997: 80–6) critique of these critical theorists’ ‘counterfeit vision of the Enlightenment’.


I am grateful to Margaret Archer for pointing this out. That structural properties themselves are traceable to the practices of past or present agents, and that emotions underpin all our actions in the world – whether latently or manifestly – suggests none the less that the risks of overstretching or overstating this proposition are, in the last instance, unlikely.


Habits, Becker (1996: 8–9) stresses – as long as they are not ‘too powerful’ – have ‘social as well as personal advantages’: individuals, in effect, help ‘guide their own destinies by exercising control over future stocks of personal capital that determine future utilities and preferences’.



Social theory and the underclass Social realism or rational choice individualism? Justin Cruickshank

My concern in this chapter is to argue that rational choice theory (henceforth: RCT) is untenable because it is a form of essentialism. Against RCT I argue for the use of social realism. The underclass issue is used to illustrate the usefulness of social realism together, conversely, with the problems associated with essentialism and RCT. The underclass concept is connected to right-wing essentialist arguments, and this leads some to reject the concept. I argue that it is useful to retain the concept, as individuals in chronic unemployment are in a position which is different from the working class. I develop this argument by using social realism to interpret empirical studies of unemployed individuals. I then move on to deal with a New Right commentator, Charles Murray, who uses RCT to discuss the underclass. The problems associated with this approach are overcome by the realist analysis, because realism can explain how individuals are enabled and constrained by an objectively real social context, whereas the RCT argument is deterministic and reductive, producing a specious stereotype.

Rational choice theory: definition and problems Definition RCT wants to replace Homo sociologicus with Homo economicus, because the former is regarded as a cultural or structural ‘dope’. To take a ‘social’ starting-point is, for RCT, to say that social factors, such as norms or structures, are antecedent to the individual, and that these factors are therefore the independent variable, with individual behaviour being the dependent variable. ‘Sociological’, or, rather, ‘oversocialized’, accounts of individuals posit conceptions of individuals whereby individuals’ attitudes and behaviour are determined by social norms and/or structures. To avoid this, RCT argues that individuals are the independent variable which is antecedent to social factors. The individual, for RCT, is conceptualized as a rational pursuer of self-interest. The rationality involved is an instrumental rationality because it concerns the most efficient

76 Justin Cruickshank means to the goal of realizing individual ‘utility’ (defined as material or ‘ideal’–normative self-interest).1 This means that the study of individuals entails dealing with how individuals seek to realize their ‘preferences’. These preferences are taken to be stable. RCT cannot conceptualize the social mediation of changing preferences, linking preferences to the influence of a particular socio-historical context, as this would violate the notion that individuals are the independent variable. If one did maintain that individuals’ preferences were socially mediated, then, for RCT, this would mean that the focus would shift from the individual to the social context, and preferences would be ‘read off’ from the context; in which case preferences would be epiphenomenal, and individuals would be the dependent variable (i.e. ‘mediation’ would really mean determination). So for RCT, the social context, meaning norms, culture and social structures, is the dependent variable. Culture would refer to the beliefs which individuals happened to hold at any particular time, rather than a system of beliefs or norms which was ‘more than’ individual beliefs. Similarly, structures are taken to pertain to patterns of action, whereby the ‘whole’ is not more than the sum of its parts. Structure would refer to the regular forms of action undertaken by individuals, and so would be reducible down to the level of individuals’ actions in the here-and-now: ‘structure’ would be no more than the aggregate of individuals’ actions. Thus to switch the focus on to the social context as somehow pre-existing individuals’ actions, and giving meaning to decisions and actions, would be to reify culture and structure, treating such entities as irreducible down to individuals, when only individuals have a real existence in their own right (unlike epiphenomenal culture or structure). In addition to having preferences which are stable, preferences are also assumed to be ranked in order of priority. As Hollis (1994: 116) summarizes, ‘Rational Choice theory begins with a single, ideally rational individual, classically Robinson Crusoe alone on his desert island. He has three components: fully ordered preferences, complete information, and a perfect internal computer.’2 Here the ‘perfect internal computer’ will seek to realize the fully ordered preferences, and thus to maximize utility, and these preferences have to be assumed to be stable, so that the ‘computer’ can work on ways of realizing the given objectives, in order of priority. Thus rational choice theories seek to study the pursuit of preferences (which is assumed to result in instrumentally rational action) and not the social formation of preferences. So, with RCT, individuals are ‘black boxes’, that is, they have fixed and fully ordered preferences, which are taken as givens, with no explanation of the socially contingent formation of changing preferences being proffered or possible. Further, such preferences are assumed to be manifest in individuals’ choices, so that, in studying individuals’ actions, one is observing the instrumentally rational pursuit of preferences. One therefore studies the means to an end, rather than the formation of the end, with individuals logically working out, and acting upon, an instrumentally rational means to an (unexplained) end. Of course, there may be irrational action, meaning action which is not based on the instrumentally rational pursuit of self-interest and is, for instance, based on emotion, or

Social theory and the underclass 77 tradition (meaning blind conforming to the norms generated by others’ actions), but this is of no explanatory import, because it is anomalous (given the definition of individuals as rational egoists). RCT thus sets itself the task of explaining ‘normal’ (read: instrumentally rational) behaviour with behaviour that is ‘not normal’ (i.e. behaviour which is not in conformity with instrumental rationality), being beyond its conception of social scientific explanation. That we are to deal with Homo economicus does not mean that the concern is narrowly ‘economic’. Of course RCT will seek to explain individuals’ actions to realize economic success, but it seeks to explain other phenomena too. So we can have a RCT sociology, by foregrounding individuals’ rational preferences, and avoiding the oversocialized structural/ cultural ‘dope’. For example, Becker uses RCT to explain the institution of the family in Western societies. He argues that people base decisions about the family on ‘cost–benefit’ analysis of the pros and cons (Becker 1996: 145–55). I will mention four issues used by Becker to illustrate this point. Firstly, marriages are more stable between richer couples, he argues, because poor people find that being married to another poor person is a disadvantage, and so divorce is more attractive. Secondly, the birth rate increases during economic slowdowns because the time spent on childcare becomes less expensive, meaning that there is less disadvantage in spending time away from work/training in order to look after children. Thirdly, it was more rational for women to stay at home, or to do work which was only parttime, and look after children, than for men to do so, because, until recent changes, it was mostly the case that men commanded higher wages. Fourthly, parents tend to offer material support to their offspring, in the hope that the offspring will, when the parents become old, spend time looking after the parents; and the offspring act in an apparently ‘altruistic’ way, to sustain receipt of material advantage from the parents. Given this, Becker (1996: 154) notes that families would drift apart, in order to maximize self-interest, if government policy increased welfare to the elderly. Thus public policy can provide incentives or disincentives for the cohesion of families, constituted by rational egoists, whose preferences, as realized in the choices mentioned above, are for individual utility.

Problems As Lukes (1968) argues, individualistic sociology will beg the question as to what the social context is, by ‘building it into the individual’. He gives the example of explaining individual acts in a banking transaction, noting that this presumes a social context (of capitalism, banking, etc.) which gives meaning to the actions of individuals, in that particular context (Lukes 1968: 125). In other words, rather than make reference to a social context in its own right, a definition of the context is implicitly imported into a description of individual acts, which begs the question about how to define the social context that gives meaning to individual acts. The purported reduction down to the level of individuals fails then, because instead of according the social context an existence in its own right, it is smuggled implicitly

78 Justin Cruickshank into any discussion of putative ‘facts about individuals’. To say that person X is cashing a cheque is as much a description of a particular context as of individual acts; and if one did (somehow) evacuate the context, then the descriptions would be meaningless. Given this, one cannot maintain that individuals are the independent variable, and that social structures, together with culture, are dependent variables which are reducible down to the level of individuals. Instead, one has to maintain that individuals and an objectively real social context (i.e. culture and structure) are interdependent, in that individuals create a social world (thus: no individuals = no social reality), but that this social reality, which is ‘more than’ individuals, then has some autonomy in its own right. One therefore has to explain how individuals interact with an objectively real social context. How this may be done is discussed in the next section, on social realism, and the structure–agency issue, which concerns how agents with free will interact with a social context that both enables and constrains their agency. If, however, one were really committed to the reductionist notion that individuals were the independent variable, then one would have to press for the reduction from the level of the ‘individual’ to something more specific, namely psychological states: one would have to argue for psychologism. In this case the individualist position would be very similar to holism, for both would maintain that individuals’ behaviour was determined by some causal factor, whether this was mental states or social structures, respectively. In both cases behaviour and preferences would be the result of some other factor which played a definitive causal role, and knowledge of this causal factor would allow one to read off behaviour and preferences from the causal factor, whether it was mental states or social structures. With neither psychologism nor holism could one explain agency, meaning the actions of agents with free will, because, in the former case, behaviour would be an epiphenomenon of mental states, and in the latter case, behaviour would be an epiphenomenon of social structures: mental states or social structures would be the independent variable, whilst attitudes and behaviour would be the dependent variable. Now RCT ends up as a form of determinism too, but this is not because it argues for a psychologistic reductionism to explain preference formation. For as we have seen, preferences are taken as givens, and the focus is upon the realization of preferences via instrumentally rational action. This is where the determinism enters, because, as Hindess (1988: 39) argues, In both cases [i.e. RCT individualism and holistic determinism], actors are creatures of their situations and they act accordingly: in one case because they pursue the most rational course of action given the situation in which they find themselves and in the other because they have internalised the appropriate norms and act on them. ... The mechanism by which individuals are subordinated to their situations may be different in the two cases, but the overall result is the same.

Social theory and the underclass 79 Thus RCT is similar to behaviourism because although RCT does make reference to ‘inner states’, which behaviourism would eschew, these inner states are taken as fixed givens, realized through instrumentally rational action. Therefore, when individuals confront a situation where course of action A will be more efficient for the realization of preference P than course of action B, then all individuals (or all ‘normal’, read: instrumentally rational, individuals) will opt for A. Or, using behaviourist terms of reference, we can argue that individuals react to positive and negative reinforcement stimuli in the external environment which determine behaviour. In this case then, the most efficient means to an unexplained end (action A) would be positive reinforcement stimuli, as acting on this would (all other things being equal) mean realizing the preference. Conversely, less efficient action (action B) would be produced by negative reinforcement stimuli as one would not realize, or would be less likely to realize, one’s preference. What this means is that RCT is a form of essentialism. A theory may be described as ‘essentialist’ if it holds (implicitly or explicitly) that individuals’ behaviour can be explained by reference to some ‘essential property/ies’ which determine/s behaviour. Behaviour would be the dependent variable, and the essential property qua causal factor would be the independent variable. Such essential properties may be structures, mental states, conceptions of human nature, conceptions of gender, and so on. In addition to being deterministic, essentialist positions are also reductionist, because complex social phenomena can be explained away as an epiphenomenon of the independent variable/essential property. Social relations are expressions of some underlying essence, and so social relations must themselves be reduced down to the level of the essential property. With essentialism, therefore, it is not possible to have new knowledge, because one already knows the essential properties of the social realm, and observation will only yield verifications of one’s theories of essential properties, rather than opening up new questions. RCT is a form of essentialism because it presumes to furnish an essence for individuals – even though it avoids a psychologistic attempt to say what determines preferences – by stating that whatever preferences individuals have, they will act in the most instrumentally rational fashion to realize their preferences, in the order of their importance. If individuals did not act in such a fashion, they would be pathological in the sense that individuals would be irrational, or less rational, than was ‘normal’, and so individuals would be beyond the remit of social scientific explanation. At a meta-theoretical level one can say that individuals will all pursue A over B if A is a more efficient way to realize preference P than B; and at an applied level, one may predict how, in a given set of circumstances, rational individuals would act (assuming they had all the relevant information), or explain action in the past by reading it in the light of RCT. RCT methodology will therefore turn on an arbitrary verificationism, reading assumptions based on its ontology of human being into observed action to verify a theory which holds that the essential property needed to understand the social realm is that individuals are rational pursuers of self-interest. As Zey argues,

80 Justin Cruickshank the ability of researchers to construct such [utility] preferences, post hoc, to account for previously observed decisions, makes rational choice theory very attractive to those who are predicting action, but are not interested in the extent to which their theory corresponds to actual preferences of actors. (Zey 1992b: 15, emphasis added) If sociological explanation is to avoid essentialism (i.e. holism, psychologism and behaviouristic conceptions of Homo economicus), and question-begging attempts to define reductively social reality as a ‘dependent variable’, then it has to explain how individuals, who have free will, are enabled and constrained by an objectively real social context. Such an explanation must not only explore how individuals act, but also explore why individuals act as they do, by exploring preference formation in a non-essentialist fashion. As Zafirovski argues, At issue is thus the social formation of the preferences and tastes of individuals. It pertains to the fact that they have different values at different times and in different societies, and that they subordinate certain values, for instance non-economic to economic ones, in a given society (a market economy) or time (modernism). Thus the major fact is not whether people prefer more or less wealth or utility – and for that matter more health, power or status – to less, but rather whether they prefer more or less wealth than status, power or health. (Zafirovski 1999a: 88, emphasis added) Whilst not all RCT is as materialist as Becker implies, meaning that individuals may use instrumental rationality to pursue altruistic preferences (as Hollis 1987: 23 notes), we need to say why some people have certain preferences (at particular points in time) and others have different preferences. This means linking individuals into a social context to explain how changing preferences are socially mediated. This needs to be connected to a study of how both culture and structure are ‘more than’ individuals, and how individual agents with free will are both enabled and constrained by culture and structure. We need to understand the social formation of preferences together with how individuals may be able to realize some preferences but not others, in particular social contexts.

Social realism Social realism develops a social ontology which avoids essentialism and question-begging notions of social reality by linking structure, culture and agency in order to say how an objectively real social context enables and constrains individuals’ agency. Whilst there are many realist texts, in this chapter I will focus upon the work of Archer (1995) because it is specifically developed in relation to the issues mentioned above.

Social theory and the underclass 81 To resolve the ‘structure–agency problematic’, that is, to specify how individual agents with free will are both enabled and constrained by an objectively real social context, Archer draws upon the notion of ‘emergent properties’. Social structures are emergent properties, because although they are created by individuals’ acts, they then exert a causal influence over individuals. Thus capitalist material relations were created by the activities of individuals in the past, and these material relations cannot be reduced down to the level of individual acts in the present: social structures are activity-dependent in the past tense rather than being activity-dependent in the present tense. Such structures may exert a causal influence over individuals, but this means that they only condition, but do not determine, the activity of individuals. Thus, over time, individuals may effect changes in structures. Rather than have behaviour which is simply epiphenomenal, individuals’ activity has a real existence in its own right, and it may exert a casual power over the context in which it is always-already located. So, individuals’ agency is always located into some social context, which may be defined in terms of social structures qua emergent properties, but agency is not determined because it can alter the context into which it is located. Archer refers to material structures, such as the structures which constitute capitalism, as SEPs, or ‘structural emergent properties’. In addition to SEPs, Archer also discusses CEPs, or ‘cultural emergent properties’. CEPs are belief systems which, as emergent properties, are irreducible to the beliefs, preferences and actions of individuals in the present tense. Thus different groups may interpret the same belief system in different ways; or draw upon different traditions of thought (which are irreducible down to what particular individuals believe in the present tense) in support of different material vested interests. An example of the former would include arguments over the interpretation of Marxism by Marxist groups; and an example of the latter would be the use of socialist ideas by ‘reformist’ political groups seeking to improve conditions for working-class people clashing with New Right ideas drawn upon by governments in effecting ‘free-market’/strong-state policies. As with SEPs, CEPs may be changed or reproduced by agency over time, depending upon the activity of agents. So individual agents have beliefs or preferences which are drawn from CEPs, and these preferences may be held by actors with an SEP vested interest (e.g. working-class people supporting reformist policies). This is not to imply that one can read off CEPs from SEPs, however. Apart from the fact that people in similar material positions may adhere to different values, with, for example, one poor person supporting revolutionary socialism, another supporting reformist socialism, and another having right-wing views, people may have values which clash with their material vested interests. For example, one might surrender a career to look after a relative if one adhered to a belief system which emphasized altruism over material self-enrichment. This acknowledges that individuals are not determined, but it does not return us to an individualism whereby we focus on the realization of (unexplained) preferences. Rather, from a realist perspective, we may explain not only how preferences are

82 Justin Cruickshank socially mediated, in that agents with free will, and who have similar material vested interests (i.e. they have a similar relation to a particular SEP), draw upon different CEPs. We may also explain the social formation of preferences, by focusing upon how individuals relate CEPs to their own lives in a way which is broader than material self-interest, and which is socially mediated, in that individuals are always-already located in some form of perspective/CEP which will condition – but not determine – their ideational choices and prejudices. In sum, structure, culture and agency are interdependent because structure and culture are activity-dependent in the past tense, and agents will reproduce or change both structures and culture with their agency. This agency may involve seeking CEPs which link to SEPs by promoting vested interests (or are perceived as promoting such interests). Or it may involve acting on the basis of a CEP with far less regard for its material felicity, with people being influenced more by ideas that they have been socialized into, which is not to make individuals the dependent variable of culture (or cultural dopes), because action and attitudes would be mediated by the belief system. Individuals’ attitudes and beliefs need not necessarily correspond directly to a CEP which they adamantly agree with. Indeed, a key problem with theories which do construct individuals as cultural dopes is that individuals have contradictory views, meaning that they add their own interpretations into their relationship to CEPs (i.e. they exercise their free will in their capacity for agency).3 Research based upon social realism would have to produce ‘analytic histories of emergence’. What this means is that the task of social science is to explain how an objectively real social context was either altered, or left unchanged, as a result of agency. To provide such an explanation, research would have to employ an ‘analytic dualism’, to separate (i) structure and culture, and (ii) agency, in order to study their interplay. So, whilst structure, culture and agency are always intertwined in reality, to study their interplay, it is necessary to use a theoretical abstraction to separate the different factors. Using such an abstraction, or theoretical artifice, one may then explain the interplay of structure, culture and agency over time, via the ‘morphogenetic cycle’. The morphogenetic cycle breaks down the interaction between structure and agency into three temporal phases, which are, obviously, theoretical constructs rather than three ontologically discrete phases where either structure, culture or agency exist in isolation. The three phases are: structural conditioning (i.e. the context in which individuals find themselves), socio-cultural interaction (i.e. what individuals do), and the resulting structural elaboration (morphogenesis, i.e. change)/structural reproduction (morphostasis, i.e. continuity). By holding that social structures and culture are irreducible to the activities of individuals in the present tense, Archer (1995) argues that we may study the interplay of structure, culture and agency over time (making an artificial distinction between the three), with agency resulting either in change or in continuity. The last phase of the cycle then furnishes the first phase of the next cycle, so that past actions influence future actions, by influencing the social context which enables and constrains others at a later time.

Social theory and the underclass 83 If one tried to argue that social realism was an essentialist position, with emergent properties being essential properties, the following points could be made. Firstly, social realism is not deterministic. It explains how SEPs and CEPs condition agency, and are the medium of agency. That is, structure, culture and agency are interdependent variables. There is no essential property which is essential because it allows social relations to be explained reductively as dependent variables which are affected by the independent variable. Agents may have their agency conditioned and mediated by emergent properties, but agents can alter these emergent properties. Secondly, the social realist ontology of emergent properties is a meta-theory which supplies some guiding precepts for methodology, meaning that a general ontology of emergent properties says nothing about specific emergent properties. Social research may provide analytic histories of emergence using a social realist ontology, but this would entail producing fallible knowledge claims about specific interplays of structure and agency over time, which may be challenged by other accounts which also draw upon the general notion of emergent properties. It certainly would not entail saying that structure, or culture, was an essential property/independent variable, which realism had definitive knowledge of, and which determined the behaviour of individuals qua dependent variables/ structural–cultural dopes. Thus social realism, unlike essentialism, can produce new knowledge, because instead of reading off behaviour and attitudes from some determining essential property, research would interpret the changing form of structural, cultural and agential interdependency over time. Given this, which brings me to the third point, a realist social explanation would avoid the homogenization that accompanies essentialism. If one reduces social relations to an expression of some fixed essential property, then it follows that instead of accepting complexity and diversity in the social realm, there would be a homogenisation, as individuals qua dependent variables all displayed identical forms of behaviour (in specific circumstances), determined by the essential property. Conversely, social realism explains the interplay of structure, culture and agency over time, and so it can account for complexity, saying, for example, how some groups may succeed in changing a structure (reforming capitalism, for instance), or reworking a cultural system (changing attitudes towards prevailing gender norms, for instance), in the face of opposition from other agents.

Explaining the underclass Biology, culture and essentialism Having discussed various theoretical positions I will now consider an issue concerning social stratification, namely the underclass issue, in order to assess the usefulness of social realism in relation to RCT. The notion of an underclass is controversial in social science because the

84 Justin Cruickshank concept has overtly right-wing political origins. In the nineteenth century the idea was mooted that there was a distinct class ‘beneath’ the working class. The purpose of positing the existence of such a class was to distinguish the ‘respectable poor’ from the ‘unrespectable poor’. Whilst the former may experience poverty or hardship through various circumstances, they would persevere to change their situation, whereas the latter, so it was maintained, were poor because they were indolent, with tendencies to commit crime. In other words, the notion of an underclass was to be used to explain how people experiencing chronic poverty were themselves to blame, because they were ‘feckless’ individuals. To explain the existence of such a group, biological notions were used, to the effect that the underclass was constituted by a ‘sub-race’ of individuals who were unable to be responsible members of civil society. This putative underclass sub-race had such appellations as the ‘substratum’ and the ‘residuum’. Attempting to define the substratum, one nineteenthcentury commentator, Henry Mayhew, listed physical and social traits such as ‘high cheek bones and protruding jaws’, ‘slang language’, ‘repugnance to continuous labour’ and ‘love of cruelty’ (cited in Morris 1994: 17). As the putative cause of the underclass was held to be biological, there could be no social/political ‘remedy’. Thus suggestions were put forward to segregate the underclass spatially and force them to work (Morris 1994: 20). Otherwise there was the fear that the underclass would corrupt the respectable working class with their indolence (although it was unclear how ‘decent’ workers could be corrupted by the ‘subrace’, if both were biologically determined). In the twentieth century the notion of an underclass was retained, although the causal factor changed from biology to culture. In the 1960s and 1970s the debate turned upon notions of a ‘culture of poverty’ (in the USA) and a ‘cycle of deprivation’ (in the UK). These arguments held that there was an underclass constituted by individuals who chose not to work, relying instead on welfare and/or crime; and the underclass was also characterized by singleparent families, meaning single mothers, who relied on welfare. The argument here was that a self-sustaining culture had developed whereby individuals born into underclass families were socialized into welfare dependency, rather than self-reliance via gainful employment, and that individuals (especially male individuals) born into such families would, given the lack of a ‘father figure’ to discipline them, often engage in crime, and fail to seek work. This view was ardently supported by Keith Joseph in the UK, who was Secretary of State for Education and Science under Margaret Thatcher, between 1981 and 1986. ‘Joseph’s central idea was that of the inter-generational transmission of poverty through a “cycle of deprivation”, where inadequate child rearing leads to failure at school, which leads to unemployment and unstable families, which continued the inadequate rearing of children’ (Bagguley and Mann 1992: 121). Such views are clearly examples of right-wing political ideology turning upon essentialist conceptions of an underclass. With both the ‘biological’ and the ‘cultural’ (or nature and

Social theory and the underclass 85 nurture) views, the argument is that there is a homogeneous group which can be explained by reference to some determining mono-causal factor. A biological cause (namely belonging to a ‘sub-race’), or a cultural cause (namely being socialized into a deviant sub-culture), would be an essential property, which was essential because it determined individuals’ attitudes and behaviour, with all individuals in the underclass being caused to have the same attitudes and behaviour. The essential property (the ‘sub-race’ or the sub-culture) would be the independent variable, and individuals’ attitudes and behaviour would be the dependent variables. This has the consequence that one need not bother to do research, as such ‘knowledge’ of the underclass is definitive, or, if one eschewed ‘armchair dogmatism’ (i.e. speculation based on right-wing negative stereotypes) to engage in empirical research, such research would be an arbitrary verificationism. Thus in the nineteenth century, one would have thought that the very notion of a respectable poor would cast doubt on the notion of an unrespectable poor, simply because it acknowledged that poverty may not be due to individual ‘failings’. However, an a priori conviction that an underclass must – and does – exist led to commentators reading such a notion into research. Thus Booth carried out a survey of London in the 1880s, finding that one third of the population lived in poverty, including members of the labouring poor (Morris 1994: 21). Yet he insisted that although the working class may endure poverty, one could still argue that there were people in poverty who refused to work, and engaged in crime, and that such people would ‘degrade’ the ‘decent’ working class (Keating in Morris 1994: 21–2). Similarly, Joseph refused to accept the findings of the Social Science Research Council after it carried out a study into the notion of a cycle of deprivation which refuted his theory. Against the notion of a cycle of deprivation, Rutter and Madge (1976: 304), drawing on the refuting data, argue that At least half the children born into a disadvantaged home do not repeat the pattern of disadvantage in the next generation. Over half of all forms of disadvantage arise anew in the next generation. On the one hand, even where continuity is strongest many individuals break out of the cycle and on the other many people become disadvantaged without having been reared by disadvantaged parents. What this outcome means is that instead of adhering to some a priori conviction that there is a homogeneous group conforming to some essential property, and that the members of this group can have their behaviour explained by reference to one definitive/essential factor (in this case a deviant sub-culture), we need to look at the interaction between individuals and their changing situations. We need to understand how structure, culture and agency are interdependent variables, in order to understand the lives of people who experience poverty and chronic unemployment.

86 Justin Cruickshank

From essentialism to social realism One way of pursuing such research would be to engage in empirical studies on the way in which individuals dealt with the structural decline of a regional economy, without relying on the notion of an underclass. Such studies could focus upon how individuals utilized social networks to get access to the remaining formal jobs (Morris and Irwin 1992), or how social networks were used to get access to informal jobs, that is, jobs undertaken whilst receiving benefit (MacDonald 1994). Such studies emphasize agency, focusing upon individuals’ choices and acts, and this avoids essentialism by avoiding a deterministic reductionism which seeks to posit the existence of a homogeneous group by reference to a mono-causal essential factor. However, such studies implicitly turn on a form of methodological individualism by dealing exclusively with individuals’ acts without explaining exactly how the social context provides enablements and constraints for agents. In other words, this work begs the question about how individuals interact with the social context: the context is read into the acts of the individual. To overcome this implicit individualism, we may turn to social realism. The use of social realism to guide empirical research would begin with an analytic distinction being made between ‘structural emergent properties’ and agency, with SEPs pertaining to economic structures, and specifically, in this example, the decline of manufacturing industry (phase one of the morphogenetic cycle). Rather than make agency epiphenomenal, by saying that passive agents simply ceased to be economically active, one would then move on to assess how the structural situation provided enablements as well as constraints. This would mean dealing with socio-cultural interaction (phase two of the morphogenetic cycle) by studying how individuals used new opportunities such as the formation of job-finding networks for formal or informal jobs. It may also mean studying how individuals can adhere to mainstream values about self-reliance and self-supporting work, whilst claiming benefit to maintain themselves and their families (MacDonald 1994). This would mean studying how individuals in different social positions use a CEP in different ways, producing contested interpretations of one set of norms. For example, those who believed in a work ethic and who claimed welfare whilst working would have a different interpretation of the norms concerning supporting oneself from those who wanted welfare to be reduced. This would bring us to phase three of the morphogenetic cycle, which in this case is morphogenesis (change), and this would then be the first phase of studying how individuals responded in this new context. From this position one may argue that it is useful to retain the concept of an underclass, which can be defined in terms of individuals who, other things being equal, would be able to work, yet who are chronically excluded from secure employment. This means that it would be incorrect to classify them as ‘working class’, for such individuals would face a different

Social theory and the underclass 87 set of enablements and constraints from the working class. Such a use of the concept ‘underclass’ would escape essentialism because it would not seek to posit the existence of a homogeneous group defined in a deterministic and reductionist fashion. Instead of reading off behavioural and attitudinal traits from a definitive ontology, a guiding ontology of emergent properties would inform empirical research into agents’ activities. This would also avoid the implicit individualism of empirical work which fails to explain contextual factors in their own right, as emergent properties.

Rational choice and the underclass: a new essentialism RCT may also claim to explain the existence of an underclass in a way that avoids essentialism, meaning in this case the deployment of ‘culturalist’ arguments about an underclass. Thus Murray (1984), whilst wishing to criticize ‘reformist’ policies in favour of a New Right emphasis on ‘free markets’ and strong government, replaces a ‘culture of poverty’ view with an argument about underclass individuals being rational agents. Instead of underclass individuals being different from ‘normal’ citizens in some way, Murray holds that underclass individuals are, like everyone else, rational agents seeking to maximize their (material) utility. He argues that: There is no need to invoke the spectres of cultural pathologies or inferior upbringing. The choices may be seen much more naturally, as the behavior of people responding to the reality of the world around them and making the decisions – the legal, approved, and even encouraged decisions – that maximize their quality of life. (Murray 1984: 162, emphasis added) This means that the development of an underclass in the USA could have been predicted (indeed, in some instances [was] predicted) from the changes that social policy made in the rewards and penalties, carrots and sticks, that govern human behavior. All were rational responses to changes in the rules of the game of surviving and getting ahead. (Murray 1984: 154–5, emphasis added) What Murray is referring to here are welfare changes by well-meaning liberals, which had the effect of creating an underclass. These changes made it more rational, Murray argues, for young men to receive welfare than to take low-paid work. He also argues that changes to AFDC (Aid to Families with Dependent Children) made in the 1960s meant that single mothers received higher benefits than before. Single mothers and their partners were now

88 Justin Cruickshank allowed to cohabit with no loss of benefit (providing that the man was not legally responsible for the child), and some income from paid employment was allowed without loss of benefit. To discuss how rational individuals would increase their utility by responding to this situation, Murray (1984: 156–62) discusses a hypothetical couple called Harold and Phyllis. This couple are described as of average intelligence, with no skills and from a poor background. Before the welfare reforms, Harold could not live with Phyllis if she received AFDC, and a low-paid job would pay more than welfare. If Phyllis married Harold, they could live together and she could supplement her income with a part-time job. After the reforms, however, they could live together, with one partner having a job if he or she chose, and the other receiving the higher benefit. In this case they would not have to get married and, indeed, if they remained unmarried, their income would be much greater, as benefits would not be cut. Further, Murray (1984: 179–84) argues that social stigma was removed from receiving welfare, with reformers failing to distinguish the undeserving from the deserving poor, presenting all poor people as victims and encouraging the view that welfare was a ‘right’ to be exercised. Murray also argues that the ‘rules’ changed with regard to crime, whereby the chances of being caught fell, whilst the penalties if caught were weakened. This argument may raise the issue of differential rationality, with less rational individuals maximizing short-term gains which are antithetical to material utility in the long term. The material situation for those on welfare may have improved (according to Murray), but to be in receipt of welfare in the long term is to be in a far worse position than to work one’s way up from a poorly paid job, which gave a lower income than welfare initially, to a better paid job.4 As we have seen with Becker’s discussion of parent–offspring relations, the notion of a longterm view can be argued for by RCT: utility may not necessarily be short-term utility. In which case, underclass individuals would be pathological in the sense that they were innately less rational than ‘normal’ individuals. Such a view would usher in essentialism again, because instead of a sub-culture being the essential property, one would have turned to some essential property concerning human being to argue that underclass individuals were in the underclass because they were innately less able than normal people. A homogeneous deviant culture which treated welfare dependency (and crime too) as normal would exist because underclass individuals were not rational enough to prioritize goals which would eventually remove them from poverty. However, as we have seen, Murray (1984) does not want to argue that underclass individuals are pathological. Instead he wants to criticize ‘dogooder’ liberal reformers for failing to understand human nature.5 His position, though, is still essentialist. For he is saying that the essential property with which to understand social relations is that individuals will seek the most instrumentally rational action to realize a preference. So in situation S, all rational individuals who have preference P will follow course of action A over B, if A is the most efficient means to realize P. Thus all (rational, i.e. non-pathological) individuals will

Social theory and the underclass 89 react in an identical fashion to the external milieu. Or, to put it another way, all individuals will react to the prevailing positive and negative reinforcement stimuli in the same way. Hence a discussion of a hypothetical couple whose ethnicity is not mentioned is used to explain the behaviour and attitudes of all individuals within the putative underclass in the USA, which is commonly held to be constituted by Black people, in inner cities. Issues such as racism and the decline of manufacturing in inner cities are irrelevant because the independent variable is individuals, or rather the trait of individuals as pursuers of instrumentally rational means to a given end, and so the explanation can be reduced down to the determining property of human being, namely that individuals are instrumentally rational. Instead of studying the interplay of different factors, such as structure (economic decline), culture (how individuals may respond to mainstream norms and racism) and agency, to explain how individuals respond to a particular configuration of enablements and constraints, in different ways, a homogeneous category is posited, with all individuals in this category reacting in identical ways to external stimuli. Given this behaviouristic conception of how individuals operate or react to the social environment, rationality must be short-term rationality, because ‘a bird in the hand is worth two in the bush’. In other words, if one were confronted by a mutually exclusive choice between living on increased welfare or working in a low-paid job which might at some point in the future lead to a higher wage, then one would be more rational to retain the immediate gratification afforded by the former option. It is more rational to seek certain gratification now than risk no gratification later (and in the present). A preference for increased utility will replace deferred gratification with immediate gratification. Of course, short- and long-term gratification may not be mutually exclusive. For example, a teenager paid a large stipend by his or her parents to go to university will (usually) be in a better position, both before and after university, than a teenager who works instead of going to university. However, when the choice is mutually exclusive, the rational seeker of increased utility cannot surrender immediate gratification for future gratification; and so one would not give up a reasonably paid job to endure poverty as a mature student to achieve a better paid job three years later (even if this job were guaranteed). If one has a preference P for increased utility (i.e. material gain), and if situation S means increased utility via welfare (course of action A) over work (course of action B), even if work may increase utility, then the rational individual will choose A over B to realize P, in situation S. All of this leads Murray to argue that S needs to be changed, so that welfare is disadvantageous to utility. If S is changed with welfare being decreased, then we can know a priori that rational individuals will leave welfare and enter work: we can predict how rational individuals will react to a change in the positive and negative reinforcement stimuli.6 So, to sum up, Murray holds that there is an underclass, and that this ought not to exist, because it entails people choosing to live on welfare (and possibly crime) when they could be

90 Justin Cruickshank supporting themselves via (legitimate) work, and that the people to blame for this situation are liberal reformers, rather than the underclass individuals, who are ‘normal’ (i.e. instrumentally rational). Therefore, it follows for Murray that, if we change the positive and negative reinforcement stimuli, or the carrots and sticks, as he puts it, we will remove the underclass by making it more rational to replace welfare dependency with work. Murray’s descriptive analysis may be internally coherent, because if we assume that individuals will seek to maximize material utility, and that A (welfare) is more propitious for this than B (low-paid job), in situation S (the prevailing policy, or carrots and sticks), then it does follow that in changing S so that B is more rational than A,there will be a concomitant change in behaviour. However RCT lacks the conceptual resources to say why we ought to change A for B, and so Murray’s prescription does follow from his descriptive framework. If individuals are acting rationally, then such a view means interfering with their rational– normal conduct, on the basis of some normative commitment to paid employment, which is not justified by the use of RCT to analyse how people are acting. In short, the ideological prescription is not justified from within RCT premises. Murray’s concern, though, is with prescription, rather than analysis, because he wants to advance New Right dogmatics, and negatively stereo-type unemployed people (as people choosing to rely on welfare instead of supporting themselves), in order to justify welfare reduction. RCT is used as a means to this end, because in saying that underclass individuals are normal, Murray can sharpen his polemic against reformers for misunderstanding human nature. RCT is used to help Murray realize his preference for reduced welfare. It may be argued, therefore, that it is unfair to associate RCT with Murray, because Murray is not primarily concerned with social scientific analysis in its own right. However, this instrumental use of RCT does illustrate the intrinsic problems with RCT. RCT, like any form of essentialist argument, buys certainty at the cost of veracity. Intellectual and conceptual violence is done to the lived reality of the complex social world, because the answer is known before any question is put. The social world is understood in terms of an essential property, which is essential because it determines behaviour, and thus explanation is to be reductive, with social relations being the result of an essential property. Given this, the theory will be taken to have a priori validity, and human behaviour is either read off from it (to predict how any and all individuals would react in a given situation), or read into a particular situation. In the latter case we have an arbitrary verificationism, which ‘proves’ the ontology of essential properties (whether this is holist, psychologistic or behaviourist, etc.) to be correct by taking any social phenomenon to be an expression of the essential property. This, in turn, leads on to a specious homogenization, as groups of individuals are all taken to have exactly the same attitudes and behaviour, because their behaviour can be explained reductively by reference to the determining essence.

Social theory and the underclass 91 Murray’s argument illustrates this by constructing a negative stereotype of unemployed people, arguing that they choose not to support themselves (and commit crime too) because this group of individuals are all reacting in exactly the same way to the external stimuli they confront. They can do nothing else, because, given the essential property of human being, namely the instrumentally rational pursuit of utility, they are determined to react in this way. Therefore policy ought to be based, for Murray, on such an a priori conviction, rather than on a study of how individuals interact in different ways within a changing social context which enables and constrains their agency. Conversely, if one is concerned with social scientific analysis, then one ought to eschew essentialism to study the complex interplay of different factors. I have argued for the use of social realism, arguing that this enables us to explain how structure, culture and agency are interdependent, with agency being enabled and constrained by emergent properties, in the form of SEPs and CEPs. As regards the underclass debate, this social realist view would allow one to explore how economic decline affected one category of individuals, who then utilized new enablements and reinterpreted main-stream norms to suit their new material situation. Such a view would thus allow us to explain why and how individuals acted as they did, by explaining how individuals’ beliefs were socially mediated, and how people were enabled and constrained by SEPs and CEPs. We may also study the preference formation of New Right ideologues, who describe people as rational egoists in need of more stick than carrot, but that is another story.

Notes 1

Thus whether my goal is to increase the profits of my company or to increase funding for a charity, I will, for RCT, pursue the most efficient means to realize this preference. As Hollis (1987: 23) argues, the RCT model of humanity as ‘bargain hunters’ may imply an exclusively materialist/egoist conception of humanity, and the rational individual ‘may turn out to be generous, far-seeing and altruistic; but, even if he does, he will still reject even slightly inferior ways of exercising these qualities’.


Owing to lack of space, I will not explore the issue of complete information here. This concerns the issue of ‘bounded rationality’, which tries to modify RCT with the clause that agents lack knowledge of all the facts required to make the most instrumentally rational decision.


This is illustrated in the argument by Abercrombie et al. (1980) that notions of a ‘dominant ideology’ fail to notice that individuals have contradictory views and do not passively imbibe such an ideology.


Wilson (1987) rejects the notion that welfare payments have increased in real terms. He argues that economic/structural decline has produced a Black urban underclass in the USA. However, he ends up making culture an epiphenomenon of structural change, with behaviour being an epiphenomenon of culture. In doing this Wilson has, as Bagguley and Mann (1992: 115–16) argue, committed the

92 Justin Cruickshank ecological fallacy of generalizing about the specific character of individuals from aggregate data. Furthermore, in a later work (Wilson 1991), he ends up detaching culture from structure, to give a basically ‘culturalist’ account of the underclass. 5

Note, however, that in Murray’s 1990 and 2000 work on the British underclass, his arguments turn from rational choice to culture, with all the emphasis being on individuals being socialized into ‘deviant’ norms.


For a discussion of how RCT cannot conceptualize deferred gratification, see Hollis (1989: Ch. 6). In this chapter, or, rather, the appendix to this chapter, Hollis discusses a tale of an ant who invested and a grasshopper who consumed. The former lives a ‘grey life’ with no pain and little pleasure, and the latter lives a life of immediate pleasure and expires early. In other words, the safe option has little pleasure because the cost of security in the future always entails compromising pleasure in the present; and to prioritize pleasure in the present precludes long-term pleasure in the future.


(Ir)rational1 choice A multidimensional approach to choice and constraint in decisions about marriage, divorce and remarriage Kay Peggs and Richard Lampard

The aims of this chapter are to examine the applicability of rational choice models to issues concerning plans and decisions about partnering, separation and repartnering. At the centre of orthodox models of rational choice is the notion that social interaction is guided by a rational choice between alternative outcomes that are considered in terms of costs and benefits (Coleman 1990). An action is taken only after its ‘benefits and costs have been weighed’ (Zey 1998: 2) and individuals seek to maximize benefits and minimize costs to produce the optimum outcome in any choice situation (Coleman and Fararo 1992a). Although rational choice models use an economic metaphor they are theoretically generalized to explain behaviour studied in nearly all social science disciplines (Zey 1992b). Criticisms of this approach maintain that this narrow definition of rationality obscures important variables present in choice situations. For example, rational choice models typically exclude emotions, because rational choice theorists suggest that when emotions are involved in action they tend to ‘overwhelm the rational mental processes’ (Elster 1985: 379). In this context we draw on the work of Milan Zafirovski (1999a), who posits a multidimensional conceptualization of rationality to include utilitarian and non-utilitarian forms. This multidimensional approach is valuable in our study of couple relationships since they are a major site of emotional involvement in modern societies, and couple breakdown is a source of conflicting emotions such as relief and anxiety. Advocates of orthodox models of rational choice postulate that self-interest is the sole motivation in decision-making (Zey 1998: 41), and thus argue that we need not investigate the motives of actors (Sciulli 1992: 162). For this reason, rational choice theories remain theoretical rather than being empirically demonstrated (Zey 1992b; see also Cruickshank in this volume). By examining qualitative data on the attitudes and decisions of a sample of 81 interviewees in relation to past and future relationships, we aim to show that emotions and rationality are not fundamentally opposed (Williams 1998), and rather that it is only through

94 Kay Peggs and Richard Lampard emotions that decision-making becomes possible (Elias 1978a; Williams 1998; and this volume). Before discussing the data collection and relevant findings, the chapter begins with an outline of the theoretical context in which we locate our analysis. Our conclusions return us to these theoretical issues since we show that a multidimensional rational choice model is more suited to the study of the formation of couple relationships and relationship breakdown. The choices that individuals make in this context are complex and multi-faceted, involving emotions, values, power relations and utility; such complexity cannot be forced into a unidimensional approach to modelling choice. We thus regard as inadequate the framework adopted by rational choice theorists such as Becker (1991), whose models can nominally incorporate non-economic and normative factors but who see such factors as contributing, alongside economic ones, to a single utility function. In our view, factors such as emotions need to be considered in parallel with (or instead of) factors contributing to the economically oriented form of utility (Etzioni 1988: 112).

(Ir)rational choice In rational choice models, rational human action is defined simply in terms of utility, profit and wealth (Zafirovski 1999a). By acting rationally an individual will maximize benefits and minimize costs (Coleman and Fararo 1992b: ix). Thus, rational human action is conceptualized as ‘profit-optimizing behavior’ (Zafirovski 1999a: 47). Criticisms from rational choice theorists themselves have highlighted how the notion of ‘optimization’ raises the problem of information overload, where we have too much information to be able to make a rational choice. When making a decision, the potential for information-processing is restricted. Moreover, in the course of making decisions, actors often must consider alternatives sequentially and decide about them as they are presented (Simon 1955). In such situations the individual will satisfice (choose the first satisfactory alternative) rather than optimize (Simon and Associates 1992: 43). So satisficing is a more valid explanation than optimizing because time constraints and human limitations prevent optimizing (Zey 1992b: 19). Satisficing is clearly more useful in explaining how individuals choose their partners since it would be impossible to meet all potential partners before making a decision. However, the notion of satisficing retains the maximization of utility as the goal of choices (Zafirovski 1999a), thus the basic premise of rational choice remains unchallenged. A more fundamental problem for our study is the role of the ‘irrational’ or ‘non-rational’ in making choices. Reason is emphasized in rational choice theories, and irrational behaviour, controlled by faith and emotion, is seen as interfering with rational decision-making. The problem here is that utilitarian rational choice models omit non-economic, ‘irrational’ influences on social action (Zafirovski 1999a) and apply the terms ‘irrational’ and ‘non-

(Ir)rational choice 95 rational’ to behaviour that cannot be constructed to fit into rational choice models (Zey 1992b: 15). Such criticisms are not new. Zafirovski (1999b: 495) suggests that, ‘In retrospect, classical sociology in its Durkheimian–Weberian version emerged as a result of dissatisfaction with utilitarianism as found ... in Herbert Spencer’s “rational choice” social theory.’ Max Weber (1968 [1922]) considered the importance of values in rational behaviour. He distinguishes between instrumentally rational (zweckrational) action and value-rational (wertrational) action. He contends that instrumental rationality is where ‘the end, the means, and the secondary results are all rationally taken into account and weighed’ (Weber 1968 [1922}: 26). Value rational action includes ‘actions of persons who, regardless of possible cost to themselves, act to put into practice their convictions’ (Weber 1968 [1922]: 25). Rational choice models deal with the first and not the second type of rationality (Zey 1992b). Chris Shilling (1997) demonstrates that, for Émile Durkheim, the rational demands of society are closely connected to ‘irrational forces’. Habit, values and emotions form the basis for the choice of ends and means in economic, political and social decision-making processes. Decisions made out of passion, fervour, fear or rage are not calculated, rational, means–endsrelated decisions in the classical economic sense (Zey 1992b: 23). We can see, therefore, that one of the major problems with applying utilitarian models of rational choice to decisions concerning couple relationships is that such models are unidimensional. In this respect Zafirovski (1999a: 57) helpfully posits a broader model of rational choice that can be defined in utilitarian and non-utilitarian terms and in micro- and macro- terms. Such a conceptualization can be used more successfully in examining considerations about and choices regarding couple relationships. This is because his multivariate model of rational choice includes economic variables such as profit and wealth, social-psychological variables such as ties and networks, cultural variables such as norms and values, and political variables such as power and domination (Zafirovski 1999a: 59). All of these variables can, and often do, play some part in decisions about couple relationships. In this way, Zafirovski (1999a: 104–5) suggests, we can conceptualize rationality as multidimensional and thus oriented towards other variables alongside the more economic aspects of utility. The cultural variables discussed by Zafirovski include norms, values, status and ‘cultural capital’ (Bourdieu 1984). For example, choices about partners are often based on ‘cultural similarities’ such as education rather than on maximizing economic utility (DiMaggio and Mohr 1985). Changing values affect couple relationships. Attitude surveys reveal that values about cohabitation are changing, with older people often disapproving of cohabitation and younger people overwhelmingly in favour of it (OPCS 1997: 45–6). Margaret Mooney Marini (1992) suggests that we need to distinguish between internalized values and normative beliefs, which are relatively unchanging, and preferences, which are more susceptible to revisions. This requires a deeper understanding of what we refer to as values,

96 Kay Peggs and Richard Lampard goals, preferences, wants and desires. She argues that an individual’s norms and values reflect the social norms and values of a society. So human action (in this case choice) is attributable, at least in part, to internalized norms and values. Thus we might choose a partner who has the physical capital (Bourdieu 1984) that we desire, based on the ‘impossible standard’ set by media images (Mirzoeff 1998: 183). In addition, social norms and values influence choices because they are internalized by significant others and thus affect the actor’s perception of others’ expectations (Mooney Marini 1992: 41). So decisions are culturally embedded, and actors share collective understandings (Zey 1992a: 79). The influence of changing societal norms and values on the nature of individual choice has been explored in work on ‘risk society’ (Beck 1992; Giddens 1991). According to Giddens (1991), transformations in traditional habits and customs have resulted in life in late modernity being characterized by uncertainty. Since habits and customs have broken down, we are confronted, on a daily basis, with ‘panoramas of choice’ (Giddens 1991: 139). For Giddens, the key to making choices is ‘reflexivity’, which involves the continual weighing up of different positions in the light of new information. Thus individuals are seen as calculative, with choices being based on rational intentions. Pierre Bourdieu’s (1984) concept of the ‘habitus’ provides a useful critique of this position since it points to the pre-conscious, non-reflexive ways in which people act. Chris Shilling (1993) describes the habitus as an internalized set of predispositions to act in certain ways, determined by various factors constituting an individual’s social location. For Bourdieu, individuals acquire their habitus as part of their personal development within a social field ‘[that] includes lifestyle, education and politics’ (Jenkins 1992: 85). Thus, the notion of reflexive decision-making can be qualified by an understanding of habitual action (Peggs 2000). In relation to couple relationships, internalized dispositions relating to ideas about ‘normal’ and ‘desirable’ relationships are likely to affect decision-making. One of our major concerns is the nature and significance of emotionally based judgements, and collective understandings of the role of emotion in couple relationships. Emotional satisfaction is important in couple formation, but this was not always the case; collective understandings of the basis of marriage have changed over time. In pre-modern Europe, marriages were often contracted and based on economic circumstances (Giddens 1991, 1992); thus these contracts could be founded on rational choice decisions in their pure utilitarian form. However, Giddens argues that modern marriage (and we can include here other forms of couple relationships) has become more associated with mutual sexual attraction (1992: 38) and emotional satisfaction (1991: 88). Norbert Elias (1978a, 1978b [1939]) argues that emotions have been excluded because rational choice theorists see them as offering nothing more than irrational conduct (e.g. Elster 1986a). In his ‘Homage to Norbert Elias’, Thomas Scheff (1992) argues that much of social science discounts the

(Ir)rational choice 97 possibility of a causal role for emotions in human conduct. Scheff (1992: 101–2) argues that rational choice overemphasizes human self-control and awareness and overlooks behaviour relating to lack of awareness and the loss of control. He maintains that many of the most important decisions an individual makes, for example choosing a spouse, are made on impulse, taking into consideration few if any of the possible options, and considering few or none of the consequences. Thus we need to consider the possibility that emotions may function as causal agents in the decision-making process. Simon Williams (1998, and this volume) argues that reason and emotion are not fundamentally opposed. He suggests that ‘rationality itself is a “passionately” held belief or cherished ideal: one which is, in large part, “irrational” or “unreasonable”’ (Williams 1998: 748). So the adherence to the idea that we behave rationally is itself based on an ‘irrational’ value. He argues that decision-making becomes difficult, if not impossible, without emotions, because emotions provide the basis for driving us in one direction or another. Thus emotions are essential to the ‘effective deployment’ of reason (Williams 1998: 748). So when leaving one couple relationship for another, our choice might be based on falling out of love with one person and falling in love with another. Although Williams (1998: 748) asserts that emotions can wreak havoc with rational decision-making (our intense feelings of love might lead us into an unsuitable relationship), the absence of emotion can be equally devastating. For Williams (1998: 757), emotions are central to reason, even when ideologically denounced as its antithesis. So emotions are fundamental to reason and are central to the decision-making process. However, decisions about couple relationships also depend on social interactions and the decisions of others. Rational choice theorists maintain that decisions are made at the micro-individual level. However, decisions within relationships are often made at a collective level. Decisions are most frequently made by couples or groups within the context of larger social collectives (Zey 1992b: 22). So decisions about whether to divorce can be made by a couple rather than an individual within a couple. However, this is not to suggest that there is an equal distribution of power within collectivities. Political variables relating to power and domination (Zafirovski 1999a) are, for example, gendered. In this respect, patriarchal relations ensure that some women remain in relationships involving domestic violence because they have few options for leaving. So inequalities in power are more complex than the rational choice notion of the gap between an individual’s benefits and the benefits of another (Munch 1992: 138). Relationships are always outside the narrower conditions of rationality, since choices and consequent costs and rewards depend on decisions of two or more individuals (Willer 1992). For co-resident couples, the longer they have lived together, the more they will be cut off from alternatives objectively and/or subjectively (Munch 1992: 142). So the decision to stay together might be costly to both of them. In addition we need to consider the role of altruism.

