Culture, Institutions, and Development : New Insights into an Old Debate [1 ed.] 9780203843338, 9780415580076

Citation preview

Culture, Institutions, and Development

Does culture matter? This question has taken on added significance since fundamentalist revivalism has recently gained ground in different parts of the world. The old controversy between Max Weber and Karl Marx, which centres around the extent to which cultural factors such as social norms and values affect economic growth is of critical importance, particularly because of its policy implications. Indeed, if culture is not an autonomous factor susceptible to influencing economic realities, it should not matter and public authorities can dispense with thinking about cultural interventions. On the other hand, if culture does have a real impact, the question arises as to whether it is conducive or detrimental to economic growth, political liberalization and the emancipation of individuals among other things. Culture, Institutions, and Development addresses this debate at a concrete level by looking at five important issues: the role of tradition and its influence on development; the role of religion, with special reference to Middle Eastern countries; the role of family, kinship and ethnic ties in the process of development; the relationship between culture and entrepreneurship; and the relationship between culture and poverty. This collection offers a nuanced view that neither denies nor exaggerates the role of cultural factors in explaining relative growth performances across countries. Instead, the contributors focus on the dynamic, two-­way relationship between culture and development in a way that stresses policy stakes and the value of multidisciplinary collaboration between economists, historians and other social scientists. This book will be of interest to postgraduates and researchers in all the social sciences, as well as to professionals working in national development agencies, international organizations and non-­ governmental organizations. Jean-­Philippe Platteau is Professor of Economics and Director of CRED (Centre for Research in the Economics of Development) at the University of Namur, Belgium. Robert Peccoud, a sociologist, is Director of the Research Department at Agence Française de Développement (AFD), Paris.

Routledge studies in development economics

  1 Economic Development in the Middle East Rodney Wilson   2 Monetary and Financial Policies in Developing Countries Growth and stabilization Akhtar Hossain and Anis Chowdhury   3 New Directions in Development Economics Growth, environmental concerns and government in the 1990s Edited by Mats Lundahl and Benno J. Ndulu   4 Financial Liberalization and Investment Kanhaya L. Gupta and Robert Lensink   5 Liberalization in the Developing World Institutional and economic changes in Latin America, Africa and Asia Edited by Alex E. Fernández Jilberto and André Mommen

  6 Financial Development and Economic Growth Theory and experiences from developing countries Edited by Niels Hermes and Robert Lensink   7 The South African Economy Macroeconomic prospects for the medium term Finn Tarp and Peter Brixen   8 Public Sector Pay and Adjustment Lessons from five countries Edited by Christopher Colclough   9 Europe and Economic Reform in Africa Structural adjustment and economic diplomacy Obed O. Mailafia 10 Post-­apartheid Southern Africa Economic challenges and policies for the future Edited by Lennart Petersson

11 Financial Integration and Development Liberalization and reform in sub-­Saharan Africa Ernest Aryeetey and Machiko Nissanke 12 Regionalization and Globalization in the Modern World Economy Perspectives on the Third World and transitional economies Edited by Alex E. Fernández Jilberto and André Mommen 13 The African Economy Policy, institutions and the future Steve Kayizzi-­Mugerwa 14 Recovery from Armed Conflict in Developing Countries Edited by Geoff Harris 15 Small Enterprises and Economic Development The dynamics of micro and small enterprises Carl Liedholm and Donald C. Mead

18 State-­Owned Enterprises in the Middle East and North Africa Privatization, performance and reform Edited by Merih Celasun 19 Finance and Competitiveness in Developing Countries Edited by José María Fanelli and Rohinton Medhora 20 Contemporary Issues in Development Economics Edited by B.N. Ghosh 21 Mexico Beyond NAFTA Edited by Martín Puchet Anyul and Lionello F. Punzo 22 Economies in Transition A guide to China, Cuba, Mongolia, North Korea and Vietnam at the turn of the twenty-­first century Ian Jeffries 23 Population, Economic Growth and Agriculture in Less Developed Countries Nadia Cuffaro

16 The World Bank New agendas in a changing world Michelle Miller-­Adams

24 From Crisis to Growth in Africa? Edited by Mats Lundal

17 Development Policy in the Twenty-­First Century Beyond the post-­Washington consensus Edited by Ben Fine, Costas Lapavitsas and Jonathan Pincus