98 Kay Peggs and Richard Lampard Mary Zey (1992b) asserts that rational choice theorists fail to acknowledge that choices might be based on the welfare of those for whom we care, for example children, as well as on our own utility. This failure is perpetuated in the work of Becker, who reduces altruism to the dependency of one person’s utility on another individual’s utility (Cigno 1991: 39). Thinking about couples with dependent children, the decisions that they make about whether to stay together have implications for themselves and for their children. Concerns for the welfare of their children might be central to their decisions. Thus it is problematic to view concern for others in terms of greed or self-interest (Zey 1992b: 20). In summary, we have shown that rational choice theory is problematic when applied to relationship decisions. The main problem with rational choice theory in this context is its unidimensional approach focusing on utility. Later in this chapter we explore the suitability of Zafirovski’s (1999a) multidimensional conceptualization of rational choice as a tool for the analysis of the formation of couple relationships and relationship breakdown. We also explore how these utility- and emotion-related dimensions apply to the lives and behaviour of formerly married or cohabiting individuals. Finally we explore how the concept of the habitus adds to our understanding of how individuals make decisions about couple relationships. The next section outlines the data used.

Data and methods The interviews drawn upon below were carried out in 1996 as part of an ESRC-funded2 multimethod study of the remarriage process. The research was designed to examine structural features of the lives and behaviour of the formerly married in Britain, using survey data from sources such as the General Household Survey (Lampard and Peggs 1999), and to provide insights into the attitudes of formerly married (or cohabiting) women and men to future couple relationships. A focus on both marriage and marriage-like relationships was necessary since marriage patterns have changed and an increasing proportion of people are living as married. This chapter is based on the qualitative aspect of the study, which aimed to identify perspectives on relationship histories and attitudes to future remarriage/cohabitation. Although the sample is not intended to be strictly representative, it provides valuable insights into the subjective understandings of a diverse range of interviewees. For this part of the study we interviewed 81 people: 50 women (62 per cent) and 31 men (38 per cent) aged between 20 and 71. The age-range was wide since it facilitated an examination of variations across the life-course (see section (a) of Table 6.1). However the range of interviewees was limited in terms of ethnicity; only one was Black and a few belonged to minority White groups (e.g. were Irish or Polish). The marital status of the interviewees is presented in section (b) of Table 6.1, and section (c) shows the time elapsed since the end of their last co-resident couple relationship.

(Ir)rational choice 99 The interviews were conducted in and around a major city in the Midlands of England. Access to a sample was relatively unproblematic and relied mainly on adverts and publicity in the local press and around the city. We also used word of mouth and a snowball technique for gathering additional interviewees. Interviewees completed short questionnaires that asked for background personal information and included some questions on attitudes. The main qualitative interviews were focused but open-ended, which allowed the interviewees to expand more fully on their attitudes. The interviews were tape-recorded and conducted over an eight-month period. They ranged in length from 45 minutes to about five hours. To preserve anonymity, pseudonyms are used below.

Table 6.1 Characteristics of the interviewees Women



7(14%) 12(24%) 16(32%) 15(30%) 0 0

4(13%) 7(23%) 9(29%) 8(26%) 2(6%) 1(3%)

11(14%) 19(24%) 25(31%) 23(28%) 2(2%) 1(1%)

12(24%) 26(52%) 8(16%) 4(8%)

4(13%) 18(58%) 7(23%) 2(6%)

16(20%) 44(54%) 15(19%) 6(7%)

7(23%) 5(16%) 3(10%) 5(16%) 9(29%) 2(6%)

19(24%) 8(10%) 6(7%) 21(26%) 21(26%) 6(7%)

(a) Age (in years) 20 −9 30 −9 40 −9 50 −9 60 −9 70 −9 (b) Marital status Used to cohabit Divorced Separated Widowed

(c) Time since end of last co-resident couple relationship Less than 1 year 1 but less than 2 years 2 but less than 3 years 3 but less than 5 years 5 but less than 10 years 10 years or more Total (N)

12(24%) 3(6%) 3(6%) 16(32%) 12(24%) 4(8%) 50



100 Kay Peggs and Richard Lampard

Results The data below facilitate a degree of understanding about the factors influencing relationship choices. The data presented indicate that choices are made in terms of utility (Coleman 1990) but that factors previously considered ‘irrational’ have an impact on the decisions. This conforms more readily to the multidimensional approach put forward by Zafirovski (1999a). The data will be discussed in terms of the dimensions (social-psychological, cultural, economic and political) outlined by Zafirovski (1999a), which will give a picture of the complexity of the decisions that people make about couple relationships. The boundaries between the dimensions are fluid. The theoretical discussion has presented them as ideal types; however, in people’s lives the dimensions overlap.

Social-psychological variables – emotions and decisions As noted by Giddens (1991, 1992), contemporary marital ‘contracts’ are based on emotional satisfaction and sexual attraction. This was emphasized by our interviewees, who generally stated that love had (or should have) been the reason for entering a couple relationship. In addition, the ‘falling out of love’ of one partner or the other was often given as the reason for relationship breakdown. Stephen:

... the marriage turned into something that was just, I guess the word would be ‘comfortable’. In other words there were not disruptions. We were comfortably off, not rich, but the mortgage was payable, in a job, comfortable, that’s all ... all the sparks had died.


I went travelling for about five and a half months and when I got back she actually finished with me and she basically said that she’d fallen out of love – she no longer loved me – so I don’t understand it really.

However, the interviewees emphasized that emotions can play havoc with the decisionmaking process (Etzioni 1992; Williams 1998). Carol:

I just feel that if somebody else was around on an emotional level it would confuse me and I’m not ready for it.


I don’t think I would be awfully great for kids now. I certainly would have to really sit and think about it. But once again it’s the power of love. It’s the power. I just don’t believe that you’re in control.

(Ir)rational choice 101 Emma:

Yes, I’ve got love. I need love and I’ve got love, but nothing is enough any more because all I do is compare him and nothing lives up to him [ex-partner] and that’s the bottom line.

Decisions to marry or cohabit are based on emotional involvement. Couple relationships based purely on economic or other non-emotional considerations were viewed as ‘abnormal’ since they counteract the ideals of romantic love evident in Western societies. For example, Mary perceived her ‘instrumental’ reasons for marrying to be unusual. I obviously didn’t love him – I knew I didn’t love him but I knew I had to get out of London and I had to get away from this chap ... I think you wouldn’t meet many people who’d got married for the reasons that they were trying to get away from somebody else.

Social-psychological considerations – altruism The role of altruism in decision-making has been much discussed by critics of rational choice theory. For example Zey (1992b) argues that altruism is present in many decisions, even though such decisions might benefit the self. A number of the interviewees suggested that altruism had had some part in their decisions. To illustrate, in response to a question about his reasons for marrying, Phil replied, None other than to make her happy. However, later in the interview he stated, I wish now that I’d actually monitored my feelings closer. It would have been a difficult time to say to somebody, ‘Look, I think the world of you, but it’s not enough feeling to get married.’ I was probably afraid of upsetting people. The extracts above indicate the important role that social-psychological dimensions have in decisions about couple relationships. However, the connection between emotions and relationship decisions is complex. Although love can damage the decision-making process, it is seen as a crucial component, since romantic love is central to Western ideals about couple relationships. In addition, concerns about the welfare of others can mean that decisions about changing or breaking up a relationship are difficult to take since they might seem harsh and too selfish.

102 Kay Peggs and Richard Lampard

Cultural norms and values The role of cultural norms and values in the decision-making process was referred to by a number of the interviewees. In reply to the question ‘So are you looking for a serious relationship?’ Paul responded, Paul:

Not really, no. It’s not something that I feel I really really need. If it hap-

Interviewer: Paul:

pens it happens, if it doesn’t it doesn’t. That sounds callous doesn’t it? When you say callous what do you mean? Callous in what way? Well it’s just that maybe it’s not normal – maybe I should be looking for a

Interviewer: Paul:

relationship. So have you got a perception of what’s normal? No, not really. I was just thinking about Mr Average – Mr or Ms Average.

Mooney Marini (1992: 41) indicates that it is not only because social norms and values are internalized that they influence choices, but also because social norms and values are internalized by ‘significant others’ and thereby affect the actor’s perception of others’ expectations, Margaret’s response illustrates this: ... I thought no, no, that is not the time, I would not want to go with Pat and I think you tend to go to your reactions of what other people [think] about [you for] what you feel about yourself. The pressures of conforming to social norms are highlighted in Carol’s response: I know you’re supposed to have sex in your life to have fulfilled like, but that can’t be true really because so many people seem to be fulfilled without it. Norms and values are, of course, socially and historically specific and change over time. The norms and values relating to marriage and cohabitation have changed (Clulow 1995), but considerations of the ‘done thing’ are still important in many decisions. For example Barbara says, in response to the question ‘So why did you marry?’: Because it was the done thing and he did ask me. It wasn’t a case of will you live with me? It’s a case of will you marry me? You just didn’t do that sort of thing. That didn’t come until twenty years later I would say. The gendered nature of social norms and values is illustrated in Davina’s discussion of why she married.

(Ir)rational choice 103 We all think we need somebody else to make us whole, that’s what we’re taught. You’re not a whole person unless you’ve got a husband and a home. People’s notions of ‘the done thing’ operate at the pre-conscious level of the habitus, pointing to non-reflexive choices. Thus Barbara and Davina direct us to the problems associated with the idea of the reflexive calculating actor making relationship decisions based on alternative outcomes. Although some form of calculation might be used, the habitus often guides individuals into socially constructed ‘acceptable’ and ‘normal’ relationships. Such ideologies do not apply in the same way to all since they are cut across by social divisions such as class, ethnicity and gender. In this case Davina points to internalized notions of women’s primary role as socially constructed around husband, home and family. Some individuals feel pressure to conform to explicitly religious norms, but strong norms relating to couple relationships extend well beyond formal religious beliefs: Fiona:

What I am saying is that we did feel that it was important to have some kind of ritual celebration as a public event, not in church, but in a way that we could hold it and that’s not an entirely secular thing, this religious notion that, a very strong thing ...

In her chapter in this volume, Margaret Archer talks about how people are differently situated and therefore have differential situational constraints as well as differential opportunity costs. Nicola highlighted the constraints in her life: I left college and I’d got two options. I could either get married or go back home to live with my mother. Janis and Mann (1977, cited in Etzioni 1992) show that all significant decision-making evokes anxiety – i.e. is an emotional strain – hence decisions often fall within one of four defective patterns: inertia (sticking to the same course), an unexamined shift to a new course, defensive avoidance and hyper-vigilance. For example, Susie:

I was determined to make a go of it and I thought we were really fine and the awful crashing reality was actually just before the wedding day or actually on the wedding day, but I was too frightened to back out. I could have backed out then, but I was too frightened about that because everybody was there and that was just my own personal feeling about him because I’d begun to think that maybe this wasn’t right, but I still was determined that I was going to be okay.

104 Kay Peggs and Richard Lampard Although some form of calculation might be used, the habitus often guides individuals into strategies of avoidance (Aldridge 1998). Thus people often stick to the decision they have made in order not to appear inconsistent (Steiner 1980). Cultural norms and values play an important role in decisions about couple relationships. These often operate at the pre-conscious level of the habitus which conflicts with the idea of the conscious rational actor making calculative decisions on the basis of self-interest.

Economic considerations Utilitarian considerations are likely to be present in the formation of most couple relationships and in many relationship breakdowns. Mary:

It cost me 7,000 quid to divorce – me – 7,000 quid, and you know if I’d have known that I’d have probably thought twice, but to me now it’s the best money I’ve ever spent.

Having caring responsibilities adds further to such considerations Catherine:

I haven’t got ambitions for marriage [as opposed to staying single] because it can be just like a sentence, but if I had a child and maybe then it would be different, but I don’t know, I can’t say.

Having children puts financial burdens on individuals. In this respect marriage (or cohabitation) might help alleviate these costs. However, given a degree of financial security, such motivations are very much secondary, as the following interviewee illustrates: Gillian:

I have responsibility for [my daughter] and ... sole responsibility for [my daughter] now because I mean I don’t get any financial help from him any more, and as I say, if her own father can’t keep his commitment going, you know, how can I expect anyone else to take responsibility for her [and later in the interview] financially I don’t need anyone else. I’m self-sufficient. I’m working, touch wood and all the rest. I’ve got my roof, I’ve got my car, I can pay my bills, I can afford to go out once in a while. I don’t need, materially, I don’t need anybody else. Emotionally, I may need someone ... emotionally I want someone ... [and a little later] ... I’m not going to be used and abused again and if that’s all there is out there then I’ll be ...

(Ir)rational choice 105 In fact, financial factors can instead act as a barrier or disincentive to repartnering: Samantha:

I suppose financially if I had more money I would probably go out and you know make an effort to meet more people, because I don’t think that this sort of distrust is ever going to go ... unless I, unless I sort of meet people who, who I do learn to trust.


I’d be cautious and I’d want to be sure that he’s bringing to the relationship an equal [in economic terms] of what I’m bringing.

Thus we can see that economic factors are influential in couple relationships and relationship breakdowns, though the connection is complex. Formation of couple relationships can be economically costly, as can relationship breakdowns. For some, caring for children can make a couple relationship seem desirable; for others, couple relationships are seen as damaging to the interests of children. The need for emotional security can be in conflict with the need to be self-sufficient. In addition, the need for financial security in terms of material wealth can be at loggerheads with the wish for financial independence based on control over more limited financial resources. Thus decisions are multi-faceted and the notion of economic self-interest is extremely complex.

Political variables: power and domination A considerable number of the women interviewed had experienced couple relationships in which they had been dominated by men but which they did not feel that they had the power to change or end, as illustrated by the following: Rosa:

... it [was] such an oppressive marriage, it was pretty ... everything was pretty controlled for me [and a little later in the interview] ... if I had thought well it’s okay now to be divorced, society accepts it and ... I wouldn’t have gone through what I went through for the last four years or even the last two years when things had really escalated when we got to the


knife stage. ... as the children grew up he was violent to the children and wouldn’t let them be who they wanted to be and that’s when you realize well that’s what he did to me. He controlled me. [And later in the interview] The latter years I think I was pretty unhappy because of everything that was put on to me by him, but you become so downtrodden that you can’t say hey, I don’t want this any more.

106 Kay Peggs and Richard Lampard Gillian:

... but you see I didn’t feel that I was in a situation to be able to say I’ve had enough of this and walk away. I couldn’t. I’d got [my daughter] and everything, but I wasn’t happy for years.

The formerly married women frequently mentioned that control of their own lives was important to them, and this sometimes made them hesitant to repartner. Claire:

... and I’m doing things that only I can control, which is only with myself and my son. So I suppose it’s a way of taking control of my life again and not putting it into somebody else’s hands which I felt I have done and it was a total disaster.


... and I like the amount of control we have to either do things or not do things and the lack of interference from somebody else, their emotions and their behaviours ... there are a lot of problems with masculine roles and behaviours.


... he was taking control and entering my life in such a way and he wanted me to give up work and everything to be at home.


... in the longer term can you see yourself looking for another relationship?


Possibly, I mean, but it would be on my terms ...

Fear and feelings of powerlessness can be seen to have shaped the past decision-making of formerly married women; a sense of control is a preference that shapes their current choices.

Choices and changes over time Rational choice theory often fails to have a thorough understanding of changes over time, not only in relation to the way in which preferences might change temporally, but also because the immediate effects of a decision might be ‘irrational’ but the decision could seem ‘rational’ in the longer term. An explicit understanding of changes over time generates a basis for a more adequate social theory (Adam 1990, 1995), and a number of the interviewees reflected on their concerns about the possible effects of time-related shifts in attitudes on decisionmaking: Mary:

I don’t think I’d ever marry anybody. I don’t think I would but I could wake up one morning just shocking myself. I mean I’m aware I change my mind and I think it does depend on the person.

(Ir)rational choice 107 Worries about ‘changing my mind’ affect current decisions. For example, Phil said, I was so afraid of committing myself and then letting her down. I can take my own upset. If I’m upset, if I had gone with her, we’d have got all involved financially and involved emotionally, which we were obviously emotionally and then after a year say ‘Oh no, I’m sorry, this isn’t working.’ Giddens (1992: 57) notes that ‘Romantic love ... is a gamble against the future.’ A major time-related consideration is, therefore, risk and the risks involved in making wrong decisions. Risk has become a key concept in modern Western societies (Lupton 1999). Giddens (1991: 123) suggests that ‘thinking in terms of risk and risk assessment is a more or less ever-present exercise’. Ray highlights the risk he feels he could be taking: I’ve had a couple of meaningless affairs which have done me the world of good and changed my attitude but there again could I cope with a serious relationship? I don’t know. I’m very, very wary. Sue Scott et al. (1998: 700) show that the negotiation of risk entails both strategies for managing actual risks and dangers and strategies for the ‘rationalisation of fear and anxiety’. Some interviewees referred to strategies used to manage needs and risk where needs can be satisfied in a less risky way. For example: Carol:

I don’t want to lose his friendship by having a relationship with him. I find it very easy to get on with men [if I think] there’s no risk or anything else.

Theories about risk range from those which suggest that risks are real, to those which focus on risks as socially constructed (for more discussion, see Fox 1998). What is important about the extracts above is the interviewees’ perceptions of risks and how these changed over time. In relation to concerns about children, Scott et al. (1998) show that risk anxiety is concerned with keeping children safe. With respect to couple relationships, risk anxiety can be about keeping children as well as oneself safe. Thus, perceptions of interpersonal risks (Lupton 1999: 14) are important in decisions about couple relationships. Since all our interviewees had been through relationship breakdown or closure (in the case of widowhood), for many the anxieties associated with new relationships are likely to be intensified.

108 Kay Peggs and Richard Lampard

Discussion and conclusions The women and men interviewed in our study present a complex picture of their relationship decisions. Such complexity breaks the fetters of the orthodox utilitarian approach. Although many referred to their weighing up of economic considerations in decisions about joining or leaving couple relationships (which were further complicated by the presence of children), such considerations were by no means the only, or the most important, ones. Especially important in relationship choices is the presence (or absence) of romantic love and sexual attraction. Interviewees referred to falling out of love as a major reason for relationship breakdown. This was often instigated by one partner feeling that they had found love elsewhere. So emotions were fundamental to the choices that they made, and formed the basis on which they chose between one person and another. Although emotions can devastate rational choices, we maintain that in the context of couple relationships emotions are often fundamental to rational choices. Societal norms and values meant that many of these women and men had married because it was considered to be the ‘right thing to do’. Norms and values about the appropriateness of cohabitation, marriage and divorce change over time and consequently interviewees presented their decisions as being time-specific. Where they would once have married, they would now prefer to cohabit. The presence of such values contradicts the orthodox model of rational choice, where values are inhibitors of rational utility. Social acceptability and pressure from others, as well as internalized values, whether religious in origin or more secular, affect our decisions about whether and with whom to enter a couple relationship. In addition, these often operate at the pre-conscious level of the habitus, which conflicts with the notion of the conscious rational actor making calculative decisions on the basis of selfinterest. Power and domination also featured in the relationship decisions that these women and men made. Feeling unable to leave an abusive relationship was an important reason for ‘choosing’ to stay, and a desire to be in control of one’s own life was a major reason for choosing not to repartner. In addition, risk assessment was fundamental to the relationship decisions of many of the interviewees since they were concerned about making decisions that they might come to regret. The experience of relationship breakdown intensifies anxieties about future relationships. Rational choice theory often fails to provide a thorough understanding of such changes over time. For a discussion of similar issues see the chapter by Procter later in this volume. Economic, social-psychological, cultural and political considerations are apparent in couple relationship decisions. Our chapter has therefore shown that the multidimensional approach to rational choice, as outlined by Zafirovski (1999a), is more suitable for the study of such decisions than the unidimensional approach advocated by orthodox rational choice

(Ir)rational choice 109 theorists. The decisions that human beings make are complex and are made in complex situations. They are made at a micro- (individual) level by social actors and are informed by the macro- (structural) level, via cultural norms, and so on. The factors affecting decisions change over time and are likely to differ nationally and culturally. A theory of choice which does not take account of such complexities and which cannot stand up to comparisons with the everyday circumstances of people’s lives is of little use in the study of the relationship decisions of the formerly married or cohabiting. While we endorse the broad, multidimensional thrust of Zafirovski’s approach to rational choice, we share some of the reservations expressed by Margaret Archer and Jonathan Tritter within the Introduction to this volume. In particular, we feel that the equation for a multivariate model of rational choice provided by Zafirovski (1999a: 59) places some unnecessary limitations on his general approach. The equation in question highlights a key feature of Zafirovski’s argument, inasmuch as it illustrates how rational social action depends upon both utilitarian variables and non-utilitarian, non-economic variables. However, the linear, additive structure of the equation appears to constrain his multidimensional approach to something more limited and less attractive than the form of multidimensional approach that we favour. More specifically, the additive nature of the model outlined by Zafirovski’s equation gives the impression that the impacts of economic, social-psychological, cultural and political factors are separate and unrelated to each other. This would seem to underestimate the extent to which the various factors are part of a complex whole, and does not allow in a satisfactory way for the interplay between structure and agency within decision-making processes. Zafirovski (1999a: 60) appears to acknowledge this limitation of the formulation of his model, and specifying interaction effects within such a model, or changing it from an additive to a multiplicative model, would also address some aspects of the model’s limitations. However, these revisions would still result in a model which treated the various explanatory factors in a uniform fashion. In general, it does not necessarily seem appropriate to view nonutilitarian factors as operating in a comparable way to utilitarian factors; for example, some cultural factors are perhaps best seen as bounding the range of possible actions rather than as affecting the choice between various possible actions. In this chapter we have shown that decisions about couple relationships depend on a diverse range of influences which are irreducible to the individual level, including economic considerations, the habitus, and fear or a sense of powerlessness. While a multidimensional approach to analysing such decision-making is a necessity, such an approach needs to be operationalized in a way which is sophisticated enough to accommodate the complexity and heterogeneity of the influences involved and to address the interplay between structure and agency in real rather than additive terms.

110 Kay Peggs and Richard Lampard

Notes 1 2

The term (ir)rational is taken from Simon Williams (1998). The interviews were carried out as part of a project focusing on the formerly married, The Remarriage Process: Structural and individual perspectives. The authors wish to express their gratitude for the funding received from the Economic and Social Research Council (Research Grant No. R000236057).


Switching allegiances Decisions by schools to ‘opt out’ to self-management Jonathan Q. Tritter

This chapter uses recent educational reforms promoting self-management in educational institutions to explore how schools make decisions to change their organizational and managerial status. Focusing, in particular, on the history of a recent British educational policy, grant-maintained status (GMS), I explore the factors involved in the decision by a school to change status. Classic rational choice theorists would argue that such decisions are a product of the utility-maximizing strategies of the individuals involved, while at least one recent reformulation by Zafirovski (1999a: 48) suggests that this can also be applied to collectivities: [R]ational choice theory postulates universal maximizing behavior – that is, not just economic agents such as consumers and firms, but also other actors like families, social movements, political parties and officials, governments, racial and ethnic groups, churches or scientists are assumed to optimize their utility functions. The shift to GMS brought with it a range of incentives, including additional resources, managerial autonomy and increased power for school managers. Despite these clear utilitarian benefits, the vast majority of schools did not change their status. It is this conundrum which, in the light of rational choice theory, is discussed in this chapter. The rational actor faced with the ‘opportunities’ provided by GMS had to decide what they should maximize. The history of GMS should demonstrate (i) what preferences ‘schools’ pursued which meant that the majority chose not to take up the incentives associated with a change of status, and (ii) for the minority of ‘schools’ which did opt out, to what extent it was the pursuit of the consequential advantages which accounted for their preferences. The answers to both questions present challenges to rational choice theory and especially as to whether it can give satisfactory responses which are framed in the individualistic terms which it advocates.

112 Jonathan Q. Tritter

Grant-maintained schools and the meaning of selfmanagement in Britain: what is ‘opting out’? The emphasis on the link between a nation’s economic performance and its school system within a global economy continues to make education a constant focus of political debate and reform. Faced both with a general public increasingly unwilling to pay higher taxes and with growing individualism, neoliberal approaches to public sector management and financing have been more readily incorporated into public policies. The growth of regulated competition is typified by the creation of quasi-markets and a reliance on contractual relationships within the public sphere built on increased monitoring of outcome measures. Internationally these have served as powerful incentives to promote certain types of school reforms which can be seen in countries which, in educational terms, are as diverse as New Zealand and Canada (Dale 1997; Kenway 1993; Waslander and Thrupp 1995). Self-managed schools have a long history in education and have received much attention on both sides of the Atlantic (see Finkelstein and Tritter 1999). In the United States, ‘charter schools’ are a classic example of this type of institution though there is a great deal of variation in the way they have been organized and funded (Clune and Witte 1990; Mohrman and Wohlstetter 1994; Murphy and Beck 1995). The creation of a new category of school, the ‘grant-maintained school’, by the 1988 Education Reform Act provided the clearest British example of self-managed schools. This category of schools received their funding directly from central government, bypassing all intervening stages of government and, in particular, the local education authority (LEA), the equivalent of the school board in the United States. The level of funding was based on the average expenditure per pupil within the LEA of which the school was no longer a part. There was additional funding, to take account of those central facilities provided by LEAs for local schools, and initially this resulted in an extra 16 per cent over what other schools in the same locality were given. There was also a better chance to bid for capital funding as well as special transition grants available to ease the shift from a ‘regular’ maintained school to a grant-maintained school. In order for a school to become grant-maintained, the board of governors1 must decide to hold a secret ballot of parents of children enrolled in the school where a simple majority of votes would allow the school to apply for the change of status. The actual change had to be agreed by the Secretary for State for Education. By 1996, 1032 schools had become grantmaintained (private communication, Funding Agency for Schools 14 July 1997), and there were few additions leading up to the general election in May 1997. By 1999, following the establishment of a Labour government, GMS was abolished and as a result previously grantmaintained schools were reintegrated into LEAs. This testifies to the extremely politicized nature of GMS, which was often front-page news, and hints at the continued debate and changing fortunes of the policy throughout its lifetime.

Switching allegiances 113 The 1988 Education Reform Act, together with the other policy changes instituted by successive Conservative governments, had, Ball (1990: 68) argues, a simple intention that ‘Schools are to become businesses, run and managed like businesses with a primary focus on the profit and loss account.’ The justification for such reforms rests on arguments of educational and fiscal flexibility linked to responsiveness and efficiency. The Conservative government’s declared interest (Carlton Club Political Committee 1991) in introducing the policy of GMS had been to empower parents to become more involved in their children’s schools and to extend parental choice in the education ‘market’. Strengthening the managerial autonomy of individual schools provided greater responsiveness to local needs. By creating a new category of schools, a greater diversity of institutions, increased marketization and consumer choice became possible, and this was underlined in the mechanism devised to decide whether a school should opt out of the LEA and adopt GMS. As the Department for Education and Employment (DFEE 1995: 1) suggested, Self-government brings clear practical benefits to schools of all kinds and the communities they serve.... It is also clear that, by giving schools full control of their own affairs, self-government helps to increase the motivation and commitment of governors, staff and parents.

Studying the decision-making process: the research project and the data This chapter is based on the analysis of data generated during an eighteen-month study funded by the Economic and Social Research Council,2 begun in Autumn 1995. We were concerned to understand how the process of ‘going grant-maintained’ required schools, and especially school managers, to manage change within the ‘loose-linked’ structure of a school (Bidwell and Rodriguez 1991). The data are based on case-studies of thirteen schools in four distinct LEAs. They are supplemented by a questionnaire-based survey of all grant-maintained schools with affiliation to either the Church of England or the Roman Catholic Church. In each case-study school, data were collected by interviews with the headteacher, various members of the board of governors, a focus group with parents, informal observation and a review of relevant documents. In order to contextualize the decisions of individual ‘schools’, we also conducted interviews with key stakeholders in the local area, including the Director of the LEA, and at a national level with the Director of the Funding Agency for Schools,3 as well as a range of different politicians and special interest groups with public views on GMS. Overall our data are based on 66 questionnaires, 99 interviews and 10 focus groups.

114 Jonathan Q. Tritter Our research approach allowed us to understand the different roles of the parties acting within the ‘contested terrain’ (Edwards 1979) of the school and how they viewed the grantmaintained issue. Thus the different perspectives of the participants and their understanding of the implications of GMS had an impact on the part they played in the decision of the school to opt out or stay put. Yet, as Jenkins (1992: 85) suggests, ‘The existence of a field presupposes and, in its functioning, creates a belief on the part of participants in the legitimacy and value of the capital which is at stake in the field.’ The continuing politicization and tension surrounding GMS testifies to the flux in this ‘field’.

Making the decision to opt out: so why didn’t all schools go grant-maintained? A shift to GMS was based on a parental ballot formally to seek a change of status, called for on the basis of either a petition by parents of children enrolled in the school or by the board of governors. In our research, in every case-study school the parental ballot was instigated by the board of governors, a section of the board of governors, the headteacher or some combination of the three. So why does the board of governors seek a parental ballot? By law, on an annual basis, the board of every school was required to consider the issue of GMS. The composition of the board varies, but always includes the headteacher and a chairperson. In our research there were no instances where the decision of the board of governors to call for a parental ballot was at odds with the view of the headteacher. That is, the board usually looks for guidance in its decisions to the headteacher (Deem and Brehony 1995).

Why didn’t all parents (consumers) vote to make their schools go grant-maintained? Once a parental ballot was organized, a series of informational meetings was arranged at which parties for and against opting out spoke; and in all the cases in our research the headteacher also spoke. Once again, the parents tended to follow the lead of the individual they saw as most identified with the school: the headteacher. There was no case in our study where the headteacher had spoken in favour of GMS but where the parental ballot opposed opting out. As one teacher-governor recalled the parental ballot: It was more a case of parents that were very into the school anyway, that their thinking was to help the school, were swayed by certain like strong points of view. That’s the way it went. Mr [headteacher’s surname] didn’t exactly direct parents but he did put his point of view in and he did think it was to our advantage and he did say that. But he said it was obviously up to parents.

Switching allegiances 115 Thus, one of the major rationales of such policies, the promotion of parental involvement at a local level as consumers, was not borne out in reality as the decision to become a selfmanaged school rested almost entirely with the headteacher. Parents did not have independent preferences of their own. They did not seek ‘perfect information’ and instead willingly accepted the information provided by an authoritative figure. Thus parents were neither maximizing nor satisficing. The ‘rational choice’ of the consumer was dictated, almost entirely, by acceding to the wishes of the headteacher. This is more a case in which a decision to maximize an individual preference is equivalent to sheep following a shepherd. Even more contentiously for rational choice theory, this shows that without the ‘manipulation of bias’ (Lukes 1974), by which the powers of the head organized the majority of interested parties out of the decision-making arena, what was legally intended to be a collective choice could never have been analysed in individualistic terms (i.e. in terms of what the heads alone sought to do). Even then, it is still necessary to ask to what extent did headteachers make choices about their school going grant-maintained that can be understood as maximizing their own individual preferences?

Why did headteachers wish their schools to go grantmaintained? GMS changed both the source and level of funding and therefore the nature of the school’s relationship with the LEA and the managerial responsibility of the headteacher. As the Department for Education and Employment (DFEE 1995: 1) presented it, It [GMS] gives a school control of its total budget, including the portion previously retained by the local education authority [LEA] to pay for centrally provided services. Because GM schools are free to buy services from any LEA or other provider, they can get the best possible value for money and translate savings into extra teachers and books. As has already been suggested, grant-maintained schools also had access to a range of grants from the Department for Education and Employment. The difference in the nature and extent of the tasks for which the management of a grant-maintained school was responsible was also recognized. Thus, transitional funding was provided for newly ‘opted out’ schools to defray the costs of restructuring the administrative/managerial team. In our survey of grant-maintained schools, two of the three most important reasons for seeking GMS were in order to obtain ‘funding for capital projects’ and ‘funding for school maintenance’. The vast majority of our respondents (88 per cent), most of whom were headteachers, reported their success in obtaining further capital funding, with grants ranging from £26,000 to £890,000. When we asked schools about the three largest grants they had

116 Jonathan Q. Tritter received, for 10 per cent of our sample these totalled more than a million pounds; and one school had received £1.94 million. Similarly, it was not merely the increased amount of funding that was important, but also the power to prioritize spending based on the school’s own agenda, rather than one set by the LEA. Opportunities for financial savings and services more appropriate for the individual school were made possible when the institution became the purchaser of contracted services. Money, therefore, was not the only advantage of GMS; there was also the personal discretionary powers acquired over its use. For many headteachers the increase in their autonomy and independence from their LEA was just as valuable an incentive for pursuing GMS. As one deputy headteacher at a grant-maintained school put it, I personally found tremendous frustrations in working as part of a local authority and I was absolutely relieved to get rid of it, I have to say. At that time I was in charge of INSET4... and the INSET arrangements [for the LEA] were like a straitjacket. I was frustrated on a daily basis. I had to ring up the office and fill in a form in triplicate every time that I wanted to spend twenty-five pounds, shall we say, of a budget that we’d been given. ...The paperwork for INSET took hours and hours and days of my time and it was a complete waste of time in my opinion and I was so frustrated. For some headteachers, pursuing a change of status was based on a reaction to a perceived external threat. During the late 1980s and early 1990s much was made of the fall in the number of pupils entering the school system. This factor, together with the greater stress on fiscal efficiency stemming from the 1988 Education Reform Act, and particularly associated with the local management of schools, National Curriculum and GMS, led to the reorganization of educational provision in many LEAs. This reorganization involved both the closure of small and undersubscribed schools and the amalgamation of schools or, more often, of sixth forms into sixth form colleges. As a governor of a Church of England grant-maintained school observed: ‘The LEA wanted to get rid of a school, and they still do, in [area] and they picked on what was an easy option or what they thought was, and in fact it didn’t turn out to be like that.’ The school in question went grant-maintained, which still left the LEA with too many school places for the number of pupils in their area. While resources and autonomy were seen as key advantages of GMS, the opportunity to compete more effectively in the local education ‘market’ was also aided by a change of status. To some extent, being a grant-maintained school was seen as having a better ‘brand’ image. With greater autonomy, it can be argued, schools are more able to respond to the needs or desires of their local consumer market and shape their schools in a way that is better able to ‘attract’ customers, or at least the parents who are choosing on behalf of their offspring.

Switching allegiances 117 Competition in the educational marketplace is further affected by the annual publication of school ‘league tables’. These provide comparative ratings of schools on the basis of a fixed set of factors, including exam results, levels of absenteeism, staff–student ratio, as well as a token measure of deprivation: the proportion of pupils receiving free school meals. The opportunity to maximize the school’s results by adopting particular policies and therefore appearing higher up in the league tables is also one of the potential advantages of selfmanagement. Some schools were not in a position to achieve high levels of academic success. For them, the opportunity to attract greater numbers of consumers was provided by seeking a particular identity or niche market. One school in the study saw an opportunity to promote the institution as particularly adept at providing for pupils with special educational needs. As the chair of governors put it, ‘We are one of the few schools in the area that can offer this [accessibility] and we are hoping to become a completely barrier-free school.’ Other schools sought a niche through technology college status, or identification with the world of work by arranging linkages with a private company (Walford and Miller 1991). But these adaptations, which were, arguably, made possible by GMS, also changed the role of the headteacher. One headteacher of a large grant-maintained school suggested his role had changed because he was able to take a strategic rather than a directive role. As his deputy put it, The head is very much in control, but knows how to delegate ... doesn’t feel that he has got to make every decision himself or have a finger in every pie so to speak. ... He gives you an area of responsibility and lets you get on with it. He expects to be kept informed of course, and when it is an important decision he obviously takes part in that. The desire for personal autonomy and the opportunity to exercise a more entrepreneurial approach to educational management were clearly part of going grant-maintained, and becoming a grant-maintained school dramatically changed the role of the headteacher (Grace 1995). A grant-maintained headteacher was responsible for a far larger budget and oversight over a much greater sphere of operations than prior to opting out. While going grantmaintained the headteacher was charged with pupil, staff and curriculum management, afterwards he or she was responsible for maintaining the school’s infrastructure as well as services ranging from school meals to the staff payroll. As one headteacher said: ‘It [GMS] is a management tool and no more or less ... [and that] has changed fundamentally our flexibility to manage.’ Considering the change between managing a school prior to opting out and following it, ‘the difference is the difference between managing fully and managing with one hand tied behind your back.’ Concomitant with this expansion of responsibilities and managerial remit were greater power and status.

118 Jonathan Q. Tritter As a consequence of going grant-maintained, a number of headteachers became national figures debating and promoting educational policy in the media. But even among those headteachers who did not attain such a high profile, many were offered increased pay and pension compensation, and some even got ‘company’ cars (NAHT 1996; Power et al. 1996). They could (and did), with the agreement of governing bodies, appoint additional administrative and teaching staff. In many ways their role changed from a teacher who managed to managing directors. Clearly, a shift to GMS brought with it a variety of rewards in both financial and managerial terms. Further, the independence from what some headteachers regarded as ‘interference’ from the LEA was another attraction of ‘opting out’. I have outlined five distinct, but linked, preferences that could be maximized by a school ‘opting out’: increased resources, strategic resource management, autonomy from the LEA, improved market position and entrepreneurial leadership role. Yet the vast majority of schools did not go grant-maintained and the vast majority of headteachers did not seek a parental ballot despite the apparent incentives. So what were the preferences maximized by the majority of headteachers by not going grant-maintained?

Why didn’t all headteachers seek GMS for their schools? The factors involved in headteachers’ decisions to seek GMS were related to a range of factors that cannot be simplified to a series of best-interest equations. Rational choice theory would argue that the ‘opting out’ decision was based on maximizing the best interests of the headteacher by optimizing his or her own preferences. The next section defines the preferences that overrode the manifest advantages promoted by the government, as linked to GMS and outlined above. Four categories of factors are proposed: structural, ideological, external and personal. The three categories of factors that define the context within which the decision is made are structural, ideological and external while the personal factors are related to managerial style, role, power and status. These personal attributes are jointly linked in the form of identity, and what I propose is that a discussion of how and why schools sought GMS is definitively related to the identity of the headteacher.

Contextualizing the decision: structural, ideological and external factors I have already suggested that there are a range of benefits associated with GMS which serve as a set of structural factors that played a part in the decision by headteachers to seek a change of status for their school. Key among these were additional resources, greater discretionary

Switching allegiances 119 powers, autonomy from the LEA and a different leadership role as a headteacher. Together these were seen to provide a potential source of competitive advantage within the educational marketplace that was too good to pass up for the schools which went grant-maintained. As one headteacher of a grant-maintained school explained, [W]e just were not prepared to go down the pan. Because that’s what would have happened and still could happen. ... We know that we are competing against schools that are able to market themselves and a lot is to do with image and how you are perceived no matter what the reality of the school. These structural attributes of GMS, however, were viewed by many schools as being associated with another set of costs that outweighed their advantages. For the majority of headteachers, with autonomy from the LEA came distance and a severing of historical links and a set of relationships based on mutual trust and support. An LEA has a legal responsibility for the education of those children who reside in the authority and for maintaining a larger strategic vision of the educational needs of the local population. By seeking GMS, a school rejected the LEA’s oversight and removed itself from the authority-wide vision. Instead it promoted its own individual strategy for recruiting pupils, not only from the entire authority but also from neighbouring authorities, thus having an impact on a far larger population of schools than simply those with which it had been traditionally linked. There is evidence from our study that LEAs reacted to the threat of or actual opting out of schools in their area by becoming more responsive and consultative with the schools that remained under their auspices, and this often took the form of an increased level of devolved resource management. As one headteacher put it, The LEA has certainly worked very hard to retain its schools and done everything in its power to develop a relationship which is different to the earlier relationship, and I think that has made us reasonably happy to be with the LEA and we don’t feel threatened by other GMS schools. Such responsiveness is often used as a way of marketing the LEA and was presented to counter the arguments of those schools interested in opting out. This suggests that as the grantmaintained policy continued to evolve, the opportunity cost of a school opting out increased. Furthermore, it also suggests that where the primary preference that a headteacher wished to maximize was to limit LEA interference and promote his or her agenda, then in responsive LEAs this would lead to not seeking GMS. The impact of schools opting out of an LEA can serve to promote the formation of closer relationships between the headteachers of the schools that remain. Often headteachers would

120 Jonathan Q. Tritter form a group that met regularly and provided a forum for strategic discussion. Typically, the LEA played a key catalysing role in the formation of these groups which served to promote a better sense of common and collective identity, and, in some cases, it made a point of excluding grant-maintained school headteachers. They [the other secondary headteachers in the LEA] came up with this arrangement that they would discuss LEA business to which I was excluded and then invite me in, and very often the business they were discussing was finances. ... The finances of [area] are my finances and whatever [area] decides and whatever become your ... budget will become my budget, so I think I should be present at those discussions. They didn’t agree. Thus there were potential negative effects associated with becoming a grant-maintained headteacher. Many found themselves isolated from the community of schools and their ‘colleagues’ in the LEA. A minority found that LEA officers became actively hostile towards them following their school opting out. Clearly, such an anticipated cost weighed heavily against the decision by some headteachers to go grant-maintained. So the decision to opt out to seek individual gain for the school was tempered by loyalty to the collective. This is further complicated for ‘church schools’, which are affiliated to and managed by a religious organization such as the Church of England or the Roman Catholic Church. For these schools, the ‘advice’ of their bishop was often an additional limiting factor or cost. A key preference that needed to be maximized by these headteachers was the resonance between their actions and the wishes of the bishop and diocesan board of education. For example the preferential funding that GM schools receive, in the form of transitional grants and more favourable capital allocation, can be seen as being at the expense of neighbouring schools, and of other schools generally who share the national cake. (Diocese of Blackburn 1994: 1) The ideological and political views of the headteacher were also key factors in decisions about GMS. As an educational policy, GMS was firmly identified with a particular political party, the Conservatives, and seen as a right-wing, neoliberal ‘reform’ of state education. Many headteachers saw themselves as liberal and left-wing politically, and were unwilling to pursue a policy that they saw as promoting a clearly political agenda which would undermine a meritocratic shift towards comprehensive schooling that had been apparent in British education since the late 1960s. In the contrasting political views of headteachers we have a source of alternative preferences, but not ones which can be reduced to individual terms. A school choosing to opt out of LEA control and become grant-maintained made an individualistic choice, maximizing its resources and having a consequent negative impact on

Switching allegiances 121 the local and national collective of schools. This apparent ethical dimension is clearly linked to the historical relationship with other schools in the locality. As has already been suggested, relationships that had developed with feeder primary schools as well as secondary schools, which often worked together within an LEA, were altered and often severed if one party chose to opt out. Pursuing GMS may have provided a way of potentially maximizing the personal individual ambitions of the headteacher, but at the same time the desire to maximize collectivist and ideological goals required a decision against opting out. What has been suggested are a range of factors that relate to the decision of the headteacher to seek GMS. But it is clear that most schools did not choose self-management. In fact, examining the schools that did go grant-maintained provides insight into the different categories of structural reasons. External factors also impacted on decisions by headteachers to pursue a change of status. In part these relate to the evolution of the policy. That is, the timing of the decision to go grantmaintained had an impact on the level of resourcing and the likelihood of approval by the Minister for Education. As a politically motivated educational policy, there were more incentives provided early in the life of the policy and therefore a higher level of resourcing. Similarly, as the Conservative government was keen to promote grant-maintained schools as an attractive and successful option, there was less attention to the detail of the application and a greater willingness to approve any school seeking a change of status, even if some were not necessarily viable (interview, Funding Agency for Schools, 25 September 1996). Thus, the incentives to seek GMS shifted over the life of the policy, altering the opportunity and extent to which preferences could be maximized.

Making a rational choice to go grant-maintained This chapter has examined a recent British educational reform that promoted self-managed schools and established a mechanism for seeking GMS. Further, it has shown that the decision to become grant-maintained was, in practice, the sole responsibility of individual headteachers. This decision, it has been argued can be understood as an outcome of the weighing up of a range of different factors. This conception is clearly one that suits a rational choice model of decision-making. As Zafirovski (1999a: 62) suggests, Like the old utilitarianism, rational choice theory defines rational behavior as ipso facto rational choice in utilitarian terms, as economic rationality or instrumentally rational action. This is done by subsuming all the ends, reasons and motives of action to egoistic ones – that is, the pursuit of self-interest. However, I have demonstrated that only a minority of headteachers saw the five reasons for opting out as sufficient to define a preference. For the majority, the costs of going grant-

122 Jonathan Q. Tritter maintained far outweighed these alleged benefits, and they, in turn, had preferences based on other factors that were structural, ideological, external and personal. Arguments could be raised that, despite the fact that going grant-maintained was, at least in our research, basically the decision of one individual, what is being analysed is the decision of a corporate body, that is, the school. Yet once again this seems appropriate as rational choice in the economic-utilitarian sense seems limited to what Weber (1968: 69) called the domain of entrepreneurial or ‘managerial action’, and only secondarily to consumer behavior largely driven by economically non-rational considerations such as status. (Zafirovski 1999a: 70) In fact, given that going grant-maintained was a case of a single decision having long-term consequences for the nature, funding and management of a school, it seems likely that ‘rational choice models will be useful when the stakes are substantial and when the actions of the individual have a significant impact on the disposition of the stages’ (Green and Shapiro 1994: 28). Although the three categories of factors I have identified clearly incorporate classical utilitarian factors (such as level of resourcing), they are dominated by non-utilitarian factors (such as ideological stance, power and status). As Hollis (1987: 23) points out: ‘Bargain hunting’ may suggest a grasping, penny-pinching, short-sighted selfishness. That is not what the basic model asserts or implies. Adam may turn out to be generous, far-seeing and altruistic; but, even if he does, he will still reject even slightly inferior ways of exercising these qualities. Recently, Zafirovski (1999a: 58) has suggested ‘a broader multivariate model of rational choice ... this model incorporates both economic and non-economic, utilitarian and nonutilitarian, egoistic and altruistic, psychological and cultural, individual and social-structural variables’. While attractive in principle, Zafirovski seems to be merely introducing a vague range of all other possible social variables and leaving a relatively unspecified generic model. Would this relatively unspecified set of ‘social factors’ explain the limited take-up of GMS? I want, instead, to suggest a further set of factors associated with the identity of the individual that may, or may not, fit within the general expanded rational choice model that Zafirovski is promoting. These additional factors relate to the individual identity of the decision-maker and play a decisive role in determining which factors are understood to have a utility and therefore the potential to be maximized. Having now established the ways in which headteachers, as

Switching allegiances 123 agents, have different preferences and assessments of costs, how is it that they can pursue such different ends in the same basic situation? I want to suggest that there are four distinct ideal types of individual headteachers that emerged from the research, and in this way examine ‘preference formation’, an area in which rational choice theory is mute.