25 The Macroeconomics of Monetary Union An analysis of the CFA franc zone David Fielding 26 Endogenous Development Networking, innovation, institutions and cities Antonio Vasquez-­Barquero

27 Labour Relations in Development Edited by Alex E. Fernández Jilberto and Marieke Riethof

35 New International Poverty Reduction Strategies Edited by Jean-­Pierre Cling, Mireille Razafindrakoto and François Roubaud

28 Globalization, Marginalization and Development Edited by S. Mansoob Murshed

36 Targeting Development Critical perspectives on the millennium development goals Edited by Richard Black and Howard White

29 Programme Aid and Development Beyond conditionality Howard White and Geske Dijkstra 30 Competitiveness Strategy in Developing Countries A manual for policy analysis Edited by Ganeshan Wignaraja 31 The African Manufacturing Firm An analysis based on firm surveys in sub-­Saharan Africa Dipak Mazumdar and Ata Mazaheri

37 Essays on Balance of Payments Constrained Growth Theory and evidence Edited by J.S.L. McCombie and A.P. Thirlwall 38 The Private Sector After Communism New entrepreneurial firms in transition economies Jan Winiecki, Vladimir Benacek and Mihaly Laki

32 Trade Policy, Growth and Poverty in Asian Developing Countries Edited by Kishor Sharma

39 Information Technology and Development A new paradigm for delivering the internet to rural areas in developing countries Jeffrey James

33 International Competitiveness, Investment and Finance A case study of India Edited by A. Ganesh Kumar, Kunal Sen and Rajendra R. Vaidya

40 The Economics of Palestine Economic policy and institutional reform for a viable Palestine state Edited by David Cobham and Nu’man Kanafani

34 The Pattern of Aid Giving The impact of good governance on development assistance Eric Neumayer

41 Development Dilemmas The methods and political ethics of growth policy Melvin Ayogu and Don Ross

42 Rural Livelihoods and Poverty Reduction Policies Edited by Frank Ellis and H. Ade Freeman 43 Beyond Market-­Driven Development Drawing on the experience of Asia and Latin America Edited by Makoto Noguchi and Costas Lapavitsas

50 Who Gains from Free Trade? Export-­led growth, inequality and poverty in Latin America Edited by Rob Vos, Enrique Ganuza, Samuel Morley, and Sherman Robinson 51 Evolution of Markets and Institutions A study of an emerging economy Murali Patibandla

44 The Political Economy of Reform Failure Edited by Mats Lundahl and Michael L. Wyzan

52 The New Famines Why famines exist in an era of globalization Edited by Stephen Devereux

45 Overcoming Inequality in Latin America Issues and challenges for the twenty-­first century Edited by Ricardo Gottschalk and Patricia Justino

53 Development Ethics at Work Explorations – 1960–2002 Denis Goulet

46 Trade, Growth and Inequality in the Era of Globalization Edited by Kishor Sharma and Oliver Morrissey 47 Microfinance Perils and prospects Edited by Jude L. Fernando 48 The IMF, World Bank and Policy Reform Edited by Alberto Paloni and Maurizio Zanardi 49 Managing Development Globalization, economic restructuring and social policy Edited by Junji Nakagawa

54 Law Reform in Developing and Transitional States Edited by Tim Lindsey 55 The Assymetries of Globalization Edited by Pan A. Yotopoulos and Donato Romano 56 Ideas, Policies and Economic Development in the Americas Edited by Esteban Pérez-Caldentey and Matias Vernengo 57 European Union Trade Politics and Development Everything but arms unravelled Edited by Gerrit Faber and Jan Orbie

58 Membership Based Organizations of the Poor Edited by Martha Chen, Renana Jhabvala, Ravi Kanbur and Carol Richards 59 The Politics of Aid Selectivity Good governance criteria in World Bank, U.S. and Dutch development assistance Wil Hout 60 Economic Development, Education and Transnational Corporations Mark Hanson 61 Achieving Economic Development in the Era of Globalization Shalendra Sharma 62 Sustainable Development and Free Trade Shawkat Alam 63 The Impact of International Debt Relief Geske Dijkstra 64 Europe’s Troubled Region Economic development, institutional reform and social welfare in the Western Balkans William Bartlett 65 Work, Female Empowerment and Economic Development Sara Horrell, Hazel Johnson and Paul Mosley