Headteacher personality: four categories and decisionmaking about GMS In terms of GMS, four distinct categories of headteacher emerge from the study which reflect different personalities and roles. The majority of the headteachers saw themselves as dedicated teacher managers. These headteachers were satisfied with the way their school was organized, and while some were afraid of change, others merely wished things to stay as they were. As collectivists, they valued the relationships they had developed with the other schools in the LEA and also the advice and support of LEA officers. For such headteachers, seemingly the vast majority, GMS, however attractive financially, was never a real option. A similar but distinct identity was exhibited by those headteachers who saw themselves as ‘hostages to fortune’ and who would have liked to stay within the LEA. These headteachers felt ‘forced’ into going grant-maintained as the sole strategy for keeping their schools the way they were. For them, opting out was merely an instrumental solution and the only way out of an impossible situation. A third category is that of ‘instrumental’ headteachers. For them, the traditional model of teacher manager is insufficient to secure a competitive advantage. They also value the work of the LEA but see GMS as an opportunity to ‘persuade the LEA to treat them differently’. When faced with a need for funding, they will use the threat of opting out to their advantage. As one headteacher put it: ‘Many a time I have said ... with somebody in the offices or the minutes from a governors meeting ... clearly if this is not approved by the LEA we will have to consider the viability of GMS.’ These headteachers would never go grant-maintained but see ‘sabre rattling’ as a useful strategy to achieve the best possible services for their school. For the instrumental headteacher it is rational not to go grant-maintained, that is, it is not their preference, as there is an alternative way of meeting their ends. The final category of headteacher is the only one which actively pursued GMS. These ‘entrepreneur’ headteachers sought a shift in their career and different challenges (Grace 1995). While they may have justified the change in status on the basis of unwanted interference from the LEA, the attraction was also about increased power, increased salary, higher profile, and a chance to be managing director. I certainly, you know, don’t doubt his integrity as acting in the best interests of the school and the pupils as he sees it, and I think that is tied in with his own personal ambition and

124 Jonathan Q. Tritter so on, and I don’t say that disrespectfully because I mean I wouldn’t want a headteacher of a school that wasn’t ambitious and I wouldn’t want a headteacher of a school not to do what he or she thought to be the best thing for the school. So I think those three things, you know, vision, entrepreneurial characteristics and, umm, personal power. This description of a high profile grant-maintained headteacher by his chair of governors clearly identifies the distinct characteristics of this category. To sum up, then, the decision to seek GMS is one that relates to a range of contextual factors but is mediated by individual characteristics and, in particular, the personality of the headteacher.

Conclusions This chapter has attempted to provide an explanation for why a minority of headteachers opted out of LEA control whilst the majority did not, based on the individual preferences of these headteachers and variability in how they evaluate costs. While this may appear to be an impeccable rational choice theory account, it raises the problem of preference formation, as I have documented ways in which different people seek to maximize different things. What I have presented is a challenge to traditional rational choice models and an attempt to clarify and specify the decision-making model outlined by Zafirovski (1999a). The decision-making underlying the policy of GMS was predicated on a demand-side approach, with little attention paid to supplyside issues. Clearly, such an approach misunderstood how the incentives that were provided related to the actual preferences of the majority of headteachers, governors and parents. Overall, the rational choice approach still left a number of discontinuities, two of which merit accentuation. Firstly, the research clearly indicates that the preferences of the vast majority of potential agents (parents and governors) were formed by the authority of the headteacher. It is only by introducing an explanation of the power of the headteacher to organize other agents out of the decision-making arena (Lukes 1974) that it is then possible to provide an individualistic account of decisions about GMS, based on the headteachers’ strategies alone. Even if the importance of collectivist variables like power is neglected and discounted, however, another difficulty then surfaces, for, secondly, unless we are willing to accept that different preferences in the same situation are merely a function of individual psychology, we are committed to saying we can provide no explanation for why different headteachers prefer different things, apart from the fact that they are different people. But individual psychology cannot provide an explanation on the grounds of political ideology for why some headteachers did not go grant-maintained. The existence of political parties, and therefore political ideologies, cannot be accounted for by methodological individualism. We need to

Switching allegiances 125 know why headteachers weighed the costs and benefits of going grant-maintained differently, unless we are willing to accept a simplistic account of psychological types. In short we need to know what weights and measures were of importance to headteachers and how they were formed. We need to know how headteachers’ preferences reflected their different evaluations of the implications of going grant-maintained and fitted their definition of what a school was and what a school should do. The first criticism, namely that without reference to power we cannot understand how heads turned a collective decision into a personal one, shows how agential doings cannot be understood without reference to structure. The second criticism points to the fact that unless individualism is to collapse into psychologism, then reference is needed to how structure enters into the formation of agency itself. In other words, the interplay between structure and agency has to be retained and explored; structure cannot be reduced to agency, which is the fundamental defect of methodological individualism, upon which rational choice theory rests.

Notes 1

The board of governors of a school has the legal responsibility for the operation of the school and is composed of members appointed by the local government, representatives elected by the parents of children enrolled in the school, and, if applicable, appointees of the organization sponsoring the school (e.g. the local Roman Catholic diocese in the case of a Roman Catholic School).


The research (ESRC Project Number R000221689) was jointly conducted with Dr Priscilla Chadwick and with the assistance of Anthony Amatrudo.


The Funding Agency for Schools was established by the 1988 Educational Reform Act as the body responsible for overseeing the funding of grant-maintained schools.


INSET is the acronym for ‘in-service training’. This is the programme of continuing professional development provided by schools for their teaching staff.


Rational choice or ‘Hobson’s choice’? Intention and constraint in UK higher education Andrew Parker

Questions of structure and agency are never far from the core of any sociological debate. Discussions concerning educational systems, policies and/or practice are no exception in this respect. The aim of this chapter is to highlight some of the fundamental aspects of structure– agency tensions within educational milieux by considering a rational choice analysis of the lifestyle experiences of a group of UK higher education students. Using empirical evidence gathered as part of a qualitative research project into student lifestyle priorities, the chapter focuses on the themes and issues highlighted by respondents as central to their negotiation of everyday life.1 Accordingly, student behaviours and actions are examined in terms of the extent to which they may be considered the result of individual intention, purpose and preference or the consequence of external constraint, influence and pressure.

Rational choice theory and educational decision-making In its purest form, rational choice theory (or rational action theory) adheres stringently to notions of utility, profit, wealth and maximization and spares little, if any, thought for those who would consider social action to be solely determined by external factors (except insofar as these factors can be reduced to and represented as ‘costs’ which agents take into account) (see Coleman 1990; Coleman and Fararo 1992a). Despite the economic roots from which these characteristics derive, rational choice models aspire to universal applicability. Hence, all social actors are seen as ‘utility maximizers’, making rational choices about the lifestyle alternatives available to them by ‘weighing-up’ the potential outcomes of any one course of action (Zafirovski 1999a). Economically speaking, ‘ideal’ rationality is a question of intention, purpose and choice optimization entailing a system of fully ordered preferences, complete information and perfectly calculated outcomes (Hollis 1987, 1994). In this sense, maximizing benefits and minimizing costs is all that really matters.

Rational choice or ‘Hobson’s choice’? 127 Education, of course, is one aspect of life where choices and preferences can, and indeed do, have profound effects on the life-course as a whole. Yet there is some contention within this field over the role of individual agency. There is a strong tradition within the sociology of education, for example, for authors dogmatically to embrace one of two explanations for the way in which educational decisions are made – either as a consequence of powerful reproductive forces social actors are exclusively ‘pushed’ (or propelled – see Elster 1979) into given destinations, or as a result of considered judgement and preference they rationally ‘jump’ (or are pulled) towards circumstances which present the greatest rewards at particular points in time (see Gambetta 1996). In terms of the impact of structural constraint on individual agents, perhaps the best-known educational writings are those of Althusser (1971) and Bowles and Gintis (1976), in which social actors are depicted more generally as cultural ‘dupes’ or ‘puppets’, disarmingly exposed to the determinant forces of the state (and, necessarily, to the forces and relations of production). Alternatively, the influential power of cultural reproduction in educational spheres (i.e. via the cultural capital hypothesis) has been outlined by Bourdieu and Passeron (1977) and by Willis (1977), who seek to establish the subcultural mechanisms by which social actors are guided towards specific courses of action (see also Hyman 1966; Lane 1972). Offering some kind of counter-argument to a preoccupation with the pushed-from-behind thesis, rational choice theory has been deployed as an explanatory tool within studies of education to highlight the role of the individual agent in the decision-making process. Several writers have pursued this line of thought. Boudon (1974), for instance, has explored notions of class differentiation in education by inferring the existence of ‘primary’ and ‘secondary’ effects of social stratification within this sphere; where ‘primary’ effects of class are taken to indicate the differences in academic ability generated by family background and ‘secondary’ effects are the educational choices made by young people and their parents at key junctures in their educational careers. Following this lead, Goldthorpe (1996) has utilized a rational action approach to describe ways in which class differentials in education have persisted over time, despite the more recent expansion of educational opportunities in advanced industrial societies. Whilst Goldthorpe (1996: 485) fervently advocates the benefits of rational action theory (RAT) over culturalist explanations of educational origin and destination, he does distance himself from ‘ideal-type’ notions of rationality, opting instead for, what he terms, a ‘weaker’ version of RAT: I assume that actors have goals, have usually alternative means of pursuing these goals and, in choosing their courses of action, tend in some degree to assess probable costs and benefits rather than, say, unthinkingly following social norms or giving unreflecting expression to cultural values. I also assume that actors are to a degree knowledgeable about their society and their situations within it – in particular, about opportunities and

128 Andrew Parker constraints relative to their goals – rather than, say, being quite uninformed or ideologically deluded. In sum, I take it that actors have both some possibility and some capacity for acting autonomously and for seeking their goals in ways that are more or less appropriate to the situations in which they find themselves. This stance is significant, Goldthorpe (1996: 485) argues, in that it provides the backdrop against which departures from a standard of ‘perfect’ rationality can be regarded as commonplace amongst social actors: I make no assumption that actors are always entirely clear about their goals, are always aware of the optimal means of pursuing them, or in the end do always follow the course of action that they know to be rational ... . Similarly, Gambetta (1996) avoids the adoption of ideal-type models of rational choice theory. His stated objective is to use the empirical research data gathered from investigations into the range of mechanisms governing individual choices within the context of Italian educational progression, in order to uncover the extent to which social actors might be pulledfrom-the-front (i.e. by individual/self will) as well as pushed-from-behind (i.e. by structural/ external forces) during their educational careers. To this end, Gambetta’s overall aim is explicitly to avoid an adherence to any specific theoretical standpoint by allowing the data themselves to demonstrate the way in which both the forces and influences of structure and the will of individual agents can be identified amidst such educational processes.2 In a similar vein, this chapter casts aside preoccupations with any one theoretical standpoint, and adopts instead a more open-ended analytical approach, whereby the educational experiences, actions and behaviours of students at one UK higher educational institution can be scrutinized in terms of both their rational and/or non-rational origins.

Context and method Whilst educational institutions, systems and practices have long since been the focus of sociological attention, higher education appears to have eluded much serious investigation, and in this sense stands out as an under-researched area. That said, issues concerning student lives and lifestyles have not gone completely unnoticed in terms of their potential as research topics. In 1995 the National Union of Students (NUS), for example, carried out investigations into student employment, revealing that as many as four out of ten students hold down some form of employment during term time: the majority of these jobs being located in the retail and hospitality sectors (NUS 1995). In turn, changes to the ways in which higher education in Britain has been financed in recent years (and the subsequent pressures for those ‘buying

Rational choice or ‘Hobson’s choice’? 129 into’ its services) have stimulated interest and criticism from a variety of sources. Between July and August 1999, The Times Higher Education Supplement published a series of weekly/ bi-weekly articles charting the plight of a number of first-year undergraduate students from a cross-section of Britain’s ‘old’ and ‘new’ universities. Representing the first cohort of feepaying students this country has seen, the respondents were asked for their views on six key areas of student life: (i) finance (including tuition fees); (ii) paid employment; (iii) the cost and quality of accommodation; (iv) academic course organization and delivery; (v) university facilities; and (vi) the learning environment. Findings from this research showed that whilst not all students needed to work to survive financially, for a significant proportion paid employment in term-time was accepted as an integral part of student life. Similarly, financial debt was not something which appeared particularly alien to the majority of those surveyed. Amongst the monetary despair over tuition fees, naïve spending habits, student loans and credit card arrears there were glimmers of hope indicating that, irrespective of the debt one might accrue whilst at university, what really mattered was finding a job at the end of it – thereby having the financial resources to pay back all that was owed (see Thomson 1999a). Of course, one might argue that these findings indicate that little has changed over recent years in terms of the pressures, tensions and dilemmas of student life. After all, a lack of money, temporary and/or part-time employment (albeit, in vacation time) and sub-standard accommodation have long since been lodged in the public psyche as key constituents of student existence – a necessary evil perhaps, to accompany the privilege of not having to ‘work for a living’. Yet to reinforce this traditional view within the present political and economic climate would be somewhat misleading. Things in and around UK student life have changed. To accrue debt, in the name of deferred gratification, through undertaking any degree course no longer guarantees the luxurious prospect of enhanced employment opportunities. Moreover, the increased demand for, and accompanying expansion of, higher education in recent years necessarily means that more people with similar qualifications will one day vie for the same occupational vacancy (see, e.g., Purcell and Pitcher 1996; Purcell et al. 1999). In turn, the pressures of holding down a job whilst attempting to construct some kind of balance between social and academic life must inevitably take its toll. Perhaps it is not surprising, then, that, according to one recent report, students are twice as vulnerable to poor mental health as their non-studying contemporaries (MacLeod and Mulholland 1999).

Stanbridge University: an institutional case-study3 Stanbridge University is one of Britain’s leading higher education institutions which proffers a long history of academic excellence. Located on the periphery of the town from which it takes its name, Stanbridge comprises a large single-site campus equipped with all the

130 Andrew Parker academic and social facilities any modern-day student might need. Over the years Stanbridge has become renowned for (and indeed, synonymous with) specific academic disciplines and vocations, and continues to offer a wide range of courses for undergraduate and postgraduate, full- and part-time students. Its promotional literature boasts a series of strengths: the maintenance of a ‘near-perfect’ record for teaching quality in assessments by the Higher Education Funding Council for England (HEFCE); a healthy departmental commitment to research activity; ‘high scores’ in Research Assessment Exercises (RAEs); strong teaching, research and recruitment links with industry; and high rates of graduate employment. During the course of the present research, many respondents made it clear that they had chosen to study at Stanbridge because of its academic reputation and the increased chances of employment which this appeared to engender. Thus, the costs and benefits of this institutional pedigree were things upon which some educational decisions were clearly based, and, as such, these factors cannot be construed in individual terms as the methodological individualism of rational choice theory necessarily suggests.

Respondent profile Respondents participated voluntarily and were recruited through Students’ Union links with chosen academic departments at Stanbridge (Chemistry, English/Drama, Social Science and Engineering), in order that a diverse range of individual experience could be explored. Initial recruits to the project were asked to identify the central lifestyle elements around which they felt modern-day student life was constructed. The following six areas were identified:

dilemmas arising from issues of financial survival and part-time work;

the prioritization and/or balancing of academic life and social life pressures;

problems and issues concerning accommodation;

the general stresses of student life relating to coursework stipulations/obligations;

public perceptions of students and student lifestyles;

the evident contrast between prior perceptions of university life and actual experience. Data collection took place between February and November 1999 via four pre-arranged

focus groups made up of students from the chosen departments. In addition, one focus group contained a mixed selection of first-year students from various academic disciplines. A total of 27 students were involved in the project, 14 of whom were female and 13 male. All respondents were between the ages of 18 and 27.

Rational choice or ‘Hobson’s choice’? 131

Finance, fees and the ‘university experience’ Alongside debates concerning the way in which notions of cultural capital might benefit some individuals with regard to educational opportunity and achievement, there are wellestablished arguments within sociological literature focusing on the relationship between economic constraint and educational choice (Ahier and Moore 1999; Boudon 1974; Gambetta 1996; Goldthorpe 1996; Hatcher 1998). Whilst higher education in Britain has benefited from government subsidy for a number of years, particularly in the form of local education authority maintenance grants, recent changes to state funding arrangements mean that those now wishing to go to university have to dig deeper to find the economic resources to do so. In this sense, money, wealth and economic prosperity emerge as key factors in determining what kind of higher educational opportunities are available to individuals in the UK. Given this situation, it is perhaps understandable that recent research findings concerning student attitudes towards money, debt and work appear to portray a general discontent towards the present-day conditions of UK student life (see Currie 1999a, 1999b; Thomson 1999a, 1999b, 1999c, 1999d). The imposition of tuition fees alongside the widespread withdrawal of maintenance grants seems to have raised the financial stakes for those entering higher education, as Thomson (1999a: 6) points out: Undergraduates are expected to pay up to £1,025 a year for their higher education, excluding any optional course extras, and to cover their maintenance with student loans ... meaning a possible £13,440 debt upon graduation. On top of this, many students acquire bank and credit card debts. Yet what is also striking from these research findings is the rate at which some students have seemingly come to accept such financial pressures as part and parcel of university life, as if they feel that ‘buying’ their education is the obligatory price one now has to pay for occupational fulfilment, job satisfaction, career development and social mobility. So, what of students at Stanbridge? How did they perceive their financial position? Opinions differed depending on which stage of their studies respondents had reached. In terms of the trade-off between ‘paying’ for higher education (via tuition fees) and gaining something worthwhile at the end of it, first years had stronger views than most. Robin, the only first-year representative in the Social Science focus group, felt that because students now had to pay for their time at university, they were more likely to work harder and take their academic commitments seriously. I think we’re gonna see the balance shifting ... with the first years this year. Course fees, first years paying £1000 a year, and it’s gonna go up each year. They’re gonna want

132 Andrew Parker something to show for over £3000 in tuition fees for a three year degree, and that’s one thing that’s gonna be driving this year’s first years. Tracy spoke from a personal viewpoint about the way in which she felt constantly ‘pressured’ by this underlying financial burden: My parents don’t really support me very much financially so it’s a pressure to myself that I think if I don’t learn things then I’m gonna come out of university with a massive debt and I’m not going to get a job ’cos I don’t know anything. And so there’s always a constant pressure that you think, ‘I really should be getting as much out of this as possible because I’m paying for it.’ These student responses indicate that structural constraints (in the form of finance) actually enter into individual preference formation rather than simply operating as ‘costs’ which curtail preferences. In this sense, student attitudes towards academic study clearly differed. But what Trudy and Robin articulated here was a more general feeling that as a consequence of the introduction of tuition fees, higher education now represented a kind of calculated risk or contractual agreement; a process of investment and reward rather than a simple matter of obligation or personal fulfilment, in which choices concerning what one intended to ‘get out of the university experience were of greater importance than they might have been previously. That is not to say that issues of finance and educational achievement were of paramount concern to students. Striking the balance between making the most of university life whilst managing to stay out of too much debt also had to be considered alongside the inevitable pressures of socializing. Negotiating the complexities of this three-part sum appeared difficult for many students. During the course of the first-year focus group several members took a retrospective look at the ‘university experience’ so far, whilst debating the detailed nuances of money management. Ben:

I think ... first years don’t know what to expect [when they arrive at university] ... don’t know how to budget and sort a lot of money out.


There are people by, sort of, Christmas ... like, twenty pounds a week [to live on], I mean if somebody says, ‘Oh, come on out, it’s not going to cost much’, that’s the budget for the whole week gone!


I think I’ve learnt very quickly because I didn’t care at first about spending absolutely loads.

Rational choice or ‘Hobson’s choice’? 133 Harriet:

You look back and think, ‘I wish I’d budgeted more and then I would have been OK now’, but now I’m really having to think I can’t go out and I can’t do this and that because of it ...

These students, it seemed, had learnt reflectively about endorsing a particular lifestyle in the light of living with certain constraints. To this end, their own experiences had led them to reformulate their preferences strategically over time. Yet when compared with the sentiments of more experienced students, there was nothing unusual about such comments. Final-year respondents also talked openly about the ways in which their social habits had been curbed both by the pressures of finance and by those more closely associated with issues of coursework and academic achievement. For some, socializing and the ‘university experience’ came second to more pressing concerns surrounding basic existence. Trudy:

Some people don’t have a choice about whether they’re gonna take a student loan out and spend it on alcohol or leave it in the bank. Some of us have to take a student loan out and actually spend it on food, on books and things like that, and for us there is no choice ...

Despite these financial fears, it was noticeable that, contrary to recent research findings (see Currie 1999a), few of the students interviewed appeared to be involved in any form of employment during academic term-time. Many were happy to rely on hand-outs from parents or to accrue more debt, safe in the knowledge that interest rates on student loans were favourable and that they would not have to begin repayments until they reached a certain level of occupational income. Juliet:

There’s a lot of scared people who worry about debt. You hear all these things – students have got this debt, students have got that debt – but if you actually look at the conditions of your student loan you don’t have to start paying it off until you’re earning something like fourteen-an-a-half grand a year.

These two student comments illustrate the disparity of individual experience with regard to economic constraint, demonstrating that it did not exert a blanket effect (i.e. a push), but was instead agentially mediated by attitudes towards it (hence, emphasizing the dynamic interplay between ‘structure’ and ‘agency’). On the one hand, Trudy articulates real fear about the whole notion of financial survival in relation to her general state of well-being, whilst, on the other, Juliet expresses an altogether more laissez-faire attitude towards money and, in particular, debt repayment. In fact, it was this latter approach which surfaced most prominently amidst student discussion. As far as a number of individuals were concerned,

134 Andrew Parker once they had got used to the idea of being in debt, it then became a matter of course for them to ‘throw caution to the wind’ and enjoy themselves whilst they could. The ‘university experience’ was not something they wanted to miss out on even if this meant blatant overindulgence in the face of financial adversity. Robin, first-year Social Science, provided a brief overview: I mean, it depends who you are. If you’re here to have a good time, you do, you spend. I mean, I know people who spend £50 a night and that’s fair enough if they’ve got the money. But I haven’t, so that’s why I don’t. ... It depends who you are and how much money you’ve got to spend and what you wanna do with it. Such revelations provide a stark contrast to some traditional assumptions surrounding student life as a time of relative poverty and financial hardship. However, they appear to reinforce the results of recent research which alludes to the more socially indulgent nature of present-day student existence (see Little 1999; Millar and Ward 1999; Pirie and Worcester 1999). According to Little (1999: 25), for example, ‘The traditional image of hard-drinking, loose-living students is alive and well.’ Citing average figures for student expenditure on such things as drink, clothes, entertainment and books, Little concludes with an exposé of one UK university Politics student who claimed, when interviewed, that spending £25 per night on drink, three or four nights a week was not uncommon for members of her immediate peer group. Similar levels of social expenditure were cited by first-year Stanbridge students as the average cost for a night out on campus. Drinking in ‘the [Students’] Union’ was seen as a central way of keeping to one’s budget due to cheaper beer prices, favourable venue entry fees and the avoidance of taxi fares to and from residence. Spending ‘fifteen [pounds] on a Wednesday and maybe on a weekend, say, twenty’ seemed the norm. Occasional nights out ‘off-campus’, however, had a tendency to cost much more, as Ben (first-year Industrial Design) described: If we went out in Hesselton [nearest large town to Stanbridge], say, it’d be twenty quid on taxis alone. ... I could easily spend about forty pounds. ... You’re looking at five or ten pounds entry into somewhere [i.e. night-club] aren’t you ... and if you’ve made the effort to go into town you might as well go out for a meal before you go out, it’s nicer ... Not only did the pursuit of such indulgences carry with them the threat of financial pressure later on, but more immediate consequences surfaced during further discussion. Many students, for example, admitted to feelings of guilt with regard to the way in which their parents might view their social behaviour – how they spent money, their excessive drinking habits and the relegation of work in relation to play. These feelings could be largely attributed to those students whose parents were making significant financial contributions towards their

Rational choice or ‘Hobson’s choice’? 135 time at Stanbridge, as Jackie (year 3 Chemistry) and Harriet (year 1 Maths and Sports Science), explained: Jackie:

My mum would freak if she knew what I was doing. If they knew half the things that went on at university, they’d never let anyone come. My mum was horrified that I got in at half past three on Saturday morning, and you’re like [trying to explain]. ... Oh, just went to the [Students’] Union, ended up having a curry, y’know ...


I think my mother would be shocked if she knew how little [work] I actually did. ... She knows I go out a lot [but] ... she’d be really disappointed in me if [she thought] I wasn’t doing my work as well ...


Do you feel ... is there a guilt thing about ... ? [Interrupted]


Oh, definitely, I feel like I’m using some of their [parents] money and just ... having a good time on it and not really doing a lot at all. ... I know that they had to put away a lot of money to get me to university ... they’re not well-off but they’ve saved for me to be at university without going in debt. ...They want to give you the best education they can so, I mean, they’re going out of their way to help you really. So the whole time you do feel guilty.

Despite these emotive concerns, overall students’ attitudes and experiences towards issues of finance, fees and university life were positive. Distinctions were evident in individual outlooks on money and social priorities, with first years adopting a somewhat hedonistic lifestyle approach whilst older students concentrated more on moderate social expenditure and academic achievement (see also Millar and Ward 1999). Heavy socializing, it seemed, was something which was an integral part of early student life yet one which most people ultimately ‘grew out of’. At that point, lifestyles changed, as did individual priorities, which, in the main, seemed to become geared more towards academic achievement and its related anxieties.

Accommodation, anxiety and academic achievement New arrivals at Stanbridge had quickly to master the art of rationally negotiating ‘social’ and ‘academic’ time and space. Central to first-year life was a strong collegial ethos based around established traditions between the various halls of residence on campus. Because the vast majority of first years ‘lived-in’ (halls), they were automatically implicated within this

136 Andrew Parker collegial ethos, one consequence of which was an obligation to be ‘socially visible’. Here, normative factors can be seen to be evident within the lives of the students concerned, and these cast doubt upon a simple cost–benefit analysis in the sense that such collective pressures cannot be recognized as a ‘cost’ in relation to the realization of a stable preference. Instead they enter into preference reformulation, which cannot therefore be construed in purely atomistic and individual terms. Thus, whilst hall life had its advantages, it also presented various forms of constraint in relation to peer group expectations and behaviours (i.e. there was always someone who wanted to ‘go out for a drink’), and because of this, it also had its drawbacks. Several students talked, for example, about the way they sometimes felt compelled to feign illness in order to fend off peer group ridicule for not socializing. Others described how it was ‘simply impossible’ to work in their rooms due to constant distractions from people knocking at their door for ‘a chat or just looking for something to do’. Although striking the balance between work and play was not always easy, there were some individuals who contravened social obligations and concentrated purely on their studies. This was seen to be acceptable by a selection of those interviewed, but such behaviours did appear to incur a degree of risk with regard to peer group ostracization: Tracy:

I think there’s some quieter people who are very focused on their course. They have heavier courses than us and they’re just in their room at their computers the whole time. We don’t really know who they are. If they come [out] they are there because everything is done as a hall. There’s very much kind of an in-crowd within a hall and there’s some people that we just don’t know because they’re the ones who are really in their rooms working.

Trudy and Jennifer (both Social Science year 3) reminisced about their days as first years: Trudy:

If you get known as a person who ‘works’ then it’s a reputation you can’t get rid of ... and people will come to you for work and not for friendship or to go out ...


It’s quite difficult to work in halls as well. I was in halls in my first year; I’ve moved out the last two years. ... When you’re in your room your door’s open, that’s how everyone knows you’re in. So you have your door open, working, and you sit down to do something and someone arrives ... and they just sit down at your desk, they wouldn’t think about you working. The fact that you’re in your room means you’re in, so you get them coming in and sitting down and that’s it, you don’t get anything done. It was worse for my friend because she shared a room ... and, I mean, they never got anything done because there was always somebody in the room.

Rational choice or ‘Hobson’s choice’? 137 For Robin (Social Science, year 1), the ongoing battle to work hard whilst protecting one’s social reputation had seemingly been enhanced by the kind of ‘hall’ he inhabited: I’ve got the same thing but it’s pretty brilliant for me. I share a room with a guy who does ... a similar course so we can help each other out. We’ve got similar kinds of work and similar work patterns so ... we’re always there for each other, to help each other out. ... I’m quite well known in hall and I’m really enjoying my hall, it’s the best thing about university I think from my point of view. But I work hard and go out hard and there’s a lot of people in my hall that do that. There’s 400 people in my hall of which about 100 go out regularly and they’re known as like the socialites that go out all the time. But amongst them there’s about 50 people that work hard a lot of the time so I’m quite happy to get the balance of having a good reputation for going out, but also having time to work hard, devoted to myself and to my studies. It is worth noting, then, that only a minority of students adhered firmly to work-oriented goals. Most succumbed to external (i.e. peer) pressure in some way and changed their lifestyle behaviours either because of reputation considerations or due to the situational dilemmas they regularly faced. Having said that, for the fortunate few, some kind of social/academic balance was possible, but this was not helped by the depth and intensity of the collegial atmosphere at Stanbridge. A good example of this ‘intensity’ was provided by the fact that each hall of residence had its own Social Committee to organize intra-collegiate competitions, and its own newsletter was regularly used as a means by which candid comment could be made about individual social habits. To be featured in a hall newsletter was some kind of public acknowledgement of one’s ‘in-crowd’ status. But where, we might ask, did academic study slot in to the lives of ‘in-crowders’ given the social obligations in play? Again, a straightforward kind of rationality prevailed whereby the minimum requirement for academic attainment was weighed up against the overall cost to other lifestyle activities. Nicola:

In the first year ... you just have to pass. Y’know, it doesn’t go towards your degree ... you just think, ‘Oh well, I just need to get 40 per cent, I’ll just go out [and socialize] or whatever.’


At this stage I think people are sort of going out a lot and thinking socialize, socialize, but at the back of my mind I’m thinking, ‘I can’t fail this year’. I really can’t fail this year. I’d be ‘gutted’ if I failed.


The thing this year, the first year, doesn’t count [towards the final degree classification].

138 Andrew Parker Samantha:

Yeah, everyone just needs 40 per cent this year I think.

It was this whole notion of ‘getting through’ year one with a ‘bare 40 per cent pass’ that appeared to provide the cushioned backdrop against which a significant number of first-year lives were structured. Those who wanted to work hard, and ‘prove themselves’ at an early stage in their courses, were able to do so, providing they could find the right kind of environment within which to study (i.e. one that was quiet and conducive to work, such as the library). In contrast, those who preferred to buy into a much more relaxed way of life were able to do so courtesy of the nominal academic ‘pass’ requirement which had no bearing on their final degree classification. Alongside the freedoms of being away from home, it was this latter scenario which first years cited as the main reason behind their heavily indulgent social culture. Indeed, discussions with final-year students indicated that they too had adopted a similar kind of lifestyle strategy in their early years at Stanbridge. From the end of the first year onwards, however, there was evidence to suggest that individual approaches to work and play changed quite dramatically, despite the knock-on affects this had for personal relationships and/or social reputations. Samantha:

In your first year you haven’t got that many worries, you’ve just met these people [fellow students], they’re your new friends, and what you concentrate on is going out and having a laugh. In your final year everyone’s got problems, everyone’s stresses, there’s a lot more stuff to work through ...


I mean, this year I just haven’t given a stuff what people think. I’ve fallen out with friends ... even this lot [gestures towards some of group] that are my best friends they’ve said ‘C’mon, we’re going out’, and I haven’t wanted to go out. Y’know, it’s caused a bit of tension now and again but I don’t really give a stuff because even if I’m not working I just wanna relax. ... I mean in the first year and the second year there’s always that, well, y’know, ‘I’ve always got next year, I can do a bit better. ... I’ve always got next semester and whatever. But this is it now for us and I just don’t care what other people think anymore.

What is the main reason behind this change of approach? Final degree classification: Trudy:

I’m really, really stressed this term because I’m desperate to get a 2:1. I’ll be absolutely ‘gutted’ if I get a 2:2. My housemate, she’s sitting on [an average mark of] about 54/55 per cent, she normally goes out three times a week. She knows she’s going to get a 2:2, she doesn’t care, she’ll be happy with a 2:2, she’s managed to have a good time for three years. Whereas I’ll be absolutely ‘gutted’ if I get a 2:2 because I’ve worked.

Rational choice or ‘Hobson’s choice’? 139 Once again, these responses highlight how academic constraints did not operate as blanket influences directing student behaviour. Rather they were taken into account and mediated in different ways by the individual agents involved. Indeed, in many ways this typified what life beyond year one was all about – the prioritization of work over play; the re-alignment of intention and purpose according to anticipated academic reward. With the pressures of social indulgence gone, a new kind of behavioural conduct kicked in, at which point the means–ends calculus changed to accommodate an altogether more serious, more moderate, more rational lifestyle approach. But to assume that this new kind of rationality did not have its own nuances and characteristics would, of course, be misguided. A number of students inevitably pursued optimum strategies of personal investment, some of which seemed as anti-academic as firstyear social habits. Opting out of lecture attendance, for example, was one such ploy which respondents suggested was more a question of time-management and efficiency than the flouting of academic obligation. Moreover, a number of individuals argued that their particular levels of absenteeism were as much a statement about their perception of teaching quality at Stanbridge as a deliberate form of institutional resistance. What intrigued me was how students managed to pass courses without attending class. Gary and Tony, two final-year Chemistry students, explained their methods, motives and objectives: AP

So if you don’t go to lectures, how on earth do you get through a module?


Reading other people’s notes. In which case the emphasis is on you to try and learn it yourself.


... It’s just as easy to not go [to lectures] and for me to photocopy Gary’s and I wouldn’t have actually missed anything to be fair. He’s [Dr Brown] that style of lecturer. You probably get more concise notes just taking it from someone else than sitting there and doing it yourself.


’Cos, while you’re sat there you don’t get a chance to understand it either.


I mean, even missing the lecture you probably save yourself an hour and you can actually concise [sic] the notes down and you get quicker and better notes in less time. The final year is a lot about time-balancing.

Styles of lecture delivery were clearly under scrutiny here, as were the support structures that went along with them. It would seem, for instance, that some members of staff were unconsciously sponsoring non-attendance by making their lecture notes available in the university library and/or via the Internet – practices which led some students to feel even more confident about missing scheduled classes. Hence, as university life progressed, student rationality surrounding academic pursuits did not simply intensify via the sacrifice of social hedonism; for some it switched focus to the point where even lecture attendance was subject to scrutiny in terms of a time/efficiency cost–benefit analysis.

140 Andrew Parker

Rational choices in higher education? It is claimed that rational choice theory can provide the basis of a unified and comprehensive explanation of social behaviour. If this is so, how might we draw upon this theoretical perspective to make sense of the data presented here? Where does our glimpse of modern-day higher educational life take us in terms of rational choice explanations? In terms of structure and agency, should we view Stanbridge students as the ‘creatures’ or ‘creators’ of their social worlds (Hollis 1994)? In providing some kind of response to these questions I wish to pursue several key themes which emerge from the data as contradictions to rational choice models of analysis. To begin with, let us consider the basic concept of ‘preference’. According to rational choice theorists, social actors develop tastes, values and/or preferences which, in the words of Zafirovski (1999a: 53), may be viewed as ‘the engines or explanatory variables of action’. Preference is something which is stable, hierarchical, transitive and based upon ‘perfect information’. Yet an adherence to stable preferences was not always in evidence with regard to Stanbridge students. Of course, it would be somewhat naïve to suggest that any of the individuals surveyed were in possession of ‘perfect information’ regarding their institutional/ social surroundings. Nevertheless, although clear preferences were expressed in terms of lifestyle choice, these appeared to change significantly as university life progressed (i.e. the shift from first-year social hedonism to final-year academic re-orientation). How and why did this shift occur? One fact it seems that we must acknowledge is that first-year lifestyle choice was heavily influenced by peer pressure and group norms which themselves were generated in accordance with the well-established and highly organized social culture existing in and around Stanbridge as an institution – and in particular its halls of residence. As we have seen, during focus group discussion some students alluded to being coerced into social activity when their own individual preferences were geared more towards academic work, saving money and/or simply ‘staying in’. Some, if not all, of these preferred actions were framed within the context of the financial pressures which students found themselves under and, therefore, should also be viewed as representing collective external pressures and forces (i.e. lack of financial assistance from the state, loyalty towards parents in relation to monetary subsidy, etc.) which cannot be reduced to the aggregate influence of other individuals. Notwithstanding these factors, the point I wish to make is that in terms of the social circumstances of first-year life, it would seem fair to suggest that individual decisions were made as a consequence of (and on the basis of) a broader collective normativity or desire. It is worth considering where the collective momentum for social decisions originated. For example, there appeared to be a general feeling amongst a number of respondents that a series of wider norms and expectations existed concerning the way in which first-year university students were supposed to behave (i.e. that the initial period of university life was

Rational choice or ‘Hobson’s choice’? 141 all about socializing, individual freedom and hedonism), and that they were simply living out these expectations, as was most clearly exemplified by the socialite ‘in-crowders’. That peer pressure played such a central part in respondent discussion raises further questions in terms of the extent to which rational choice models might be applied to student life at Stanbridge, given their methodologically individualistic groundings. If we consider the variable nature of preferences displayed over time (and in particular throughout the duration of the university period), the end of first year clearly brought with it the opportunity for many individuals to escape the intensity of external/group social pressures and to re-assert their own individual tastes, desires and values over these normative influences. In this sense, the move out of residence appeared to facilitate its own kind of individual freedom for some via which preference reformulation took place. What then of this new preference structure whereby academic orientation was hierarchically placed above socializing? How might this be viewed in accordance with rational choice theory? Choice optimization and utility maximization are clearly visible here. In terms of calculated outcomes, interview discussions indicate that greater academic input on the part of the student was seen by him or her as necessarily linked to the likelihood of attaining a better classification of degree, which, in turn, was seen as a means of increasing occupational opportunity and choice. This shift in preference was commensurate with a shift in allegiance from the socialite collective (i.e. peer group norm) to the individual. Indeed, academic orientation appeared to be entirely driven by self-interest. Once formal methods of academic assessment became intimately tied in with final degree classification, students put their own needs, desires and livelihoods firmly before those of their associates. Whilst this increased the risk of peer group ostracization, for many respondents, collective endeavour was no longer a priority when the calculated implications and repercussions of its continuance were seen as detrimental to individual life-chances. That does not mean to say that utility maximization ended there. Even within the context of academic orientation there were more specific utility decisions to be made in relation to the costs and benefits of individual effort. In this sense, lecture non-/attendance was cited by a small number of respondents as just one example of the lengths to which students might go in scrutinizing the costs and benefits of their own actions in order further to increase their chances of academic success over and above those of others.

Conclusions This chapter has outlined the lifestyle behaviours, the shifting priorities and preferences of a cohort of UK higher education students in the late 1990s. In doing so, it has highlighted that, far from being a straightforward issue of structural, external constraint or rational purposive action, student lifestyles are influenced and informed in different ways at different times by

142 Andrew Parker a range of different factors. What this suggests is that within this context social action cannot be described as exclusively rational or irrational. Evidence presented here demonstrates how, in terms of the educational decisions they made, students were both pushed and pulled in a variety of ways. The central aspects of these decision-making processes are now summarized. The student cohort was explicit about the economic constraints which they were under, irrespective of their stage/year of study. Some found the pressures of accruing debt and financial hardship particularly difficult to deal with, and this inevitably influenced individual priorities in terms of lifestyle choice. That said, a number of those interviewed (mainly firstyear students) appeared to ignore financial/economic constraint, preferring instead to enjoy ‘university life’ to the full, despite the increased threat of financial adversity. Powerful subcultural norms and expectations existed around notions of sociability at Stanbridge, so much so that preference was often given to social endeavour over and above issues such as financial difficulties and academic study. According to respondents, such exploits were aided by the relatively moderate assessment requirements in place, which appeared to facilitate the lack of academic commitment shown. This hedonistic period of social indulgence, however, was relatively short-lived. The onset of year two witnessed a more serious academic outlook courtesy of the ‘degree-related’ nature of academic performance. Assumptions concerning career and occupational progression were based largely upon notions of degree classification, and this, in itself, appeared to influence social behaviours. Once academic work began to ‘count’ in this way, students realigned their lifestyle priorities in accordance with a more studious academic approach where the costs of socializing were explicitly weighed up against (and exchanged for) the benefits of scholarly discipline. So what does all this say about rational choice theory? Quite simply, it says that contemporary student life demonstrates evidence of external constraint as well as individual choice, and that the kinds of ‘choices’ made by students at Stanbridge University (centred on their their main lifestyle dilemmas/areas) differed depending on their stage/phase of higher educational life. This, in turn, indicates that whilst rational choice models are indeed useful, they do not sufficiently explain the complex interplay of factors impinging upon students’ lives. This investigation shows that rational choice theory is clearly applicable to higher educational contexts -students thought for themselves during their negotiation of lifestyle priorities and opportunities and were directly motivated/attracted by the future rewards of final degree outcome. However, other factors (i.e. economic, sub/cultural) were influential upon their everyday existence. Hence, respondents were pushed from behind as well as being pulled from the front. More specifically, the push factors were mediated by the pull factors – by individuals who assigned their own weights and measures to these two elements – thus leading to the existence of observable regularities and individual diversity. Needless to say, contemporary UK student life may be seen to be as much about structural constraint as it is about individual will. Indeed, to suggest that it was more about one of these factors than the

Rational choice or ‘Hobson’s choice’? 143 other would be decidedly misleading, just as it would be to assume that they exert a constant influence from year to year. Structure and agency are in dynamic interplay over time. Agential preferences are not a stable point in the students’ moving world. Instead, they are formulated and reformulated as the stringency of various constraints themselves vary over time. Ultimately this means that preferences are part of the ‘vexatious fact of society’; they pertain to individuals, but they cannot be explained in purely individualistic terms.

Notes 1

For a more detailed outline of the research from which this data is derived, see Parker (2000).


For further insight into the way in which writers within the sociology of education have utilized rational choice/action theory, see Ahier and Moore (1999), Goldthorpe (1998) and Hatcher (1998).


To preserve anonymity, pseudonyms have been used throughout this chapter.



‘I do’ A theoretical critique of Becker’s rational choice approach to marriage decisions Ian Procter

Gary Becker has been hugely influential in developing rational choice theory (Blaug 1992). This is because he has engaged with substantive matters which at first glance seem unamenable to the application of such a theory (Becker 1996: ch. 7; Coleman 1996: 171). This applies particularly to his analysis of the family in the two editions of A Treatise on the Family. Here he develops ‘an economic or rational choice approach to the family’ (Becker 1991: ix), making it clear that this emphasizes the theoretical approach rather than the ‘economic’ being restricted to the material aspects of life, as in the common-sense meaning of the ‘economy’. This chapter will provide a theoretical critique of Becker’s approach, focusing on his conception of the rational actor and how s/he takes decisions. It will concentrate on one area of decision-making in particular, that of whether to marry and whom to marry. In his theory Becker is clear that his model of the human actor is indeed a model – an abstract construct rather than a rounded description. This critique asks whether his model can work, not in the sense that it is an impoverished conception of the full complexity of human decision-making, but in the sense of whether the process of choice which Becker’s model implies is possible. The next section of the chapter outlines rational choice theory and Becker’s application of the approach in relation to the marital decision-making process. This is followed by a section criticizing two crucial aspects of Becker’s theory: the claim that people can compare the utility to be gained from alternative marital situations; and his conception of how people process information. The last main section turns to Becker’s 1996 book Accounting for Tastes, and asks whether the developments in rational choice theory proclaimed there answer the criticisms made earlier in the chapter.

Rational choice theory and its application to marriage Hollis (1987) has usefully analysed the crucial components of rational choice theory. It explains by showing the rationality of actions within its limiting assumptions. If empirical actions do not conform to this, then the theory identifies where other factors beyond the rational are intervening and thus where further explanation is necessary. Rationality is

148 Ian Procter defined in a precise and minimalist way (Hollis 1987: Ch. 2). An act is rational if the agent has ordered preferences, has perfect information about how to achieve preferences and what the costs are, and correctly calculates the relationship of preferences and costs and thus maximizes the benefits to be achieved. The rational agent’s preferences must be complete and consistent, but need not have any content or shape. To call an agent rational is to say merely that he reasons correctly in identifying the action likeliest to satisfy his preferences. (Hollis 1987: 60) Let us take a woman who has just two preferences, which, for the sake of argument, we will assume are mutually exclusive: to have a child, which would bring her satisfaction measured as 100 ‘utiles’;1 and to have a job, which would carry 80 utiles. She also has perfect information about the costs associated with each preference. Let us suppose that these are only the following. To have a child involves a risk to health in pregnancy, which will cost −10 utiles. There are also the long-term costs of bringing up a child, which will cost −20 utiles. She also knows that being a mother carries low social status, which will cost her −20 utiles. On the other hand, getting a job will involve training costs of −10 utiles, the long-term costs to health of being employed are −10 utiles, but being employed will bring considerable social status of +15 utiles. So the correct calculation of the balance of utility is presented in Table 9.1. In these circumstances the rational choice is the job. The woman might want the baby more than the job, but taking into account the balance of preferences and costs the overall maximization of utility is gained from the job. Becker’s analysis of marriage decisions exemplifies this concept of rational decisionmaking. One of his ‘simple principles [at] the heart of the analysis’ is ‘persons marrying ... can be assumed to expect to raise their utility above what it would be were they to remain single’ (Becker 1973: 814; see also Becker 1996: 149). People thus make a decision whether to marry on the basis of a calculation of the relative utility to be gained

Table 9.1 Calculating a utility balance Utility Preference Information about costs and benefits



To have a child (a) Pregnancy

+100 −10

To have a job (a) Training

+80 −10

(b) Upbringing (c) Low status

−20 −20 +50

(b) Health risk (c) High status

−10 +15 +75

‘I do’ 149 from marriage as against remaining single, in terms of their given preferences. The same reasoning applies as to whom to marry. A woman’s decision to marry Joe rather than Fred is based on her calculating the relative utility to be gained from the two unions. So Becker (1991: 108) says: an efficient marriage market assigns imputed incomes or ‘prices’ to all participants that attracts them to suitable ... marriages. Imputed prices are also used to match men and women of different qualities: some participants, we have seen, choose to be matched with ‘inferior’ persons because they feel ‘superior’ persons are too expensive. Three features of rational choice theory should be drawn out for particular emphasis: 1.

The theory is strictly ‘methodologically individualistic’ (Smelser and Swedberg 1994: 5). The theory of course recognizes that individuals cooperate and conflict with each other, but claims these are explainable in terms of the rational choices of individuals. These include wishing to be with others (or not) as preferences and taking into account information relating to others. So, for example, with reference to marriage, people have preferences for the kind of person they may wish to marry, and they will have information about marital law and other people’s expectations of married people. But the theory claims that there should be no need to go beyond this to any appeal to ‘social relationships’, ‘structures’ or ‘institutions’ (Clark 1991; Delphy and Leonard 1992). Becker (1991: 15) states that ‘institutions are assumed to evolve in the direction dictated by rational individual responses to changed conditions’. Note that it is the ‘rational individual responses’ which ‘dictate’ the evolution of institutions. So when considering the relative decline of polygamy, Becker (1991: 81) puts aside the impact of ‘the spread of Christianity and the growth of women’s rights’ in favour of the relative gains of men and women from different marriage forms and how these have shifted toward monogamy.


The theory is amoral: it sees no need to appeal to moral categories in explaining action. This applies both at the individual and at the societal level. Descriptively, an individual may ‘believe in’ marriage, but this is not why people marry; the explanation comes from an overall analysis of rationality in which the moral attachment to marriage is only one item of preference. If it is outweighed by the costs of marriage, such as the experience of violence, then a married person will divorce. Again, descriptively, a culture may place great weight on the centrality of marriage, but this is the aggregate result of a myriad of individual choices, not an explanatory category for marital

150 Ian Procter patterns, other than as one item of information for individuals. This is nicely illustrated in Becker’s work by his comment on search behaviour in the marriage market: If [participants in marriage markets] could search as ‘cheaply’ for other mates when married as when single, and if marriages could be terminated without significant cost, they would marry the first reasonable mate encountered, knowing that they would gain from even a less-than-optimal marriage. They would then continue to search while married. (Becker 1991:325) Here there is no sense that courtship and marital behaviour are governed by social rules. They might be, but such rules are merely items of information for utility-maximizers rather than having some determinate effect in themselves. 3.

The theory is agnostic as regards the worth of different preferences. These are taken as given. If a woman wishes to marry in order to have a big wedding ceremony, or if a man wishes to marry in order to gain cheap domestic labour, these are neither silly nor wrong. They are as they are; rationality arises from the calculation of preferences and costs, not from the intrinsic worth of motives or preferences. Becker’s motivational agnosticism is illustrated by his approach to such qualities as love or altruism in marriage. Such qualities are simply part of a person’s utility function. A man loves a woman ‘if her welfare enters his utility function’. The man ‘can benefit from a match with [her], because he could then have a more favorable effect on her welfare – and thereby on his own utility’ (Becker 1991: 124). Altruism is defined in the same way (Becker 1991: 278). A person may approach the marital decision with preferences for love or sado-masochism, altruism or selfishness. They are simply preferences in the calculation of utility. One further aspect of rational choice theory needs comment. This is not a theory of how

people actually make choices. That would be a concern, for example, of the psychology of decision-making, what Simon (1987: 26) calls ‘procedural rationality’. Such a theory would have to take account of elements which are self-evidently missing from rational choice theory; an example would be the place of emotion in decision-making. Rational choice theory claims that in investigating patterns of social action, such as regularities in marital behaviour, a minimalist model of the human actor is all that is required. This will not give an accurate account of how particular individuals come to particular decisions, but will allow explanation of the aggregate pattern arising from large numbers of such decisions. On these grounds, rational choice theorists reject criticisms that their model of decision-making is ‘unrealistic’

‘I do’ 151 in being contrary to evidence from actual decision-making. In developing this claim, reference is often made to Friedman’s essay on ‘The methodology of positive economies’ (1966: Part 1; see, e.g., Frank 1997: 6; Hogarth and Reder 1987: 10). Friedman sees theory as hypotheses yielding predictions which must be tested. To do this involves building abstract models which make certain assumptions about the world, such as perfectly competitive markets or utility-maximising behaviour (Friedman 1966: 21). One confusion that has been particularly rife and has done much damage is confusion about the role of ‘assumptions’ in economic analysis. A meaningful scientific hypothesis or theory typically asserts that certain forces are, and other forces are not, important in understanding a particular class of phenomena. It is frequently convenient to present such a hypothesis by stating that the phenomena it is desired to predict behave in the world of observation as if they occurred in a hypothetical and highly simplified world containing only the forces that the hypothesis asserts to be important. (Friedman 1966: 40) As this quotation implies, Friedman (1966: 14, 31, 40) strongly rejects the idea that theories can be assessed in terms of the ‘realism’ of their assumptions rather than in terms of the validity of their predictions. This is because theories ‘are not intended to be descriptive; they are designed to isolate the features which are crucial to a particular problem’ (Friedman 1966: 36). The example is given of the expert billiard player (Friedman 1966: 21). In order to predict which shots will be chosen, the hypothesis is advanced that expertise will be displayed only by those players whose play obeys the laws of mathematics. Such a player acts as if s/he were fully cognizant with these laws, but it is highly unlikely that many expert billiard players do indeed know advanced mathematics. It is, however, important to note that billiard players could, in principle, base their play on mathematical knowledge. The theory, to repeat the phrasing above, isolates the features which are crucial to a particular problem. However, in other instances which Friedman discusses, this is not the case. He gives the example of the different density of leaves around a tree, denser on the southern than the northern side, hypothesizing that ‘the leaves are positioned as if each leaf deliberately sought to maximize the amount of sunlight it receives’ (Friedman 1966: 19). Friedman is clear that leaves do not self-consciously decide where to obtain most sunlight, but he denies that this is relevant, so long as the predictions generated by assuming they do are valid (Friedman 1966: 20). But this seems to contradict the idea that theories abstract the features of complex phenomena to isolate the key features at work. A practical knowledge of geometry is vital to the billiard player even if s/he is unaware of it; it is this which accounts for success in the game and which is highlighted in the assumptions of the theory. It is entirely different in the leaves example, where the assumptions cannot be the case. Assumptions have to be realistic, not in

152 Ian Procter the sense that they describe a particular reality, but in the sense that the simplified model could, in principle, work. This point is crucial to examining Becker’s theory. He does not claim that present and prospective family members take decisions solely on the basis of rational choice principles. But by assuming that they act as if they do, Becker claims that we abstract the recurrently key features of family behaviour.2 But for this to be the case, then this assumption about decisionmaking must be logically possible otherwise one is not abstracting what is essential but imagining the impossible, as in the leaves example above.

Becker’s theory of marital decision-making Having outlined Becker’s general approach, let us proceed to a more detailed examination of his analysis of marital decision-making. ‘Participants [in marriage markets] prefer to marry if, and only if, their utility from marriage exceeds their utility from remaining single’ (Becker 1991: 82). Becker notes that this claim only holds under certain limiting conditions. The first of these is ‘if the commodity outputs of households can be combined into a single homogeneous commodity, such as the quantity of children’ (Becker 1991: 82). This condition concerns people’s preferences and how they might estimate whether their utility from marriage might be better or worse than that from singledom. There must be some homogeneity in output which allows people to compare the different kinds of possible household in terms of their preferences. In the first instance (only, see later), Becker assumes a single criterion, using the example of the quantity of children. As we shall see in a moment, he considerably elaborates on this point. The second limiting condition is ‘if the output of marriages is known with certainty’ (Becker 1991: 82). This relates to the information available to people in making their decisions. Initially information is assumed to be perfect; again we will consider later Becker’s elaboration of this. The third condition is ‘if the output is distributed as income to mates’ (Becker 1991: 82). This relates to the distribution of the output of different possible households amongst their participants. It thus refers to social relationships between them. Such a ‘relationship’ has no existence in itself. It is only the result of the configuration of individuals’ rational choice-making processes. So, people marry (or not) on the basis of a calculation of utility. This theory holds under certain specified conditions of ceteris paribus. These identify where other factors might come in – people may be confused about their preferences, not have accurate information, or be subject to distributional rules based upon ‘force or fraud’ (Parsons 1968: 89–94) rather than the configuration of rationally deciding individuals. The chapter will proceed by considering how Becker elaborates the first two of these conditions. The third will not be further discussed because it relates to his approach to social relationships through his conception of the marriage market. This takes us rather away from actors’ decision-making processes and would introduce a host of further problems which lack

‘I do’ 153 of space prohibits considering here. Clearly, the first two conditions are highly improbable: people patently do not calculate utility simply in terms of the quantity of children, and they rarely, if ever, have perfect information. In the next section we will critically consider how Becker deals with decision-making under more elaborate ‘as if’ conditions.

Critique: comparing the utility of different marital statuses People can only rationally maximize utility if the output of different possible households, such as single-person and marital households, can be compared. In his first condition above, Becker approaches this problem by using the very simple assumption that just one criterion is invoked, such as the quantity of children. Thus, if we assume that only marital households produce children, then such households will generate more utility than single households. This is clearly a highly restrictive condition because Becker is well aware that households have many different kinds of outputs. Of course, the commodity output maximized by households is not to be identified with national output as usually measured, but includes the quantity and quality of children, sexual satisfaction, and other commodities that never enter into measures of national output. (Becker 1991: 112) So the question arises: how are these different commodities to be compared if there is no single over-riding criterion, as in the first limiting condition? Becker’s solution to this is indicated when he argues: In particular, [man A] and [woman B] would marry each other if their total output exceeded their combined aggregate incomes from marrying other persons or from remaining single, no matter how radically their preferences differed. (Becker 1991: 123) Remember that output and incomes here are not used in the conventional sense of moneymeasurable things such as bank savings or wages, but include non-marketed output such as children, sexual satisfaction or fidelity. So, in deciding whether to marry man A, woman B is assessing the total output of a union with A and comparing it with the output from unions with other men and with remaining single. Let us take a woman, Freda, and assume that she has three preferences: she wants to have a job; to maximize money income; and to have a baby. She is currently single and in a job which gives her a moderate income, but believes that if she remained single and had a baby then she could not continue with her current job. Freda also has two suitors, Joe and Fred. Joe

154 Ian Procter is rich but has strong views against working mothers. Fred is poor but is supportive of working mothers. Freda also has good information that both men are equally potent. So three scenarios are considered by Freda (as set out in Figure 9.1). Figure 9.1 Freda’s three scenarios Stay single

Marry Joe

Marry Fred

Full-time job No baby Moderate income

No job Baby High income

Part-time job Baby Low income

How does Freda choose? Well to begin with she has to order her preferences (see Figure 9.2). But this is to assume that preferences are all ordered with equal weight and overlooks utility as a balance of preference and cost in assuming that each preference carries a uniform cost. Insofar as having a job is first preference, then this may far outweigh the other possibilities. If income is first, this may be only marginally more important than the other two. Having a job and having a baby incur costs to health but in different ways and potentially to different degrees. An income generated by two people possibly involves the costs of arguing over its distribution as compared to a single income. So ordering preferences has to be matched by attributing costs to each preference. This can only be done if there is a common criterion for measuring preferences and costs. Let us suppose that there is such a thing. (The matrix of possibilities appears in Table 9.2.) Figure 9.2 Freda’s preference order and choice Job





then she stays single






then she stays single






then she marries Joe






then she marries Fred






then she marries Joe






then she marries Joe

What could this criterion possibly be? How can the costs and benefits of having a baby and a job be assigned quantitative values relating to their respective worth? Becker’s answer to this comes right at the beginning of his book. It is that time is such a criterion.3 [E]ach person allocates time as well as money income to different activities, receives income from time spent working in the marketplace, and receives utility from time spent

‘I do’ 155 eating, sleeping, watching television, gardening, and participating in many other activities. (Becker 1991: 21) As this implies, money and time are not two measures but one, as money is no longer ‘given’ but is determined by the allocation of time, inasmuch as earnings are determined by the time allocated to work. Therefore, goods and time-budget constraints are not independent and can be combined into one overall constraint. (Becker 1991: 22) Table 9.2 Freda’s matrix of possibilities Stay single Full-time job No baby Moderate income

50 0 40

health cost health bonus bonus for autonomy

−5 10 5 Total

45 10 45 100

Marry Joe No job Baby High income

0 50 60

health bonus health cost cost of dependency

10 −5 −10 Total

10 45 50 105

Marry Fred Part-time job Baby Low income

20 50 20

health cost health cost cost of shared income

−5 −5 −5 Total

15 45 15 75

From this proposition Becker goes on to develop the ‘shadow prices’ on non-market commodities, which depend on their ‘costs of production’ in terms of time spent. All of this depends on the idea that time can be a universal criterion to assess costs, including money, as a special form of representing costs, but also other activities which are not directly represented in money terms. So, let us consider Freda’s preference for a baby. What utility will she gain from either singledom or marriage? If she is married, then her spouse may share some of the time required to care for the child. Her spouse may also bring in income which will save some of Freda’s time. If she is single, then more of Freda’s time will have to be devoted to childcare, and she will have to find sufficient time to gain an income. In these terms Freda can compare utility in terms of time costs. Once again it bears repetition that Becker is not saying that this is how

156 Ian Procter actual people make decisions in all their complexity but that this is the essential feature of decision-making abstracted by the theory. But within this caveat, is it feasible? Is it parallel to the billiard player or the leaves on the tree? Scholars working in a number of disciplines have theorized ‘time’, noting its cultural and historical variability (Adam 1995; Moore 1963; Thompson 1974). While the same can be said of money (Chown 1994; Davies 1994), this does not exclude it from its use as a measure of cost. Both have become ‘disembedded’ from local contexts and form abstract systems. So Giddens (1990: 22) refers to money as a ‘symbolic token’ in modern society, and Nowotny (1994: Ch. 1) to ‘standardized time’ in the form of the calendar and clock. Are they then the same kind of abstraction which can act as a universal criterion in comparing relative utilities? Let us suppose that Freda is making a choice between A and B. Everything about the two is equal except that A costs £5 and B costs £1. Freda chooses B whatever A and B are. If it is something that Freda likes, then choosing B allows her to have more of it or something else she likes. If she does not like A and B, but has to choose one of them, then why waste scarce resources on what she does not like? Now let’s do the accounting not in pounds sterling but in hours. Again A and B are equal in all respects but A costs five hours and B costs one hour. Here it is impossible to make a, choice without knowing what A and B are. If it is eating black pudding (which Freda detests), then one hour is the choice. If it is sex or a holiday (which she enjoys), then five hours is the choice. Wherein lies the difference? Both money and time are symbols – pounds or dollars, hours and minutes. But money is only a symbol; it is worthless in itself. As a symbol system it can be manipulated, it can be saved, decreased or increased. Time is symbolized abstractly in hours and minutes, but it is more than a symbol system, it is the ‘flow of events’ (Elias 1992), which we grasp through our symbol system. This cannot be manipulated; stopping the clock does not stop the flow of events. When we ‘save’ time, it is in a different sense from accumulating money, because time is intrinsically limited – we all have just twenty-four hours in a day.4 Thus whilst money is worthless in itself, time is precious in itself. Thus if A costs £5, then the £5 is merely a neutral symbol; spending it gives us A, but the ‘spending’ is not part of the utility of A. If A costs five hours of time, then the time itself is worth something, either positively or negatively. Time cannot be only a neutral symbol. Let us suppose that A and B are forms of childcare that Freda is choosing between. In the first instance they are a day nursery costing £5 and a childminder costing £1. If everything else is equal, then she chooses the childminder. In the second instance, Freda is choosing between herself looking after the child, which will take five hours of her time, and her mother looking after the child, which will take one hour of Freda’s time. Again everything else is equal but we simply cannot say whether five or one is preferable because the time is not a neutral symbol but is worth something to Freda in itself. If she values looking after the child, then five is chosen; if not, then one is chosen.

‘I do’ 157 In Becker’s theory, time acts as a uniform basis for comparing the utility to be gained from possible marital choices. It is intended to provide the solution to the problem that many marital benefits and costs do not correspond to market goods with prices which allow such comparisons. But we have seen that time is not a standard measure which allows such comparisons. As Glucksmann (1998: 244) puts it: It is possible ... for time-use studies and time-budgets ... to operate with a simplistic notion of measurement in terms of standard units, whereby one hour of assembly-line working, for example, is equated with one hour of, say, walking the dog, as if this conversion were unproblematic. But the expenditure of time in different economic spheres or social relationships may well be incommensurable. We can thus conclude that it is not possible for people to choose whether to marry or not and whom to marry on the basis of comparisons of the relative utility to be gained. Money cannot act as a measure of the relative utility of marital costs and benefits, as Becker well knows, but neither can time, his proposed alternative. But, as Becker (1996: 149) might remind us, people do say that ‘I will be better off married than single’ or ‘I can’t afford to have a baby’. Note that these statements include a consideration of monetary costs and benefits but they are not restricted to these. To be unable to afford a baby includes an estimation of the cost of a child, of loss of income, but also of the standard of provision and care felt necessary for the child and the emotional cost of reduced living standards to the parent. In other words, these judgements of the relative worth of different possibilities are in terms not of utility but of normative criteria (Duncan and Edwards 1999; Himmelwelt 1999). Jobs, babies and incomes do not carry more or less utility but are more or less important to a person in terms of their normative standards of what things are worth. Thus if our Freda’s moral commitment is to motherhood, then no amount of utility might persuade her to have a job. On the other hand, if she is committed to both motherhood and a job, then all sorts of compromises might be made to accomplish the two. This may involve all sorts of disutilities, such as being constantly stressed, but it is ‘worthwhile’. To summarize, Becker states that people decide to marry on. the basis of a comparison of the utility to be gained from remaining single or marrying. He recognizes that this implies a criterion whereby the various outputs of both, including those to which a monetary value cannot be attributed, can be compared. He claims that time is such a criterion. This must be doubted because time is not a neutral measure but is itself of worth. People no doubt do compare what they gain and lose from marriage, but the comparison is not only in terms of quantifiable costs and benefits but also in terms of their criteria of what is important to them. As in other versions of rational choice theory (Holton 1995), Becker’s amorality cannot be sustained.

158 Ian Procter

Critique: information and preferences Turning now to Becker’s second condition relating to information. The initial assumption that people have perfect information is again clearly restrictive. Becker (1991: Ch. 10) goes on to modify the assumption and in fact to make a virtue out of this for the applicability of his theory in his consideration of divorce patterns. Here he is clear that prospective marriage partners do not have perfect information because of ‘limited information about the traits of [potential] mates’ (Becker 1991: 325). Information can be improved by both ‘extensive’ searching for additional possible mates and ‘intensive’ investigation of those considered. Nevertheless, gathering accurate information about relevant issues to do with temperament and future partnership is difficult, and people have to use proxies such as family background to indicate honesty and amiability (Becker 1991: 326). However, both forms of search involve costs to the searcher. Thus a rational person would search until there was a balance between the marginal utility to be gained from marriage to an already known prospect and the marginal cost of further search. In particular, rational persons marry even when certain of eventually finding better prospects with additional search, for the cost of additional search exceeds the expected benefits from better prospects. (Becker 1991: 325) The fact that decisions to marry are made on the basis of imperfect information naturally carries with it the possibility that people will make mistakes. If they are then allowed to rectify mistakes, many marriages will lead to divorce. Becker (1991: 327–9) claims that the pattern of divorce bolsters his theory of marriage. There are three arguments here: 1.

A high proportion of divorces occur in the first few years of marriage. The implication is that the imperfections of earlier information quickly become manifest.


The grounds for divorce, when this occurs soon after marriage, develop the first point. They tend to be ‘“difficult” spouses and value conflicts’ (Becker 1991: 328). These are qualities of the other individual which are more difficult to assess prior to marriage but which quickly become clear in marriage. On the other hand, issues which are easier to assess before marriage, such as physical appearance or education, tend not to be grounds for divorce because the pre-marital partner had good information about them.


Looking at divorces which occur later in marriages, the reasons for divorce are to do not with corrected information but with changed preferences. For example, they include adultery and financial conflict, which Becker sees as changes in preferences for partners and lifestyle.

‘I do’ 159 But is this just a matter of (i) poor information about the individual traits of potential mates whilst (ii) preferences remain the same in the short term? An alternative reading of high rates of divorce soon after marriage is as follows: (a)

The decision to marry (and to divorce) is not only about judging the traits of individuals, it is also about the traits of a relationship, how the two individuals will blend as marital partners.


This, then, cannot be just about poor information as compared to an ideal of perfect information because there cannot be perfect information about just how Freda and Fred will blend. Only through experiencing marriage can they come to know this information; all other information is about how other people have blended, not the specifics of Freda and Fred. As Granovetter (1985: 486) comments on the application of the model of Homo economicus to non-economic contexts: The interpersonal ties described in these arguments are extremely stylized, average, ‘typical’ – devoid of specific content, history or structural location. Actors’ behavior results from their named role positions ... [one example given is husbands and wives] ... but these relations are not assumed to have individualized content beyond that given by the named roles.


But in gaining this information, Freda and Fred may well change; the information they generate may change their preferences for marriage. Support for this view comes from the pattern of initiation of divorce. Becker is correct to

say that a high proportion of divorces occur early in marriages, but he only hints at another feature of this – that a high proportion of these are initiated by women. In Britain two and a half times more divorces are granted to women as compared to men (Newman and Smith 1997: 17). This may be because their experience of marriage is more dramatically different from singlehood than that of men, and their preference change is more dramatically affected by marriage. In research conducted by Maureen Padfield and myself (Procter and Padfield 1998), we asked two women on the verge of marriage how their career-minded ideas had changed since their marriage decision. Diana,5 a business manager, said: It’s so different. I have somebody else to consider who comes first in everything now. My whole outlook on life has changed, my future has changed. In what appears to be such a short space of time, ... my position has changed so radically, urrmmm ... I think yeah it’s affected everything really, the way I feel about everything. (Procter and Padfield 1998: 230–1)

160 Ian Procter Carol, a secretary, had a similar presentiment of the significance of her coming marriage: My whole general outlook about what my future offers now. Whereas before it was what I would like of the future, now I always envisaged that I would be married, but who to and what to and when ... I never really knew. But now I’ve actually got that. Come August next year I will be married and then, when we have children I will be a mum and I’ll have kids to look after and ... (Procter and Padfield 1998: 233–4) Under this alternative, the divorce pattern does not back up the original model of decisionmaking. Rather, it suggests that prior to marriage people cannot know that it is preferable to singlehood on the basis of a calculation of utility, as the only way to know is to do. Further, it suggests that there are not stable preferences plus imperfect information. Rather, the coming to know changes the preferences themselves. ‘Difficult spouses’ and ‘value conflicts’ are not just traits of individuals not easily detected prior to marriage. For Freda, the difficulties are not just about ‘his habits’ but about the relationship between Freda and Fred. The conflicts in value are not about individual traits somehow not known before marriage, but about the values applicable to this marriage. Thus in Mansfield and Collard’s (1988) study of newly wed couples, it is noted that the spouses brought with them conceptions of what was involved in being a ‘husband’ or ‘wife’. One source of these was the roles played by their own parents: And sometimes the novice wife did not compare very well with her husband’s mother, as Mr Monkhouse explained: Do you think their marriage influenced your attitude to marriage at all? Well, in a way I suppose – because you expect your wife to be the same as your mother. Or to do the same sort of things. But you have a rude awakening. (Mansfield and Collard 1988: 127) Here Mr Monkhouse’s ‘rude awakening’ was not simply because he lacked information about his wife’s behaviour, but because he and his wife had to work out their relationship from the differing preconceptions they brought to the marriage: ‘for the majority of couples (who had not cohabited), setting up home involved the merging of differing approaches to housekeeping and different standards about tidiness and cleanliness’ (Mansfield and Collard 1988: 128).

‘I do’ 161 Insofar as information impacts upon preferences in marriage, then this may occur prior to marriage, in the ‘search’ process. Becker portrays individuals surveying the prospects and calculating their relative utility but he does not consider that this process itself affects who people are and what their preferences might be. An illustration of this comes from Holland and Eisenhart’s (1990) analysis of the peer system in their study of two American universities. This is a sharp example of Becker’s (1991: 81) ‘marriage market’, or, as Holland and Eisenhart (1990: 212) call it, the ‘sexual auction block’. Within the student peer group, men and women are ranked by the participants in terms of their ‘attractiveness’ and are socially placed in terms of their position. However, attractiveness has two aspects: ‘both as an essence of a person and as an effect on another person’ (Holland and Eisenhart 1990: 102). The first is Becker’s individual traits, a person’s appearance and demeanour. The second arises from a person’s relationship with another. If a woman enters a relationship with a man who is deemed in the peer group to be attractive, and he treats her well, then she becomes attractive. Alternatively, if a woman dates an unattractive man, then her attractiveness diminishes. Or if she ‘catches’ an attractive man but is treated badly, then she loses attractiveness. In each case this is not simply a matter of her collecting information, but the action of doing this involves interaction in a social system which impacts upon who she is and what her preferences and utility may be. As Smelser and Swedberg (1994: 5) put it: ‘[Microeconomics] assumes that actors are not connected to one another; [economic sociology] assumes that actors are linked and influenced by others.’ In summary, Becker claims that people choose to marry if they expect increased utility on the basis of the information available to them. This is inadequate, in two ways. Firstly, this is not simply a matter of degrees of perfection of available information. For some information about a future marriage is, in principle, not available because it is about a particular future marriage and can only be gained when that marriage is created through the relationship of the spouses-to-be. Secondly, the gaining of this information, whether in the ‘search’ process for a spouse or in a marriage itself, shapes people’s preferences, their concepts of who they are. In Hollis’s (1987: 118–119) striking example: The young Catholic who needs to know more what it would be like to become a priest cannot find out in the public library. To acquire and understand the information he must invest enough of himself to change his life, even if he finally decides against the priesthood.

Further development of Becker’s theory In 1996 Becker published his book Accounting for Tastes. Here he claims to take on board past experiences, future anticipations and social influences. This potentially deals with the

162 Ian Procter issues raised above. His general strategy when tackling a new problem is to include new items in the actor’s utility function. Thus, he tackled the issue of racism by incorporating ‘tastes for discrimination’ (Becker 1996: 141, 155). When approaching the question of altruism, the individual’s preference for seeking the welfare of others is included in the overall listing of preferences. In Accounting for Tastes two further items are included – personal and social capital. The first is the actor’s experience, the second the influence of others. This book retains the assumption that individuals behave so as to maximize utility while extending the definition of individual preferences to include personal habits and addictions, peer pressure, parental influences on the tastes of children, advertising, love and sympathy, and other neglected behavior. (Becker 1996: 4) Becker thus refers to the ‘extended utility function’, which he contrasts with the ‘subutility function’, which economists usually consider, including only preferences for goods rather than his additional inclusion of personal and social capital (Becker 1996: 4-9). We thus have to inquire just what it means to extend the utility function to include ‘capital’ as well as our preferences for goods. Two different nuances of Becker’s discussion can be identified. In the first, capital seems to be used to ‘account’ for tastes or preferences. Here past experience and social influence indicate why actors have the preferences they have. Becker’s (1996: 8–9) examples of habit formation, the influence of parents on children, or advertisers on consumers (Becker 1996: 10), and the general impact of ‘culture’ (Becker 1996: 16–17) suggest this usage. In the example of marriage, Freda is considering whether to marry, and whether to marry Fred. She has a preference for having a baby and calculates whether to marry (Fred) in terms of the benefits and costs in satisfying her desire for a baby. Where does capital come in? In this usage it is as an explanation of her preferences, why she wants a baby. We might understand this by knowing about her experience of family life and/or the impact of Freda’s peer group. Whilst Becker sometimes gives the impression of this meaning of capital, it cannot be his intention. For a moment’s reflection indicates that this fundamentally undermines the concept of Freda as a rational actor. For if we explain Freda’s behaviour in terms of her socialization or peer group influence, then we effectively discount her own calculation of utility or, more widely, her own reasons for marriage. Put another way, accounting for Freda’s tastes by explaining their causes does not add to the rational choice model; it only undermines that model. So we must move to a second sense of incorporating capital into the utility function. This focuses not on how an observer, such as Becker, accounts for taste, but on how the actor herself takes account of past experience, social influences and future expectations. This can

‘I do’ 163 be explored by considering what Becker says about the past and the future. With reference to the past, we now imagine Freda considering her preference for a baby and coming to understand the influence of her family on this. As Becker (1996: 9) says, this implies some degree of freedom for Freda: ‘individuals, in effect help to choose their own preferences’. Once Freda realizes that it is the influence of her family which has shaped her desire to have a baby, she can think of the alternative of not having a baby. Becker’s (1996: 9) example is as follows: [A] woman who fears and loathes men, perhaps because she was sexually abused as a child, may try to change her attitudes towards men by undergoing psychotherapy treatment and by taking other actions, or she may decide to accept these feelings and seek relations only with other women. Here, as well as the outsider accounting for the woman’s taste (explanation in terms of sexual abuse), the woman herself comes to realize her own capital endowment and to consider alternatives. The key point here is that this seems to push the decision-making process back a stage. Rather than actors having given preferences, they can reflect on their preferences and decide which they want. In Becker’s example, the woman can decide whether to go for therapy or to accept her existing capital stock and its preferences from the past. For Freda, the question is does she maintain her preference for a baby or decide to change it? So far we have considered actors reflecting on their past capital endowment, but anticipations of the future are also included. Current choices are made partly with an eye to their influence on current capital stocks, and hence on future utilities and choices. ... Mary may choose to date Tom rather than Bill – even though Bill is handsomer and smarter – because she believes Tom has a better character and will make a better husband if they married later on. (Becker 1996: 6) Once again, this implies a decision-making process in choosing between preferences. Does Mary here value the gratification to be had from a relationship now with the handsome and smart Bill more than or less than future utility from a possible marriage to Tom? Note that there is a choice of preferences here. This is not a calculation and comparison of utility given stable preferences. This would mean that Mary had a preference, say for marriage, and was estimating the utility to be gained from relationships with Bill and Tom in terms of that preference. In Becker’s example, the inclusion of the impact of a choice on future capital means that she is choosing between a preference for the vicarious pleasure of a relationship

164 Ian Procter with Bill and a preference for marriage with Tom. As he himself says, the introduction of ‘capital’ into the preference listing means that he sees ‘people mak[ing] fundamental choices about overall lifestyles’ (Becker 1996: 17). Becker is aware that he has made a significant move here. What he has done is to break with what were termed earlier in this chapter the ‘amorality’ of his position and his ‘agnosticism’ about preferences. Moral rules are no longer simply items of information which the actor takes account of but are included, as preferences, in the extended utility function (Becker 1996: 17–18). People may, for example, have a preference to act honestly. This may be an endowment of their personal capital or may be considered as future capital, say in effecting their future reputation. However, Becker (1996: 18) also says that ethics and culture affect behavior in the same general way as do other determinants of utility and preferences. In particular, considerations of price and cost influence moral choices – such as whether to act honestly – just as they influence choices of personal goods. This is not a persuasive claim. Yes, matters of price and cost do influence ethical and moral choices, but not alone. If Freda is considering marriage, she might consider the utility to be gained, she might have an understanding of the influence of her religious upbringing on her capital stock, she might wish to maintain her capital stock amongst her religious co-believers now and in the future, but the fact remains that belief in, or commitment to, marriage (or not) is a part of her decision-making process distinct from all these other calculations. This is what Weber (1964 [1947]: 115–18) called ‘substantive rationality’, and it remains distinct from instrumental rationality even when Becker has extended the utility function to include what he calls personal and social capital. His use of the following example illustrates this: A young man may drink heavily because he does not anticipate that he will become addicted to alcohol. ... Of course, if he turns out to be wrong and he does become addicted later on in life, he would wish he had not drunk so much as a young man. He might decide to fight his addiction by joining Alcoholics Anonymous and in other ways; on the other hand, continuing to drink heavily could be a way of maximizing utility if his preferences ‘shifted’ greatly in favor of alcohol. (Becker 1996: 9) The last clause is crucial. How does the man choose between whether his preference is for continuing use of alcohol or abstinence? Not only by a calculation of utility, but also, to use Becker’s own phrase, by a fundamental choice about ‘overall lifestyles’.

‘I do’ 165

Conclusion This chapter has given close attention to Becker’s approach to how people decide whether or not to marry. For him this decision is not exceptional but typical of how people make decisions, not only about family matters but in any human action. The failings detected here are thus indicative of wider flaws in his general rational choice theory. Becker’s basic idea is that, in considering marriage, people compare different possibilities (staying single or marriage; marriage to Fred or Joe) in terms of a comparison of the utility to be gained. Utility involves the calculation of costs and benefits not only of money-valued goods but also of other outputs of households such as children and sexual satisfaction. Rational choice theory claims to be not a descriptive account of concrete decision-making but an abstraction of the key elements in this process which are necessary to arrive at explanations of social phenomena such as marriage, divorce and many other recurrent forms of social behaviour. A crucial part of my argument is that the elements abstracted must be theoretically coherent, not in the sense that they reflect the full complexity of decision-making, but in the narrower sense that it is conceptually possible for the abstraction to work. My claim is that this is not the case in Becker’s theory. For people to compare utility, there must be some ‘exchange mechanism’6 whereby people can compare diverse outputs (gains/ losses of income, social status, sexual satisfaction ...) from diverse possible households (single, married, different marriage partners ...). Becker claims that time is such a common mechanism, but this presupposes that time is a symbol system worthless in itself, akin to money. But time differs from money in that it is intrinsically scarce and thus valuable in itself and cannot act as a neutral measure of utility. My second criticism picks up Becker’s treatment of how people process information. He assumes that information about future states can be known and does not interact with people’s preferences in the short run. This chapter argues that crucial information about future possible marriages cannot be known because it depends upon the relationship developed once the marriage is made, and that this process of relationship development involves the information that people gain interacting with their preferences. Most of this discussion considers Becker’s arguments as presented in A Treatise on the Family. Becker has subsequently published Accounting for Tastes, which takes up some of the issues raised here through the concepts of personal and social capital being included in the actor’s utility function. This is found not to be adequate as Becker moves from calculating utility given consistent preferences, to choosing between preferences, without recognizing the fundamental difference between the two. He does not show how people can make a rational choice between, for example, wanting to be single and wanting to be married as ends in themselves. The critical analysis developed here has not intended to generate an alternative, but two points are emergent which are consistent with other chapters in this book. The first is the place

166 Ian Procter of norms in defining ‘rationality’ in a wider sense than their status as items of information. Norms define what is worthwhile and reasonable. Secondly, preferences are not simply lists of (un)desirables but must be viewed more widely as concepts of self-definition of the human actor. Acting rationally involves instrumental calculations, adherence to socially negotiated norms and acting coherently in terms of who someone is.

Notes 1

A utile is a hypothetical measure of the positive or negative gain arising from an action. It implies that there is one consistent criterion which can measure preferences and costs. Whether this can conceivably be the case is taken up later.


For examples of Becker using the ‘as if’ formulation, see Becker (1974: S16; 1996: 193).


It is, of course, not only Becker who uses this concept. See Cherry (1998).


Or whatever units one uses to measure the flow of events. Becker (1996: 140) recognizes the scarcity of time, but to him this is simply a constraint, not the basis of the value of time.


The names of the two women are pseudonyms.


Thanks to Richard Lampard for this phrase.

10 Decision-making as a process over time The careers of home-located cultural workers Carol Wolkowitz

Much of the debate about home-located work has revolved around the issue of the degree of choice people working at home have been able to exercise over their place of employment. Some have argued that their choices are explicable in terms of a clear preference between narrowly conceived goals. Catherine Hakim (1991, 1996), for instance, sees women as working at home because they are oriented around their families rather than paid employment, a view which is entirely congruent with rational choice theory. Others do not see those opting for home-based work as without agency, but stress rather the ideological and practical constraints on decision-making (Allen and Wolkowitz 1987; Huws et al. 1996; Phizacklea and Wolkowitz 1995). Such a view is not necessarily incongruent with rational choice theory either, but only insofar as constraints are seen as merely the opportunity costs which individual preferences confront. A volume debating the adequacy of rational choice theory provides an opportunity to focus on the complexity of the decisions which have been made by some people who work at or from home. What this chapter fundamentally challenges is the notion of stable preferences themselves. At least in its original versions, rational choice theory rests on an economic model of consumption which envisions actors acting as individuals choosing between alternatives defined in terms of utilitarian values (Heap et al. 1992; Zafirovski 1999a; see also Procter, this volume). Given the same preference ranking, any rational individual would make the same choice. Instead, the data presented here show preference formation and reconsideration to be a trajectory over time. Far from being fixed, preferences are malleable, a key factor in their reshaping being the experience gained from a previous decision. Moreover, there is considerable evidence that decisions are matters of collective household deliberation, rather than being simple expressions of individual preferences. To set decision-making within the context of the needs and desires of other family members is particularly useful for illuminating men’s decisions, since the usual assumption

168 Carol Wolkowitz is that to explain men’s employment decisions it is sufficient to consider only the work-related aspects of the jobs to which they have access (Feldberg and Glenn 1979). Indeed, it could be argued that rational choice theory itself is constructed in line with an implicitly gendered model of decision-making, not only because of the conventional association of rationality with men and emotion with women (but see Williams, this volume), but also because, historically, the acceptance of unfettered utility maximization as the basis of men’s decisions in the public sphere was accompanied by the invention of the wifely ‘angel in the house’, whose actions were guided by a morality of self-sacrifice (Trudgill 1976). Moreover, in this division of labour, men became free to operate as careerist maximizers because their wives bore the brunt of life’s other responsibilities. Those like Hakim present women’s domestic preferences and activities as a voluntary goodness of fit, as against seeing them in the context of a division of labour which, because it is located within a particular material and ideological framework, cannot be construed in individualist terms. While rational choice theory challenges the assumption that decisions taken in relation to domestic and public considerations are based on different rationales, it does so not by challenging this male bias but by assuming its universality. Indeed, Hakim (1995) seems to take it as axiomatic that women should only be recognized as committed workers if their employment patterns (and employment-related decisions) mirror men’s. Certainly, current theories of decision-making would look very different if they had taken women’s enmeshment in caring relationships as their model, rather than first developing a model based on economic rationality and then applying it to the family and personal relationships. Unfortunately, the data I will be examining cannot support the development of an ambitious argument about the nature of decision-making (or its relation to gender), but they do suggest the importance of widening our frame of analysis beyond homo economicus. After briefly describing the research design, the chapter is organized around the trajectory of the careers of one group of home-located workers, as this best indicates shifts in their preferences over time, especially in relation to their reactions to experiences. It challenges the model of the lone maximizer by showing the dependence of individuals, pursuing a certain hierarchy of preferences, upon collaborative decision-making within the family unit. It also suggests that recognizing the agency of these and other actors is not (as Hakim’s use of rational choice theory, for instance, sometimes seems to imply) incompatible with recognizing the embeddedness of the family–employment nexus in changing employment structures.

Cultural workers The informants whose decisions I am examining are twenty-one cultural workers interviewed in 1994 in the course of a small-scale study of home-located work in North London. The two-

Decision-making as a process over time 169 to three-hour long interviews formed part of a cross-class study, initially directed towards comparing men’s and women’s experiences. In order to include both men and women and a range of social class and ethnic identities, home-located work was defined very broadly, so as to include those doing paid work at or from home for a substantial portion of the week.1 It focused on three contiguous, socially heterogeneous neighbourhoods, using the parents of children at one state primary school as the starting-point for snowball sampling. Although the idea for a locality-based study originated as a way of obtaining access to workers otherwise hard to identify, place may be a significant variable in its own right. Urban areas in the process of gentrification, like this one, have been seen as the heart of new middle-class formation, urban cultural regeneration and postmodern culture (see Ley 1997 among others). Although the concept of ‘cultural industries’ has its origins in Frankfurt School writings, terms like ‘cultural industries’ and ‘cultural workers’ have been circulating more recently as part of attempts to document the economic significance of production in the arts, performance, media and design and to foster government investment in it (Roodhouse 2000). O’Brien and Feist’s (1995) analysis of 1991 Census data found cultural occupations to be a growing sector, particularly but not only in the London labour market. Among those with cultural occupations in the cultural sector, the increase in the number of women and of the self-employed was particularly pronounced. Indeed, wider sociological interest in cultural workers in the cultural industries has arisen partly because they are believed to be at the forefront of shifts related to the development of a post-industrial economy, especially the associated breakdown of long-term employment security (Lash and Urry 1994; Lury 1993). Of the professional and managerial home-located workers identified by my small-scale survey, nearly all were engaged in cultural occupations, as defined by O’Brien and Feist (1995), and twenty-one of these twenty-six individuals were interviewed. Their economic and cultural capital gave them rather more scope for identifying and deciding between alternatives than was the case for the manual workers I interviewed, or for many workers interviewed in other studies of home-located work (Felstead and Jewson 2000; Grainger et al. 1995). The occupations of the nine women interviewees comprised: television producer, architect, publishers’ publicity agent, freelance journalist, freelance editor, graphic designer, singer, sub-editor and a director of fringe theatre. The twelve male interviewees included a graphic designer who employed two workers in his own home, the co-owner of a piano renovation business (with employees) which had originally been started by a husband and wife in the shop below their flat, a TV producer, an administrator for a national concert orchestra, a film director, a folk music promoter, two freelance journalists, a photographer and a part-time clothing designer. (I have altered the descriptions of some of my informants’ occupations very slightly, to avoid identifying the individuals involved.) The sample also included two men whose main earnings came from jobs outside the home, but who spent one

170 Carol Wolkowitz or more weekdays every week at home on creative work which they hoped to sell.2 Apart from one Black African man and one Asian man, the interviewees identified themselves as White.

The initial decision: individually made or householdbased? Rational choice theory assumes purely individual preferences, and needs to do so because goals have to be fixed as internal to each agent, otherwise expected utilities cannot be calculated. These data question the individualization of the decision to start working from home, because the needs of others are not just externalities which are taken into account, but instead form an active part of collaborative decision-making. Although my informants’ decisions were formulated in the context of shifting opportunity costs, as reflexively acknowledged, nevertheless this is only part of the story, the part which can be told in standard individualistic terms. My informants describe their initial decision to work from or at home partly in terms of their wish, as individuals, to start or continue careers which were seen as a source of selfidentity as well as income. Some began businesses in their own homes, or had moved their businesses into the home during a lean period. Others had left jobs as employees to work at home as freelances. There were sometimes precipitating circumstances in their previous employment which had led them to leave, especially difficult interpersonal relations and/or deteriorating terms and conditions of employment, but some informants were not able to pinpoint (or did not wish to reveal) particular circumstances of this kind. A few did not perceive themselves as having taken a decision to work from home as such, since home-based freelance work has always been (or has recently become) the norm in their chosen profession, thus indicating the importance of temporality in the very definition of a decision-making act. So far, this is compatible with a rational choice theory account dealing with the exercise of individual preferences in the context of variations in opportunity costs. Most informants were aware of changing patterns of employment in the cultural industries. The number of people who must put together their own ‘portfolios’ (Cohen and Mallom 1999) of remunerated work, or whose enterprises depend on a succession of relatively short-term contracts, seems to be growing, especially in the cultural industries (Lash and Urry 1994). In an analysis of employment in cinema production, for instance, Lury (1993: 66) suggests that long-term arrangements ‘in which once-and-for-all contracts were coupled to sustained role commitments’ have been ‘largely replaced by hiring on a pet-project basis’. My informants see themselves as making active decisions to cope with this changed environment, and they are well aware that their own decisions are part of a growing trend. According to a film director and screenwriter, with one commercial success so far:

Decision-making as a process over time 171 In the past, before the 1970s, directors used to be given work to do by a studio, so they could just move on to the next thing. Now the whole idea is different – the director has to find projects and spend time developing them, particularly in England. He thought it normal for this kind of work to be done at home, even though in many cases there would be a development fee of two months’ salary in advance, which would be subtracted from one’s eventual salary if the project went through. People with gaps between projects find that maintaining an office or studio has become uneconomic, especially when they can rent studio space temporarily or meet various production needs in other ways. Between production contracts, much of their time is spent on communication (on the phone, fax, formal meetings and professional socializing), reading scripts, or other activities involved in trying to get a ‘speculative project’ started for which a home office is sufficient. As one TV producer says, nowadays ‘all you really need is a place where bills come in ... [because] a sizeable set of film and TV production is organized around key individuals’. Rather than working at home representing either the push of unemployment or the pull of entrepreneurial ambitions, which, as Grainger et al. (1995) comment, is the usual assumption, for many in my sample working at home reflects a process whereby the externalization of production means that ‘a creative career has to be built out of a succession of temporary projects’ (Lury 1993: 67).3 Not only were informants’ attitudes towards these changes influenced by their age and previous work experience, younger workers had the advantage of having received training in dealing with a new terrain. As the film-maker says, In film school they encourage you to look after yourself. You can’t wait for something to happen. Even at film school you have to raise money for your films, you weren’t helped. If you can’t do it in film school, you can’t expect to walk into a job now. No one from my year in film school has walked into a sure job. ... Older people have an idea about what should be happening – younger people are steamrollering over the top. These younger workers see shifts in the organization of cultural careers as also having widened access to them. A single male journalist, also aged 35, says that the changes in technology and organization ‘have had brutal consequences but it has given [me] an opportunity to enter journalism which I wouldn’t have had fifteen years ago. ... Technology has broken up the traditional fetters of employment.’ Although he started off editing reference books for a large publisher, he left to freelance from home in his own field, continuing to edit a series for the publisher, but then adding on more work of his choice until ‘the balance changed’.

172 Carol Wolkowitz Older people are more cynical: less concerned about accessing cultural occupations than worried about staying in them, as well as about the personal and wider costs of job insecurity. A long-established 49-year-old producer of documentaries says that the way broadcasting is now organized is ‘wearing’: ‘I feel at my age, will I be happy doing this in ten years? It’s a young person’s game to some extent ...’ A woman journalist of 40 sees the costs of increasing reliance on freelance work in more general terms. Since leaving a full-time, in-house job, her scope for developing investigative articles and campaigns has declined. At the time of the interview, her work consisted of articles of 1000 words or less, often less than 500, each of which had to be individually placed with editors she knew. She sees the increase in freelancing as representing a shift in power between media owners and journalists: [It is] a bad thing on the whole. It could affect content, it gives more power to the owners if you sell your wares without a contract. ... Especially if you feel vulnerable and do things you wouldn’t do otherwise. However, here we reach the limits of a model in which the lone decision-maker confronts the market and its costs. Instead, decisions were usually also taken in relation to household responsibilities, as members of couples and families. They were not atomistic individuals but concerned about a range of needs and wishes of other family members. Although there was one case in which a woman said that she had always preferred working from home (a partnered woman without children who transferred her public relations agency to her home as a way of avoiding literary intrigues), both men and women cited domestic considerations. As has been found in most other studies of both white-collar and manual home-based work, children were cited as a factor by women more frequently than they were by men. However, in this sample the proportion of women and men mentioning domestic considerations was more evenly balanced than has been reported in other studies. Six of the nine women, that is, all the women with children, mentioned family factors as one of their reasons for beginning to work from or at home. Although one of the men with children did not see caring responsibilities as a factor, two of the seven fathers mentioned caring for children as playing a part in their initial decision to work from home, and four more said that it was one of the advantages or a reason why they had continued to work at home for so long. This is the first intimation of the impact of learning from experience on preference re-formation. Informants’ comments also raise other questions regarding the adequacy of rational choice theory, particularly its foundation in economic decision-making. In my sample both women and men could be seen to be doing what in rational choice theory is termed ‘satisficing’, although the decision to combine parenting responsibilities with employment is usually viewed as more typical of women’s employment decisions than men’s (Crompton and

Decision-making as a process over time 173 Harris 1998a, 1998b). In the version of rational choice theory developed by Simon (1978), satisficing (as against optimizing) behaviour develops as a way of dealing with uncertainty; actors hedge their bets because information about alternatives is incomplete and expensive to obtain. However, I would argue that the ‘satisficing’ behaviour of many women challenges rational choice theory, insofar as uncertainty is not its main explanation. As Chafetz and Hagan (1996) point out, many women ‘satisfice’ because of a positive desire to obtain the rewards of both career and parenting. Another way of saying this is that they feel that the very different intrinsic rewards that parenting and employment offer are not interchangeable (they cannot be reduced to a common denominator, whether time, money or the portmanteau term ‘utility’). Increasingly, they therefore refuse to choose between them or to accept that they have to evaluate them in terms of a hierarchical preference schedule. Indeed, one suspects that one reason that committed employment was long equated with (or thought to presume) an unambiguous and stable preference hierarchy was because so long as wives assumed responsibility for raising the children, men did not have to see family and career as alternatives or to have to choose between them. One might impute a decision to them after the event, but in fact none had to be taken. The fact that traditionally men have not been expected to adjust their employment behaviour in order to care for children means that although both male and female informants take the decision to begin or continue working from home partly in order to meet their children’s needs, their decisions are not strictly speaking the same. Most of the women in my sample saw working at home as an option which enabled them to meet existing obligations, or, as one woman put it, ‘to fulfil my criteria’. Her decision to work at home is a good example of the kinds of considerations which women have to balance. This divorced journalist, the managing editor for eight years of a widely read magazine, decided to freelance from home when work relations in her office deteriorated but she was unable to find another job which gave her the same flexibility: I wanted to leave, I had been looking [for another job]. ... But it is hard to fulfil your criteria in relation to family needs. That’s why I stayed there for so long, there was no problem about putting your child first so long as I did my work. You could take your child to work [e.g. during sudden temporary school closures]. Initially there were women working there. ... When there was a new editor, there were games. ... Before, the magazine was controlled by journalists, then there was a take-over by a group of male engineers. ... I asked for voluntary redundancy when I clashed with a particular engineer who wanted control over my position.

174 Carol Wolkowitz Although this example is compatible with a view of the actor satisficing between two apparently incompatible aims, the woman herself suggests that the incompatibility was itself due partly to the gender politics of her workplace, especially its willingness to recognize workers’ other obligations. Men would not normally be expected to alter their place of work to look after their children, suggesting that when they in fact do so they have in a sense ‘chosen to choose’. Their decisions help to create new forms of familial co-operation, based on concern for their wives as much as for their children. When men with children took day-to-day responsibility for caring for their children, they usually assumed obligations for which their wives had previously taken responsibility, in order to enable these women to advance their careers. The clearest example was an administrator who sourced musical texts for a national orchestra. He had made an arrangement to work at home in the afternoons so that he could collect his son from school, something which permitted his partner to accept a demanding appointment as a secondary school headteacher. In addition, he had already written two books on composers, so knew that he enjoyed working at home. Another example was a home-based worker whose wife was able to travel for work because he was available in case of emergency during the day, when the child was cared for by a nanny, and could continue his paid work in the evenings while looking after the child. In other cases men at home for part of the week, trying to develop careers as creative writers or artists, took over responsibility for the household on the days they were at home, thus enabling their wives to return to part-time or full-time work. Here we can see the dynamics of collaboration emerging over time and from experience. Interestingly, the only woman who saw her working at home as a choice in the same way men did was the woman whose employment situation was most similar to professional men’s; this television producer had the highest earnings of the women, she had the most secure job and a quite senior position at work. She saw working at home as adding to her life, because, as she said, she already had a successful career. Even so, for her there was an element of personal obligation: ‘Family is of the utmost importance,’ she said. ‘We have no family people to help out, the children really need us.’ What about informants without children? The very absence of children means that these informants could conform more closely to the model of an individual maximizer who alone assesses costs in relation to preferences. In truth, however, it makes conformity to the rational choice model contingent upon being able to act as a lone individual. As home-located workers, informants without children have definite advantages, in that they can move back and forth between home and outside workplace as it suits them, although only insofar as the employment structure in their profession permits. Not only did the workers with no children find it easier to work outside the home, their adult-only homes provided work-friendly environments. One man who used the spare room of his flat for his files and the living room

Decision-making as a process over time 175 as an office said that he would never have decided to work at home if he had children. Those without care responsibilities are also much freer to shift their routines if they feel lonely. While a widowed woman editor with a child of 6 had to arrange a baby-sitter in advance for meetings which took place after school hours, and was able to talk to supportive colleagues only by telephone, the single people said they used meetings as a springboard for socializing. Child-free informants without the financial responsibility for children were also freer to persist in creative careers, even when their poor financial rewards might have discouraged others: my informants without children were over-represented among those coping with low earnings and long periods of unemployment. Whether or not this was an advantage in the long run is difficult to say: were such people freer to enjoy an artistic lifestyle, or had the absence of a spur to seek alternatives left them unwittingly ‘stuck’ in the metaphorical rut? The association of structure with constraint by both advocates and certain challengers of rational choice paradigms overlooks the fact that in other areas, studies of unemployment, for instance, structure is seen as enabling, acting as a mechanism through which people pace their day or lives and measure their progress.