66 The Chronically Poor in Rural Bangladesh Livelihood constraints and capabilities Pk. Md. Motiur Rahman, Noriatsu Matsui and Yukio Ikemoto 67 Public-­Private Partnerships in Health Care in India Lessons for developing countries A. Venkat Raman and James Warner Björkman 68 Rural Poverty and Income Dynamics in Asia and Africa Edited by Keijiro Otsuka, Jonna P. Estudillo and Yasuyuki Sawada 69 Microfinance: A Reader David Hulme and Thankom Arun 70 Aid and International NGOs Dirk-­Jan Koch 71 Development Macroeconomics Essays in memory of Anita Ghatak Edited by Subrata Ghatak and Paul Levine 72 Taxation in a Low Income Economy The case of Mozambique Channing Arndt and Finn Tarp 73 Labour Markets and Economic Development Edited by Ravi Kanbur and Jan Svejnar

74 Economic Transitions to Neoliberalism in Middle-­Income Countries Policy dilemmas, crises, mass resistance Edited by Alfedo Saad-­Filho and Galip L. Yalman 75 Latecomer Development Innovation and knowledge for economic growth Banji Oyelaran-­Oyeyinka and Padmashree Gehl Sampath 76 Trade Relations between the EU and Africa Development, challenges and options beyond the Cotonou Agreement Edited by Yenkong Ngangjoh-­Hodu and Francis A.S.T. Matambalya 77 The Comparative Political Economy of Development Africa and South Asia Edited by Barbara Harriss-­White and Judith Heyer 78 Credit Cooperatives in India Past, present and future Biswa Swarup Misra 79 Development Economics in Action (2nd edition) A study of economic policies in Ghana Tony Killick

80 The Multinational Enterprise in Developing Countries Local versus global logic Edited by Rick Molz, Cătălin Ratiu and Ali Taleb 81 Monetary and Financial Integration in West Africa Temitope W. Oshikoya 82 Reform and Development in China What can China offer the developing world Edited by Ho-­Mou Wu and Yang L. Yao 83 Towards New Developmentalism Market as means rather than master Edited by Shahrukh Rafi Khan and Jens Christiansen 84 Culture, Institutions, and Development New insights into an old debate Edited by Jean-­Philippe Platteau and Robert Peccoud

Culture, Institutions, and Development New insights into an old debate

Edited by Jean-­Philippe Platteau and Robert Peccoud

First published 2011 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Simultaneously published in the USA and Canada by Routledge 270 Madison Avenue, New York, NY 10016 Routledge is an imprint of the Taylor & Francis Group, an informa business This edition published in the Taylor & Francis e-Library, 2010. To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk. © 2011 Selection and editorial matter, Jean-­Philippe Platteau and Robert Peccoud; individual chapters, the contributors All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Culture, institutions, and development: new insights into an old debate/ edited by Jean-­Philippe Platteau and Robert Peccoud. p. cm. Includes bibliographical references and index. 1. Economic development–Sociological aspects. 2. Economics– Sociological aspects. 3. Culture. 4. Social institutions. I. Platteau, J. P. (Jean-­Philippe), 1947– II. Peccoud, Robert. HD75.C86 2010 306.3–dc22 ISBN 0-203-84333-9 Master e-book ISBN

ISBN13: 978-0-415-58007-6 (hbk) ISBN13: 978-0-203-84333-8 (ebk)