Cushioning career progress: further effects of collaboration on decision-making The careers of home-based workers are cushioned by the support they gain from significant others, along with other resources, such as life insurance or income support. So far as I know, the importance of private sources of support for home-located workers has not been investigated in previous surveys, because the assumed separation of familial relations and work obscures its relevance. Yet any attempt to foster economic development through cultural work needs to recognize the interdependence of public and private spheres. In contrast, for rational choice theorists, of course, any form of financial cushion reduces the opportunity costs involved in a decision by reducing risks. Among cultural workers, partners’ earnings may be a very important source of financial security. The successive waves of the British Film Institute tracking survey of broadcasting workers clearly show that workers who live with employed partners are more confident of remaining in broad-casting than are those who do not (Dex et al. 2000). In my survey, informants’ households depended on partners’ earnings in thirteen cases, as well as on a former husband’s child support (one case), a widow’s pension (one case), inheritance or private income (at least four cases), a parental loan (one case) and redundancy payments or income support (three cases). This dependence of the self-employed cultural workers on other sources of income security has similarities to the dependence of young rock musicians on state income support, which came in for debate in spring 1998 when new regulations

176 Carol Wolkowitz threatened to remove it. Although mention of the (partial) dependence of married female parttime workers on their husbands’ earnings is a commonplace (Hakim 1996), it is clearly also relevant to the careers of self-employed men. However, something much more collaborative than financial risk reduction is involved in some cases of dependence on others’ income. As we saw above, some men have been willing to take up a share of the family work to help their wives progress in their careers, and they do not see this in purely pragmatic terms. Moreover, co-operation across the wider kin network in making their decisions is also evident. An example is a single man who works as a promoter in the folk music industry. His capital investment goes toward funding ‘his’ folk groups’ development, advance venue bookings and tour travel arrangements, in the expectation that his percentage of the proceeds will eventually make him a profit. He had established his first office based on experience of this kind of work for his university student union, but, after a financial crisis six years later, he was bailed out by his parents, closed his office and moved his work into his home. One reason why working from home is acceptable, he says, is that at his parents’ insistence it is his own brother, an accountant, who now manages the finances, whereas he would have been reluctant to have someone with whom he had a purely professional relationship coming to his flat. Another aspect of their social position which enables these cultural workers to take the decisions they do, I would argue, is the availability and use of paid domestic services. This again is fully consonant with rational choice theory’s notion of costs, but this does not mean that maximization strategies remain constant over time. Seventeen of the twenty-one informants employed a weekly cleaner, and all those with children under 10 employed a childminder or shared a nanny with another family. As Gregson and Lowe (1994) have argued, in Britain the employment of domestic workers by dual-career households replaces at least some of women’s work and, in households characterized by a more egalitarian ethos, some of the men’s. However, and here is the caveat, for the men in my sample, working at home typically involves taking on rewarding care for their children, food shopping, and, sometimes, weekday evening meal preparation, although they do not need to take on housewifely jobs like cleaning or ironing as part of the package. Here the actuality of sharing – together with freedom from the more feminized tasks – seems to reinforce an egalitarian ethos over time, as (some) men realize the rewards of closeness to their children and become more involved. Collaboration is correspondingly reinforced. The women working as childminders and cleaners in this neighbourhood tended to be older women who had been employed in local factories or as homeworkers in the garment industry until this work disappeared. As migrants from Ireland, Scotland, the Caribbean, the Mediterranean or South Asia, the nannies and childminders in the neighbourhood were available for childminding, another kind of home-located work, in part because other paid

Decision-making as a process over time 177 work was not available.4 I would argue that in Britain paid domestic help is crucial to the maintenance of the kind of dual-career households I studied. Of course the availability of childcare workers and cleaners in this neighbourhood is compatible with rational choice theory, because a bigger pool of childminders reduces the costs of home-located work. It does, however, highlight the shifting nature and contingency of costs, which, because dependent on global factors, cannot be assumed to remain constant.

Changes in attitude over time: the trajectory of preference redefinition The more important point is that the experiences encountered by those working from or at home mean that we need to conceptualize their involvement in home-located work as a process of reflexive review, in which preferences become revised in the light of experience. Their preferences were not fixed and hierarchical but subject to change as they changed and learned new, sometimes unexpected, lessons about their choice. Many have the resources to overcome difficulties as they become aware of them, so in this respect their situation may be quite different from low-paid manual homeworkers. Most of those who continued to be satisfied with working from home had over time found or maintained success in their careers, and, in addition, had come to appreciate a closer relationship to their children. Only one respondent manifested a stable hierarchy of preferences which seemed indifferent to the tough experience of her faltering career. A single woman without children, she had been directing fringe theatre productions for many years, but opportunities had declined over the past twelve. She drew income benefit, coming off it for a nine-week spell each year when she did sessional teaching of direction at a private acting school. She also did private tutoring in her home, and she said, on being asked directly, that she had had ‘a little private income at some times’ in her life. When one of her projects got off the ground, she could earn Equity rates, but this rarely happened. Most of her time was spent trying to organize new productions, or on her painting: I always try to have some kind of plan or project up my sleeve – otherwise life is depressing. I create work for myself if I don’t have any. ... I try to earn a living and keep work I don’t enjoy out of the picture. ... Yes, I could get a menial job, but I feel that after eighteen years in the theatre I should avoid that if I can. However, most of the informants whose careers had failed to develop had become very dissatisfied. Although they suffered inadequate earnings, it was their loss of confidence, exacerbated by the isolation of home-based work, which affected them most. For example,

178 Carol Wolkowitz one man said that after leaving art college he began working from home as a photographer – he says that he never thought of anything else – and got small jobs from various newspapers and magazines. Later he obtained regular work from a fashion magazine, and along with other work this produced a reasonable income, certainly as compared to his student days. When the design of the entire magazine was revamped, however, and his drawings were dropped, he was unable to replace this work by obtaining other commissions. This led to what he calls a nervous breakdown. Then, after counselling for depression, he decided to stop working from home, let the room in his house where he used to work to a lodger, and used the money to rent work space in a converted factory. At the time of the interview he still had no work coming in, but in the meantime his partner was paying most of the bills. Analysing the informants’ stories, it seems clear that, whether or not they continued to enjoy working from home, their experience made them aware of facets of it that they did not know about in advance. Some of these were fairly predictable; for instance, the producer in broadcasting worries that ‘I don’t have a public face. ... No one drifts in ...’, and felt that being unable easily to maintain his social networks made obtaining new contracts more difficult. But others pointed to aspects of working at home which they could not have predicted and which were intensely personal. For instance, one man said he had discovered he concentrated best at home, sitting at the kitchen table in the hub of family life, because it mirrored studying for exams as a child. For another man, working at home had the opposite effect, bringing back feelings of social isolation from years before: ‘I was at home again like a child – a complete replica of my childhood. I was not aware of it before, I just did it.’ In my sample, continuing to work from or at home seemed to reflect not so much stable or fixed preferences as having the resources and innovation to make changes, a situation quite unlike that of many manual homeworkers. For example, the woman who ran a publicity agency from home said she did not feel isolated, as she had the company of her assistant and a colleague she had invited to share her home-based office. Even so, she said, she had become depressed at being in the same environment, weekday and weekend alike, and in the year before the interview had looked for premises outside the home. When she realized how high the rent would be, she and her husband decided to use the same money to pay the mortgage on a country cottage, which allowed her to get away from the London house at weekends and have the garden she and her husband both wanted. In other cases, informants dealt with isolation by finding ways to socialise. Like the illustrator mentioned above, who took workspace outside the home to deal with depression, a male journalist had taken desk space at the office of a journal for which he did a lot of writing. He remembered being ‘delighted to be going into work, on time, having that sense of belonging. You can go into the office and feel professional, feel grown up.’ However, after eighteen months

Decision-making as a process over time 179 I didn’t feel there was any reason to have it any more. I didn’t feel I needed to get out of the house because work was becoming less solitary. Also my lodger moved so I had a spare room for a study. Neither the publicity agent, photographer nor journalist has children, suggesting that even within this relatively privileged sample the flexibility to shift their place of work is more available for those with few other responsibilities. Such stories also suggest that the balance between working at home and seeking outside work is very delicate and resonates differently not just in terms of social situation but also with very individual considerations, namely how informants had been affected by experiences, something which they could not have predicted prospectively.

Who stays in home-based work, and why? It is a pity that a longitudinal study was not undertaken, so as to identify some of the career risks which home-located work may involve, especially regarding the ability to return to outside work once children are older. From what information I have been able to gather, it appears that once experience has been accumulated, this not only modifies preferences but also their attainability, partly because the people themselves have been transformed by their experiences. Thus we come to the last stage of this dynamic trajectory. Information gathered during the interviews, along with anecdotal knowledge I later obtained about some of the informants, suggests that the men have tended to return to work outside the home, or to go back and forth between them, to a greater extent than have the women. More of the men with children in my sample seem to be working outside the home now than they did in earlier years. A writer has taken up a better paid, full-time post and given up his literary ambitions, and an artist fought a tribunal case in order to obtain a permanent, full-time employment contract which recognized his increasing teaching input. At the time of the interview, the male graphic designer, who had employed two staff in his home-based studio for eight years, was in the process of moving the workplace to an office block. The piano restorer had already moved his business from the storefront below their flat to larger (but nearby) premises as the enterprise grew. Three of these families moved to more expensive houses in neighbourhoods with ‘better’ secondary schools, suggesting that as their children got older not just childcare but concern about schooling replaced or took priority over their earlier creative ambitions. What we can only hypothesize is that greater experience of closer involvement with their children, through working at home, meant that they had come to give priority to their well-being over other individual preferences. There was one case in which a man’s career may have faltered because of his decision to work at home: the orchestra administrator, who had been willing to work at home so as to care

180 Carol Wolkowitz for his son after school, and said he had turned down offers of better jobs where he would not have been able to make the same arrangements. At the time of the interview he had thought himself privileged because he enjoyed the autonomy of the writer working at home, but without the economic insecurity usually associated with it. He was eventually made redundant due to arts funding cuts, something the development of a more conventional career earlier on might have forestalled. Less is known about the later careers of the women informants. One of the journalists who had been freelancing from home is still working at home, even though her son is now 18. She is keen to increase her income through some form of commercial activity, rather than continuing to depend on freelance journalism. The woman part-time sub-editor has never been able to obtain either another in-house publishing job or the academic post she would prefer. It seems as though enterprises can be moved to an outside workplace if the amount of custom permits, but both men and women may find getting back into (or even retaining) a job as an employee in the workplace hierarchy more difficult. Those whose careers have been side-tracked, when children’s needs meant stepping off the career ladder, may find it difficult to get back on. Although this clearly indicates that lack of paid daytime childcare for young children is an insufficient explanation of women’s career patterns, it suggests that the career disadvantages precipitated by childcare responsibilities are not readily reversed, even for men. My informants do not necessarily choose a life in which family comes before career, as a fixed orientation (Hakim 1995), it is rather that decisions taken at one stage in the life-cycle may have unforeseen, long-term implications. The woman architect I interviewed felt that women’s careers were impeded partly because of their openness to children’s demands for attention. Hers again is a somewhat different interpretation of women’s career patterns than a long-term individual preference schedule. While any such willingness to notice and/or respond to children’s wishes could merely reflect a wider ideology of ‘maternal sacrifice’ (Whitehead 1986), it could also echo the relational ethics which some feminists have seen as characteristic of women’s moral choices (Gilligan 1982). Whether or not responsiveness to others’ needs is linked to gender, as it is for this informant, it must presume an individual whose decisions are enmeshed in interpersonal relations and are not simply the expression of prior individual preferences.

Conclusions The study of employment has tended to treat domestic and work worlds as separate, at least in the case of men. The failure of surveys of middle-class home-located workers to include questions on home ownership is no doubt part of the assumed division between public and private spheres, which shapes ideas about what it is relevant to ask about. It appears that the

Decision-making as a process over time 181 externalization of risk in the cultural industries depends partly on the private relations of the people employed in them, especially partners’ earnings and independent incomes, and their willingness and ability to replace some of their domestic work with paid domestic services. In terms of grasping the relevance of the relation between private and public worlds, rational choice theory’s attempt to bring decision-making in different aspects of life into a common framework is in one way laudable. The problem is that it reduces all decisions to the logic of neoclassical economic behaviour. This chapter has looked at aspects of the careers of a small set of metropolitan cultural workers who worked entirely or partly at or from home. Although they obviously weigh up costs and benefits in making their decisions, this is not the whole story. The two main features discussed which are incompatible with rational choice theory are, firstly, that their decisions were made in the context of relationships with partners and children, cushioned by personal networks as well as institutional sources of income. The interdependence of spouses’ careers was especially visible in the case of one estranged couple, who continued to share a home because their finances and childcare routines were so interlocked. Informants without children, who had only themselves to consider, were more likely to maintain their focus on a creative career, even if their economic rewards were meagre. Secondly, many people’s attitudes toward working at home changed over time. The fact that some in this relatively privileged sample suffered from isolation, or marital breakdown, or have been deflected from successful professional careers suggests the existence of insecurity and other costs which are different from those associated with highly structured conventional promotion ladders, and which cannot necessarily be foreseen by people deciding to work from home. Finally, knowledge of the pay-offs and drawbacks of working from home can only be gained by experiencing it, but experience, in turn, changes people themselves, and often in ways which make their revised preferences hard to realize. In short, we are dealing with a dynamic process of decision-making which is neither individualistic, nor proof against the contingency of individual experience or the shifting social context.

Notes I would like to thank the University of Warwick’s Research and Innovations Fund for financial support for the survey on which this chapter is based, and the editors of this volume for their comments on a previous draft.


I am using the term ‘home-located work’ to cover much the same broad range of activities as ‘homelocated production’, a concept proposed by Felstead and Jewson (2000). Their recent book reviews many of the existing studies of the nature and extent of home-located work to date. Both terms refer

182 Carol Wolkowitz to a much wider range of paid work than the narrower concept of homeworking, which has come to refer mainly to low-waged assembly or routine clerical work in the home, usually remunerated through piecework rates. 2

Those who refused to be interviewed tended to be more ‘successful’ than the interviewees, especially the two whose work has gained a national reputation.


More recent developments in the organization of broadcasting have accentuated these developments in Britain (BFI n.d.; Dex et al. 2000; Saundry 1999).


Not only has London’s factory sector declined, outwork in the garment industry was being given to those recent migrants working as homeworkers in other neighbourhoods (Mitter et al. 1993). Lash and Urry (1994) point out that absorption of low-paid labour into the personal services sector is more extensive in gentrifying city neighbourhoods than in the suburbs.

11 The decision to commit a crime against humanity Robert Fine and David Hirsh

The topic of this chapter is the decision to commit crimes against humanity. By the term ‘crimes against humanity’ we refer not so much to its specific legal definition as to the social phenomenon of individuals committing mass murder and other atrocities on behalf of a state or some other organized political movement. The concept of ‘crimes against humanity’ was first legally articulated in the Nuremberg Charter of 1945. The Charter defined it in terms of certain specific acts (murder, extermination, enslavement and deportation), other non-specific ‘inhumane acts’, and persecution on political, racial or religious grounds. The limiting factor in all cases was that these acts had to be committed against civilian populations, have some connection with war and be carried out as part of systematic governmental policy.1 The concept has now been amended in international law to broaden its application; for example, the nexus with war has been watered down and systematic rape has been incorporated in its list of specific acts. The Charter introduced (as international law continues to advance) a strong notion of individual responsibility. It held that individuals acting within the legality of their own state can nonetheless be held responsible for crimes against humanity; the fact that defendants act under orders from their own government or superior officers cannot free them from responsibility; leaders and organizers who participate in the formulation of plans to commit crimes against humanity are as guilty as those who execute them; and the official position of defendants, whether as heads of state or as officials in government, cannot free them from responsibility. Not least, atrocities committed against one set of people – be it Jews, Poles, Roma, Tutsi or Muslims – are an affront not only to these people but to humanity as a whole, and humanity has the authority and duty to punish those who commit them. International law establishes a definite link between individuals and their actions, by treating so-called ‘cogs in a murder machine’ as perpetrators and by refusing the excuse of service to the state. It presupposes that choices are available to the perpetrators of such crimes. If no such choice is in fact available, that is, if the situation is one of ‘kill or be killed’, then

184 Robert Fine and David Hirsh this would constitute a legitimate defence or mitigation. This juridical presumption of choice and responsibility opens the door to sociological investigation into how decisions to commit crimes against humanity are made, even if the law is not itself directly concerned with this question. The prevailing sociological explanation of such decisions is that provided in Zygmunt Bauman’s (1993) genuinely path-breaking study, Modernity and the Holocaust. Bauman himself may not think of his work precisely in this way, but his basic proposition is that it is the dominance of ‘rational choice’ over moral response in the modern age that is the key to understanding how ordinary men and women commit such extraordinary crimes. Conversely, the key to overcoming this potentiality is seen to lie in the development of a postmodern ethics which subordinates the imperatives of ‘rational choice’ to a reconfigured ‘moral point of view’. Bauman does not like a world structured around ‘rational choice’, but he accepts that this is the actuality of our present world. He looks to a way of thinking which overcomes the constraints of ‘rational choice’ and that in its place revives our suppressed capacity to act in a moral rather than rational way. From this perspective, ‘rational choice’ appears as a form of human decision-making which arises in the modern epoch and which has as its consequence the exclusion of ethical concerns. And ‘rational choice theory’ appears as a form of reified consciousness which hypostasizes rational choice, as a natural presupposition of social life, and blinds us to its historical preconditions and de-moralizing consequences. Our interest, in this chapter, is not to defend rational choice theory against this very sharp line of criticism, but rather to argue that this line of criticism is over-dependent on the rational choice model which it attacks. We want to argue, firstly, that the reduction of ‘modernity’ to the imperatives of an amoral and instrumental rationality paints a one-sided picture of modernity which obscures the inner connections between modernity and the development of moral consciousness itself; secondly, that the reduction of reason to instrumental or technical or technological rationality distorts the meaning of reason and severs its connections with thinking, understanding, willing and judgement; thirdly, that the decisions of individuals to participate in crimes against humanity (including those synthesized under the name of the Holocaust) cannot be adequately explained within this framework; and, fourthly, that the moral point of view itself is far from being a purely innocent or suppressed factor in decisions to commit crimes against humanity. Most of all, although we recognize that Bauman and those who think like him have undoubtedly revealed something extremely important about the nature of organized violence in the modern age, we must also be alert to the dangers of forcing the empirical phenomena into an overdetermined theoretical straitjacket.

The decision to commit a crime against humanity 185

Modernity and the Holocaust: Bauman’s critique of rational choice In Zygmunt Bauman’s conception, the modern world really does run the way rational choice theory says it runs: in terms of short-term and instrumental preferences set within a given domain. Every aspect of social life encourages, coerces and impels individuals to act in accordance with their own short-term, narrow, selfish interests. We live according to that ‘principle’ alone; we do that which we find rational in terms of immediate self-interest. This includes the search for means, obedience to orders and conformity with social norms, regardless of their moral content. It also means the prioritizing of self-advancement or selfpreservation regardless of moral cost. We become a new type of bourgeois: not the Kantian who thinks and judges for herself, but the ‘mass man’ (to use a phrase borrowed from, among others, Hannah Arendt) who can kill without passion or enmity, simply as a job or in service to the state, because it is an efficient means to a given end, or because he is commanded so to do, or because that is what everyone else is doing. The making of merely ‘rational choices’, without regard for ethics, is the very mark of this social type. It is, Bauman argues, through the combination of many such ‘rational choices’ that the Jews of Europe were rounded up and murdered. As long as we remain within this ‘rational’ template, we are destined to play our part in the genocide. The frighteningly domestic image through which Bauman portrays modernity is that of a ‘garden culture’ in which the extermination of weeds is the necessarily destructive aspect of the gardener’s productive and aesthetic vision. A gardener has an image of how he wants his garden to be. He wants it to be well ordered and to conform to his own dreams of beauty and serenity. He likes certain plants and breeds them to fit in with his plan. He does not like other plants, which he designates as weeds, and then poisons or incinerates. In this scenario, the gardener sees the elements of nature instrumentally, in terms of how they effect him and may be affected by him, rather than as things endowed with an intrinsic value of which he is guardian (Bauman 1993: 91–2). In modernity, human beings are themselves stripped of intrinsic value. Some are defined as weeds, others are selectively bred. Genocide is a kind of social weeding, and Hitler and Stalin were but ‘the most consistent, uninhibited expressions of the spirit of modernity’ (Bauman 1993: 93). If the technologization of conception is one aspect of the spirit of modernity, the other is the technologization of execution. In this reading of the situation, it was the bureaucracy which executed the Final Solution, and even the ‘political master’, Hitler, found himself in the position of the ‘dilettante’ standing opposite the expert and facing the trained official (Bauman 1993: 15).2 There is no decision, as such, to commit crimes against humanity, simply the normal functioning of a bureaucratic state. In his discussion of Claude Lanzmann’s Shoah, Bauman tells us that ‘by far the most shocking among Lanzmann’s messages is the rationality of evil (or was it the evil of rationality?)’ (Bauman 1993: 202). For the bureaucratic

186 Robert Fine and David Hirsh form of administration, which prevails in modern society, has a machine-like quality in which each bureaucrat follows detailed written rules unthinkingly and without responsibility for what the machine is doing as a whole. Bureaucracy is a machine for the exclusion of moral responsibility. Bauman argues that the defining features of modern bureaucracy were not only well established in Germany during the Holocaust, but made the Holocaust possible. Government was conducted through a centralized, hierarchical and bureaucratic state; respect was afforded to science, knowledge and expertise; rational behaviour was valued over irrational behaviour; the breaking down of tasks into small parts was prevalent; the technology of factories and railways was well established. The Nazi regime appears in this reading as an extreme form of the modern state, and the administration which carried out the Holocaust as but an extreme form of modern bureaucracy. Even the choice of extermination ‘was an effect of the earnest effort to find rational solutions to successive “problems”’, and at no point did the Holocaust come into conflict with the principles of rationality: The ‘Final Solution’ did not clash at any stage with the rational pursuit of efficient, optimal goal-implementation. On the contrary, it arose out of a genuinely rational concern, and it was generated by bureaucracy true to its form and purpose. (Bauman 1993: 16–17) In Max Weber’s exposition of modern bureaucracy, Bauman sees ‘no mechanism ... capable of excluding the possibility of Nazi excesses ... nothing that would necessitate the description of the activities of the Nazi state as excesses’ (Bauman 1993: 10). If it were the case that modern rational bureaucracy reduces the individual to nothing more than a cog in a machine, a blind applicant of rules, an actor only in the narrowest sense of making rational choices on exclusively instrumental grounds – if all this were true, then we could only conclude with Bauman that the condition of modernity robs people of any significant sense of moral responsibility and that it is this negation of moral responsibility that is the condition of the possibility, as Bauman might put it, of the decision to commit crimes against humanity. Bauman implies that neither the abstract conceptions of individual responsibility found in law, nor the lack of any conception of responsibility in sociology, offers a remotely adequate response to the enormity of the issue. In this context, legal notions of individual responsibility are only a legal fiction imposed on a recalcitrant technological reality, and in any event a court is itself a bureaucratic, rule-bound institution which judges questions of criminal guilt by abstracting them from the complex reality of three-dimensional events. Putting the blame on a particular individual does little to confront the system of ‘rational choice’, for it is ‘modernity’ rather than individual killers that is primarily at fault. If perpetrators are guilty of not breaking free from this system, such is also the fate of the vast majority of people. Only

The decision to commit a crime against humanity 187 the few have the courage and vision to risk everything by stepping out of society and confronting their unconditional responsibility for others. As far as sociology is concerned, Bauman argues that it typically mimics the society which it purports to understand. The general absence of the concept of moral responsibility in sociology, Durkheim’s identification of morality with conformity to social norms, Weber’s rationalization of bureaucracy, the reification of rational choice by rational choice theory – all this in effect reflects the conditions of modern society. In opposition to the unheroic ‘mass man’ who succumbs to the pressures and constraints of rational choice, the only way to save ourselves from complicity is to hear the call to Being with Others, the call of alterity, the call to act morally, the call to go beyond the ‘morality-silencing’ bounds of reason and society and rediscover the presocial sources of ethical life in the face of the other.

Rationality and the Holocaust reconsidered Two particularly problematic areas in Bauman’s critique of ‘rational choice’ concern his focus on bureaucracy. Firstly, in Weber’s conception of bureaucracy, individual officials are responsible for their actions, and part of the immense power of bureaucracy is based on this responsibility for decision-making and rule interpretation which is distributed throughout the hierarchy. If the Nazi organization of terror and extermination constituted a typical modern bureaucracy, as Bauman argues, then individuals would have been expected to take responsibility for the tasks assigned to them, and the leadership could not have relied on its employees to perpetrate murder simply as ‘cogs in a machine’. As Weber recognized, the process of following a rule is always mediated through mind and consciousness, and the ethos of public service is the oil that allows the machine to run.3 Secondly, the social organizations which conceived and executed the Holocaust were so different in both ideology and organization from the ‘Weberian’ model of bureaucracy that they should not rightly be called ‘bureaucracy’ at all. They expressed a mode of rule which inherited elements of bureaucratic authority but reconfigured them in a way that cannot simply be understood in terms of Weber’s analysis of rationality. The Holocaust was organized neither by typically ‘modern formations’ nor by anything approximating an ideal Weberian bureaucracy. Certainly, the analysis of totalitarianism in power offered, notably, by Hannah Arendt (1975) and Franz Neumann (1957, 1963) paints a profoundly different picture of totalitarian rule. Totalitarianism was not the final culmination of the power of the modern state, but a revolution against the structures of the modern state. Movement rather than structure was its essence. Totalitarian rule was organized on the basis of the intermeshing of various state and party institutions and the proliferation of organizations within the party. Duplication was particularly apparent within the many police apparatuses, which all did similar work, spying

188 Robert Fine and David Hirsh on the population and on each other, without any clear knowledge of who would be rewarded and who would be purged.4 In the complex duplication of organizations involved in the Final Solution, all were ‘equal with respect to each other and no one belonging to one group owed obedience to a superior officer of another’ (Arendt 1994a: 71).5 The only ‘rule’ according to the Führerprinzip was that formulated by Hans Frank: ‘Act in such a way that the Führer, if he knew your action, would approve it’ (in Arendt 1994a: 11). This ‘categorical imperative’ is the opposite of clear, rational, written rules. The ‘leader principle’ is not that of a bureaucracy organized on the basis of formal rules within a structured hierarchy, for the allegiance of the official is not owed to his or her immediate superior but to the leader himself. The individual responsibility of the official is arguably even greater under the leader principle than in a regulated hierarchical bureaucracy in which responsibility and authority are distributed according to plan. On the one hand, to grasp the will of the Führer demands zeal and creativity far in excess of the old-fashioned plodding bureaucrat; and wide latitude is given to sub-leaders for the execution of policies. On the other hand, each holder of position is held responsible for all the activities of his subordinates, even in cases of disobedience and failure. The perpetrators were not generally forced into the formations which implemented the Holocaust. Eichmann was keen to win promotion on his particular ‘front line’ and the members of the murderous police battalions (the Einsatzgruppen) were given the opportunity to withdraw from the killing actions (Browning 1993). When they accepted the authority of these outfits, they chose to do so even if the parameters of their choices were limited. Authority in the modern sense of the term is not the same as power. People choose to defer to authority. To be sure, choices are never completely free. They are made within the limits of what is possible and of what alternatives are possible; there are always external constraints. Yet rarely are those constraints so rigid that there is no choice; rarely is the structure so dominating that it removes all agency. Under the leader principle, authority works through the will of every member to know and act in accordance with the will of the leader and to take responsibility for all the decisions taken in their field of operation. Bauman was right to tie his analysis of responsibility to the actual ways in which decision-making was organized in the planning and execution of the Holocaust, that is, not to remain exclusively at the level of political philosophy or legal theory but to link such concerns with a sociology of decisionmaking. However, the presumption of rationality in the substance of his analysis obliterates what Arendt called ‘the horrible originality’ of totalitarian rule. A striking feature of the Holocaust commonly remarked upon was its industrialization of death. The Holocaust was of its time; it used the methods of its time, and particularly important to Bauman are the methods of modern management through which the genocide was in part carried out. We say ‘in part’ lest the ‘industrial’ image of Auschwitz overtake our imagination of the Holocaust as a whole. We should remember that the Nazis devised two basic strategies for the annihilation of Jews: mass shooting and mass gassing. Special duty

The decision to commit a crime against humanity 189 troops of the SS’s Security Service and Security Police, called Einsatzgruppen, were assigned to each of the German armies invading the Soviet Union and were given the task of rounding up Jews and killing them through crude and primitive methods of shooting. These methods were the antithesis of Bauman’s image of clean and dispassionate white-coated technicians introducing gas into gas-chambers. These were methods which confronted the killers with the blood, faces and screams of their victims. It is estimated that some 2 million Jews were murdered in this way. To murder the rest of European Jewry the Nazis built ‘camps’ with large-scale gassing and sometimes crematorium facilities (Auschwitz, Belzec, Chelmo, Majdanek, Sobibor and Treblinka), but also many other ‘camps’ which were designed to work their inmates to death. The technology used here was often barely more sophisticated than the brute violence of the Einsatzgruppen, and it was only when death camps were combined with labour camps (such as at Auschwitz) that architectural relics of ‘industrial killing’ were left behind. All in all about 3.5 million Jews were murdered in this way. A further half a million Jews or so were killed through hunger, disease and exhaustion in the ghettos and as victims of random terror and reprisal. In short, we should be wary of the contemporary synecdoche which substitutes ‘Auschwitz’, or rather an industrialized representation of Auschwitz, for the whole. Some elements of bureaucracy certainly existed in the Third Reich: people were sometimes numbered, processed by bureaucratic-type machines, placed under systems of surveillance; there were papers, form filling, official stamps and files of information kept on individuals. But there was no bureaucratic hierarchy of command or system of rules that would be recognizable to a student of Weber. Officials who were technically in positions of authority could be denounced and replaced by their juniors; one apparatus was liable to be liquidated in favour of another; the stability and hierarchy of genuine bureaucracy were absent. What was most significant about the execution of the Holocaust was not the presence of bureaucratic authority but rather the reconfiguration of these elements to construct a principle of rule such as the world had not experienced before. When Bauman turns rational choice into a modern fatality, he also reduces it to its basest elements. He declares that ‘most scientists would be prepared in exchange [for research grants] ... to make do with the sudden disappearance of some of their colleagues with the wrong shape of nose or biographical entry’ (Bauman 1993: 109). He says that rational individuals would play their part in gassing millions, if it meant holding on to a good job. The rational individual would look the other way, stand by and refrain from intervening into affairs that were none of his business, that were not in his job description. This is not the individual who would devote her life to making sense of the world in all its boiling complexity. The shame we feel when we live in a world in which the Holocaust has happened is represented as the antithesis of reason. It is as if morality and reason are opposed armies or

190 Robert Fine and David Hirsh that the opposition of morality and reason which Bauman discerns under Nazism is true of modernity itself. Bauman also totalizes rational choice to explain the behaviour of those who conceived the genocide, those who organized it, those who perpetrated it and those who stood by without intervening. And the same mechanism also appears to have governed the behaviour of the victims: ‘The Jews could ... play into the hands of their oppressors, facilitate their task, bring closer their own perdition, while guided in their action by the rationally interpreted purpose of survival’ (Bauman 1993: 122). Bauman argues that the regime in power is always in control of the ‘game’ in such a way that the ‘rational choice’ from the point of view of the subordinates is also the preferred choice from the point of view of the regime. So it was that the Jewish administrators and police of the Ghettos were enticed to co-operate with the Nazis in the deportation of Jews on the grounds that, however many Jews they produced, they were saving or at least delaying the transport of the rest. The Nazis were able to rely on the Jews to act ‘rationally’ and thus collaborate in their own extermination: ‘In [the world of Auschwitz], obedience was rational; rationality was obedience. ... Rational people will go quietly, meekly, joyously into a gas chamber, if only they are allowed to believe it is a bathroom ...’ (Bauman 1993: 203).6 Here the ‘rationality’ of the Jewish response, which looked to make an accommodation with the Nazis, is contrasted with the ‘irrationality’ of the Warsaw Ghetto uprising. But the choice was not between unreason and reason. We may prefer the heroism of the Warsaw Ghetto uprising to the conformity of the Jewish councils, but in both cases Jews were faced with an impossible choice. The accommodation strategy seemed reasonable to a conservative Jewish leadership who understood the Nazi threat as a continuation of an age-old antisemitism with which a modus vivendi could eventually be found. It was an attempt to give a little in order to save more. The ‘rebellion’ strategy adopted in Warsaw seemed reasonable when it became clear that the Nazis planned to kill everyone and that there was no exit. It does not increase our understanding of events to assign the epithet of ‘rational’ to one strategy and ‘ethical’ to the other. There can be two interpretations of Bauman’s overall thesis. The weak one may be summed up by his observation that ‘modern civilization was not the Holocaust’s sufficient condition; it was, however, most certainly its necessary condition’ (Bauman 1993: 13). This interpretation brings to the fore the fact that the Holocaust was modern both in its conception and in its execution and that the conventional view of Nazism as simply ‘anti-modern’ cannot hold. The strong interpretation of Bauman’s thesis is that the dynamics of modernity push towards genocide, that there is nothing in modernity that pulls away from genocide, that even when genocide is not actual, its potentiality is ever-present. Bauman himself vacillates between these theses, but between them there is a lot of ground. The weak thesis reminds us that the Holocaust happened in a ‘civilized’ European country that was technologically and

The decision to commit a crime against humanity 191 culturally advanced and cannot be written off as an aberration or just another example of man’s inhumanity to man. The strong thesis is that modernity brings with it the uncoupling of human beings from moral choice and the tying of human beings to a narrow, short-term instrumental rationality. People are made into unthinking cogs in the all-powerful structures of modernity. Bureaucracy brings us the human being who is incapable of seeing the bigger picture. Science brings us ‘a rule forbidding the use of teleological vocabulary’ (Bauman 1993: 190).7 Rational choice becomes our fate. Doubtless the technical-administrative success of the Holocaust was due in part to the skilful utilization of ‘moral sleeping pills’ made available by modern science and technology; but it was also due to the skilful use of moral imperatives. The appeal by Nazi leaders to duty over private passion, economic utility and military need is now well established – whether in overcoming the resistance of ‘ordinary men’ to slaughtering other human beings or in overcoming the resistance of army generals to wasting much needed military resources on the killing of Jews. The ‘moral point of view’ was neither an innocent nor an excluded party in the decision to commit atrocities.

The case of Police Battalion 101 In his book Ordinary Men, Christopher Browning (1993) tells the story of Police Battalion 101, which was one of the formations which followed the German front as it invaded Russia in 1941 in order to kill the Jews who lived there. The personnel for the battalion was recruited from Hamburg during the war, after the youngest and fittest men, as well as the most politically committed, had already been drafted. Browning explores how these middle-aged citizens of Hamburg were transformed into mass killers. In interrogations after the war, the men of Police Battalion 101 identified a number of factors which led them to become killers: the wish to conform, to yield to peer pressure and to obey authority. They told of their desire not to be designated cowardly and not to evade their part in the dirty work that had to be done. Neither political indoctrination nor antisemitism seems to have been a major factor in these decisions. The first assignment for the battalion was the rounding up of the Jews of Josefow. The men were to be sent to work camps and the women and children were to be shot. The Commander, Wilhelm Trapp, made it clear that no member of his battalion would be compelled to participate in the shootings: about a dozen of the men immediately decided not to take part and others opted out later. However, about 80 per cent of the men decided to participate. At first they found their task difficult to perform, but Browning argues that there was a ‘toughening up’ process which hardened the men to killing once they had already taken part. The ‘decision’ to commit crimes against humanity seems to have followed the first killings rather than to have preceded them. Once these men found themselves implicated in

192 Robert Fine and David Hirsh massacres, the group acquired an esprit de corps of mutual guilt. As we find recently among perpetrators of atrocities in Bosnia, the group regularly drank large amounts of alcohol in the evenings to ‘blank’ out their days and avoid having to think about their actions. The members of Police Battalion 101 seem to fit Bauman’s model better than that controversially advanced by Goldhagen (1996): that they were driven by an antecedent and virulent antisemitism. They decided to commit crimes against humanity under the influence of the command-structure to which they were subordinated. In private life, they were no more predisposed to violence than any other randomly selected group. Yet this genocidal formation was able without much difficulty to incorporate most of them and use them as its agents. There was a role for deference to authority and for the unthinking following of orders. The individuals were explicitly given a choice, and most of them made a positive choice to kill. Social factors, such as esprit de corps, peer pressure, the wish not to stand out, and so on, were all present in the making of these choices. However, the ‘hands-on’ massacres in which these men participated had nothing to do with ‘social distancing’ from unseen and faceless victims.8 The overcoming of their initial repulsion to their new-found jobs – and of the nausea they suffered as a consequence – were presented by their senior officers as a demonstration of their exceptional being, their moral superiority in overcoming natural inclinations, their toughness in doing what they naturally found repugnant. Many of the men seem to have accepted this proof of their extra-ordinariness. It was perhaps their way of not being ‘ordinary men’.

The case of Eichmann Adolf Eichmann was not only a key bureaucrat and engineer of the genocide of Jews; he was the man in charge of the whole programme of Jewish extermination. He was not, of course, forced into his job. In his case, there was no question of ‘kill or be killed’. On the contrary, he was ambitious, keen to win promotion, and personified unquestioning recognition of the authority of the Führer (Arendt 1970: 45). In his trial he said that, though he bore no ill feelings toward his victims, he simply could not have acted otherwise. He acted according to his conscience, and apparently his conscience would have troubled him only if he had questioned orders – a thought which happily seems never to have occurred to him (Arendt 1994a). If we are to believe what Hannah Arendt wrote about him, and her evidence is compelling, he was a rather pedestrian individual, with few motives beyond his diligence in looking out for his own career advancement. He had no ambition ‘to prove a villain’ (to quote Shakespeare’s Richard III), nor did he appear as even a convinced antisemite. He presented himself as simply a bureaucrat rooted in an everydayness that made him incapable of critical reflection or moral judgement. As Arendt put it, he was marked more by ‘thoughtlessness’

The decision to commit a crime against humanity 193 and ‘remoteness from reality’ than by any streak of ‘Satanic greatness’. It was sheer thoughtlessness which predisposed him to become one of the greatest criminals of the modern age. The lesson Arendt (1970: 3–4) took from Jerusalem is that we have to come to terms with the fact that the man responsible for the execution of the Holocaust was terrifyingly normal: ‘the deeds were monstrous but the doer ... was quite ordinary, commonplace, and neither demonic nor monstrous’. Eichmann appears in this account as the very personification of Bauman’s ‘rational actor’ driven by a narrow and petty self-interest to push aside any consideration of the moral substance of the job he did. When he offered the improbable defence that he had nothing to do with the killing of Jews, he seems not so much to have been lying as revealing that ‘he merely never realized what he was doing’ (Arendt 1970: 287). Since he conceived himself as a man who was ‘only doing his job’, acting not out of inclination but only in a professional capacity, he could not regard himself as a murderer. He saw himself merely as a ‘cog in a machine’ and, like any other cog, without moral responsibility. He was an archetype of what Arendt called the ‘mass man’: the new type of bourgeois who presents himself simply as an ‘employee’.9 Eichmann stands at once as the exemplar of the claim that the perpetrators of the Holocaust were ‘men like ourselves’ who merely followed the norms of rational decisionmaking, and as a rejoinder to conventional images of a world dichotomized between our own absolute innocence and the unspeakable Nazi beast. He was living proof of what Arendt and Jaspers (1993: 62) termed the ‘banality of evil’: that the perpetrators were endowed more with ‘prosaic triviality’ than with ‘satanic greatness’. On the face of it, the case of Eichmann offers a strong case for Bauman’s ‘rational choice’ argument. It also highlights, however, a major difficulty with his formulation of the problem. Hannah Arendt mentions one moment in the trial when Eichmann suddenly declared that he had lived his whole life according to Kant’s moral precepts, and especially according to a Kantian definition of duty. Arendt comments that this was outrageous, since Kant’s philosophy was bound up with the human faculty of judgement (i.e. with thinking for oneself) and so rules out blind obedience. However, when pressed further, Eichmann revealed that he had read Kant’s Critique of Practical Reason and he came up with a roughly correct version of the categorical imperative: ‘I meant by my remark about Kant that the principle of my will must always be such that it can become the principle of general laws’ (in Arendt, 1994b: 136). He added that, from the moment he was charged with carrying out the Final Solution, he knowingly ceased to live according to Kantian principles. Arendt comments that Eichmann did not merely cease to follow Kant’s categorical imperative, but rather that he distorted it in line with Hans Frank’s formulation which we mentioned above: ‘Act in such a way that the Führer, if he knew your action, would approve it.’ This meant that duty was duty, a law was a law, and there could be no exceptions, not even for one’s own friends. But when Eichmann said that he had given up on Kant, this also meant in effect that he put his own self-

194 Robert Fine and David Hirsh advancement before any ethical concerns, and placed blind obedience to the Leader before his own practical reason and reflective judgement. In saying this, he must have recognized at some level his own descent into thoughtlessness, lack of reflection, unreason. This account reveals the inversion of ‘reason’ and ‘passion’ in Bauman’s reformulation of Kant. In place of Kant’s identification of ‘practical reason’ with larger moral concerns and ‘passion’ with self-interest, self-advancement and self-preservation, Bauman reverses this order of association. Reason is now identified with self-interest, self-advancement, selfpreservation, and so on, and ethics with one’s emotional response to the face of the suffering Other. In Kant’s hierarchy of reason and passion, passion is subordinated to the demands of ‘reason’ but it is not denounced or damned. Bauman’s hierarchy is more severe: it does denounce ‘reason’ (that which Kant calls ‘passion’) in favour of postmodern ethics (that which Kant calls ‘reason’). The neo-Kantian turns out to be more Kantian than Kant. The effect of this inversion is not only to accept the disconnection of rational choice from ethics, but also to sever the relationship between thinking and understanding, on the one hand, and moral judgement and decision-making, on the other.10 There are many moments in the text when Bauman writes of the separation of reason and ethics under Nazism. This may well be true, though we would continue to insist that the Holocaust had more to do with the ‘eclipse of reason’ (whether conceived in terms of economic, political or military utility) than with the triumph of reason, and with the triumph of a horrible kind of racist morality rather than with the eclipse of morality. The main point, however, is not to turn this opposition of rational and moral choice into an unalterable fact of ‘modernity’, still less into a fact of life as such. This is the slippage which seems to us to dog Bauman’s extraordinary analysis. The case of Eichmann reveals to us a man who, when he became a Nazi, self-consciously gave up on ‘practical reason’ (thinking for himself, developing his reflective capacities, judging on the basis of universal criteria) and replaced it with mere obedience to orders, social conformity, rigid duty to order. This was his choice. It was a terrible one in the circumstances. But it had nothing to do with the effacement of a pre-social moral consciousness by the technical-rational norms of modernity. For the individual’s capacity to think and judge for herself is as much a feature of ‘modernity’ as is the awesome power of ‘society’ over the individual.11

The case of Andrei Sawoniuk Let us end this exploration with a brief discussion of a recent case.12 In March 1999 Andrei Sawoniuk was found guilty at the Old Bailey of taking part in the genocide of the Jews in 1943 in Belorussia (contemporary Belarus). He was born and lived his childhood in a small town. His family did not have land to tend and so his mother made a living by doing laundry and other casual work for Jews, and when he was old enough, he also worked for Jews, doing odd

The decision to commit a crime against humanity 195 jobs wherever he could find them. He never knew his father, and his mother died when he was a child. He was regularly called a ‘bastard’ and subjected to bullying on that account. After his mother died, he lived with his grandmother and brother. He must have experienced a certain amount of alienation from the society in which he lived, and in which he did not really have an established place. He left school at the age of 14. He was known by everyone in Domachevo simply as Andrusha, a diminutive of Andrei. He was still known by this diminutive when he was the commandant of the local Nazi-organized police force. Within a very few days of the occupation in 1941, the Germans organized a local police force which Sawoniuk and a handful of other local men joined enthusiastically. He was 20 years old and had experienced two years of very difficult times under Russian occupation. For the first time in his life he had a job and a place in the world. Quickly the policies of ghettoization and starvation of the Jews were put into place, and it was the local police who were given the main responsibility for their implementation. Sawoniuk’s brother also joined the police, but he did not like having to carry out these duties and did not want to hurt and kill Jews. He left the police, using the excuse of poor health. The two brothers seem to have fallen out over this decision and Andrei Sawoniuk stayed in the police force. Witnesses at his trial tell of a number of brutal and cruel acts that Sawoniuk committed, such as beating a Jewish woman for attempting to smuggle some potatoes into the ghetto. A German Einsatzgruppe massacred the Jews of Domachevo on Yom Kippur 1942. It was the local police, however, who knew the local Jews and the local geography and had responsibility for hunting and killing those who had managed to escape the slaughter. By this time Sawoniuk was secondin-command of the local force and took a primary role in the search-and-kill operation. At his trial, witnesses tell of seeing him kill and beat different groups of Jews who had been found in the operation. Later, when the Germans retreated from the advancing Russians, Sawoniuk retreated with them and joined the Belorussian section of the Waffen SS, before deserting and joining the Free Polish Army and coming to the UK with them. Sawoniuk has lived in Britain since 1946. He worked for British Rail, and retired in 1986. He married twice after the war, both short marriages, and had a son with his second wife, but they parted shortly after his birth. In Britain he has been, as the police testified at his trial, ‘of good character’. It is difficult to estimate how many Sawoniuk killed. Maybe it was as few as 50 or 100 or 200 Jews. There were many tens of thousands just like him who took part in the genocide and were a necessary part of the machine which committed it. Some aspects of Sawoniuk’s transformation into a mass killer are in tune with Bauman’s framework, yet others are in contradiction with it. Both before and after the Holocaust, he was an ordinary, law-abiding, unexceptional person, but between 1941 and 1944 he was a sadistic, brutal mass murderer. It is probable that, had he not found himself in a social structure which was committing genocide, then he would never have become a killer. It is also true that, believing the Nazis were destined to win the war, his strategy of becoming a policeman and behaving in such a

196 Robert Fine and David Hirsh way as to be trusted and promoted by the occupying power had a certain logic to it from the point of view of his own narrow self-interest. It gave him a job, a living, power and the possibility of promotion. But it is stretching the facts to suggest that Sawoniuk’s decision to become a genocidaire was simply an example of rational decision-making. Firstly, since his brother chose to leave the police force and is still living quite happily just across the river from Domachevo, it is clear that Sawoniuk could have made the same choice if he had wanted to. He chose a different course, and it was a free and conscious decision. He had an argument with his brother; he chose to kill Jews and his brother chose to take his chances outside the police force. Secondly, a decision such as whether or not to become a mass killer must involve factors other than rational choice. It is only possible to speculate about Sawoniuk’s early life and what kind of a person he was when he chose to become a killer. It is clear enough that he was not brought up in a loving family and that he was poor. It also seems that he suffered as a child from some bullying. None of this, of course, can explain a man becoming a brutal mass murderer, but it is not irrelevant that he was an excluded, alienated, unloved young man. He found a way to improve his social prospects and also, perhaps, an outlet for his anger. But Sawoniuk was in no way a Weberian bureaucrat, who just obeyed orders and carried out professional duties. He chose to become a killer and he chose to kill and beat with more brutality than the efficient pursuit of a bureaucratic goal could possibly require.

Conclusion Although the aim of this chapter is to cast doubt on the rational choice model of modernity which Bauman surprisingly sets up in order to destroy, we are not in a position to draw any definitive positive conclusions from our critique or our case-studies. We can, however, make some tentative suggestions as to a way forward. Firstly, we are not dealing with individuals who had prior (fixed) preferences for antisemitism and were thus ‘just on the look-out’ for propitious circumstances in which they could maximize these preferences at low opportunity costs to themselves. The making of a mass murderer is a social process in which there is an interplay between the act and the actor in which the commission of the deed may precede both its signification and its justification by the actors involved. Rather than the motive leading to the act, it was often the case that complicity in atrocity, torturing and murdering innocent human beings led to the search for good reasons – perhaps on the basis of the Pascalian principle that if you kneel first, then prayer will follow. We must emphasize the malleability of preferences, how experience changes them, how ‘ordinary men’ turn into hardened monsters or at least become hardened in their monstrous acts. Secondly, the making of a mass murderer is a social process in which there is also an interplay between structure and agency. Regarding structure, the Führer principle

The decision to commit a crime against humanity 197 represented a new context (contra Bauman) in which ‘ordinary people’ are given new carrots to become ‘extraordinary’ by committing vile deeds. We see here a kind of ordinary (conformist and officially validated) extraordinariness. Once these incentives disappear, some became (like Eichmann) obedient servants to the authority of the court that tries them and most become ordinary ‘democratic’ civil servants and businessmen. Regarding agency, it is clear that some people walked away from the ‘incentives’ to murder and exercised their own moral judgement. Such judgements were not entirely ‘reflective’ in the sense that there were no rules or standards to guide them, for individual subjects could still appeal beyond the particular normative order of the so-called Volk to a humanist tradition – of thinking for yourself, of the right to subjective freedom, of universal equality – that is as much part of ‘modernity’ as instrumental rationality. We see here strong confirmation of our argument (again contra Bauman) that not even this totalitarian epoch could reduce all action to instrumental rationality. On the contrary, totalitarian terror demonstrated ultimately the subordination of instrumental rationality to a certain ‘moral’ point of view in which (as Hannah Arendt has in our view correctly argued) questions of economic, political and military utility were self-consciously subsumed to the end of killing and degrading Jews (Arendt 1994b). Put at its strongest, it is difficult to escape the conclusion that in the modern/postmodern world the moral point of view is always a crucial element of decision-making, and that no rational choice can be understood solely in terms of instrumental rationality. Without reference to moral concerns, we cannot explain how some people monitor their preferences, refuse all incentives to violate them and resist to the end lending themselves to the horrible processes we have described in the text. The wertrational works not just as an accidental or subordinate ingredient within preference formation and expression, but as a constitutive aspect of how ‘we’ – i.e. individuals thrown into a world without absolute foundations – make sense of, understand and judge the preferences we endorse.

Notes 1

The Charter of the International Military Tribunal was signed on 8 August 1945 by the governments of the USA, the French Republic, the UK and the USSR and designed for the trial and punishment of the major war criminals of the European Axis (Avalon Project 1999).


Compare with Alan Bullock (1983: 381): ‘[Hitler] had a particular and inveterate distrust of experts. He refused to be impressed by the complexity of problems, insisting until it became monotonous that if only the will was there any problem could be solved.’


Rules are nothing without interpretation. Bureaucracies are machines made up of people, each of whom takes decisions within given parameters. Weber writes: ‘a system of rationally debatable

198 Robert Fine and David Hirsh “reasons” stands behind every act of bureaucratic administration, that is, either subsumption under norms or a weighing of ends and means’ (Weber 1991: 220). 4

See also Alan Bullock (1983: 381): There was always more than one office operating in any field. A dozen different agencies quarrelled over the direction of propaganda, of economic policy, and the intelligence services. Before 1938 Hitler continually went behind the back of the Foreign Office to make use of Ribbentrop’s special bureau or to get information through Party channels. The dualism of Party and State organizations, each with one or more divisions for the same function, was deliberate. In the end this reduced efficiency, but it strengthened Hitler’s position by allowing him to play off one department against another.


In September 1939, the Security Service of the SS, a party organization, was fused with the regular Security Police of the State, which included the Gestapo, to form the Head Office for Reich Security (RSHA), commanded by Heydrich. The RSHA was one of twelve Head Offices in the SS, two others of which were the Head Office of the Order Police, which was responsible for rounding up Jews, and the Head Office for Administration and Economy, which ran concentration camps and later the ‘economic’ side of extermination. The RSHA contained Section IV, the Gestapo, divided into Section IV-A, dealing with ‘opponents’, and Section IV-B, dealing with ‘sects’. The higher SS and police leaders were in a different command structure to the twelve offices of the RSHA, while the Einsatzgruppen were under the command of the RSHA, but were not one of the twelve offices (Arendt 1994a: 70).


He adds: [T]here are no scientific methods to decide whether the well-off residents of the Warsaw ghetto could have done more to alleviate the lot of the poor dying in the streets of hunger and hypothermia, or whether the German Jews could have rebelled against the deportation of the Ostjuden, or the Jews with French citizenship could have done something to prevent incarceration of the ‘non-French Jews’. 7(Bauman 1993: 205)


Take the case, which Bauman (1993) cites, of Dr Arthur Gütt, the Head of the National Hygiene Department in the Ministry of Interior, who argued for selective breeding of human beings. Bauman comments that Gütt had no doubt that the policy he envisaged of ‘selection-cum-elimination’ was a logical extension, if not culmination, of the advancement of modern science. But Bauman does not discuss whether the theories of Dr Gütt actually constituted a logical extension of the work of the celebrated scientists, or indeed whether there was any scientific basis whatsoever for his theorizing. Gütt and his colleagues may have been recognized by the Nazis as genuine scientists, but that does

The decision to commit a crime against humanity 199 not mean that we have to accept this recognition. The problem with eugenics was not that it was scientific but that it was not scientific. Bauman seems to accept that Nazi doctors are doctors: that their talk of hygiene, cleansing, blood and purification were genuinely within a medical tradition. But this is to take rhetoric at its face value. 8

Bauman (1993: 26) acknowledges this point: At the Einsatzgruppen stage, the rounded-up victims were brought in front of machine guns and killed at point blank range. Though efforts were made to keep the weapons at the longest possible distance from the ditches into which the murdered were to fall, it was exceedingly difficult for the shooters to overlook the connection between shooting and killing. But Bauman (1993: 26) immediately goes on to claim that this was why the administrators of the Holocaust found the methods inefficient and dangerous to morale: Other murder techniques were therefore sought – such as would optically separate the killers from their victims. The search was successful and led to the invention of ... gas chambers; the latter ... reduced the role of the killer to that of the ‘sanitation officer’. It seems to us that this account not only misconstrues the order of succession between the face-toface and the distanced (what about the ‘death marches’ at the end of the war?), but also misconstrues the organization of murder in the camps (as if the executioners did not have face-to-face contact with those they humiliated, tortured and killed).


Alain Finkielkraut (1992: 3–4) argued, in relation to the Barbie trial, that the Holocaust was ‘from Eichmann to the engineers on the trains ... a crime of employees’, and that it was ‘precisely to remove from crime the excuse of service and to restore the quality of killers to law-abiding citizens ... that the category of “crimes against humanity” was formulated’.

10 The interconnections of thinking, willing and judging and the dangers inherent in the separation of thinking from willing and judging became the subject-matter of Hannah Arendt’s later investigations in The Life of the Mind (1978). 11 There is now considerable evidence that the image which Eichmann presented of himself at the trial, which by and large Arendt accepted, that he was a bureaucrat of mass murder, concealed the far more active role he actually played in the extermination process. 12 The information on this case is drawn largely from notes taken by David Hirsh at the trial of Sawoniuk, at the Old Bailey, London, 1998–9.

12 ‘Race’,1 ethnicity and housing decisions Rational choice theory and the choice–constraints debate Peter Ratcliffe

Introduction: ‘Race’, ethnicity and rational choice theory Applications of rational choice theory in the sociological literature on ‘race’ and ethnicity are relatively scarce, and have tended to be implicit rather than explicit. They have been confined largely to the work of ethnographers and social anthropologists, and have been showcased in edited volumes such as Watson (1977) and Wallman (1979). In a more theoretical vein, there are the significant contributions of Banton (1983) and Hechter (1986). Over the past decade, the approach has to all intents and purposes disappeared, in terms of macro-sociological work at least, although vestiges of it remain in particular substantive areas, as we shall see. Even its strongest adherents, such as Hechter (1986: 265), have expressed some serious misgivings about its usefulness. One admirably clear and succinct explanation of the theory suggests why it has had a certain seductive appeal to many writing about the salience of ethnic and ‘racial’ divisions: Rational choice considers individual behaviour to be a function of the interaction of structural constraints and the sovereign preferences of individuals. The structure first determines, to a greater or lesser extent, the constraints under which people act. Within these constraints, individuals face various feasible courses of action. The course of action ultimately chosen is selected rationally. ... When individual preferences are assumed to be known, transitive and temporally stable, behaviour can be predicted in the face of a combination of structural constraints. (Hechter 1986: 268) Institutionalized discrimination against minorities, on the basis of religion, culture and phenotype, constitutes a very obvious set of structural constraints, over and above those faced by members of majority society (however the latter is defined). Individual hostile acts,

‘Race’, ethnicity and housing decisions 201 grounded in racism or some other exclusionary ideology, also clearly impose constraints on the actions of the ‘societal other’. In a very obvious sense, then, the ‘preferences’ of minorities will be conditioned by such external forces. What this chapter examines is: (i) the derivation of fixed preferences from the ‘constancy’ of discrimination; and (ii) the similarity then presumed to characterize the preferences within and between ethnic ‘groups’, because they are held to belong to something with similar characteristics which confronts similar contextual constraints. The rational choice model raises a number of questions, however. It is based on means– ends decision-making processes at the level of individual actors. But the concept of the sovereign individual is culturally specific, and even if it were not, one would have to concede that decision-making in practice often assumes a collective form. Further, there is the issue of how sociologists come to read/understand the framework of actors’ ‘preferences’, and as Mason (1986: 17) asks: ‘Does ... the predictive power of the theory depend on the actor’s conception of the means–ends relationship being the same as the analyst’s?’ One could also ask whether it makes sense to draw such a rigid distinction between constraints and preferences. For example, do not the ‘shared preferences’ implicit in Islam also represent a system of normative constraints on the social actors concerned? All of these issues will be explored later in the context of urban decision-making processes. The fact that the rational choice model gives primacy to the role of (individual) social actors (albeit subject to structural constraints) largely explains its appeal to ethnographers. The argument is that individuals are concerned with ‘boundary maintenance’, acting (rationally) so as to preserve both their (distinctive/separate) ‘ethnic identity’ and the global integrity of the collectivity (or ‘community’) of which they are a member. This raises a further question: is there not a danger here of essentializing ethnicity, and also of failing to recognize change in social/cultural patterns as a shift in ‘preferences’ which are processually changeable, given that contextual changes allow new goals to be pursued?

Explaining urban inequalities: the utility of rational choice theory In evaluating the utility of rational choice theory, this chapter reflects on some of the key debates in the ‘race’ and ethnicity literature. The focus is on current explanations of social change and urban inequality (more especially housing inequality) between putative ‘ethnic groups’; not only between majority and minority segments of the population but also between various minority ethnic communities. For the moment, I shall leave aside some difficult definitional issues and concerns about ethnic/cultural essentialisms, and simply spell out the sorts of substantive questions which need to be addressed.

202 Peter Ratcliffe However one divides up a given population, it is clear empirically that the communities so defined are differentially placed in terms of housing position. In Britain, as in other advanced industrial societies such as the United States, there are clear, persistent and collective variations in overall housing quality, as well as in tenure, property type and locational patterns (Brown 1984; Jones 1993; Karn and Phillips 1998; Modood et al. 1997; Ratcliffe 1997; Smith 1989, 1991). Associated with these residential patterns are differences, in some cases quite marked, in household size and structure distributions, which in turn often lead to concerns about differential rates of overcrowding and associated health implications. The question is: what can we as sociologists make of all this? In particular, to what extent might rational choice theory illuminate, or even provide an adequate explanation of, the observed differences in housing position? This chapter takes as its starting-point the analytical framework which has dominated the UK literature. Usually known as the ‘choice–constraints debate’, this is in many ways rational choice theory writ large. At its simplest it is, as we shall see, a conflict between the proponents of structural, systemic explanations of difference as against accounts which give primacy to individual agency. It will be argued that both of these approaches lead to explanatory accounts which are overly static, narrow and ultimately naïve. In addition, the underlying processes pose severe challenges to rational choice theory, by virtue of tensions between decisionmaking seen as operating at an individual, as distinct from a collective (household/family), level. What then of alternative models? The chapter follows up one attempt, ostensibly employing Giddens’ structuration theory, which claims to solve some of these problems. It will be argued that none of the existing models, rational choice included, comes close to explaining existing differences, let alone ongoing social change processes. Such a model would need to go far beyond the form of explanation which sees an essentialized ‘race’ and/ or ethnicity as pivotal. On the choice side of the equation, we need to ask what we mean when we say that ethnicity explains substantive differences. More concretely, in what sense does ‘ethnicity’ per se determine patterns of housing choice and ultimately outcomes? We need to assess the efficacy of other forms of explanation which give a higher priority to questions of wealth, occupational class and gender. Finally, an adequate theoretical framework must by definition reject static mono-causal explanations, either of the culturally based ‘choice’ variety or of the constraints school, especially where the latter employs naïve notions of racism(s), whether of the systemic/ institutionalized variety or as evidenced by individual acts of discriminatory behaviour. The latter provokes a rather obvious, yet intriguing, question: might these acts (whether individual or individualized representations/expressions of the ‘institution’/bureaucracy) be conceived of as the expression of ‘choice’ grounded in an alternative frame of ‘(ir)rationality’?

‘Race’, ethnicity and housing decisions 203

Explaining housing differentials: the choice–constraints debate Whether it is avowed or not, the focus on ‘choices made under constraints’ is a clear version of rational choice theory, since it is about the pursuit of supposedly fixed preferences in the face of opportunity costs, which are often presented as being so high that agents cannot maximize desires and are reduced to behaving according to the maxim ‘take the housing you can get’. The essence of the ‘constraints’ model is that minorities occupy the housing they do because they have little choice in the matter. Thus their sense of individual (or household) agency is denied, or at least minimized. As to the source of the constraint, this can be structural/systemic (housing institution-based) and/or resulting from face-to-face discrimination. This is seen to affect not only where people live but also the type and quality of property they occupy. In other words, whatever the preferences of the people concerned may be, structural constraints represent opportunity costs which are far too high for preferences to be realized. The consequence is simply that their preferences are then ignored: this is a population to which things happen, not a set of agents who make things happen. Perhaps the earliest UK example of this approach, and certainly the most celebrated, was the study of Sparkbrook in Birmingham by Rex and Moore (1967). Employing a Weberian approach, they developed the notion of ‘housing class’. This focused on the (housing) market position of minority households, but also linked outcome to identity. That is, the disadvantaged position of minorities did not simply stem from their poor economic position; they were subjected to systematic discriminatory forces which, say, the White working class did not have to face. At the time Rex and Moore’s research was undertaken, there was no legislation to outlaw discrimination on grounds of ‘race’/ethnicity. As is well documented elsewhere, minorities were consistently denied rented housing in the private sector both by estate agents and by individual landlords (CRE 1990; Smith 1989, 1991). Local authorities imposed ‘residence requirements’ which effectively prevented minorities from even gaining access to waiting lists, and those who satisfied these requirements were often subject to adverse assessments from housing visitors, on the grounds of poor ‘house-keeping standards’ (Henderson and Karn 1987; Smith and Whalley 1975). Those who came through all these hoops successfully were invariably ‘rewarded’ with tenancies in hard-to-let property; usually deck-access maisonettes, flats in high-rise blocks or ‘patch-and-prop’ terraced houses, that is, older property purchased by local authorities and made habitable, but often with a short life expectancy. These factors led Rex and Moore to talk about those who were ‘forced to buy’. The problem here was that housing market institutions also found ways of preventing minorities

204 Peter Ratcliffe from buying property. Estate agents frequently marked record cards relating to certain properties as ‘not to be shown to a Black applicant/family’. Financial institutions either refused loans outright, on the (individualized) grounds of alleged ‘poor risk’, or by a formal collective (i.e. institutionalized) process known as ‘red-lining’, resulting in what US commentators label ‘racial steering’. Certain inner urban areas therefore effectively became exclusion zones, unless alternative sources of funding were found. This obviously constituted a major problem for poor migrants of working-class or peasant origin who, simply on financial grounds, were ‘constrained’ to seek housing in such areas. Research suggests that loan finance came from a pooling of resources within kin groups and/ or from the ‘fringe’ banking sector. The problem with the latter was that repayment periods were often very short, and these companies set levels of interest far above base rate. This, when combined with higher prices resulting from the increased demand for (invariably poorquality) dwellings, meant that this particular group of low-income home-owners was subjected to a double ‘colour tax’ (Fenton 1977). Thus, a combination of discriminatory factors set prohibitively high opportunity costs for obtaining good-quality housing, and these alone were held to account for the well-known regularities in the poorer housing of minority ethnic groups. Such an explanation consequently made no reference to the agential preferences of those involved. Alternatively, those adopting the ‘choice’ model have tended, in Britain at least, to come from a fairly small band of social anthropologists. In some ways, this came as a reaction to overly crude and deterministic accounts based upon constraints which suggested that racism alone accounted for the experiences of migrants to Britain. Watson’s (1977) book gave voice to these communities, if mediated, or some would argue distorted, by the social anthropologists concerned. He argued, along with Piore (1979), that it was wrong to treat migrants like ‘shirts’: commodities to be shunted to and fro at the whim of capitalist markets and associated ideological forces. Ballard and Ballard (1977) argued that although discriminatory forces were very real, the aims and strategies of Sikh migrants needed to be recognized. Dahya (1974) went rather further by prioritizing agency over structure. He invoked the notion of an ethnic enclave as a desired goal, that is, as central to a strategy of preserving a distinctive and separate identity, which was designated as the preference to be maximized. The position adopted by the constraints ‘school’ was that, although the maintenance of spatially distinct communities might represent a rational collective response in terms of culture maintenance, especially in the light of high opportunity costs precluding alternatives, no-one would ‘choose’ to live in the particular areas occupied by these groups if given a real ‘choice’. Thus, the ‘choice’, if not ‘Hobson’s’, is at the very least a heavily circumscribed one. People are therefore held to make the best of the worst circumstances. If they cannot exercise real choice, they can at least look

‘Race’, ethnicity and housing decisions 205 on the bright side of their constraints. For instance, they can reflect that to live outside such an ‘enclave’ would be to risk constant harassment and violence. Although subsequent accounts, over the last two decades, have moved on somewhat because the world they are attempting to describe has changed, there have been few attempts to break with this general framework. However, there is a need for an examination of preference reformulation within the changing context of the housing market. Specifically, the Race Relations Acts passed in 1968 and 1976 eventually outlawed some of the overtly discriminatory acts such as ‘red-lining’ and the assessment of ‘house-keeping standards’, although this is not the same thing as saying they ended. Limited empowerment of minority communities through political representation, voluntary agencies and statutory bodies, such as the Commission for Racial Equality (CRE) and local Racial Equality Councils (RECs), also enhanced the efficacy of agency, both individual and collective. Both structure and agency had changed: did not the process of their interplay also change accordingly?

The dynamics of inequality: an alternative framework Despite the shortcomings of the choice–constraints model, most notably its blanket assertions about collective aims and objectives and also negative treatment, writers in this area have tended simply to tinker with its various elements. One of the few explicit attempts to move the debate on was contained in a book based on a study of Bedford by Sarre et al. (1989). They argued that a more flexible, dynamic model could be developed by applying Giddens’ theory of structuration. Unfortunately, there appears to be a considerable disjuncture between ‘the theory’ and the ‘substance’ of the work. One rather gets the impression that the former was tacked on at the end of the study rather than guiding the research, as was suggested by the text. Some interesting ideas emerge, nevertheless, particularly about the impact of change on the interplay between modified structures and modified agents. This point is best illustrated by a couple of examples, one contained in the book and one not. Sarre et al. (1989) discuss the problems faced by Italian migrants in gaining home loan finance by assessing the attitudes of estate agents over time. It appears that one particular ethnic stereotype was common in the town, namely that Italians were not very good with money and were therefore not a good ‘risk’ for lenders.2 It was clear that, despite this negative image of Italians, some lenders did take the ‘risk’ and advanced loans. Subsequent experience proved extremely positive, in that the stereotype was negated by the evidence, and Sarre and his colleagues then witnessed a major sea-change in institutional behaviour. Exchange professionals could be deemed to be acting ‘rationally’ in responding to the ‘facts’, as they now saw them.

206 Peter Ratcliffe The key point here, though, is that it is only the policy which changes: the structure of individual institutions and therefore the physical ‘market’ remain the same. A rather more instructive use of the dynamics of changes in both ‘structure’ and ‘agency’ comes from a look at market changes. The constraints argument tells us that minorities have been consistently discriminated against, for example by ‘racial steering’ and the refusal on the part of estate agents to let, or even sell, property to them, thus forcing many to use word-of-mouth techniques to locate suitable property. Now, it does indeed remain the case that, although formally outlawed by the current legislation (Race Relations Act 1976), these practices continue, albeit in a more covert form than hitherto. The major failure of the constraints model is that it treats all minorities as if they were without agency, and because of this some would argue that it also represents a discourse of disempowerment. The empirical evidence also shows that, far from constituting a universal source of disadvantage, the White-dominated market has provided a significant niche for some minority entrepreneurs. Over the last decade there has been a marked increase in the number of minority-owned and -run estate agencies and finance houses. Although mainly dealing currently with the cheaper end of the market, their presence has ensured a ready source of property relatively free from the ‘colour tax’, alluded to earlier. The point of this example is that the market has changed, and is continually changing, influencing the context within which choices and strategies are conceived and realised.3 In other words, choices and constraints are mutually influential: preferences are not fixed; instead their horizons are themselves conditioned by what is possible. New strategies emerge selectively from those who are not fundamentally constrained, and these in turn contribute towards changing the market, which then becomes the new context for the preference formulation of others. This suggests that developing an effective model of housing outcomes requires us to think about the interaction between a changing institutional structure and the agency of minority individuals and households. It will be helpful at this stage to consider each side of this equation separately.

Preparing the ground for an alternative theoretical approach. I: Questioning the rationality of housing supply The choice–constraints debate implies that ‘choice’, usually assumed to be rational irrespective of its source, is made only by the consumer of housing. But, clearly, refusing to sell or rent property to a particular individual/household, or withholding financial support which would enable a transaction to proceed, also involves ‘choices’. As implied earlier, the

‘Race’, ethnicity and housing decisions 207 exclusionary practices were both varied and widespread, and the institutions involved were multi-layered. The literature criticizes institutions such as estate agents for ‘steering’ minority customers to certain areas and properties: it likewise condemns banks and other providers of housing finance for failing to behave in an even-handed fashion, by instituting policies such as ‘redlining’. It has long berated local authorities for imposing residence requirements, which have a disproportionate impact on incoming minority populations, and for instituting assessments of ‘housekeeping standards’ which are infused with cultural bias if not overt prejudice and bigotry. What the literature fails to ask, however, is why they do it. Are these bureaucracies, and the individual actors within them, whilst acting unethically, nevertheless making ‘rational choices’? This invokes classic sociological debates, initiated most notably by Weber, as to the nature of bureaucracies. Focusing here first on estate agents, one assumes that their rational ‘end’ is to maximize profits. This presumably concurs with the individual employee’s position, in that the retention of one’s post, and the subsequent maximization of income, depends on the number of successful transactions made. How, therefore, can discriminatory behaviour possibly be deemed rational as a means of satisfying the ends of either the individual or the institution in this case? There is some evidence to suggest that discriminatory actions were deemed to be in the best interests of estate agents in multi-ethnic urban locations, on the grounds that selling properties to minorities in ‘White areas’ would impact adversely on prices in those locations, thus affecting their income and general confidence in the market. ‘Tipping-point’ arguments from the United States fed this view. Thus, in this context, the ‘business case’ accorded with the normative behaviour of the self socialized in a racialized milieu. However, two points arise which question the assumed economic rationality of these actions. Firstly, there was little, if any, empirical evidence to support the theory of adverse market effects: if anything, competition from a new source actually boosted the market. Secondly, such actions are illegal under the Race Relations Act 1976 (RRA 1976) and the company therefore risks the imposition of financial penalties (CRE 1988), though the company may feel the attendant publicity might not in this case harm, and might even boost, its reputation amongst Whites. Nevertheless, the assumed and underlying economic rationality does surface in adverse market conditions. Phillips and Karn (1992) and Sarre et al. (1989) conclude that under buoyant market conditions, agents are ‘choosy’ as to their customers, behaving as was suggested in the previous paragraph. In less happy times, agents, their sales personnel and individual vendors/landlords are seen to be more ‘flexible’. This would suggest that whilst acting on the basis of imperfect information, heavily tainted by negative images of minorities

208 Peter Ratcliffe stemming from a socialization steeped in an endemic racism/ethnocentrism, agents do retain a clear rational-economic goal. Lenders have consistently claimed that ‘red-lining’ is undertaken purely on business grounds. Risks are greater, they would argue, in poor inner urban areas, in terms both of the soundness of financial investment and of the reliability of the (almost inevitably less wealthy, probably working-class) client. If anything, they might reluctantly admit that their policy discriminates in terms of class. However, businesses have consistently denied discriminating consciously on grounds of ‘race’; despite being forced to concede the legal point that such policies constitute ‘indirect discrimination’ under the RRA 1976, on the grounds that incoming poor minority households are disproportionately affected by them. In the Bedford case reported by Sarre et al. (1989), lenders were seen to rank minority groups in terms of their character and trustworthiness. Italians, having ‘proved themselves’, were elevated above ‘Indians’ and ‘West Indians’ in the financial pecking order. Thus, these institutions, allegedly on the basis of local empirical evidence, formed a ‘rational’ ranking of groups in relation to the financial risks they were held to represent, but this was buttressed by social constructions of national character imbued with what actors have become socialized to ‘know’ (in other words, they accord with ‘common-sense’ racist imagery). Hence: The Italians are very thrifty people and they always have been. ... We’ve got some terrific savers amongst the Italians ... they always put a good deposit down. (Sarre et al 1989: 285) You have to watch [the Asians] very closely. ... One hates to put it in these words, but one would probably say they were the most devious. (Sarre et al. 1989:287) The West Indians are something altogether different [compared with the White British]. I think they have an entirely different mentality to life from the rest of us. It’s very much mañana. ... They are pretty cavalier about it all. They have a total lack of any sense of urgency. ... It’s hard work to get the money, although you get it in the end. (Sarre et al. 1989:289) These images clearly influenced the way minority clients were treated in the private sector. ‘Steering’ towards certain areas and properties was justified on the grounds that agents ‘knew’ where their clients would want to live, that is, certain locations would constitute a ‘rational choice’. People would ‘fit in’ better in areas where ‘their type’ already lived. In the public sector, similar steering has been well documented (Henderson and Karn 1987). In this

‘Race’, ethnicity and housing decisions 209 case, the ‘choices’ made by housing applicants were found to be heavily conditioned by council officers who were ostensibly there to record the views of clients. Councils would defend their imposition of ‘housekeeping standards’ as a way of reducing maintenance costs, and bolstering the reputation of an area (hence reducing void levels). Underpinning the policies were deeply ethnocentric constructions of what constitutes a ‘high standard’ (Smith and Whalley 1975). The obvious conclusion here is that there are multiple interlocking institutions which adopt policies which are ‘rational’ in the sense that the actors involved, both in making and in executing the policy, conceive of these as being in the best economic interests of the institution concerned. They are irrational to the extent that they are based on false, essentialist, constructions of identity which might lead to legal challenges and, possibly, the imposition of financial sanctions. Rational choice theory runs into difficulty here on the grounds that it cannot deal with the core issue of socialization (Holton 1995) which operates to distort perfect information. If we ignore the vicarious sense of pleasure which individuals might derive from their exclusionary powers, their goal maximization is blurred by irrational belief systems held by the unreflexive self. It would, after all, be difficult to argue that the ethnic stereotypes they hold emerge solely as a result of market interactions. Such actions in the private sector do, however, leave a niche in the market for institutions which deploy more equitable service delivery policies, as we saw earlier. Insofar as these newer, often minority-run, institutions cut into existing markets, they may also undermine the apparent economic ‘rationality’ of the discriminatory practices adopted elsewhere. This once again emerges in the Bedford study, where ever larger numbers of exchange professionals are competing for a market share, and the very existence of some is threatened when economic conditions are harsh.

Preparing the ground for an alternative theoretical approach. II: Questioning the rationality of housing demand We now turn to the demand side of the housing equation, focusing on consumer choices. Thus far, we have largely relied on the concept of ‘minority’ (effectively shorthand for ‘minority ethnic group’) as a global identity construct. This was done partly to avoid reifying the concept of ‘race’, but its use nevertheless invokes ontological claims in respect of ‘groups’: groups comprising individuals/households with a common identity who make choices and/ or are discriminated against on the grounds of their identity, whether phenotype or culturally based. Herein lies the central problem of most past and contemporary analyses of urban inequality, namely essentialism.

210 Peter Ratcliffe For pragmatic reasons, a conventional ‘ethnic group’ schema, such as that provided by the decennial Census of Population, tends to be used in empirical analyses, despite serious concerns both about the meaningfulness of the labels and about the way they have been operationalized and measured (Ratcliffe 1996a). ‘Ethnicity’ is therefore, in the UK, effectively reduced to ‘White’, ‘Black-Caribbean’, ‘Indian’, ‘Pakistani’, ‘Bangladeshi’, and so on. Subsequent analyses effectively reify these collectivities in the sense of implying that they represent mutually exclusive and internally homogeneous entities. Even more important, and less obviously, when related to housing position, they attain an implicit explanatory role in a causal chain. The key question for sociologists to explore is that of what it means to say that ethnicity explains housing choices and housing position. In other words, can individual or even sub-sectional preferences simply be ‘read off from ethnicity, by imputing them from the ‘ethnic group’ to its component members – a move which implies essentialism and commits the ecological fallacy? Here, ‘conventional wisdoms’ abound. (South) Asians are said to prefer to buy property (hence their universally high rates of owner occupation), whilst those of Caribbean origin tend to prefer or to rely upon social rented housing. Both of these positions were endorsed by Sarre et al. (1989). The former also prefer, we are told, to live in close proximity to one another, and segregated from the White population, whereas the latter are more spatially dispersed, at least partially out of choice. But what is it about being, say, ‘Indian’ or ‘Pakistani’ which propels one towards home ownership and segregated settlement? It is implied that it has to do with their ‘cultural traditions’, but this simply invokes the question: what is meant here by ‘cultural’ and ‘tradition’? Might it not simply be that migrants bring with them a stock of knowledge about the social and economic norms and expectations in their home country, that is, the product of a temporally and historically specific set of socialization processes? Moreover, why are these assumed to be immune from change? In fact, the research evidence seems to suggest that for a variety of reasons the initial question is based on a false premise. Higher rates of home-ownership amongst South Asian households do not suggest a higher preference, cultural or otherwise, for this form of tenure (nor, incidentally, do they suggest that there are differentially discriminatory practices at work). There are at least two key additional factors to be taken into account: wealth/ occupational class and household structure. Using the 1 per cent Household Sample of Anonymized Records from the 1991 Census of Population, I have shown (Ratcliffe 1997) that by controlling for household structure differences, most of the tenure differentials between ‘ethnic groups’ disappeared, especially when taking sampling error into account. In the course of research on the workings of the ‘right to buy’ scheme, introduced by the first Thatcher administration in its Housing Act 1980, Peach and Byron (1993) showed that, controlling for income/class and household structure, there was no difference between the

‘Race’, ethnicity and housing decisions 211 Black (Caribbean and African) and White communities in terms of the propensity to buy their council dwellings. Thinking more generally now, it may well be that some minority households are still ‘forced to buy’, as Rex and Moore (1967) argued. In terms of the contemporary situation, though, a much more important issue is that of housing quality. Although those of ‘Indian’ origin display similar levels of owner occupation to those defined as ‘Pakistani’, a much higher proportion of the former live in good-quality detached or semi-detached property outside the inner urban areas which typified the initial settlement areas. Indeed, there is research evidence to suggest a steady movement away from the latter areas (Phillips and Karn 1992), albeit into contiguous areas with a relatively modest increase in the standard of built environment (rather than into the higher quality areas occupied by the White middle class). The key issue, as suggested by the previous comment and paragraph, is ability to pay. Choice may be constrained by familial obligations, cultural ‘preferences’, the fear of harassment and continued resistance from some housing market institutions, but ‘money whitens’. Any analysis of housing differentials and mobility therefore needs to take into account wealth and occupational class position. It also needs to take on board issues such as household structure and social change processes, as will be discussed shortly. Not only do the ‘ethnic group’ labels conceal disparate origins and identities, they also lure the unsuspecting researcher, and more importantly policy-maker, into treating the ‘group’ as if it was the bearer of some homogeneous ‘traditional’ culture with fixed preferences, the essence of which is frozen for all time (hence the dangers of ‘ethnic managerialism’, see Law 1996, 1997). Anticipated regularities at the meso level are predicated on the assumption of agential balkanization. The rapidly expanding literature on cultural hybridity (see Back 1996; Bauman 1996) has yet to influence in any meaningful way the literature on ‘race’, ethnicity and urban inequalities. So, how should we model ‘housing choices’, and does rational choice theory provide an adequate basis for such a model? One has to recognize that all of the above arguments on differentials in housing position imply a probabilistic form of explanation. They do not explain why, say, two households in ostensibly ‘similar’ positions make markedly different decisions. Sociologists obviously have to live with this difficulty unless they adhere to something approaching the extreme empiricist position that if we make a model sufficiently complex, we can explain all decisions. This said, we can certainly do better than we have thus far. There are a number of key issues which do not feature in current accounts of decisionmaking in housing. In this chapter, I have space to explore only two, which, it will be seen, are intimately interrelated. The first is raised elsewhere in this volume, namely the question as to whether the basic unit of analysis is the household, the individual, or both. The second has to

212 Peter Ratcliffe do with the extent to which cultural norms and patterns change over time, and, with them, preferences. Attitudes of first-generation migrants may change (or may, on the contrary, remain rigidly fixed in a way which no longer even reflects the contemporary social and cultural attitudes or norms in their country of origin, thus representing an ‘over-socialized’ self seemingly concretized by a hostile social environment). It is almost certainly the case that the second and succeeding generations, of whatever origin, will not share the elder generation’s approach to life in Britain (irrespective of whether they retain the commitment to honour a ‘traditional’ sense of social and moral obligation). Our two factors are now clearly interrelated, in the sense that the younger generations currently living within the parental or grand-parental home may well have their own independent housing plans: they may seek to maximize something different. To illustrate the complex interplay of the factors involved in the decision-making process, we turn to a study of Bradford undertaken in 1995 (Ratcliffe 1996b).

Housing choices: a case-study of Bradford Future decisions, whether of the household unit as a whole or of a sub-set(s) thereof, have a number of possible outcomes. The first option is clearly the status quo, which for many minority households, especially poorer Pakistani and Bangladeshi owner-occupiers, means continuing to suffer appalling housing conditions and overcrowding in pre-1919 terraces or ‘back-to-backs’. In Bradford, the fact that around half of all Pakistani and Bangladeshi households were found to contain no-one in full time work (Ratcliffe 1996b: 32) meant that there was often little choice but to remain. The failure of successive phases of urban renewal to make a significant impact on the built environment (in particular by providing renovation grants) placed properties on a downward spiral from disrepair to unfitness. The only realistic hope of positive change for the poorer households is via the social housing sector, and this move is often difficult due to size/design and locational problems (of which more later). For those wishing to move, there are a number of considerations. Thinking first of the potential movement of whole households, the key elements of the decision-making process involve tenure, size/design and location. Again, much depends on financial position, which reminds us that housing position cannot be dissociated from economic position and associated inequalities. But there are clearly other areas of concern. Insofar as the local economic as well as social, cultural and religious infrastructure is well developed, households may decide to move locally. They may, of course, also do so for the reason noted earlier, that of avoiding possible ‘racial’ attacks, or at least harassment. The decisions they make have obvious implications for planners, but they also raise important sociological questions concerning the ways in which households negotiate the means–ends nexus. Insofar as

‘Race’, ethnicity and housing decisions 213 households wish to leave an area of concentrated settlement, and perhaps even move some distance from kin, what impact will this have on social/kin networks? Do those of (South) Asian origin reinforce existing tenure patterns by buying property, or do they consider social housing? In the Bradford study, the moving intentions of the African and Caribbean populations suggested that there would be little change from the current situation, either in tenure or in locational terms. Few had the resources to move from social housing to owner-occupation, and the population looks likely to remain fairly dispersed in spatial terms. The moving intentions of the South Asian groups, however, raised some interesting questions. In terms of tenure, around half of those who planned, or wished, to move expected to quit owner-occupation, this group being evenly split between the two principal alternative options: renting from the council or a housing association (now known collectively as Registered Social Landlords). Were this even to be partially realized, we would be witnessing a major change in tenure patterns, which would severely dent the ‘ethnic stereotype’ noted earlier. It appears that a changing housing market, involving new actors on the scene in the shape of minority-run housing associations, combined with an increasing level of knowledge about housing options, was resulting in shifts in the way in which local people approached the question of housing provision, that is, the preferences which they now entertained. There are clearly echoes here of the earlier structure–agency debate, in the sense of an ongoing interaction between a changing structure (in this case housing market) and agency (constantly subject to a process of preference re-evaluation). Thereafter, the increasing popularity of certain forms of housing, and crucially people’s first-hand experience of them, will provoke further changes in the market, possibly through new forms of shared-ownership, specialist provision for the minority elderly, and so on. In terms of location, it was ironically the younger cohorts of home-owners who were more likely to wish to stay within, or remain very close to, the existing areas of settlement; ironic because many of the mainstream texts have suggested a tendency for spatial integration, as part of a move towards social integration. The most likely explanation is a combination of factors: to develop and reinforce an identity which is in a sense becoming consciously and ‘by choice’ hybridized (though not necessarily totally divorced from ‘tradition’); to fulfil obligations to family/kin; to maintain cultural/religious and economic links (given a wider context where formal paid work is scarce); and knowledge of potentially hostile social environments elsewhere. It should perhaps be added, however, that as many as three in ten South Asian householders regarded the ethnic composition of a location as unimportant, a finding which reminds us of the heterogeneity of the ‘groups’ concerned (Hamzah and Harrison 2000).

214 Peter Ratcliffe If only implicitly, the discussion thus far has focused on household choices. There are many situations, however, where decision-making becomes devolved or fragmented. The Bradford research was original in that it attempted to assess the housing strategies of individual adults within households who were neither the ‘household representative’ nor her/ his spouse/partner. This was to assess the likelihood of significant shifts in ‘headship rates’ and to throw some light on the question of ‘concealed families’. It is important in sociological terms, and especially so in the current context, because it generates crucial data on decisionmaking processes. The basic idea behind the notion of ‘concealed family’ is quite simple, but also highly problematic in the context of research on minority communities. It constitutes a normative framework with the nuclear (or at least single unit) family at its core. As such it leads to the generation of official statistics which are heavily imbued with Eurocentric and even Anglocentric ideas and culture. The underlying assumption is essentially that household units which are outside the core family share a given dwelling out of necessity (usually deemed to be financial), and when circumstances permit they will wish to leave to form a separate household. This, then, forms one of the bases on which household projections are computed. The definition offered by the Department of the Environment (now the Department of Environment, Transport and the Regions) specifies three types:

concealed married couple family: a married couple family living within a household where another person is household representative;

concealed cohabiting couple family: a cohabiting couple family living within a household where another person is household representative:

concealed lone person family: a lone parent with dependent children living within a household where another person is household representative.

The obvious point to be made here is that although this might ‘work’ quite well in the case of a mono-ethnic (Anglo) population, such a population does not exist outside the minds of government statisticians. The Bradford data suggested first that (vertically or horizontally) extended or joint households were very much in a minority even among the South Asian population. Secondly, only one in five was likely to fragment over the foreseeable future, irrespective of financial considerations. Thirdly, sub-units which might conceivably break away from the ‘core household’ were likely to remain in the locality, so that the original unit would effectively spread across contiguous, or near contiguous, property/streets. In short, the official household projections falsely inserted a presumed move towards more individualistic decision-making, which the Bradford data cannot support.

‘Race’, ethnicity and housing decisions 215

Rational choice theory and urban inequality in polyethnic societies: some brief concluding thoughts The minority ethnic populations on which this chapter has focused have, in varying degrees, endured sustained racist abuse and discriminatory behaviour during their life in Britain, irrespective of wealth and class background (Dipanitta Basu, personal communication, November 1997). They have been targeted not on the grounds of their self-identity but because of an ascribed identity, rooted in a history of colonial exploitation, with phenotype and/or cultural differentiation as signifiers. Most of the points made about housing ‘decisions’ need to be seen against this background, in that ‘choices/preferences’ are invariably constrained by these forces. In the case of some other minority groups, say of Old Commonwealth or European origin, these forces either will not apply, or will be somewhat different in form. Rational choice theory has an immediate, even seductive, appeal in the context of housing. People have housing preferences, make choices on a rational basis, and in doing so encounter external structural forces which impose constraints on their actions in the form of prohibitively high opportunity costs. In the case of minorities, the latter constraints are all too clear. Discriminatory behaviour at both individual and institutional levels, combined with the fear of harassment, inevitably influences housing outcomes. Moreover, as we saw earlier, these expressions of exclusionary behaviour constitute a further form of (ir)rational choice on the part of the actors involved on the supply side. This superficial resonance with common-sense reasoning should not, however, lure us into accepting the theoretical package as a whole. There is a static reductionism which belies the complexity of the world which it is purporting to represent. One cannot apprehend the subtle nuances of means–ends schemata for individuals, let alone groups. And individualism, as we have argued, is in itself culturally laden. Decision-making is a complex dynamic process: to project individual (or for that matter household/kin-based) preference schedules and resulting action frameworks onto putative ‘groups’ or ‘communities’ is deeply problematic. Furthermore, although the use of such collective constructs is commonplace, as in the current chapter, one should be aware that this is little more than a heuristic device, a convenient shorthand. One does not have to buy into the postmodernist project to reject essentialisms, ‘ethnic’, cultural or otherwise. The dangers of the rational choice paradigm become particularly evident when one translates it into the urban policy sphere. Assuming the existence of a liberal regime which wishes to meet the needs of ‘minority communities’, it leads unerringly towards a form of ‘ethnic managerialism’. Here, the ‘needs’ of particular communities are simply read off a policy impact assessment chart which embodies assumed collective means–ends constructs.

216 Peter Ratcliffe Thus, ‘Asians’ prefer to buy property, especially in certain localities because of the availability of ethnic support networks, thus implying little demand for social housing; ‘Black-Caribbeans’ prefer to rent social housing, owing to the prevalence of lone parent households, and so on. Such reductionism based on rigid, and unchanging, models of preferences and action frameworks demonstrates scant regard for ‘difference’ in ethnic/ cultural terms (Hamzah and Harrison 2000), and for the process of shifting preference formation in the face of both market and policy changes. Perhaps even more important, both in theoretical and in practical policy terms, such essentialism tends to ignore other forms of social identity cleavage, most notably class, wealth and gender. In short, an effective theory of rational choice would, quite apart from clarifying the notion of rationality, need to deal with the question of intersecting identity components which spell real heterogeneity in what are presumed to be homogeneous groupings. It would also need to deal with the influence of socialization on decision-making, and with the obvious point, demonstrated in this chapter, that decision-making is often as much a collective as an individualized process. It is not easy to see how this might be achieved within a framework which denies the processual nature of preference formation, but regards these preferences as fixed, by invoking ethnic essentialism.

Notes 1

‘Race’ is placed in inverted commas in order to draw the reader’s attention to its deeply contentious ontological status. In so doing, I also conform to probably the most common European convention. I wish to deny the term any claim to scientific status, or as representing anything ‘real’, but to reaffirm its position as an ideological construct which continues to invoke significant material effects.


This notion could be seen as a simple empirical manifestation of the sort of ‘common-sense ideology’ analysed by Lawrence (1982).


Much the same argument could be employed in the context of the local economy in areas with significant levels of minority concentration, in that new market opportunities are available to the ‘ethnic entrepreneur’ with the appropriate knowledge and awareness of consumer needs.


13 ‘When the battle’s lost and won’ James A. Beckford

One could be forgiven for thinking that a pitched battle is raging in the social sciences. One side appears to be composed of monarchists who are fighting to impose an absolutist sovereign on the throne of theory. I’ll call them ‘Royalist Cavaliers’ (or ‘rational choice’, as some of them call themselves). Their opponents are assorted defenders of a cause that I’ll call ‘the Roundhead Tendency’ (or ‘rounded theorists’, as some of them claim). As in all good battles, there is more posturing and gesturing than actual fighting. Nevertheless, salvoes fly back and forth from time to time. Champions from each side occasionally grapple with one another in public view. Momentary truces are called. Conferences are held, experts are summoned, soothsayers are consulted, and new recruits are called to the colours. But, unlike in real battles, not a single Cavalier or Roundhead ever dies; no hostages or prisoners are taken; there are no mutinies or defections; and the ammunition never runs out.1 Only the width of no-man’s-land changes with the flux of combat. The reason why victory and defeat never really occur in the battle between rational choice and rounded theory is that each side is fighting against a caricature of its enemy that represents its own worse fears. Royalists fear a world of chaos or sloppiness where no single principle of theoretical order is observed. Roundheads fear a bleak world of colourless uniformity where a single principle of instrumental rationality is all-powerful. Each side will go on battling with its caricatural enemy until, perhaps one day, they both realize that their fears bear little resemblance to the reality of their opponents’ actual beliefs and conduct. They may also come to borrow so many of their opponents’ ideas that the causes of war will be all but forgotten over time. Paralysis by a thousand subordinate clauses may bring hostilities to a virtual cessation, if not resolution, eventually. Alternatively, a more threatening battlefront may open up elsewhere (e.g. against socio-biology or evolutionary psychology). The new call to arms may then make allies of former enemies; and aggression can be turned against a new foe. The Cavalier cause has no true believers among the contributors to this volume. Most of us are Roundheads by inclination and training. Some have been wounded in skirmishes or

220 James A. Beckford hardened in battle. But all report faithfully on the state of the fighting in their particular sectors. The strategy of some is to mount a head-on attack against the very principles on which their opponents operate (Wagner, Archer, Cruickshank). Others specialize in guerrilla tactics, sniping against isolated targets (Procter, Fine and Hirsh). A variant on this tactic is to probe specific weaknesses and gaps in the enemy’s defences, exposing particular omissions and inadequacies (Tritter, Peggs and Lampard, Williams). Finally, a few contributors are less interested in defeating an enemy than in securing a more defensible redoubt from which to conduct their own research (Parker, Wolkowitz, Ratcliffe). My role is analogous to that of the military artist or chronicler: it is to try to capture the significance of the struggle and to compare it with other theoretical disputes in sociology. My perspective is unavoidably partial. My theoretical sympathies are Roundhead rather than Royalist, but this will not prevent me from appreciating the merits of rational choice. Indeed, I shall argue that there is a place for rational choice theorizing in sociology, but that it is considerably less foundational or indispensable than some of its advocates claim. Examples from the sociological study of social movements and religion will support my argument. They suit my purposes well because they are not domains in which rationality is widely credited by social scientists with a major part to play. This is not the place to discuss all the varieties of theoretical approach that share the rational choice theory (RCT) label (but see Coleman and Fararo 1992a; Goldthorpe 1998; Hogarth and Reder 1987; Zafirovski 1999b). For my purposes it is only necessary to stress the wide range of possibilities and the family resemblance among them. Finer distinctions are required for other purposes, but I simply wish to stipulate that most versions of what I refer to as ‘RCT’ share at least the following basic assumptions:

Sociological explanation should, in principle, be expressible in terms of the properties, beliefs, dispositions and actions of individual human beings.

Human agents normally try to act in accordance with their own sense of what is rational or reasonable in the light of their knowledge and preferences. Their choice of action may take into account the likely actions of other actors.

Attempts to act rationally at the level of individuals generate patterns of social relations and cultural meanings that may or may not be rational by the standards of the actors concerned.

These three assumptions amount to a ‘prophylactic postulate’ (Andreski 1964) in the sense of being rules of thumb designed to pre-empt the unthinking or uncritical reification of collective phenomena. They do not question or challenge the reality of collective and emergent social phenomena; they merely hold that, at some point in sociological analysis, the

‘When the battle’s lost and won’ 221 articulation of individual-level and collective-level phenomena should be clarified (either by imaginative modelling or by empirical investigation), and that a useful starting-point for this process of clarification is the assumption that human actors normally try to attain their goals by the most efficient means available to them. But I do not find it helpful to use the language of ‘optimization’ in this connection, contra Coleman and Fararo (1992b: xi–xii).