2010010439

Contents



Notes on contributors Acknowledgements

Part I

xiii xvii

Overview of issues

1

  1 The role of culture in development: an overview

3

J ean - ­P hilippe P latteau

  2 Culture and development: the continuing tension between modern standards and local contexts

20

H é la Y ousfi

  3 Culture and development: do social struggles make a difference?

65

J ean - ­F ran ç ois B ayart

Part II

Religion, family and ethnicity

85

  4 Economic underdevelopment in the Middle East: the historical roles of culture, institutions and religion

87

T imur K uran



Comments on Timur Kuran’s chapter

103

M ona M akram - ­E beid

  5 Family and kinship ties in development: an economist’s perspective E liana L a F errara

107

xii   Contents   6 The demand for disadvantage

125

R ohini S omanathan

Part III

Culture and entrepreneurship

141

  7 Markets and the diffusion of institutional innovations

143

M arcel F afchamps

  8 Culture, management and development

160

P hilippe d ’ I ribarne

  9 The invention of traditions and entrepreneurship: a critical perspective

171

J ean - ­P ierre W arnier

Part IV

Culture and poverty reduction

193

10 Culture matters for poverty, but not because of a culture of poverty: notes on analytics and policy

195

M ichael W alton



Comments on Michael Walton’s chapter

234

S tuti K hemani



Comments on Michael Walton’s chapter

239

S t é phanie M ahieu

Part V

Conclusion

245

11 Revisiting the role of culture

247

J ean - ­P hilippe P latteau and R obert P eccoud



Index

259

Contributors

Jean-­François Bayart is a professor at the Ecole des Hautes Etudes en Sciences Sociales (EHESS). He specializes in comparative politics and the historical sociology of the state, with a special focus on sub-­Saharan Africa, Turkey and Iran. He is the author of several books including his celebrated The State in Africa. The Politics of the Belly (Longman, London, 1993), which was originally written in French and later translated into Spanish and Arabic. He was also a permanent consultant for the Policy Planning Department of the Ministry of Foreign Affairs (1990–2005). Philippe d’Iribarne is French, born in Casablanca, Morocco, on 7 March 1937. He is Managing Director of Gestion et Société (Management and Society), at CNRS (National Center for Scientific Research) in Paris. Gestion et Société is devoted to research training and consultancy, on the adaptation of management to national cultures. He has been a member of the scientific councils of Renault and Gaz de France. He is a member of the scientific council of Agence Française de Développement and of the International Editorial Board of the International Journal of Cross Cultural Management. He has authored 11 books including La Logique de l’honneur (Seuil, 1989; translated into German, Dutch and Chinese; translation into English and Spanish in progress); Cultures et Mondialisation (Seuil, 1998); Le Tiers-­monde qui réussit (Odile Jacob, 2003; translated into English as Successful Companies in the Developing World); Penser la diversité du monde (2008; translation into Arabic in progress); L’épreuve des differences (Seuil, 1989; translation into Chinese in progress). Moreover, he has published more than 100 articles and chapters of books in French, British, Chinese, German, Portuguese, Spanish, Swedish, Japanese, etc. In the first part of his career, he held various positions in the French Civil Service, including advisor to President Georges Pompidou. His education includes the Ecole Polytechnique, Ecole des Mines and Institut d’Etudes Politiques. He is Docteur honoris causa of Mons University. Marcel Fafchamps is Professor of Development Economics at Oxford University and Deputy Director of the Center for the Study of African Economies. He has also taught at Stanford University and worked for the International Labour Organization. His research focuses on market institutions and social

xiv   Contributors networks in sub-­Saharan Africa. Marcel holds a PhD from UC Berkeley and degrees in Law and in Economics from the Catholic University of Louvain. His regional interests are Africa and South Asia. Stuti Khemani is a senior economist in the Development Research Group of the World Bank. She joined the Bank in 1999 in the Young Professionals Program. Her area of research is the political economy of public policy choices, and institutional reforms for development. Her work is published in leading economics and political science journals, including the American Economic Journal, Journal of Development Economics and American Political Science Review. She has studied the impact of electoral politics on fiscal policy and intergovernmental fiscal relations; and analysed political constraints to efficient allocation of public resources for health and education services. She is currently examining the role of mass media and local elections in addressing political incentives for development policies. Her research and advisory work spans a diverse range of countries, including Benin, Bolivia, China, India, the Philippines and Nigeria. She holds a PhD in Economics from the Massachusetts Institute of Technology. Timur Kuran is Professor of Economics and Political Science and Gorter Family Professor of Islamic Studies at Duke University. His research focuses on social change, including the evolution of preferences and institutions. He has just completed a book, The Long Divergence: How Islamic Law Held Back the Middle East (Princeton University Press), which explores the role that Islam played in the economic rise of the Middle East and, subsequently, in the institutional stagnation that accompanied the region’s slip into a state of underdevelopment. His earlier works include Private Truths, Public Lies: the Social Consequences of Preference Falsification (Harvard University Press) and Islam and Mammon: The Economic Predicaments of Islamism (Princeton University Press). Eliana La Ferrara received her PhD in Economics from Harvard University in 1999. She is currently Professor of Economics at Bocconi University, Milan. She is Research Fellow of the Centre for Economic Policy Research (CEPR) and the Innocenzo Gasparini Institute for Economic Research (IGIER); Member of the European Development Research Network (EUDN); Affiliate of the Bureau for Research in Economic Analysis of Development (BREAD); Member of the Council of the European Economic Association. Her fields of interests are development economics, public economics and political economics. She has published original research on the following topics: conflict and financial markets, ethnic diversity, political participation and trust, economics of the family, informal institutions in developing countries. Her publication outlets include the American Economic Review, The Quarterly Journal of Economics, Journal of Public Economics and Journal of Development Economics. Stéphanie Mahieu (García-Pelayo Fellow, Centro de Estudios Políticos y Constitucionales, Madrid, Spain) received her first degree and Masters in Social

Contributors   xv Anthropology from the Free University Brussels and her PhD at the École des Hautes Études en Sciences Sociales (Paris) in 2003. She has held post-­ doctoral fellowships at Viadrina University (Frankfurt/Oder), the Max Planck Institute for Social Anthropology (Halle/Saale), the European University Institute (Florence), and the Centro de Estudios Políticos y Constitucionales (Madrid), where she is currently García-Pelayo Fellow. Her research interests include Eastern Catholicism (she recently co-­edited a volume on the Greek Catholic churches in postsocialist Europe), religion and development, and religion and migrations. Mona Makram-­Ebeid’s career spans the academic, education, political and writing field. She graduated from Cairo University, the American University in Cairo, and the John Kennedy School of Government (at Harvard University). A political sociologist by training, she is currently Distinguished Lecturer in the Department of Political Science at the American University in Cairo, where she has been teaching for the past 17 years in both the sociology and political science departments. She has also been lecturing at several other universities, among them Harvard’s Institute of Politics, Georgetown, Maryland and Northeastern universities, the Georgia Law School, and the American University of Washington. Moreover, she has been a dynamic and active member of Egypt’s Parliament from 1990–5, where she has served on several committees. She worked as an advisor to the World Bank for the MENA region from 1992 to 1996, and was appointed as an expert to the UN Committee for Policy Development (CPD) from 2000 to 2003. Besides numerous publications in the form of contributions to books, she writes regularly in several Arabic newspapers. Robert Peccoud, doctor in sociology, began his professional career at the University of Paris 1 and at the State University of New York, as well as at SUNY, Buffalo, where he taught until 1975. He then entered the French Ministry of Cooperation where he soon became the advisor to the Minister, the Head of Mission, and lastly the Sub-­director of the Education and Research department. He joined the Agence Française pour le Développment (AFD) in 1995. After having been the director for Togo and Tunisia, he created in 2002 the research department of which he is director to this date. Jean-­Philippe Platteau is Professor of Economics and Director of CRED (Centre for Research in the Economics of Development) at the University of Namur, Belgium. His main field is development economics and most of his work has been concerned with the understanding of the role of institutions in economic development, and the processes of institutional change. The influence of non-­economic factors and various frontier issues at the interface between economics and sociology are a central focus of his research projects. He has written numerous articles in academic journals and published several books, including Halting Degradation of Natural Resources – Is There a Role for Rural Communities? (Clarendon Press, Oxford, 1995) with J.M. Baland,