‘The centre cannot hold’ Disputes about RCT bear a striking resemblance to some other theoretical controversies in sociology. In the first place, firm positions are taken and defended in publications that become authoritative, if not iconic. The clearest and most exaggerated arguments tend to prevail in these pioneering works. A small number of champions thrust themselves forward to speak for a much larger number of interpreters and practitioners. The debate then takes place largely through the medium of the champions’ canonical works. Secondly, the strain towards definitive, triumphant formulations of theoretical purity elicits iconoclastic responses or rebuttals that also seek to establish impregnable positions. Again, the tendency is for the sharpest rebuttals to monopolize the field and to serve as the main vehicles of dissent. Polarization of opinion is the outcome. Nuanced, moderate or complicated contributions are obscured by the blunter battering rams propelled by would-be champions. Thirdly, a sectarian inwardness and concern for purity spreads through the opposing theoretical camps. Camp followers gather for safety around their champions on condition that they ‘walk the walk and talk the talk’. Seeking common ground with the uninitiated could be construed as fraternizing with the enemy. Only conformity is rewarded. Scepticism and indifference are equally heretical. Can this sectarian dynamic ever be broken? The history of sociological theory is littered with the relics and remnants of sects, the true believers of which have either abandoned them or been converted into followers of a new faith (Mullins 1973). But this rarely happens as a direct result of a decisive blow delivered by one of the champions. It is much more likely to occur imperceptibly as the work of ideological purists is bypassed by more modest scholars whose watchwords are caution and curiosity. That is to say, their suspicion of triumphant mantras makes them curious about the applicability of formulaic theories to the aspects of the social world that interest them most. For example, they may want to examine things in greater detail than was ever attempted by the theoretical champions. Alternatively, they may want to see what happens when concepts from different theoretical languages are combined. Interesting questions generate the kind of careful analysis, theoretical as well as empirical, that has the potential to challenge sectarian attitudes and to foster deeper understanding.

222 James A. Beckford As with other theoretical disputes in sociology, the passage of time has already begun to foster less polarized and more constructive attitudes towards RCT (e.g. Goldthorpe 1998; Zafirovski 1999b). The stimulus for these changes has arisen in many cases from attempts to subject stark versions of theoretical logic to empirical test. That is, scholars seeking to discover how far their empirical investigations or their conceptual elaborations could benefit from being framed in RCT terms have felt the need to modify some of the theory’s assumptions, to emphasize contextual considerations or to import insights from other perspectives. This is true of many contributions to this volume. When the intention is no longer primarily to construct and defend a theoretical bastion simply for its own sake or for the sake of destroying a competing theory, the opportunities for dealing with complexities, exceptions, extensions and special cases can expand. In the process, the theory’s analytical usefulness comes to count for more than its distinctiveness, coherence or elegance. This is not to say that the underlying concerns with, for example, ontological and epistemological assumptions merely evaporate into insignificance. On the contrary, they tend to remain unresolved because they are irresolvable metaphysical issues. Meanwhile, curiosity about the social and cultural world has inspired the contributors to this volume to search for the limits of the applicability of RCT and/or to enhance its capacity, in carefully specified conditions, to throw explanatory light on puzzling phenomena. Although the results of this type of work are often highly critical of aspects of RCT, they nevertheless contribute towards the theory’s refinement in the long term.

Ring doughnuts Some people prefer their doughnuts to be like jam-filled, solid dumplings. Others prefer ring doughnuts. On the principle of de gustibus non disputandum, however, I have no intention of comparing the relative costs and benefits of these two types of confection. I merely wish to use the analogy between doughnuts and sociological theories in order to characterize the contribution that this particular collective volume makes to thinking about decision-making and RCT. Some theoretical debates progress outwards from a solid centre of widely understood assumptions, definitions and propositions towards an outer layer of empirical and conceptual tests or modifications. Examples include Marxian theories of class struggle, normative-functionalist theories of religion or social interactionist theories of crime and deviance. But it seems to me that debates about RCT have taken a different form. The central assumptions are simple but so contentious that they never solidify. The middle of the doughnut is virtually empty. The main substance lies in the surrounding ring where attempts to apply RCT to the world of social and cultural life are concentrated. There is certainly philosophical interest in questions of rationality; and the principles of utility maximization in conditions of imperfect information are of great concern to many game theorists and

‘When the battle’s lost and won’ 223 economists. As far as sociologists are concerned, however, RCT is more of a target of negative criticism and general scepticism. Their preoccupation with, for example, social change, collective identity and conflict, and the normative bases of social order is bound to be in tension with a theoretical approach that is methodologically individualist and primarily focused on competition and calculation. Yet, contributors to this volume show that even a ring doughnut can be tasty. Many of their arguments, far from being fixated on the problematic ‘centre’ of RCT, grapple with the question of how well this particular theoretical approach can make sociological sense of a wide range of social and cultural phenomena. Their answers can be summarized under two general and interrelated headings.

Degree of complexity Firstly, strong claims about the status of RCT as an, if not the, indispensable foundation for sociological explanation receive no encouragement in this volume. Whilst RCT can be helpful in exposing some of the procedural aspects of decision-making, the price paid for this benefit is high. For the process of explaining how individual human actors seek an optimal relation between their goals, preferences and means tends to simplify or reduce matters to such a degree that the very ‘humanness’ of social action is lost. Nobody denies that calculation of means–ends rationality runs through many spheres of life and social situations. Nobody claims that sociologists should abandon the examination of instrumental rationality. But a theoretical approach that reduces social complexity and richness to a single dimension, even for analytical purposes, is an unproductive distortion of reality. In Peter Wagner’s memorable phrase, ‘the messy richness of the world is exchanged for some clean rags’ in RCT. A recurrent theme of this volume is, therefore, the need for sociologists to take seriously the texture and intensity of social life by maintaining a focus on precisely those aspects that are ignored, under-valued or marginalized by RCT. Passion, emotion and moral seriousness are high on the list of omissions or exclusions from RCT’s range of considerations. But, according to Margaret Archer, these missing elements cannot simply be factored into an improved or ‘multidimensional’ model of rational choice because they are incompatible with methodological individualism and the logic of self-interest. What is required, then, is an approach to sociological understanding of decisionmaking that preserves the rich and complex texture of social life. This entails due concern with questions of rationality, of course; but they can only be properly considered in conjunction with the ‘ultimate concerns’ that have the capacity to inspire human agents to be critical of their circumstances and to act as independent agents in, for example, social movements. The pursuit of moral ends for their own sake is therefore no less a constant of

224 James A. Beckford decision-making than is the pursuit of rational efficiency. And the latter cannot be reduced to the former. As a result, RCT is fundamentally flawed for leaving out emotion and morality from its account of decision-making. Archer’s case is supported by the conclusion of Robert Fine and David Hirsh’s chapter. Against Zygmunt Bauman’s allegation that Nazi crimes against humanity were the product of the type of rational action that is enshrined in RCT, Fine and Hirsh argue that, whereas evidence of short-term rationality and of banal bureaucratic attitudes towards the systematic persecution of Jews and other minorities is plentiful, Nazi bureaucracy actually departed in highly significant ways from ideal types of modern bureaucracy. They also show that it was rational for some Jews to resist the Nazis and for others to comply with the procedures of the Holocaust – just as some German and Polish citizens exercised rational and moral judgements about the extent of their willingness to be involved in the atrocities. In other words, the appearance that rational choice served as the motivation, the vehicle and the sanctification for crimes against humanity is deceptive. This appearance is not without foundation but it conceals the importance of distinctly non-rational factors in the Holocaust; and it downplays the force of the moral resistance against such crimes. Thus, just as Archer denies that RCT could ever provide an adequate sociological account of decision-making because it omits emotion and morality, so Fine and Hirsh find that Bauman attributes too much importance to the role of rational bureaucracy and too little importance to moral considerations. Again, the implicit charge is that RCT cannot obtain a proper purchase on the complexity of social life. The tendency for RCT to reduce complex social realities to the rational choices of individuals smacks of essentialism, according to Justin Cruickshank. His analysis of debates about the existence of an ‘underclass’ in advanced industrial societies leads to the conclusion that RCT obscures the dynamic relations between agents and social structures. It does so by treating social structures or emergent collective properties as the dependent effects of individual agents’ independent decisions. But this abstraction of individual agents from their cultural and social contexts is allegedly a sleight of hand because it illicitly assumes that these actors are fully socialized without explaining how they can be simultaneously independent and shaped by their social and cultural environment and upbringing. In other words, social and cultural effects are smuggled into RCT models of supposedly independent agents, thereby conflating the analytical levels of agency and structure. The assumption that they seek to maximize their utility is therefore a claim about their essential nature but it is also in contradiction with the implicit assumption that human beings are normally socialized into collective patterns of thought, feeling and action. According to Cruickshank, this contradiction comes to the surface in Charles Murray’s (1984) argument that members of the ‘underclass’ are somehow condemned to make a rational choice for welfare benefits over low-paid employments. Put differently, members of the ‘underclass’ allegedly exercise their

‘When the battle’s lost and won’ 225 essential rationality as independent individuals to choose to be dependent on collective resources. Cruickshank concludes that such contradictions could be avoided if sociologists routinely distinguished between agents, on the one hand, and social and cultural structures, on the other, for the purpose of analysing the dynamic relations between them over time instead of conflating them into a single, essentialist notion of the instrumentally rational, independent actor. This is a further argument for the need to preserve the complexity of social life against RCT’s tendency to reduce complexity to artificially simplified models. Not all of the failings of RCT identified by other contributors are at such a fundamental level, however. For example, Jonathan Tritter’s analysis of the process of deciding whether schools should opt for grant-maintained status (GMS) emphasized the crucial role of headteachers’ estimates of the personal benefits that would accrue to them in terms of professional power and managerial kudos. It also made the point that perception of the costs and benefits of GMS changed over time and was affected by the decisions being made by other schools. Many different, non-utilitarian considerations entered into the calculations of costs and benefits. Now, much of Tritter’s argument could be expressed in terms of RCT; but the argument would only make sociological sense if all the contextual and historical complications were added. The paradox and the irony would certainly be lost if the decisionmaking process were reduced to its instrumental essentials. A similar point emerges from other chapters. People making decisions about marriage and choice of partners, according to Ian Procter, face the inescapable fact that they cannot know in advance how their marital relationships will develop and how their marriages will affect each partner separately. These unknown and unknowable factors are beyond the control of individuals but are nevertheless crucial determinants of the long-term outcomes of marriage. The changing popularity, frequency and duration of marriages cannot therefore be explained exclusively in terms of the choices that individuals make. Models of marriage that take account of nothing but individual-level decisions not only simplify the process but also obscure the larger picture. The importance of keeping the larger picture in mind is also the burden of Andrew Parker’s chapter on the choices confronting higher education students today. Without dismissing the advantages of RCT, he insists on the need for sociological analysis to take account of the varied and shifting character of the external circumstances affecting students’ choices. Instrumental rationality is virtually a constant in his analysis, but he argues that it is unhelpful to ignore the effects of economic and cultural factors, at the societal level, in precipitating the particular issues to which students have to respond and in shaping their responses. Parker’s argument is that the limitations of RCT should be supplemented by a more rounded analysis of the contexts in which decision-making takes place.

226 James A. Beckford

Level of agency Secondly, RCT tends to assume that it is individual human beings who take decisions in the light of their preferences, goals, knowledge and resources. This is another aspect of the tendency to reduce complex social phenomena to a dimension that is convenient from the sociologist’s point of view. Some rational choice theorists pursue this particular approach to the point of analysing the psychological dispositions of individual actors insofar as they have a bearing on decision-making. It is more common, however, for analysis in the rational choice tradition to proceed on the basis of what can be known about actors’ beliefs and preferences rather than their psychological states. Yet several contributors to this volume have good reasons for regarding even this practice as unhelpfully narrow. Of course, nobody denies that, in some senses, it is individuals who take decisions. But this apparently commonsensical view actually conceals the fact that the effective level of decision-making can be higher than that of individuals. The social character of decision-making is a major theme of Carol Wolkowitz’s chapter on ‘cultural workers’ who decide to combine their place of residence with their place of work. For, while many of their decisions reflect impeccably instrumental rationality and high levels of self-reflexivity, these ‘home-located cultural workers’ also live in social worlds where the division of domestic labour remains systematically gendered and where individual decisions are ‘cushioned’ by inheritances that helped them to obtain mortgages, help from kin and neighbours or the employment of paid domestic services. Moreover, their preferences change over time in accordance with career developments, economic fluctuations and life-cycle changes among cultural workers and their social networks. The complex interactions among these changing circumstances constrain individual decisions in ways that could not be plausibly explained merely in terms of individual-level rationality. The need for sociologists to take proper account of the changing circumstances in which decision-making occurs is also heavily emphasized in Peter Ratcliffe’s analysis of decisions made about housing among minority ethnic groups in Bradford. Neither external circumstances nor personal preferences are ever static. Changes in legislation, local government policies, housing markets and economic climate are normal; and personal preferences change in accordance with, for example, household composition and phases of the life-cycle. Consequently, the balance between structural constraints and personal preferences is too delicate and elusive to be reducible to a matter of the instrumental rationality of individuals. In any case, decisions about housing are often taken at the level of households – not just because ‘joint family’ structures are sometimes normative, but also because of the economic exigencies of disadvantaged groups contending with prejudice and discrimination. In these circumstances, ‘rational’ choice is important but not necessarily confined to the self-evident calculations of costs and benefits made by would-be

‘When the battle’s lost and won’ 227 householders alone. Public policies about housing that are built on RCT assumptions therefore run the risk of failing to appreciate the complexity and the time-sensitive character of decision-making. In addition, decisions about marriage, divorce and remarriage are also inescapably social, according to Kay Peggs and Richard Lampard, and cannot be adequately explained if they are reduced to matters of individual choice alone. This is partly because such decisions are taken against a backdrop of shared norms and values, and partly because those norms and values help to constitute the individual actors’ definitions of costs and benefits. In other words, there is an internal, but not necessarily deterministic, relation between prevailing social norms and individuals’ views about the costs and benefits of marriage and divorce. Marital decisionmakers are not simply dispassionate accountants: they employ a currency that carries an emotional charge reflecting collective, but not universally shared, evaluations of particular forms of intimate relationships. It follows that to abstract individuals from this collective setting is to distort social reality to the point where the very possibility of sociological understanding is jeopardized. In particular, such an approach would exclude from consideration unequal distributions of power and other social resources between the individual parties to a decision. Gendered relationships, for example, often involve inequalities that shape individuals’ assessments of the costs and benefits associated with decisions about marriage, divorce and remarriage. A similar point emerges from Simon Williams’ argument that human emotions have a social character and that, far from being antithetical to reason or rationality, they are an integral part of decision-making (Barbalet 1998). Indeed, rationality has its own shared emotional register and cannot therefore be adequately approached by means of individuallevel expressions on their own. Emotions are relational phenomena that are not external to decision-making but are constitutive of the very things that human beings feel strongly about. To separate emotions from the process of decision-making is therefore to remove one of its social properties and thereby virtually to disqualify it as ‘social action’. Although I have grouped assessments of RCT’s capacity to explain decision-making under only two general headings, the range of critical responses is impressively wide. The need to avoid distorting the complex reality of decision-making by reducing it to utility maximization is agreed by all contributors. They also agree on the importance of examining the social contexts of decision-making by taking account of factors and settings that operate at a level above that of individual human beings. Yet there are significant differences of opinion among the contributors about the actual or potential value of RCT as an approach to the explanation of decision-making. Some regard RCT as fundamentally flawed on grounds of ontology or epistemology. Others regard it as, at best, an aid to partial understanding. Some see no prospect of remedying the fatal flaws; others nevertheless hold out the prospect of building elements of rational choice into more rounded theorizing in sociology. In the final

228 James A. Beckford section of this chapter I shall discuss the conditions in which it seems to me appropriate and defensible to regard some insights of RCT as potentially helpful to sociological analysis, whilst bearing in mind all the strictures and reservations expressed by other contributors.

A bird in the hand and a baby in the bathwater The intellectual ancestry of RCT owes much, as Peter Wagner has explained, to the philosophical and political currents that fed into ideas of modernity – with all the ambiguities surrounding the key notions of ‘individuality’, ‘rationality’ and ‘freedom’. Indeed, these ideas have such a strong grip on many fields of social science that they amount to the default option in social theory. This means that individualist rationalism is widely assumed to provide the basis for explanations of modern social phenomena in all circumstances except those which call for special explanation because they are considered exceptional departures from the ‘normal’ patterns of modernity. Exceptions, according to Wagner, occur when social configurations are being destroyed and/or created, for example in times of revolution and warfare. Far from challenging the dominant assumptions about the normality of individualist rationalism, however, the exceptional only confirms the assumptions’ correctness by showing that major upheavals occur when individuals fail to exercise rationality. Furthermore, individualist rationalist assumptions have often recommended themselves not only as the best hope for an explanation of social upheavals but also as the best foundation for public policies designed to (re-) create modern social order in the face of, for example, massive migratory movements, disruptions caused by world wars, and rapid urbanization. In Peter Wagner’s view, then, RCT is an outgrowth of, and a contribution towards, specific social and cultural developments, the effect of which is to constitute human beings as instrumentally rational individuals and their societies as the products of myriad individual quests for self-interest and optimal utility. In this perspective, social order is not to be understood primarily as the product of tradition, shared values, metaphysical assumptions or system requirements. Rather, RCT presents social order as a provisional, shifting equilibrium produced by the pursuit of self-interest on the part of instrumentally oriented individuals. I agree with Peter Wagner and with most of the other contributors to this volume that RCT redescribes social life in drastically stripped-down and flattened-out forms which omit much of human life’s complexity and moral texture. And I recognize that RCT is not an objective, morally neutral, timeless and universally valid metric for gauging all social activity. A bird in the hand may not, after all, be worth two in the bush unless all avian qualities are reduced to a single comparator. But I now wish to add some countervailing considerations that go some way towards preserving a place, albeit a limited one, for RCT-related ideas in the repertoire of workable theoretical perspectives. These considerations stem in large part from the conviction that social and cultural life in advanced industrial societies has evolved in a

‘When the battle’s lost and won’ 229 symbiotic relation with prevailing social science theories because they have all been influenced and shaped by the forces that gave rise to forms of modernity, including its ambiguities and contradictions. For example, the pressure to adopt rational means of controlling and mastering the natural environment, the social world and individual minds is not simply an ideological reflex; it is also a real aspect of the human experience of modernity. The long tradition of critical theory in its many guises is testimony to the seriousness with which thinkers have responded to this pressure. Similarly, the increasing sophistication, scope and power of formal organizations in all spheres of life are further evidence of the extent to which highly instrumental forms of rationality impinge on us. In short, the impact of instrumental rationality is too immediate and powerful to be safely ignored by good social science. What reasons do I have for thinking that there is a baby that should not be thrown away with the bathwater? My first main reason for wanting to resist the temptation to reject RCT out of hand is therefore that it promises to address an important feature of daily social life by asking how far the instrumental rationality that is especially strongly embedded in the organizational realms of advanced industrial societies is reflected in the dispositions of individual human beings. An example relevant to two chapters in this volume concerns marriage and ‘partnering’: to what extent are decisions about choice of would-be long-term partners affected by the growth of bureaucratic processes for labelling and registering certain kinds of ‘unions’ (legal marriage, common law marriage, pre-marital contracts, ‘civil pacts’ in France, etc.), the marketing ploys of banks, insurance companies and building societies, local government policies and the state’s fiscal regime? In view of this ‘force field’ of formal organizations, it is interesting and legitimate to ask questions about changes that might be occurring in the range of considerations that wouldbe (and deterred) partners take into account. The degree of instrumental rationality that they deploy in their decision-making, consciously or otherwise, would provide a valuable insight into one aspect of social and personal life that is often presented as being beyond the realm of rationality. There is a strong affinity between this first reason for investigating the instrumental rationality of individuals or groups and Jürgen Habermas’s (1985) notion of the ‘colonization of the lifeworld’ and Alain Touraine’s (1971) depiction of the ‘programmed society’. These theoretical interests converge at the point of asking how far the dominance of formal organizations has instilled, impelled or elicited correspondingly instrumental responses in or from individuals. They also explore the possibilities of resistance to ‘colonization’ and ‘programming’. In making this argument, I am not endorsing RCT’s assumptions. And there are many other ways of approaching the study of ‘organizational society’. But I simply wish to make the point that RCT raises some important questions about it. Moreover, the battery of concepts and empirical indicators that have already been refined in the process of elaborating RCT may incidentally provide some of the tools required for investigation of individuals’ responses to formal organizational

230 James A. Beckford rationality. The scope for other, more direct, normative challenges to ‘organizational society’ remains wide and undiminished by the relatively slight contribution of RCT. My second reason for wanting to preserve a place for RCT in the repertoire of sociological theoretical perspectives is closely related to the previous point. It arises from the realization that instrumentally rational forms of organization and efficient use of resources can help to mobilize support for causes that are centrally concerned with non-instrumental norms and values. Beginning in the 1970s with the pioneering work of McCarthy and Zald (1977), the so-called resource mobilization (RM) perspective sought to explain the differential success of protest campaigns and social movements in terms of, for example, their success in overcoming the ‘free-rider’ problem and in activating social networks of supporters who were not self-evidently likely to be direct beneficiaries of the movements’ success. The new perspective recognized that the revitalization of activism in long-running movements for human rights, feminism, peace, environmentalism and new religions in the 1960s could not be adequately explained in terms of the simple convergence of people suffering from the same problem or holding the same values. Instead, the RM argument was that successful mobilizations of sentiment, material resources and people required rational organizational strategies for persuading ‘conscience constituents’ and ‘patrons’ to overcome their supposedly rational reluctance to incur costs on behalf of ‘victims’ when they knew that other people could be counted on to make the necessary sacrifice. According to RM theory, with some assistance from rational choice thinking, the free-rider problem can be overcome partly by offering ‘selective benefits’ to those who contribute and partly by tapping support from pre-existing, imbricated social networks. In this way, according to the theory, rational strategies are indispensable aids to the aggregation of the resources required to sustain successful campaigns and movements in an organizational era. The professionalization of social movement organizers is a necessary corollary. Consequently, concern with ‘hearts and minds’, ‘bleeding heart liberalism’ or moral motivation seems to give way to a hard-headed, economistic approach that claims to explain why the number of serious grievances in any society is always high while only a relatively small number of protests are ever successful. It goes without saying that the RM perspective on social movements commits precisely the kinds of reduction and ‘flattening out’ of which many contributors to this volume have complained (see also Jenkins 1983). And, just as some critics have tried to place RCT in the category of ‘American exceptionalisms’, so some European critics of the RM perspective have argued that it is inapplicable outside the USA (Klandermans 1986). But in my view the growing realization that some grievances and social problems have a global dimension (Cohen and Rai 2000), and that transnational, if not global, social movements are responses to them (Smith et al. 1997), will eventually render the study of formal organizational strategies indispensable. This is not to say that the grievances and moral issues are irrelevant; it is merely to acknowledge that these emotional and normative considerations are ‘framed’

‘When the battle’s lost and won’ 231 or ‘packaged’ by rational strategists for transmission to individual agents and groups by means of sophisticated technologies of communication. Since a ‘market for compassion’ clearly exists, then, sociologists cannot afford to ignore its possibly distasteful implications. If the point is to understand this market, the organizational context must be analysed; and the responses of individual ‘consumers’ must also be scrutinized. It would be a mistake to ignore the contribution of instrumental rationality to these responses. The RM perspective on social movements, benefiting from some of RCT’s economistic features, deserves some credit for broaching uncomfortable but important questions about support for morally motivated campaigns and protests. My third reason for believing that insights from RCT can enhance the sociological understanding of complex social phenomena, including those involving emotions and moral values, returns to the concept of ‘market’. I have already mentioned the ‘market for compassion’, in which the staff of social movements and charitable organizations have long been active in competing for donations of money, goods and free labour. There is no mystery about the relation between levels of disposable income and levels of charitable donations. The impact of variables such as wealth, occupational status, age, sex and religious affiliation is also well understood by specialist researchers and charity organizers. One of the virtues of RCT is therefore that it sensitizes social scientists to the importance of the competition for advantage that functions in spheres of life customarily considered to lie in the realm of noninstrumental values. Of course, markets are not necessarily the most important context in which social movements compete, but sociological analysis would be considerably weaker if this dimension were ignored. Indeed, research on religious organizations has gone one stage further. As there is understandable reluctance to analyse such an apparently non-rational matter as religion in rational choice terms, a lively debate is currently taking place between the protagonists and the antagonists of a so-called ‘new paradigm’ (Warner 1993) in the sociology of religion. The new paradigm turns on the view that, contrary to the widely held assumptions about the inevitability of secularization in advanced industrial societies, religion is actually thriving in the USA. The main reason given for this non-materialization of secularization is that the separation of religion from the US state and the consequently long-standing competitiveness of American religious organizations have combined to foster both a pluralistic ethos and a constant process of religious renewal. The argument holds that religious organizations that become complacent or lazy lose their market position to more energetic competitors with relatively low start-up costs. This process of continuous decay and revitalization is believed to keep aggregate levels of religious vitality high and to keep secularization at bay – in sharp contrast to most European countries, where the ‘interference’ of the state and the conferral of official privileges on state churches have supposedly distorted what would otherwise have been more buoyant religious markets.

232 James A. Beckford This economistic paradigm has roots in exchange theory and in RCT. On the demand side, it constitutes the religious activity of individuals as an exchange of various costs (time, money, discipline) for the presumed benefits of psychological security, social support and, possibly, eternal life. The problem of the free-rider syndrome is said to be solved by reference to the selective benefits on offer exclusively to religious believers. And participation in highdemand or sectarian religious groups is explained in rational choice terms: believers who are prepared to pay higher costs will be the only ones to reap correspondingly higher rewards. On the supply side, the new paradigm claims that aggressive ‘outreach’ and strictness, in theological and moral terms, are likely to attract and retain new ‘consumers’. By contrast, complacency or timidity about market share is likely to undermine the potential for organizational growth. At the level of both individuals and organizations, then, the search for rational advantage appears to explain differential religious vitality. Critical response to the application of RCT to religion is largely polarized in the form that I characterized as a ‘battle’ at the beginning of this chapter, although there are small signs of moderation or fatigue on both sides (see Young 1997).2

Last post Attentive readers will have noticed that I have not tried to defend RCT against its detractors. Indeed, I share Margaret Archer’s scorn for the theoretical imperialism that is evident in the work of some of its advocates. It also seems to me most unlikely that the philosophical difficulties with individualistic rationalism, skilfully exposed by Peter Wagner, will ever be resolved to the satisfaction of most sociologists. This is why I couched my three reasons for not throwing away the baby with the bathwater strictly in terms of the ‘insights’ or ‘features’ of RCT that could assist sociological analysis. In other words, my claim is that RCT is flawed and ‘floored’ as an attempt to furnish sociology or other social sciences with an indispensable, all-purpose theoretical foundation. But, on the other hand, I also want to argue that RCT can still be useful in practical terms. Firstly, RCT is a rich source of questions, hypotheses and factual claims that represent a challenge to the exponents of other theoretical perspectives. It can function as an imaginative experiment, akin to Max Weber’s use of ideal types, in the sense of deliberately selecting and exaggerating certain features of social and cultural reality for the purpose of demonstrating their partial contributions to the perceived whole. This purely instrumental use of RCT ironically entails no commitment to ontological notions of human beings as instrumental utility-maximizers. Secondly, RCT sensitizes sociologists to the importance of instrumental rationality as a feature of the formal organizations that dominate much of life in advanced industrial societies. RCT is certainly not the only theoretical perspective capable of raising questions

‘When the battle’s lost and won’ 233 about ‘organizational society’, but it does so in a peculiarly pointed fashion – often shocking us with what some critics feel to be an iconoclastic, if not sacrilegious, disdain for fondly held assumptions about human societies. For example, it is highly provocative to claim that the success of social movements depends at least as much on the extent of their resources as on the strength of the moral and ideological commitment of their supporters. Yet such claims force sociologists to look more closely at the material dimension of ostensibly normative phenomena. There is indeed an elective affinity between RCT and modernity. They partly define each other. But this does not prevent us from using insights from the former to explain aspects of the latter. Why should the Devil have all the best tunes?

Notes 1

This is true only for the world of theoretical battles. In the ‘real’ world, however, the application of RCT can have material effects on life-chances.


For example, Rodney Stark and Roger Finke, whose writings have become virtually synonymous with RCT, at least in the opinion of their peers, now deny that their work properly belongs in the category of RCT because they consider the notion of ‘maximization’ as ‘too precise’. Instead, they assume only that human beings make ‘subjective efforts to weigh the anticipated rewards against the anticipated costs, although these efforts usually are inexact and somewhat casual’ (Stark and Finke 2000: 37).


Abercrombie, N., Hill, S. and Turner, B. (1980) The Dominant Ideology Thesis, London: Allen and Unwin. Adam, B. (1990) Time and Social Theory, Cambridge: Polity. —— (1995) Timewatch: The social analysis of time, Cambridge: Polity. Ahier, J. and Moore, R. (1999) ‘Post-16 education, semi-dependent youth and the privatisation of inter-age transfer: Re-theorising youth transition’, British Journal of Sociology of Education 20, 4: 515–30. Aldridge, A. (1998) ‘Habitus and cultural capital in the field of personal finance’, Sociological Review 46, 1: 1–23. Allen, S. and Wolkowitz, C. (1987) Homeworking: Myths and realities, London: Macmillan. Althusser, L. (1971) ‘Ideology and ideological state apparatus’, in B. Cosin (ed.) Education, Structure and Society, Harmondsworth: Penguin. Andreski, S. (1964) Elements of Comparative Sociology, London: Weidenfeld and Nicolson. Archer, M.S. (1995) Realist Social Theory: The morphogenetic approach, Cambridge: Cambridge University Press. —— (2000) Being Human: The problem of agency, Cambridge: Cambridge University Press. Arendt, H. (1970) On Violence, New York: Harcourt Brace. —— (1975) The Origins of Totalitarianism, San Diego, CA: Harvest. —— (1978) The Life of the Mind, San Diego, CA: Harcourt Brace. —— (1994a) Eichmann in Jerusalem: A report on the banality of evil, Harmondsworth: Penguin. —— (1994b ) Essays in Understanding, New York: Harcourt Brace. Arendt, H. and Jaspers, K. (1993) Correspondence 1926–1969, Orlando, FL: Harvest. Aristotle (1987) Poetics I with The Tractatus Coislinianus, A Hypothetical Reconstruction of Poetics II, The Fragments of the On Poets, Indianapolis/Cambridge: Hackett. AUT(2000) , 24 April. AvalonProject(1999) , 3 November. Back, L. (1996) New Ethnicity and New Urban Culture, London: UCL Press.

Bibliography 235 Bagguley, P. and Mann, K. (1992) ‘Idle, thieving bastards? Scholarly representations of the underclass’, Work, Employment and Society 6, 1: 113–26. Baker, K.M. (1975) Condorcet: From natural philosophy to social mathematics, Chicago: University of Chicago Press. Ball, S. (1990) Politics and Policy Making in Education, London: Routledge. Ballard, R. and Ballard, C. (1977) ‘The Sikhs: The development of South Asian settlements in Britain’, in J.L. Watson (ed.) Between Two Cultures: Migrants and minorities in Britain, Oxford: Basil Blackwell. Banton, M. (1983) Racial and Ethnic Competition, Cambridge: Cambridge University Press. Barbalet, J. (1998) Emotion, Social Theory and Social Structure, Cambridge: Cambridge University Press. —— (1999) ‘William James’s theory of emotions: Filling in the picture’, Journal for the Theory of Social Behaviour 29, 3: 251–66. Bauman, Z. (1991) Modernity and Ambivalence, Cambridge: Polity. —— (1993) Modernity and the Holocaust, Oxford: Polity. —— (1996) Contesting Culture, Cambridge: Cambridge University Press. Beck, U. (1992) Risk Society, London: Sage. Becker, G. (1973) ‘A theory of marriage: Part I’, Journal of Political Economy 81, 4: 813–46. —— (1974) ‘A theory of marriage: Part II’, Journal of Political Economy 82, 2: S11–S26. —— (1976) The Economic Approach to Human Behavior, Chicago: University of Chicago Press. —— (1991) A Treatise on the Family, Enlarged Edition, Cambridge, MA: Harvard University Press. —— (1996) Accounting for Tastes, Cambridge, MA: Harvard University Press. BFI (n.d.) ‘British Film Institute Television Industry Tracking Studies, First and Second Interim Reports’, unpublished, London: BFI. Bidwell, C. and Rodriguez, P. (1991) Organizational Control in the High School Work-place: A theoretical argument, Chicago: Ogburn Stouffer Discussion Paper Series. Black, M. (1990) Perplexities, Ithaca, NY and London: Cornell University Press. Blaug, M. (1992) The Methodology of Economics: Or how economists explain, Cambridge: Cambridge University Press. Bohman, J. (1992) ‘The limits of rational choice explanation’, in J. Coleman and T.J. Fararo (eds) Rational Choice Theory: Advocacy and critique, Newbury Park, CA: Sage. Boltanski, L. (1990) L’amour et la justice comme compétences. Trois essais de sociologie del’action, Paris: Métailié. Boltanski, L. and Thévenot, L. (1991) De la justification. Les economies de la grandeur b, Paris: Gallimard. Boudon, R. (1974) Education, Opportunity and Social Inequality, New York: Wiley. Bourdieu, P. (1977) Outline of a Theory of Practice, Cambridge: Cambridge University Press. —— (1984) Distinction: A social critique of the judgment of taste, London: Routledge. —— (1990) The Logic of Practice, Cambridge: Polity. —— (1998) Acts of Resistance: Against the new myths of our time, Cambridge: Polity.

236 Bibliography Bourdieu, P. and Passeron, J.-C. (1977) Reproduction in Education, Society and Culture, London: Sage. Bowles, S. and Gintis, H. (1976) Schooling in Capitalist America, London: Routledge. Brian, E. (1994) La mesure de l’État: Administrateurs et géomètres au XVIIIe siècle, Paris: Michel. Brown, C. (1984) Black and White Britain: The third PSl study, London: Heinemann. Browning, C. (1993) Ordinary Men, New York: HarperCollins. Brubaker, R. (1984) The Limits of Rationality: An essay on the social and moral thought of Max Weber, London: Allen and Unwin. Bullock, A. (1983) Hitler: A study in tyranny, Harmondsworth: Penguin. Burke, E. (1993 [1790]) Reflections on the Revolution in France, ed. L.G. Mitchell, Oxford: Oxford University Press. CAF (2000) , 5 May. Carlton Club Political Committee (1991) Conservative Government: Policy and practice: ‘Education in the 1990s’ A current assessment of requirements for the manifesto, London (March). Castoriadis, C. (1990) Le monde morcelé: Les carrefours du labyrinthe III, Paris: Seuil. Chafetz, J. and Hagan, J. (1996) ‘The gender division of labor and family change in industrial societies: A theoretical accounting’, Journal of Comparative Family Studies 27, 2: 187– 219. Cherry, R. (1998) ‘Rational choice and the price of marriage’, Feminist Economics 4, 1: 27– 49. Chown, J.F. (1994) A History of Money from A.D. 800, London: Routledge. Cigno, A. (1991) Economics of the Family, Oxford: Clarendon Press. Clark, D. (ed.) (1991) Marriage, Domestic Life and Social Change: Writings for Jacque-line Burgoyne (1944–88), London: Routledge. Clulow, C. (1995) ‘Marriage: A new millennium?’, in C. Clulow (ed.) Women, Men and Marriage, London: Sheldon Press. Clune, W. and Witte, J. (eds) (1990) Choice and Control in American Education, vol. 2: The practice of choice, decentralization and school restructuring, London: Falmer Press. Cohen, L. and Mallom, M. (1999) ‘The transition from organisational employment to portfolio working: Perceptions of “boundarylessness”’, Work, Employment and Society 13, 2: 329–52. Cohen, R. and Rai, S. (eds) (2000) Global Social Movements, London: Athlone Press. Coleman, D. (1996) ‘European patterns of fertility’, British Society for Population Studies meeting on ‘Family and Household Formation in Europe’, London School of Economics, London, 11 December. Coleman, J.S. (1990) Foundations of Social Theory, Cambridge, MA: Harvard University Press. Coleman, J.S. and Fararo, T.J. (eds) (1992a) Rational Choice Theory: Advocacy and critique, Newbury Park, CA: Sage. —— (1992b) ‘Introduction’, in J.S. Coleman and T.J. Fararo (eds) Rational Choice Theory: Advocacy and critique, Newbury Park, CA: Sage.

Bibliography 237 Collins, R. (1981) ‘On the micro-foundations of macro-sociology’, American Journal of Sociology 86, 5: 984–1014. —— (1990) ‘Stratification, emotional energy, and the transient emotions’, in T.J. Kemper (ed.) Research Agendas in the Sociology of Emotion, New York: State Univer-sity of New York Press. Craib, I. (1995) ‘Some comments on the sociology of emotions’, Sociology 29, 1: 151–8. CRE (1988) Investigation into Estate Agents, London: Commission for Racial Equality. —— (1990) Sorry It’s Gone, London: Commission for Racial Equality. Crompton, R. and Harris, F. (1998a) ‘Explaining women’s employment patterns: “Orientations to Work” Revisited’, British Journal of Sociology 49, 1: 118–36. —— (1998b) ‘Gender relations and employment: The impact of occupation’, Work, Employment and Society 12, 2: 297–315. Crossley, N. (1998) ‘Emotions and communicative action: Habermas, linguistic philosophy and existentialism’, in G. Bendelow and S.J. Williams (eds) Emotions in Social Life: Critical themes and contemporary issues, London: Routledge. Currie, J. (1999a) ‘Missing out on “the experience”’, The Times Higher Educational Supplement, 16 July: 9. —— (1999b) ‘Debt dampens spirit of protest’, The Times Higher Educational Supplement, 20 August: 6. Dahya, B. (1974) ‘The nature of Pakistani ethnicity in industrial cities in Britain’, in A. Cohen (ed.) Urban Ethnicity, London: Tavistock. Dale, R. (1997) ‘The state and the governance of education: An analysis of the restructuring of the state–educational relationship’, in A. Halsey, H. Lauder, P. Brown and A. Wells (eds) Education: Culture, economy, society, Oxford: Oxford University Press. Damasio, D. (1994) Descartes’ Error, New York: Putnam. Davies, G. (1994) A History of Money: From ancient times to the present day, Cardiff: University of Wales Press. Deem, R. and Brehony, K. (1995) Active Citizenship and the Governing of Schools, Buckingham: Open University Press. Delphy, C. and Leonard, D. (1992) Familiar Exploitation: A new analysis of marriage in contemporary western societies, Cambridge: Polity. Descartes, R. (1931 [1649]) ‘The passions of the soul’, in Philosophical Works of Descartes, vol. I, Cambridge: Cambridge University Press. Desrosières, A. (1998) The Politics of Large Numbers: A history of statistical reasoning, Cambridge, MA: Harvard University Press. Dex, S., Willis, J., Paterson, R. and Sheppard, E. (2000) ‘Freelance workers and contract uncertainty: The effects of contractual changes in the television industry’, Work, Employment and Society 14, 2: 283–305. DFEE (1995) Consultation Paper on Self-Government for Voluntary Aided Schools, London: Department for Education and Employment. DiMaggio, P. and Mohr, J. (1985) ‘Cultural capital, educational attainment, and marital selection’, American Journal of Sociology 90, 6: 1231–61. Diocese of Blackburn (1994) Diocesan Guidance on GMS, Blackburn.

238 Bibliography Duncan, S. and Edwards, R. (1999) ‘Lone mothers, gendered moral rationalities and New Labour’s rationality mistake’, ESRC Seminar Series, ‘Parenting, Motherhood and Paid Work: Rationality and ambivalence’, South Bank University, London, 16 April. Edwards, R. (1979) Contested Terrain: The transformation of the workplace in the twen-tieth century, New York: Basic Books. Elias, N. (1978a) What is Sociology?, London: Hutchinson. —— (1978b [1939]) The Civilizing Process: vol. I: The history of manners, Oxford: Basil Blackwell. —— (1982 [1939]) The Civilizing Process, vol. II: State formations and civilization, Oxford: Basil Blackwell. —— (1991) ‘On human beings and their emotions: A process sociological essay’, in M. Featherstone, M. Hepworth and B.S. Turner (eds) The Body: Social process and cultural theory, London: Sage. —— (1992) Time: An essay, Oxford: Basil Blackwell. Elster, J. (1979) Ulysses and the Sirens, Cambridge: Cambridge University Press. —— (1983) Sour Grapes: The subversion of rationality, Cambridge: Cambridge University Press. —— (1985) ‘Sadder but wiser? Rationality and the emotions’, Social Science Information 24, 2: 375–406. —— (ed.) (1986a) Rational Choice, Oxford: Basil Blackwell. —— (1986b) ‘Introduction’, in J. Elster (ed.) Rational Choice, Oxford: Basil Black-well. —— (1988) The Multiple Self, Cambridge: Cambridge University Press. —— (1989) Nuts and Bolts for the Social Sciences, Cambridge: Cambridge University Press. —— (1999a) Alchemies of the Mind: Rationality and the emotions, Cambridge: Cambridge University Press. —— (1999b) Strong Feelings, Cambridge: Cambridge University Press. Engels, F. (1958 [1857]) The Condition of the Working Class in England in 1844, Oxford: Basil Blackwell. Etzioni, A. (1988) The Moral Dimension: Towards a new economics, New York: Free Press. —— (1992) ‘Normative-affective factors: Toward a new decision-making model’, in M. Zey (ed.) Decision Making: Alternatives to rational choice models, Newbury Park, CA: Sage. Feldberg, R.L. and Glenn, E.N. (1979) ‘Male and female: Job versus gender models in the sociology of work’, Social Problems 26, 5: 524–38. Felstead, A. and Jewson, N. (2000) In Work, at Home: Towards an understanding of homeworking, London: Routledge. Fenton, M. (1977) Asian Households in Owner Occupation: A study of patterns, costs and experiences of households in Greater Manchester, Social Science Research Council Research Unit on Ethnic Relations, University of Bristol. Finkelstein, N. and Tritter, J. (1999) ‘Managing Education or Managing Money? Findings from Great Britain and the United States on the partial implementation of site-based management’, unpublished, American Educational Research Annual Meeting, Montreal, Canada, 22 April.

Bibliography 239 Finkielkraut, A. (1992) Remembering in Vain: The Klaus Barbie trial and crimes against humanity, New York: Columbia University Press. Flam, H. (1990a) ‘Emotional “man”: The emotional “man” and the problem of collective action’, International Sociology 5, 1: 39–56. —— (1990b) ‘Emotional “man”: Corporate actors as emotion-motivated emotion managers’, International Sociology 5, 2: 225–34. (1993) ‘Fear, loyalty and greedy organizations’, in S. Fineman (ed.) Emotion in Organizations, London: Sage. Fox, N. (1998) ‘“Risks”, “hazards” and life choices: Reflections on health at work’, Sociology 32, 4: 665–87. Frank, R.H. (1997) Microeconomics and Behavior, Boston: McGraw-Hill. Frankfurt, H. (1988) The Importance of What We Care About, Cambridge: Cambridge University Press. Freud, S. (1984 [1915]) On Metapsychology: The theory of psychoanalysis (Pelican Freud Library, vol. 11), Harmondsworth: Penguin. Friedman, M. (1966) Essays in Positive Economics, Chicago: University of Chicago Press. Gambetta, D. (1996) Were They Pushed or Did They Jump? Individual decision-making mechanisms in education, Boulder, CO: Westview Press. Game, A. and Metcalfe, A. (1996) Passionate Sociology, London: Sage. Gellner, E. (1968) ‘Holism versus individualism’, in M. Brodbeck (ed.) Readings in the Philosophy of the Social Sciences, New York: Macmillan. Gibbard, A. (1986) ‘Interpersonal comparisons: Preference, good and the intrinsic reward of a life’, in J. Elster and A. Hyllan (eds) Foundations of Social Choice Theory, Cambridge: Cambridge University Press. Giddens, A. (1990) The Consequences of Modernity, Cambridge: Polity. —— (1991) Modernity and Self-Identity: Self and society in the late modern age, Cambridge: Polity. —— (1992) The Transformation of Intimacy: Sexuality, love and eroticism in modern societies, Cambridge: Polity. Gilligan, C. (1982) In a Different Voice, Cambridge, MA: Harvard University Press. Glucksmann, M.A. (1998) ‘“What a difference a day makes”: A theoretical and historical exploration of temporality and gender’, Work, Employment and Society 32, 2: 239–58. Goldhagen, D. (1996) Hitler’s Willing Executioners: Ordinary Germans and the Holocaust, London: Little, Brown and Company. Goldthorpe, J.H. (1996) ‘Class analysis and the reorientation of class theory: The case of persisting differentials in educational attainment’, British Journal of Sociology 47, 3: 481–505. —— (1998) ‘Rational action theory for sociology’, British Journal of Sociology 49, 2: 167– 92. Gordon, S. (1990) ‘Social structural effects on emotions’, in T. Kemper (ed.) Research Agendas in the Sociology of Emotions, New York: State University of New York Press. Grace, G. (ed.) (1995) School Leadership: Beyond education management, London: Falmer Press.

240 Bibliography Grainger, B., Stanworth, J. and Stanworth, C. (1995) ‘Self-employment Career Dynamics: The case of “employment push”’, Work, Employment and Society 9, 3: 499–516. Granovetter, M. (1985) ‘Economic action and social structures: The problem of embeddedness’, American Journal of Sociology 91, 3: 481–510. Green, D. and Shapiro, I. (1994) Pathologies of Rational Choice: A critique of applications in political science, New Haven, CT and London: Yale University Press. Greenwood, J.D. (1994) Realism, Identity and Emotion: Reclaiming social psychology, London: Sage. Gregson, N. and Lowe, M. (1994) Servicing the Middle Class, London: Routledge. Habermas, J. (1985) The Theory of Communicative Action, vol. 2, Boston: Beacon Press. Hakim, C. (1991) ‘Grateful slaves and self-made women: Fact and fantasy in women’s work orientations’, European Sociological Review 7, 2: 100–21. —— (1995) ‘Five feminist myths about women’s employment’, British Journal of Sociology 46, 3: 429–55. —— (1996) Key Issues in Women’s Work, London: Athlone. Hamzah, M. bin and Harrison, M. (2000) Islamic Housing Values and Perspectives: Insights from influentials in Malaysia, Leeds: ‘Race’ and Public Policy Research Unit, University of Leeds. Harsanyi, J. (1977) ‘Rule-utilitarianism and decision theory’, Erkenntis 11: 25–53. Hatcher, R. (1998) ‘Class differentiation in education: Rational choices?’, British Journal of Sociology of Education 19, 1: 5–24. Heap, S., Hollis, M., Lyons, B., Sugden, R. and Weale, A. (1992) The Theory of Choice, Oxford: Basil Blackwell. Hechter, M. (1986) ‘Rational Choice Theory and the study of race and ethnic relations’, in J. Rex and D. Mason (eds) Theories of Race and Ethnic Relations, Cambridge: Cambridge University Press. Heller, A. (1990) Can Modernity Survive?, Cambridge: Polity. Henderson, J. and Karn, V. (1987) Race, Class and State Housing: Inequality and the allocation of public housing in Britain, Aldershot: Gower. Hillier, S. (1987) ‘Rationalism, bureaucracy, and the organization of health services: Max Weber’s contribution to understanding modern health care systems’, in G. Scambler (ed.) Sociological Theory and Medical Sociology, London: Tavistock. Himmelwelt, S. (1999) ‘The limits of rational choice, or why do economists become parents?’, ESRC Seminar Series, ‘Parenting, Motherhood and Paid Work: Rationalily and Ambivalence’, South Bank University, London, 16 April. Hindess, B. (1988) Choice, Rationality and Social Theory, London: Unwin Hyman. Hobbes, T. (1962 [1651]) Leviathan, ed. Plamenatz, London: Fontana. Hochschild, A.R. (1983) The Managed Heart: The commercialization of human feeling, Berkeley: University of California Press. —— (1998) The sociology of emotion as a “way of seeing”’, in G. Bendelow and S.J. Williams (eds) Emotions in Social Life: Critical themes and contemporary issues, London: Routledge.

Bibliography 241 Hogarth, R.M. and Reder, M.W. (1987) ‘Introduction: Perspectives from economics and psychology’, in R.M. Hogarth and M.W. Reder (eds) Rational Choice: The contrast between economics and psychology, Chicago: University of Chicago Press. Holland, D.C. and Eisenhart, M.A. (1990) Educated in Romance: Women, achievement and college culture, Chicago: University of Chicago Press. Hollis, M. (1987) The Cunning of Reason, Cambridge: Cambridge University Press. —— (1989) ‘Honour among thieves’, Proceedings of the British Academy LXXV: 163–80. —— (1994) The Philosophy of Social Science: An introduction, Cambridge: Cambridge University Press. Holton, R.J. (1995) ‘Rational choice theory in sociology’, Critical Review 9, 4: 519–37. Hume, D. (1969 [1739/40]) A Treatise of Human Nature, ed. with intro. by E.C. Mossner, Harmondsworth: Penguin. Huws, U., Podro, S., Gunnansson, E., Weijers, T., Arvanitaki, K. and Trova, V. (1996) Teleworking and Gender, Brighton: Institute for Employment Studies. Hyman, H.H. (1966) ‘The value system of different classes’, in S.M. Lipset and R. Bendix (eds) Class, Status and Power, London: Routledge and Kegan Paul. Jaggar, A. (1989) ‘Love and knowledge: Emotion in feminist epistemology’, in S. Bordo and A. Jaggar (eds) Gender/Body/Knowledge: Feminist reconstructions of being and knowing, New Brunswick, NJ and London: Rutgers University Press. James, W. (1956 [1897]) The Will to Believe and Other Essays in Popular Philosophy, New York: Dover Publications. Janis, I. and Mann, L. (1977) Decision-Making: A psychological analysis of conflict, choice and commitment, New York: Free Press. Jenkins, J.C. (1983) ‘Resource mobilization theory and the study of social movements’, Annual Review of Sociology 9: 527–53. Jenkins, R. (1992) Pierre Bourdieu, London: Routledge. Joas, H. (1996a) The Creativity of Action, Cambridge: Polity. —— (1996b) The Genesis of Values, Cambridge: Polity. Johnson, M. (1987) The Body in the Mind: The bodily basis of meaning, imagination and reason, Chicago: University of Chicago Press. Jones, T. (1993) Britain’s Ethnic Minorities, London: Policy Studies Institute. Karn, V. and Phillips, D. (1998) ‘Race and ethnicity in housing: A diversity of experience’, in T. Blackstone, B. Parekh and P. Sanders (eds) Race Relations in Britain: A developing agenda, London: Routledge. Kaul, I., Grunberg, I. and Stern, M. (eds) (1999) Global Public Goods: International cooperation in the 21st Century, Oxford: Oxford University Press on behalf of the United Nations Development Programme. Kemper, T.D. (1990) ‘Social relations and emotions: A structural approach’, in T.D. Kemper (ed.) Research Agendas in the Sociology of Emotions, New York: State University of New York Press. Kenway, J. (1993) ‘Marketing education in the postmodern age’, Journal of Education Policy 8, 2: 105–22.

242 Bibliography Klandermans, B. (1986) ‘The new social movements and resource mobilization: The European and the American approach’, International Journal of Mass Emergencies and Disasters 4: 13–37. Lamont, M. and Thévenot, L. (eds) (2000) Rethinking Comparative Cultural Sociology: Polities and repertoires of evaluation in France and the United States, New York: Cambridge University Press. Lampard, R. and Peggs, K. (1999) ‘Repartnering: The relevance of parenthood and gender to cohabitation and remarriage among the formerly married’, British Journal of Sociology 50, 3: 443–65. Lane, M. (1972) ‘Explaining educational choice’, Sociology 6, 3: 255–66. Lash, S. and Urry, J. (1994) Economies of Signs and Spaces, London: Sage. Laver, M. (1997) Private Desires, Political Action, London: Sage. Law, I. (1996) Racism, Ethnicity and Social Policy, Hemel Hempstead: Prentice Hall/ Harvester Wheatsheaf. —— (1997) ‘Modernity, anti-racism and ethnic managerialism’, Policy Studies 18, 3/4: 189– 205. Lawrence, D.H. (1962) ‘The spirit of place’, in A. Arnold (ed.) The Symbolic Meaning: The uncollected versions of studies in classic American literature, Fontwell: Centaur Press. —— (1987 [1926]) The Plumed Serpent (Quetzalcoatl), Cambridge: Cambridge University Press. Lawrence, E. (1982) ‘Just plain common sense: The “roots” of racism’, in Centre for Contemporary Cultural Studies, The Empire Strikes Back: Race and racism in 70s Britain, London: Hutchinson. Lazarus, R. (1984) ‘Thoughts on the relations between emotion and cognition’, (reprinted) in K.R. Scherer and P. Ekman (eds) Approaches to Emotion, Hillsdale, NJ and London: Lawrence Erlbaum Associates. Ley, D. (1997) The New Middle Class and the Remaking of the Central City, Oxford: Oxford University Press. Little, A. (1999) ‘Undergrad, oversexed’, Daily Express, 7 August: 25. Lloyd, G. (1995) The Man of Reason: ‘Male’ and ‘female’ in Western philosophy, 2nd Edition, London: Methuen. Lukes, S. (1968) ‘Methodological individualism reconsidered’, British Journal of Sociology 19, 2: 119–29. —— (1974) Power: A radical view, London: Macmillan. Lupton, D. (1999) Risk, London and New York: Routledge. Lury, C. (1993) Cultural Rights: Technology, legality and personality, London: Routledge. McCarthy, J.D. and Zald, M. (1977) ‘Resource mobilization and social movements: A partial theory’, American Journal of Sociology 82, 6: 1212–41. MacDonald, R. (1994) ‘Fiddly jobs, undeclared working and the something for nothing society’, Work, Employment and Society 8, 4: 507–30. MacLeod, D and Mulholland, H. (1999) ‘I can’t take any more’, Guardian Higher, 23 March: 11. Macmillan (2000) < http://www.macmillan.org.uk/framed.html >, 26 April.

Bibliography 243 Maffesoli, M. (1996) The Time of the Tribes, London: Sage. Mandelbaum, M. (1973) ‘Societal facts’, in J. O’Neill (ed.) Modes of Individualism and Collectivism, London: Heinemann. Mansfield, P. and Collard, J. (1988) The Beginning of the Rest of Your Life, Basingstoke: Macmillan. Mason, D. (1986) ‘Introduction: Controversies and continuities in race and ethnic relations theory’, in J. Rex and D. Mason (eds) Theories of Race and Ethnic Relations, Cambridge: Cambridge University Press. Mellor, P. and Shilling, C. (1997) Re-forming the Body: Religion, community and modernity, London: Sage. Merleau-Ponty, M. (1962) The Phenomenology of Perception, London: Routledge and Kegan Paul. Mestrovic, S.G. (1997) Postemotional Society, London: Sage. Millar, S. and Ward, D. (1999) ‘Students play hard but want to get on’, Guardian, 7 August: 7. Mills, C.W. (1959) Sociological Imagination, New York: Oxford University Press. Mirzoeff, N. (ed.) (1998) The Visual Culture Reader, London: Sage. Mitter, S., Phizacklea, A.M., Tottendil, P. and Wolkowitz, G. (1993) ‘Tower Hamlets Homeworkers’ Cooperative: Feasibility report’, unpublished, London: Tower Hamlets Homeworkers’ Cooperative. Modood, T, Berthoud, R., Lakey, J., Nazroo, J., Smith, P., Virdee, S. and Beishon, S. (1997) Ethnic Minorities in Britain: Diversity and disadvantage, London: Policy Studies Institute. Mohrman, S. and Wohlstetter, P. (1994) School-Based Management: Organizing for high performance, San Francisco: Jossey-Bass. Mooney Marini, M. (1992) ‘The role of models of purposive action in sociology’, in J.S. Coleman and T.J. Fararo (eds) Rational Choice Theory: Advocacy and critique, Newbury Park, CA: Sage. Moore, W.E. (1963) Man, Time and Society, New York: John Wiley. Morris, L. (1994) Social Divisions: Economic decline and social structural change, London: UCL Press. Morris, L. and Irwin, S. (1992) ‘Employment histories and the concept of the underclass’, Sociology 26, 3: 401–20. Mouzelis, N. (1995) Sociological Theory: What went wrong?, London: Routledge. Mullins, N. (1973) Theory and Theory Groups in American Sociology, London: Harper and Row. Munch, R. (1992) ‘Rational Choice Theory: A critical assessment of its explanatory power’, in J.S. Coleman and T.J. Fararo (eds) Rational Choice Theory: Advocacy and critique, Newbury Park, CA: Sage. Murphy, J. and Beck, L. (1995) School-Based Management as School Reform: Taking stock, Thousand Oaks, CA: Corwin Press. Murray, C. (1984) Losing Ground: American social policy 1950–1980, New York: Basic Books.

244 Bibliography —— (1990) The Emerging British Underclass (Choice in Welfare series no. 2), London: Health and Welfare Unit, Institute of Economic Affairs. —— (2000) ‘Baby Beware’, Sunday Times (News Review), 13 February. NAHT (1996) Bulletin, May. NCVO (2000) < http://www.ncvo-vol.org.uk/main/press/releases/2000/april/aprpr_1.html >, 12 May. Neumann, F. (1957) The Democratic and Authoritarian State: Essays in political and legal theory, New York: Free Press of Glencoe. —— (1963) Behemoth: The structure and practice of National Socialism, 1933–1944, New York: Octagon Books. Newman, P. and Smith, A. (1997) Social Focus on Families, London: The Stationery Office. Newton, T. (1998) ‘The sociogenesis of emotion: A historical sociology?’, in G. Bendelow and S.J. Williams (eds) Emotions in Social Life: Critical themes and contemporary issues, London: Routledge. Nowotny, H. (1994) Time: The modern and postmodern experience, Cambridge: Polity. NUS (1995) Students at Work: A report on the economic conditions of students in employment, London: National Union of Students/GMB. Nussbaum, M. (1986) The Fragility of Goodness: Luck and ethics in Greek tragedy, Cambridge: Cambridge University Press. O’Brien, J. and Feist, A. (1995) Employment in the Arts and Cultural Industries: An analysis of the 1991 Census, London: Arts Council of England. Olsen, M. (1965) The Logic of Collective Action, Cambridge, MA: Harvard University Press. OPCS (1997) Social Trends 27, London: The Stationery Office. Parker, A. (2000) Student Lives and Student Lifestyles: University experience in the late 1990s, A Report prepared for the National Students’ Union Research Partnership, Coventry, CEDAR: University of Warwick. Parsons, T. (1968) The Structure of Social Action, London: Collier-Macmillan. Peach, C. and Byron, M. (1993) ‘Caribbean tenants in council housing: “Race”, class and gender’, New Community 19, 3: 407–23. Peggs, K. (2000) ‘Which pension? Women, risk and pension choice’, Sociological Review, 48, 3: 349–64. Phillips, D. and Karn, V. (1992) ‘Race and housing in a property owning democracy’, New Community 18, 3: 355–69. Phizacklea, A. and Wolkowitz, C. (1995) Homeworking Women: Gender, racism and class at work, London: Sage. Piore, M.J. (1979) Birds of Passage: Migrant labour and industrial societies, Cambridge, New York and Melbourne: Cambridge University Press. Pirie, M. and Worcester, R. (1999) The Next Leaders? London: MORI/Adam Smith Institute. Power, S., Halpin, D. and Fitz, J. (1996) ‘The Grant Maintained schools policy: The English experience of educational self-governance’, in C. Pole and R. Chawla-Duggan (eds) Reshaping Education in the 1990s: Perspectives on secondary education, London: Falmer Press. Procter, I. and Padfield, M. (1998) Young Adult Women, Work and Family, London: Mansell.

Bibliography 245 Purcell, K., and Pitcher, J. (1996) Great Expectations: The new diversity of graduate skills and aspirations, London: CSU/CST/IER. Purcell, K., Pitcher, J. and Simm, C. (1999) Working Out: Graduates’ early experiences of the labour market, London: CSU/CST/IER. Rapoport, A. and Chammah, A.M. (1965) Prisoner’s Dilemma, Ann Arbor, MI: University of Michigan Press. Rawls, J. (1970) A Theory of Justice, Cambridge, MA: Harvard University Press. —— (1993) Political Liberalism, New York: Columbia University Press. Ratcliffe, P. (1996a) ‘Social geography and ethnicity: A theoretical, conceptual and substantive overview’, in P. Ratcliffe (ed.) Social Geography and Ethnicity in Britain: Geographical spread, spatial concentration and internal migration (Ethnicity in the 1991 Census, vol. 3), London: HMSO. —— (1996b) ‘Race’ and Housing in Bradford: Addressing the needs of the South Asian, African and Caribbean communities, Bradford: Bradford Housing Forum. —— (1997) ‘“Race”, ethnicity and housing differentials in Britain’, in V. Karn (ed.) Employment, Education and Housing Among the Ethnic Minority Populations of Britain (Ethnicity in the 1991 Census, vol. 4), London: HMSO. Reinicke, W.H. (1998) Global Public Policy: Governing without government?, Washington, DC: Brookings Institution Press. Rex, J. and Moore, R. (1967) Race, Community and Conflict: A study of Sparkbrook, London: IRR/Oxford University Press. Roodhouse, S. (2000) ‘The future of the arts as cultural industries’, unpublished conference paper, ‘The Long-Run: Long-term developments in the arts and cultural industries’, Rotterdam, February. Rorty, R. (1989) Contingency, Irony, Solidarity, Cambridge: Cambridge University Press. Rutter, M. and Madge, N. (1976) Cycles of Disadvantage, London: Heinemann. Samuelson, P. (1938) ‘A note on the pure theory of consumers’ behaviour‘, Economica 5, 19: 353–4. Sarre, P., Phillips, D. and Skellington, R. (1989) Ethnic Minority Housing: Explanations and policies, Aldershot: Avebury. Sartre, J.-P. (1971 [1939]) Sketch for a Theory of the Emotions, London: Methuen and Co. Satz, D. and Ferejohn, J. (1994) ‘Rational choice and social theory’, The Journal of Philosophy 91, 2: 71–87. Saundry, R. (1999) ‘Casualisation, creativity and collectivism: Employment relations in British television’, unpublished paper. Scambler, G. (1987) ‘Habermas and the power of medical expertise’, in G. Scambler (ed.) Sociological Theory and Medical Sociology, London: Tavistock. Scheff, T.J. (1990a) ‘Socialization of emotions: Pride and shame as causal agents’, in T.D. Kemper (ed.) Research Agendas in the Sociology of Emotions, New York: State University of New York Press. —— (1990b) Microsociology: Discourse, emotion and social structure, Chicago and London: University of Chicago Press.

246 Bibliography —— (1992) ‘Rationality and emotion: Homage to Norbert Elias’, in J.S. Coleman and T.J. Fararo (eds) Rational Choice Theory: Advocacy and critique, Newbury Park, CA: Sage. Schwartz, T. (1982) ‘Human welfare: What it is not’, in H.B. Miller and W.H. Williams (eds) The Limits of Utilitarianism, Minneapolis: University of Minnesota Press. Sciulli, D. (1992) ‘Weaknesses in rational choice theory’s contribution to comparative research’, in J.S. Coleman and T.J. Fararo (eds) Rational Choice Theory: Advocacy and critique, Newbury Park, CA: Sage. Scott, S., Jackson, S. and Backett-Milburn, K. (1998) ‘Swings and roundabouts: Risk anxiety and the everyday worlds of children’, Sociology 32, 4: 689–705. Seidler, V. (1994) Unreasonable Men, London: Routledge. Sen, A. (1973) ‘Behaviour and the concept of preference’, Economica 40, 159: 241–59. Shilling, C. (1993) The Body and Social Theory, London: Sage. —— (1997) ‘Emotions, embodiment and the sensation of society’, Sociological Review 45, 2: 195–219. Simon, H.A. (1955) ‘A behavioural model of rational choice’, Quarterly Journal of Economics 69, 1: 99–118. —— (1978) ‘Rationality as process and as product of thought’, American Economic Review 68, 2: 1–16. —— (1987) ‘Rationality in psychology and economies’, in R.M. Hogarth and M.W. Reder (eds) Rational Choice: The contrast between economics and psychology, Chicago: University of Chicago Press. Simon, H.A. and Associates (1992) ‘Decision making and problem solving’, in M. Zey (ed.) Decision Making: Alternatives to rational choice models, Newbury Park, CA: Sage. Smelser, N.J. and Swedberg, R. (1994) ‘The sociological perspective on the economy’, in N.J. Smelser and R. Swedberg (eds) The Handbook of Economic Sociology, Princeton, NJ: Princeton University Press. Smith, D. and Whalley, A. (1975) Racial Minorities and Public Housing, London: Political and Economic Planning. Smith, J., Chatfield, C. and Pagnucco, R. (eds) (1997) Transnational Social Movements and Global Politics: Solidarity beyond the state, Syracuse, NY: Syracuse University Press. Smith, S.J. (1989) The Politics of ‘Race’ and Residence: Citizenship, segregation and white supremacy in Britain, Cambridge: Polity. —— (1991) ‘“Race” and housing in Britain: A review and research agenda’, unpublished paper prepared for the Joseph Rowntree Foundation Race and Housing Workshop, University of York, 4–5 April. Speth, J.G. (1999) ‘Forward’, in I. Kaul, I. Grunberg and M. Stern (eds) Global Public Goods: International cooperation in the 21st Century, Oxford: Oxford University Press on behalf of the United Nations Development Programme. Stark, R. and Finke, R. (2000) Acts of Faith: Explaining the human side of religion, Berkeley: University of California Press. Steiner, I.D. (1980) ‘Attribution of choice’, in M. Fishbein (ed.) Progress in Social Psychology, Hillsdale, NJ: Lawrence Erlbaum Associates.

Bibliography 247 Taylor, C. (1989) Sources of the Self: The making of modern identity, Cambridge: Cambridge University Press. —— (1995) ‘Irreducibly social goods’, in C. Taylor., Philosophical Arguments, Cambridge, MA: Harvard University Press. Thompson, E.P. (1974) ‘Time, work-discipline and industrial capitalism’, in M.W. Flinn and T.C. Smout (eds) Essays in Social History, Oxford: Oxford University Press. Thomson, A. (1999a) ‘Living on borrowed time’, The Times Higher Educational Supplement, 9 July: 6. —— (1999b) ‘At home with the Young Ones’, The Times Higher Educational Supplement, 23 July: 7. —— (1999c) ‘Quids in or just out of pocket?’, The Times Higher Educational Supplement, 30 July: 8. —— (1999d) ‘College libraries brought to book’, The Times Higher Educational Supplement, 6 August: 8. Tönnies, F. (1922) Kritik der öffentlichen Meinung, Berlin: Springer. Touraine, A. (1971) Post-industrial Society, London: Wildwood Press. Trudgill, E. (1976) Madonnas and Magdelens: The origins and development of Victorian sexual attitudes, London: Heinemann. Wagner, P. (1994) A Sociology of Modernity. Liberty and Discipline. London: Routledge. —— (2000) Theorizing Modernity: Inescapability and attainability in social theory, London: Sage. Walford, G. and Miller, H. (1991) City Technology College. Milton Keynes: Open University Press. Wallman, S. (ed.) (1979) Ethnicity at Work, London: Macmillan. Warner, R.S. (1993) ‘Work in progress toward a new paradigm for the sociological study of religion in the United States’, American Journal of Sociology 98, 5: 1044–93. Waslander, S. and Thrupp, M. (1995) ‘Choice, competition, and segregation: An empirical analysis of a New Zealand secondary school market, 1990–93’, Journal of Education Policy 10, 1: 1–26. Watkins, J.W.N. (1968) ‘Methodological individualism and social tendencies’, in M. Brodbeck (ed.) Readings in the Philosophy of the Social Sciences, New York: Macmillan. Watson, J.L. (ed.) (1977) Between Two Cultures: Migrants and minorities in Britain, Oxford: Basil Blackwell. Weber, A. (1925) Die Krise des modernen Staatsgedankens in Europa, Stuttgart and Berlin: DVA. Weber, M. (1964 [1947]) The Theory of Social and Economic Organization, London: CollierMacmillan. —— (1968 [1922]) Economy and Society: An outline of interpretive sociology, vol. 1, ed. G. Roth and C. Wittich, New York: Bedminster Press. —— (1991) From Max Weber: Essays in sociology, ed. H.H. Gerth and C.W. Mills, London: Routledge.

248 Bibliography Wentworth, W.M. and Yardley, D. (1994) ‘Deep sociality: A bioevolutionary perspective on the sociology of emotions’, in W.M. Wentworth and J. Ryan (eds) Social Perspectives on Emotion, Greenwich, CT/London: JAI Press. Whitehead, A. (1986) ‘I’m hungry, Mum’, in K. Young, R. McCullagh and C. Wolkowitz (eds) Of Marriage and the Market: Women’s subordination in international perspective, London: Routledge and Kegan Paul. Willer, D. (1992) ‘The principle of rational choice and the problem of a satisfactory theory’, in J.S. Coleman and T.J. Fararo (eds) Rational Choice Theory: Advocacy and critique, Newbury Park, CA: Sage. Williams, R. (1961) The Long Revolution, Harmondsworth: Penguin. Williams, S.J. (1998) ‘Modernity and the emotions: Corporeal reflections on the (ir)rational’, Sociology 32, 4: 747–769. Williams, S.J. and Bendelow, G. (1998) The Lived Body: Sociological themes, embodied issues, London: Routledge. Willis, P. (1977) Learning to Labour, Farnborough: Saxon House. Wilson, W.J. (1987) The Truly Disadvantaged: The inner-city, the underclass, and public policy, London: University of Chicago Press. —— (1991) ‘Studying inner-city dislocations: The challenge of public agenda research’, American Sociological Review 56, 1: 1–14. Wouters, C. (1992) ‘On status competition and emotion management: The study of emotions as a new field’, Theory, Culture and Society 9, 1: 229–52. Young, L.A. (ed.) (1997) Rational Choice Theory and Religion, New York: Routledge. Zafirovski, M. (1999a) ‘What is really rational choice? Beyond the utilitarian concept of rationality’, Current Sociology 47, 1: 47–113. —— (1999b) ‘Unification of sociological theory by the rational choice model: Conceiving the relationship between economics and sociology’, Sociology, 33, 3: 495–514. Zajonc, R.B. (1984) ‘The interaction of affect and cognition’, in K.R. Scherer and P. Ekman (eds) Approaches to Emotion, Hillsdale, NJ and London: Lawrence Erlbaum Associates. Zey, M. (ed.) (1992a) Decision Making: Alternatives to rational choice models, Newbury Park, CA: Sage. —— (1992b) ‘Criticisms of rational choice models’, in M. Zey (ed.) Decision Making: Alternatives to rational choice models, Newbury Park, CA: Sage. —— (1998) Rational Choice Theory and Organizational Theory: A critique, Thousand Oaks, CA: Sage.

Name index

Adam, B. 106, 156 Adorno, T. 67 Ahier, J. 131 Aldridge, A. 104 Allen, S. 167 Althusser, L. 127 Andreski, S. 220 Archer, Margaret S. 1–16, 36–55, 59, 69, 70, 71, 81, 83, 109, 220, 223, 224, 232 Arendt, Hannah 21, 185, 187, 188, 192, 193, 197 Aristotle 61, 63, 65 Back, L. 211 Bagguley, P. 85 Baker, Keith M. 28 Ball, S. 113 Ballard, R. and C. 204 Banton, M. 200 Barbalet, J. 65, 66, 67, 69, 70, 71, 227 Basu, Dipanitta 214 Bauman, Zygmunt 58, 184, 185–7, 188, 189, 190, 191, 192, 193, 194, 196, 197, 211, 224 Beck, L. 112 Beck, U. 96 Becker, Gary 3, 11, 41–2, 46, 49–50, 59, 70, 77, 80, 88, 94, 98; critique of marriage decisions 147–66

Beckford, James A. 16, 219–33 Bendelow, G. 62 Bidwell, C. 113 Black, M. 47 Blaug, M. 147 Bohman, J. 68 Boltanski, Luc 22, 23, 32 Booth, Charles 85 Boudon, R. 127, 131

Bourdieu, Pierre 14, 15, 46, 70, 95, 96, 127 Bowles, S. 127 Brehony, K. 114 Brian, E. 29 Brown, C. 202 Browning, Christopher 188, 191–2 Brubaker, R. 59 Burke, Edmund 27 Byron, M. 210 Castoriadis, Cornelius 25 Chafetz, J. 173 Chammah, A.M. 52 Chown, J.F. 156 Cigno, A. 98 Clark, D. 149 Clulow, C. 102 Clune, W. 112 Cohen, L. 170 Cohen, R. 231 Coleman, D. 147 Coleman, J.S. 60, 68, 93, 94, 100, 126, 220, 221 Collard, J. 160–1 Collins, R. 69 Condorcet, Marie Jean Antoine Nicolas Caritat 26, 28 Craib, I. 62, 70 Crompton, R. 173 Crossley, N. 66, 67 Cruickshank, Justin 9, 13, 75–91, 93, 220, 224–5 Currie, J. 131, 133 Dahya, B. 204 Dale, R. 112 Damasio, D. 61, 63, 64–5, 70 Davies, G. 156

250 Name index Deem, R. 114 Delphy, C. 149 Descartes, René 58, 68 Desrosières, A. 29 Dex, S. et al. 175 DiMaggio, P. 95 Duncan, S. 157 Durkheim, Emile 45, 46, 95, 187 Edwards, R. 114, 157 Eichmann, Adolf 180, 192–4, 197 Eisenhart, M.A. 161 Elias, Norbert 62, 68, 94, 96, 156 Elster, Jon 21, 23, 39–41, 44, 53, 58, 59–60, 61, 63, 71, 93, 97, 127 Engels, Friedrich 29 Etzioni, A. 94, 100, 103 Fararo, T.J. 60, 68, 93, 94, 126, 220, 221 Feist, A. 169 Feldberg, R.L. 168 Felstead, A. 169 Fenton, M. 204 Fine, Robert 11, 12, 183–97, 220, 224 Finkelstein, N. 112 Flam, Helena 44, 50, 69 Fox, N. 107 Frank, Hans 188, 194 Frank, R.H. 151 Frankfurt, H. 54 Freud, Sigmund 62, 70 Friedman, M. 151 Gage, Phineas 58 Gambetta, D. 49, 127, 128, 131 Game, A. 58, 71–2 Gellner, E. 8, 11, 45–6 Gibbard, A. 37 Giddens, A. 57, 70, 96, 100, 107, 156, 202, 205 Gilligan, C. 180 Gintis, H. 127 Glenn, E.N. 168 Glucksmann, M.A. 157 Goldhagen, D. 192 Goldthorpe, J.H. 127–8, 131, 220, 222 Gordon, S. 61 Grace, G. 117, 123 Grainger, B. et al. 169, 171 Granovetter, M. 159 Green, D. 122 Greenwood, J.D. 51, 62, 71

Gregson, N. 176 Habermas, Jürgen 3, 67, 229 Hagan, J. 173 Hakim, Catherine 167, 168, 176, 180 Hamzah, M. bin 213, 215 Harris, F. 173 Harrison, M. 213, 215 Harsanyi, J. 38–9 Hatcher, R. 131 Heap, S. et al. 167 Hechter, M. 200 Hegel, Georg Wilhelm Friedrich 24 Heidegger, Martin 20 Heller, A. 61, 65–6, 67 Henderson, J. 203, 208 Hillier, S. 59 Himmelwelt, S. 157 Hindess, B. 78–9 Hirsh, David 11, 12, 183–97, 220, 224 Hitler, Adolf 185 Hobbes, Thomas 19, 25, 27, 33 Hochschild, A.R. 62, 69, 71 Hogarth, R.M. 151, 220 Holland, D.C. 161 Hollis, M. 5, 43, 48, 52, 53, 54, 76, 80, 122, 126, 140, 147, 148, 161–2 Holton, R.J. 60, 158, 209 Horkheimer, M. 67 Hume, David 36, 65 Huws, U. et al. 167 Hyman, H.H. 127 Irwin, S. 86 Jaggar, A. 63, 66–7, 68 James, W. 66 Janis, I. 103 Jaspers, K. 193 Jenkins, J.C. 230 Jenkins, R. 96, 114 Jewson, N. 169 Johnson, M. 66 Jones, T. 202 Joseph, Keith 85 Kant, Immanuel 58, 185, 193–4 Karn, V. 202, 203, 207, 208, 211 Kaul, I. et al. 15 Kemper, T.D. 69 Kenway, J. 112 Klandermanns, B. 230 Kuhn, Thomas 3, 4

Name index 251 La Bruyère, Jean de 58 La Rochefoucauld, François, duc de 58 Lakatos, I. 4 Lamont, M. 33 Lampard, Richard 8, 10, 93–109, 220, 227 Lane, M. 127 Lanzmann, Claude 185 Lash, S. 169, 170 Law, I. 211 Lawrence, D.H. 30, 31 Lazarus, R. 61 Leonard, D. 149 Ley, D. 169 Little, A. 134 Lowe, M. 176 Lukes, S. 8, 45, 77, 115, 124 Lupton, D. 107 Lury, C. 169, 170, 171 MacDonald, R. 86, 87 MacLeod, D. 129 Madge, N. 86 Mallom, M. 170 Mandelbaum, M. 7 Mann, K. 85 Mann, L. 103 Mansfield, P. 160–1 Marcuse, Herbert 67 Marx, Karl 26, 31, 32, 50, 222 Mason, D. 201 Mayhew, Henry 84 Merleau-Ponty, Maurice 21, 58, 62, 70 Metcalfe, A. 58, 71–2 Millar, S. 134, 135 Miller, H. 117 Mills, C.W. 72 Mirzoeff, N. 96 Modood, T. et al. 202 Mohr, J. 95 Mohrman, S. 112 Montaigne, Michel Eyquem de 58 Mooney Marini, Margaret 95, 96, 102 Moore, R. 131, 203, 210 Moore, W.E. 156 Morris, L. 84, 86 Mouzelis, N. 68 Mulholland, H. 129 Mullins, N. 221 Munch, R. 97, 98 Murphy, J. 112 Murray, Charles 75, 87–8, 89, 90, 91, 224

Neumann, Franz 187 Newman, P. 159 Newton, T. 62, 68 Nowotny, H. 156 Nussbaum, M. 63 O’Brien, J. 169 Olsen, M. 43 Padfield, Maureen 159–60 Parker, Andrew 8, 10, 126–43, 220, 225 Parsons, T. 152 Pascal, Blaise 58, 197 Passeron, J.-C. 46, 127 Peach, C. 210 Peggs, Kay 8, 10, 93–109, 220, 227 Phillips, D. 202, 207, 211 Phizacklea, A. 167 Piore, M.J. 204 Pirie, M. 134 Pitcher, J. 129 Plato 58, 62 Popper, Karl 66 Power, S. et al. 118 Procter, Ian 11, 12, 108, 147–66, 167, 220, 225 Purcell, K. 129 Rai, S. 231 Rapoport, A. 52 Ratcliffe, Peter 11, 13, 200–16, 220, 226 Rawls, John 27 Reder, M.W. 151, 220 Reinicke, W.H. 15 Rex, J. 203, 210 Rodriguez, P. 113 Roodhouse, S. 169 Rorty, Richard 22 Rutter, M. 86 Samuelson, P. 37 Sarre, P. et al. 205, 207, 208, 210 Sartre, Jean-Paul 62, 70 Sawoniuk, Andrei 195–6 Scambler, G. 67 Scheff, Thomas J. 62, 68, 69, 97 Schwartz, T. 38, 39 Sciulli, D. 93 Scott, Sue et al. 107 Seidler, V. 68 Sen, A. 38

252 Name index Shakespeare, William 193 Shapiro, I. 122 Shilling, Chris 95, 96 Simon, H.A. 95, 150, 173 Smelser, N.J. 149, 161 Smith, A. 159 Smith, Adam 26 Smith, D. 203, 208 Smith, J. et al. 231 Smith, S.J. 202, 203 Spencer, Herbert 95 Speth, J.G. 14 Spinoza, Benedict 63, 65 Stalin, Josef 185 Steiner, I.D. 104 Swedberg, R. 149, 161 Taylor, Charles 22, 58 Thatcher, Margaret 85 Thévenot, Laurent 22, 32, 33 Thompson, E.P. 156 Thomson, A. 129, 131 Thrupp, M. 112 Tönnies, F. 29 Touraine, Alain 229 Trapp, Wilhelm 191 Tritter, Jonathan Q. 1–16, 109, 111–25, 220, 225 Trudgill, E. 168 Turgot, Anne Robert Jacques 28 Urry, J. 169, 170 Wagner, Peter 5, 6, 19–34, 220, 223, 228, 232 Walford, G. 117

Wallman, S. 200 Ward, D. 134, 135 Warner, R.S. 231 Waslander, S. 112 Watkins, J.W.N. 10 Watson, J.L. 200, 204 Weber, Alfred 29 Weber, Max 5, 6, 26, 29, 31, 32, 54, 57, 58– 9, 63, 95, 122, 164, 186, 187, 189, 196, 203, 207, 232 Wentworth, W.M. 61 Whalley, A. 203, 208 Whitehead, A. 180 Willer, D. 97 Williams, Raymond 31 Williams, Simon J. 6, 57–72, 93, 94, 97, 100, 168, 220, 227 Willis, P. 127 Witte, J. 112 Wittgenstein, Ludwig 22 Wohlstetter, P. 112 Wolkowitz, Carol 11–12, 167–81, 220, 226 Worcester, R. 134 Wouters, C. 61 Yardley, D. 61 Young, L.A. 232 Zafirovski, Milan 8–9, 60, 71, 80, 93, 94, 95, 97, 100, 108, 109, 111, 121, 122, 124, 126, 140, 167, 220, 222 Zajonc, R.B. 61 Zey, Mary 57, 60, 70, 80, 93, 94, 95, 97, 98, 101

Subject index

addictions 39, 53, 60 agency 4–13, 36–7, 39–40, 47–54, 58, 60, 62, 69, 78, 80–4, 86–7, 89, 91, 97, 109, 111, 122, 124–8, 133, 139–40, 143, 148, 167–8, 170, 188, 197, 202–7, 213, 220, 223–8 aggregation 7, 10, 41, 45–6, 68, 76, 140, 149–50, 230 agnosticism 150, 164 alienation 26 altruism 1–2, 13–14, 16, 38, 77, 80, 82, 150, 162 America 29–32 amorality 149, 158, 164, 184 analytic histories of emergence 82–3 antisemitism 12–13, 190–2, 196 atomism 7–8, 24, 26, 45, 50, 136, 172 attitudes 8, 75, 82, 85, 87, 91, 93, 95, 106, 133, 135, 171, 177–9, 181, 211 attraction, sexual 96, 100, 108, 161 autonomy 25–6, 30, 39, 45, 111, 113, 116– 19, 128 behaviourism 4, 10, 29, 37–8, 79–80, 89, 91 beliefs 14, 38–9, 42–3, 45, 60, 63, 66, 76, 81–2, 91, 95, 97, 220, 226 biology 61–2, 84–6 bonds, social 31–2 bureaucracy 59, 185–9, 191, 196, 207, 224, 229 capital: cultural 95, 131; personal and social 49–50, 162–4, 166; social and economic 169

careers 175–81, 226 case studies: Bradford housing 212–14; crimes against humanity 191–6;

cultural workers and home location 168– 80; grant-maintained schools 113–24; irrationality and remarriage 98–108; student behaviour 129–40 character-planning 40, 43 charitable organizations 1–2, 13–14, 231 children 42, 77, 85, 98, 104–5, 107–8, 152– 7, 162–3, 165 class, social 69, 75–91, 127, 168–9, 180, 202–4, 207–8, 210–11, 214–15 cognition 61–4, 66 coherence 19, 22 collectivity 5, 7–10, 12, 14–16, 25–8, 38, 41, 43–5, 50–1, 68–9, 96–7, 111, 115, 120–1, 123–5, 136, 140–1, 167–8, 170, 174, 176, 201–2, 204–5, 215, 220–1, 223–5, 227 colonization of policy 1–2 commensurability 16, 21–5, 42, 44, 47 commitment 2, 6, 53–4 commodification 42 concealed family concept 213–14 conformity 12, 26, 77, 102, 174, 185, 187, 191–2, 194, 197 consciousness 62, 184, 187, 194 consistency 12, 19, 22–4, 32 constraints 41, 45, 118, 121–2, 126, 128, 137, 142, 167, 175, 187–8, 200–16, 225–6 context 5–6, 8, 10–13, 15, 20–1, 23–8, 33, 36, 41, 45–9, 51, 53, 60, 64, 67–8, 71, 75–83, 86–7, 89–91, 93–4, 103, 124, 127–8, 167–8, 170, 175, 181, 197, 201, 206, 223–7, 231 cost–benefit: cultural workers and home location 12, 167, 170, 172, 174–7,181;

254 Subject index emotions and reason 60, 64; ethnic housing 226; grant-maintained schools 118–20, 122–4, 225, 232; homo economicus 42, 51, 55; irrationality and remarriage 93–4, 97; marriage and decision-making 148–50, 154–8, 162, 164–5, 227; rationality 5–6; social realism and RCT 77; student behaviour 126, 130, 135–6, 141 crime 84–5, 88–91, 183–97, 224 critical theory 4, 7, 32, 67–8, 229 cultural emergent properties 81–2, 87, 91 culture 7, 20, 33, 43, 50, 52–3, 63, 69, 76–8, 80–6, 89, 91, 95–6, 100, 102–4, 108–9, 127, 140, 149, 162, 164, 201, 204, 207, 210–11, 214–15, 220, 222–5, 228–9, 232 deprivation, cycle of 84–6 desires 37–41, 43, 53, 58–9, 62, 96, 140–1, 167, 171, 191, 203 determinism 9, 78–9, 83, 86–7, 227 differential rationality 88 discrimination 200–12, 214–15, 226 divorce 8, 12, 158–60, 165, 227 domains 6, 68 doves, hawks and RCT concepts 23–5, 32 economic factors 19, 45, 60, 70, 77, 93–6, 100, 104–5, 108–9, 126, 129, 131–3, 142, 147, 167–9, 172, 175–8, 181, 225– 6, 230 economic rationality 121–2, 207, 209, 212 economics 26, 28, 43, 46, 60 education 8, 111–43, 225 Education Reform Act (1988) 112–13, 116 embodiment, human 66, 71 emergent properties 7, 9–10, 11, 68–9, 71, 81–4, 86–7, 91, 224 emotions 6, 37–45, 57–72, 77, 93–8, 100–1, 103–4, 108, 150, 168, 194, 223–4, 227, 231 employment 14, 75, 85–91, 128–30, 133, 167–80, 225–6 Enlightenment 3, 25, 28, 36 essential properties 79–80, 85–6, 89, 91 essentialism 9, 13, 75, 79–81, 83–91, 201– 2, 209–10, 215–16, 224–5

ethics 52, 60, 121, 164, 184–5, 187, 190, 194 ethnicity 13, 89, 169–70, 200–16, 224, 226 ethnomethodology 4 Europe, America and social theory 29–32 evaluation 5, 22–3, 25, 33 exchange theory 232 exile and emigration 31–2 expansionism 3 experience 11–12, 71, 130–5, 159, 162–3, 167–9, 172, 174, 177–9, 181 family 12, 41–2, 77, 85, 147, 152, 163, 165, 167–8, 171, 173–6, 178, 180, 202, 211, 213–14, 226 Frankfurt School 32, 169 free-lancing 170–3, 178, 180 free-riding 43, 50, 52, 68, 230, 232 freedom 25, 27–8, 31, 39–41, 163, 228 French Revolution 26–7, 29, 33 game theory 49 gender 97, 103, 159, 167–9, 171–6, 179–80, 202, 215, 226 globalism 1, 15–16, 112, 231 goals 6, 13, 48, 50, 60, 64–5, 68, 70, 75, 89, 96, 121–3, 128, 167, 170, 186, 204, 206–7, 209, 221, 223, 228 grant-maintained schools 7–8, 111–25, 225 habit 49, 59, 60–1, 70, 95–6 habitus 61, 96, 98, 103–4, 108–9 holism 78–80, 91 Holocaust 184–91, 193–4, 196, 224 homo economicus 8, 13, 60, 75, 77, 80, 159, 168; sociologicus and sentiens 36–55 homogenization 83–7, 89, 91, 152, 209, 211, 215 households 152, 165, 170–7, 202–3, 206, 209–15, 226–7 housing and ethnicity 13, 200–16, 226 ideal preferences 38–9 identity 54, 71, 117, 120, 122–3, 170, 201, 203–4, 209, 211, 213–15, 223 ideology 3, 118, 120–2, 124 individualism 3, 5, 10–12, 14–16,22–30, 32–3, 45, 50, 52, 54, 60, 71, 75–91, 93–

Subject index 255 109, 111–43, 149, 159–62, 167–8, 170– 5, 179–81, 183–4, 186–90, 192, 194, 197, 200–3, 205–6, 209–10, 220–1, 213–15, 223–9, 232 individuality 7–10 individualization 15, 31–2, 170 indoctrination 42 influence 162–3 information 12, 22, 24, 29, 39, 47, 76, 80, 94, 96, 115, 126, 140, 147–9, 152–3, 158–62, 164–6, 173, 207, 209, 223 instrumental rationality 2–3, 5–6, 12, 15, 25–6, 32, 36–7, 42, 54, 57, 59, 66–7, 70, 75–7, 79–80, 89–91, 95, 121, 123, 164, 166, 184, 191, 197, 219, 223, 225–6, 228–31, 233 intentionality 24, 62 irony 20–1 irrationality 36, 38–41, 43, 47, 57–9, 62–3, 67, 72, 77, 79, 93–109, 142, 186, 190, 202, 209, 215 irreducibility 5, 7, 22, 25, 54, 57, 59–61, 69, 71, 76, 78, 81, 83 judgements 12, 22, 24, 53–4, 97, 127, 157, 159, 184, 193–4, 197, 224 knowledge 12, 19, 54, 66–7, 72, 79, 83, 151, 220, 226 law 183–4, 186, 203, 205–8, 226 leadership 117–18, 188, 191 liberalism 3, 28 life-chances 50–1 lifestyles 12, 126, 129–30, 132–5, 137–40, 142–3, 164–5, 175, 180, 226 linguistics 19–22 location 167–81, 202–3, 207–8, 210–15 love 6, 23, 41–2, 97, 100–1, 108, 150, 152 Macmillan Cancer Relief 2 market 113, 116–19, 150–3, 161, 172, 203, 205–7, 209, 211, 213, 215, 226, 231–2 Marxism 81 mass murder 13, 183, 185, 189–92, 195–7

mastery 25–6, 29, 58 materialism 80, 82, 88 maximization 3, 5–6, 9, 13–15, 40–1, 47, 59–60, 76–7, 87, 90, 93–5, 111, 115, 117–22, 124, 126, 141, 148, 150–1, 153, 162, 168, 174, 176, 196, 203–4, 209, 212, 223–4, 227, 233 means–ends rationality 33, 36, 53–4, 59– 60, 70, 75–6, 95, 139, 201, 207, 212, 215, 223–4 metaphysics 21, 30, 32–3 methodological individualism 3, 7, 9–11, 41, 45, 50, 86, 125, 130, 141, 149, 223 micro–macro linkages 57–8, 60, 68–9, 71, 95 modernity 1, 25–6, 29–32, 36, 43–4, 68, 70, 96, 228–9, 233 morality 37, 53, 58, 60–1, 70, 149, 164, 184–7, 190–1, 193–4, 197, 224 morphogenetic cycle 69, 71, 83, 86–7 multidimensional rationality 9, 57–8, 60–2, 70–1, 122, 223 multiple self 44–5 Nazism 12–13, 186–7, 189–90, 193–4, 196, 224 neoclassicalism 26, 28, 60, 181 neoliberalism 1–3, 14–16, 112, 120 New Right 81, 87, 90–1 norms 8–9, 13, 23–4, 29–30, 36, 42–5, 51– 3, 75–6, 79, 87, 90, 95–6, 102–4, 108–9, 135, 140–2, 157, 166, 185, 187, 207, 210–11, 213–14, 227, 230–1 Nuremberg Charter (1945) 183 opportunity costs 48, 103, 119, 167, 170, 175, 196, 203–4, 215 optimization 9, 40, 60, 65, 93–4, 111, 118, 126, 128, 139, 141, 173, 221, 223, 228 order of preferences 19, 24, 29–30, 33, 46– 7, 60, 76, 79–80, 126, 148, 154, 167, 173, 177, 223, 228 organizational society 229–33 outcomes 11, 15, 23, 93, 103, 126, 141–2 parents 8, 42, 46–7, 77, 88, 113–16, 124, 127, 134–5, 172–7, 179–81

256 Subject index partner choice 94, 147, 158–61, 225, 229 passion 36–9, 43, 50, 58–9, 65–8, 71–2, 95, 194, 223 passivity 9, 48–51, 53 peers 8, 136–7, 140–1, 161–2, 191–2 policy 1–3, 13, 77, 112–13, 115, 118, 120– 1, 124, 126, 205, 207–9, 211, 215, 226– 7, 228 politics 27–9, 33, 60, 70, 85, 95, 97, 100, 105–6, 108–9, 112, 114, 120–1, 124, 129 positivism 37 postmodernism 169, 184, 194, 197, 215 poverty and underclass 84–90 power 7–9, 11, 94–5, 97, 105–6, 108–9, 111, 115–18, 122–5, 187–8, 190, 196, 225, 227 predictions 15 preferences 2, 5–6, 9–13, 19–25, 33, 36–43, 45–8, 50, 52–4, 59–60, 70, 76–82, 89– 91, 96, 106, 111, 115, 118–127, 132–3, 136, 140–3, 148–50, 152–5, 158–168, 170, 172, 174, 177–81, 185, 190, 196–7, 200–1, 203–6, 210–11, 213–16, 220, 223, 226 Prisoner’s Dilemma 38, 43, 49, 52 procedural rationality 150 psychology 68, 78–80, 91, 95, 100–1, 108– 9, 124–5, 150, 226 public goods concept 14–15 public–private spheres 15–16, 168, 175, 180–1 quantity/quality 19–22 race 13, 200–16 Race relations Act (UK 1968, 1976) 205–8 rational action theory 127–8 rationality 4–6, 19–34, 36–55, 57–72, 75, 89, 93, 95, 97, 126–8, 137, 139–40, 142, 147–50, 166, 168, 185–91, 220, 223–32 rationalization 1, 26, 31–2, 39, 59 reason 5, 20, 22, 25, 27–30, 32, 36, 57–72, 94, 97, 148, 184, 187, 190, 194, 227 reductionism 7, 11, 28, 78–9, 86–7, 91, 215, 230

reflexivity 10, 13, 64, 70–1, 96, 103, 133, 163, 177, 197, 226 relations, social 8, 26, 31, 54, 61–3, 69–72, 79, 83, 91, 93–6, 98, 100–5, 107–9, 119–21, 149, 152–3, 159–61, 163–5, 168, 177, 180–1, 220, 225, 227 religion and RCT 220, 231–2 remarriage and divorce 8, 93–109 resource mobilization 230–1 resources 33–4, 45–9, 111, 116, 118–22, 129, 175, 177–8, 204, 212, 230, 233 risk 107–8, 175–6, 179–80, 207–8 risk society concept 96 role-taking 52–3 routinization 49 rules 49, 150, 152, 187–9 satisficing 9, 94, 115, 171–2, 174 self-interest 38, 50–4, 65, 75–7, 80, 82, 93, 98, 104–5, 108, 118, 121, 141, 185, 193–4, 196, 223, 228 social change 5, 7, 15–16, 26–7, 31–3, 39– 41, 43–5, 51, 53–4, 60, 70, 95, 100–1, 108–9, 149–50, 201–2, 211, 222–9, 232 social constructionism 4 social contract theory 3 social movements 41, 50–1, 220, 224, 230– 1, 233 social reality 9, 43, 46, 75–91, 224, 227 social sciences 16, 26, 28, 80, 82, 84, 91, 93, 97, 228–9, 231–2 social theory 3–4, 19–34, 48, 75–91, 106, 228 socialism 81–2 sociality 61–2 socialization 8, 11–12, 54, 82, 85, 132–3, 135–7, 141–2, 162, 207–11, 215, 224 solidarity, social 44–5 somatic markers 64–5 status 95, 153–8, 223 structural emergent properties 81–3, 86, 91 structuration theory 4, 7, 202, 205 structure 4, 7–11, 13–14, 43, 58, 62, 69, 71, 75–8, 80–4, 86, 89, 91, 98, 109, 118–19, 121–2, 125–8, 132–3, 140, 143, 149, 175, 187, 197, 200–6, 213, 224–6 subjectivity/objectivity 4–6, 37–9

Subject index 257 substantive rationality 5, 12, 28–9, 36–8, 59, 164 technologization 185–6 temporality 10–13, 71, 147–81, 183–97, 200–16 time 10–13, 81–3, 106–9, 127, 154–8, 163– 5, 227 totalization 2–3, 26, 187–8, 197 transaction costs 22 transitivity 47 underclass 75–91, 224–5 urban areas 169, 201–2, 205–6, 209 utilitarianism 3, 36, 38–9, 59–60, 62, 70–1, 93–6, 104, 108–9, 111, 121–2, 167–8, 170, 173, 225

utility 3, 5, 11–12, 19–22, 37, 39–41, 43, 47, 59, 75, 77, 80, 87–91, 94–5, 98, 100, 108, 111, 122, 126, 141, 147–58, 160–6, 201–2, 223–4, 227–8, 233 values 2, 5–6, 8–9, 29, 42, 80, 82, 94–7, 102–4, 108, 140–1, 158, 160, 167–8, 185, 227, 230–1 verificationism 80, 85, 91 voluntarism 1–2, 13–14, 16, 45 wars, revolutions and social theory 25–9 wealth and housing 202, 207, 210–11, 214– 15 welfare 37, 39, 42–3, 45, 85–90, 101, 225 Wertrationalität 5–6, 54, 59, 95, 197 Zwekrationalität 6, 57, 59, 95