xvi   Contributors and Institutions, Social Norms, and Economic Development (Routledge, London, 2000). Rohini Somanathan received her PhD from Boston University in 1996 and is currently Professor of Economics at the Delhi School of Economics. Her research interests lie at the intersection of development economics, public economics and political economy. A major strand in her research explores mechanisms through which public institutions and community behavior influence patterns of mobility and group inequality. This includes work on the politics of caste identities in India and racial segregation in the United States. She is also actively engaged with examining the effects of a range of public programs on household behavior and poverty in India. Michael Walton is Lecturer in International Development at the Harvard Kennedy School and Senior Visiting Fellow at the Centre for Policy Research, Delhi. He has recently been Visiting Professor at the Delhi School of Economics, and VKRV Rao Professor at the Institute for Social and Economic Change, Bangalore. From 1980–2004, he worked at the World Bank, including as Regional Chief Economist for East Asia and the Pacific, Director for Poverty Reduction, Chief Economist for Human Development, and Advisor on Poverty and Human Development in Latin America and the Caribbean. Prior to joining the World Bank, he worked for the Central Planning and Development Office of the Government of Lesotho. Major publications include: director of World Development Report 1995 on labour (World Bank); co-­editor of Culture and Public Action (Stanford University Press, 2004), co-­ author of Inequality in Latin America and the Caribbean: Breaking with History (World Bank, 2004); co-­director of World Development Report 2005/06 on Equity and Development (World Bank); and co-­editor of No Growth without Equity? Inequality, Interests and Competition in Mexico (Palgrave Macmillan and the World Bank, 2009). Michael Walton has a BA in Philosophy and Economics and an MPhil in Economics, both from Oxford University. Jean-­Pierre Warnier (PhD, Anthropology, University of Pennsylvania, 1975) taught anthropology in Nigeria, Cameroon, and, since 1985, at the University Paris-­Descartes. He is now professor emeritus, and associated researcher at the Centre d’Etudes africaines, Paris. Since 1972, he did research on the economic and political history of the Cameroon Grassfields. He later shifted to the study of bodily and material cultures as technologies of power. One of his latest books is The Pot-­King. The Body and Technologies of Power (Brill, Leiden and Boston, 2007). Héla Yousfi is Associate Professor of Management and Organization at the University of Paris-­Dauphine. She teaches graduate courses on cross-­cultural management, strategic management and organization theory. She specializes in the field of sociology of organizations. Her work has centred on issues such as democratization, good governance, public–private partnerships and socioeconomic transformations in North Africa and Middle East regions.

Acknowledgements

We are grateful to the Agence Française de Développement (AFD), and to Aymeric Blanc in particular, for having funded and hosted the conference (Paris, December 2007) for which the contributions presented in this book have been originally written. The European Development Network (EUDN) was a key partner of AFD in the preparation of this event and could rely on the effective assistance of Tatiana Goetghebuer. Our editors’ work was greatly facilitated by the prompt and adequate responses given by all the contributors to our requests for the production of the book. Finally, the editorial assistance of Marie-­Hélène Mathieu (University of Namur) has been an invaluable help in the finalization of the project.

Part I

Overview of issues

1 The role of culture in development An overview Jean-­Philippe Platteau

A short retrospective view The study of the role of culture has long been the almost exclusive province of sociology and cultural anthropology. Not surprisingly, development economists have lent relatively little attention to it. Yet, it is interesting to notice that, during the 1950s and 1960s, the most prominent among them have usually discussed cultural and social aspects of development, albeit briefly, and expressed (sometimes crude) opinions about whether traditional institutions, attitudes and values are likely to block or to promote economic growth. Some of them, eager to identify determinants or prerequisites of modern economic growth, have made statements very close to the modernization view originated in the field of sociology (see Chapter 2). This applies to Walter Rostow (1960, 1963), Stephen Enke (1963), Simon Kuznets (1966, 1968), Henry Bruton (1965) and Benjamin Higgins (1968) who tended to look at development as the result of a ‘Big Push’ driving traditional societies out of secular stagnation into the era of self-­ sustaining growth. What characterizes their writings is a good amount of optimism concerning the pace at which traditional culture and institutions, which are ill-­suited to the new system of growth, can adjust to its requirements. Other contemporary development economists, however, envisioned institutional and sociocultural change in pre-­modern societies as a much less radical step (Meier and Baldwin 1957; Bauer and Yamey 1957; Hirschman 1958). For example, Gerald Meier and Robert Baldwin adopted a resolutely gradual approach to sociocultural change: Not only must economic organization be transformed, but social organization . . . must also be modified so that the basic complex of values and motivations may be more favourable for development. . . . To avoid human discontent, [however], changes should be introduced in ways that will disrupt the existing culture as little as possible: the cultural change should be selective . . . more rapid progress will come by utilizing as much as possible existing attitudes and institutions rather than by attempting a frontal breakdown of the culture. (Meier and Baldwin 1957: 356, 359)

4   J.-P. Platteau In addition, they explicitly warned against the danger of ethnocentrism in discussions of sociocultural problems: ‘In considering the social and cultural requirements for development, a Western student should not make the mistake of ethnocentrism, that is, assuming that, because the West is developed, Western values and institutions are therefore necessary for development’ (ibid.: 355). Still other authors emphasized the co-­evolutionary nature of cultural and economic development. Thus, Peter Bauer and B.S. Yamey (1957) did not seem to believe that traditional culture ought to be seriously undermined before growth may become possible. Revealingly, they pointed out that ‘the economic history of Japan demonstrates the compatibility of rapid economic change and growth with the preservation of traditional attitudes and social relationships, recast or re-­emphasized as these may be to suit the needs of a new economic order’ (p. 68). In the same vein, Arthur Lewis (1955) was of the opinion that religious beliefs, for example, may evolve and be reinterpreted depending on the economic environment confronting societies. In other words, traditional values and attitudes, whenever they are hostile to economic advancement, will eventually adapt themselves to new economic opportunities (p. 106). As for Alfred Hirschman (1958), he thought that traditional images of change will remain a critical bottleneck of constructive action for economic development until experience modifies them in the appropriate direction. The diversity of views among development economists mirrored the wide spectrum of opinions and perspectives offered by other social scientists. At one extreme, we encounter the ‘hypermodernist perspective’ (Rao and Walton 2004: 10) exemplified most recently by the works of Lawrence Harrison (2000), Samuel Huntington (2000) and David Landes (2000). Through the lens of this approach, laggard countries appear as societies deeply immersed in traditional cultures that are unsuited to market-­oriented development. Without a profound reform of their cultural characteristics, typically initiated from without, no change in their people’s behaviour, norms, habits and collective rules will be possible so that the challenge of modernity will not be met. At the other extreme, are the cultural critics of development for whom economists have shaped modernization perspectives that have had the effect of reifying the distinctions between developed and developing countries originating in the colonial era (see, for example, Escobar 1995). In short, while the former view tends to see culture as a set of external constraints, the latter considers culture as an endogenous product of neo-­colonial conceptions and praxis. In between these two extremes positions, more nuanced approaches have been proposed by scholars such as Mary Douglas, Clifford Geertz, Pierre Bourdieu, Arjun Appadurai, Jan Breman, Louis Dumont, James Scott and others. Resting on the holistic methodology that is the landmark of sociology and cultural anthropology, their approach treats culture as ‘one of the realms of everyday life’ (Rao and Walton 2004: 11), and offers the advantage of shedding interesting light on concrete issues that arise in the course of development. For example, Bourdieu (1990), using his deep knowledge of Kabylia (Algeria), has defined a notion of ‘cultural capital’ that is very useful to understand the perpet-

The role of culture in development   5 uation of inequality since it shows that opportunities have a subjective dimension in the sense that they are perceived differently by people occupying different positions in the society. As a matter of fact, perceived opportunities are transformed into individual aspirations or expectations that people then internalize in actions and choices. Since these actions and choices themselves tend to reproduce the objective structure of life chances, inequality is reproduced. Another illustration is provided by the work of Scott (1976, 1985), who has described in considerable detail how poor people, such as Asian peasants, view themselves in the subtle hierarchies to which they belong, how this worldview shapes their objectives and their perceptions of the constraints they are facing, and how they conceive of possible means of action and forms of resistance against the dominating groups. Far from being passive agents, they try to assess their position in an asymmetric world, and they perform actions in the realistic hope of moderating the authority of their patrons rather than attempting to overturn the system that oppresses them yet allows them to subsist. As argued by other sociologists/anthropologists (see, for example, Breman 1974; Alexander 1982), ‘small people’ tend to be immersed in a culture that has inculcated in them the values of loyalty, deference, respect and gratitude to the ‘big people’ who provide for their livelihood. Thus, while economists lay stress on the exchange, give-­and-take character of the relationship between a patron and his clients – under conditions of highly unequal wealth endowments, the clients need insurance against the risk of hunger (and the patron may need a compliant workforce available at their beck and call) – sociologists/anthropologists draw a lot of attention to the cultural underpinnings of patronage relationships, in particular, to the father-­and-son analogy projected as their correct representation. A last illustration is borrowed from the literature dealing with common property resources. This is a field in which anthropologists have made important contributions to our understanding of the mechanisms of local cooperation and the role of cultural and social norms in particular. Since economists have become interested in producing a theory of collective action on the commons, the insights from other social sciences are especially valuable. More precisely, in order that villagers refrain from excessive extraction of available natural resources, as they would if they were allowed to give free rein to their self-­interest, they must face some form of punishment whenever they depart from cooperative practices. One mechanism that has often been stressed in anthropological writings is the operation of what may be called a ‘social exchange game’ in which a member of the community earns social prestige and public respect if he or she conforms with the normative pattern of social cooperation, but experiences social shame and opprobrium if he or she does not. For example, from his work about the Moose society of Burkina Faso, Jean Badini concludes: Activated by social rebuke and the accompanying public humiliation, the feeling of shame appears as the most formidable weapon in the service of the traditional Moose pedagogy (the moaga). Above the individuals, indeed,

6   J.-P. Platteau this feeling asserts the supremacy of social judgment and constitutes a powerful regulating mechanism to which everybody submits. . . . Since a person can exist only through collective opinion, it is collective opinion that rates people and rare are those who are willing to incur the risk to defy it. The point is that its verdict is merciless and without appeal. (Badini 1994: 146–7) Looking more precisely at water-­access rules in the lagoon fishery of Bahia, Brazil, John Cordell and Margaret McKean find that they are essentially enforced through a decentralized mechanism based on an ethical code of honour and social respect: It is impossible to fish for long in a given community without receiving and showing respeito. People honor each other’s claims because of respeito, which is created, bestowed, and reaffirmed through sometimes trivial and sometimes substantial acts of benevolence bordering on self-­sacrifice. . . . Failure to cooperate in these practices can be much more devastating for a fisherman than would be breaking a government law. Respeito is a cognitive reference point to the community conscience. It influences how fishermen evaluate each other’s actions on and off the fishing grounds. It is a yardstick for measuring the justice of individual acts, especially in conflicts. Collective social pressure to conform to the ethics of fishing is reflected in the ôlho do povo (watchfulness of the community’s eye, or sense of justice), reminiscent of the forceful moral and ethical standard in Palauan fishing, ‘words of the lagoon’. Reputations rise and fall in terms of the ôlho do povo. The ôlho do povo determines whether territorial competition in fishing is deliberate or accidental, and whether it is antagonistic enough to require counteraction. (Cordell and McKean 1986: 94, 98) The above contributions, which are often anchored in considerable amounts of fieldwork, are a far cry from the works of academic anthropologists and sociologists who have chosen to remain confined to a critical perspective and to deconstruct development concepts in very general and abstract terms, thus carefully avoiding becoming involved constructively in any development policy debate. This perspective has led their proponents to argue for a ‘post-­development’ discourse that emphasizes cultural specificities so much that any generalization across societies becomes impossible. Diversity is extolled because an intrinsic worth is attributed to anything ‘local’, partly seen as a move of resistance against the hegemonic West. Since any general statement becomes invalid, the post-­ development approach places itself outside the scope of science: inquiries into societies can only consist of highly specific case studies that verge on the anecdotal. The absence of theoretical underpinnings that is implied in this way of approaching the study of culture is radically different from recent evolutions that have occurred in the field of economics, and therefore creates an unbridgeable gap of misunderstanding between adherents of post-­modernism and economists.

The role of culture in development   7 In economics, indeed, the search is for patterns in economic and social life ‘that, while not universal, are widely generalizable’ (Bardhan and Ray 2008b: 16).

Recent contributions by economists If the relationship between culture and development stopped being a concern of economists during most of the 1970s and 1980s, it is interesting to note that, when a surge of renewed interest occurred in the subsequent decades, the response was much more systematic and articulated than it was in the immediate post-­war period. In those times, indeed, as witnessed by the above citations, expressions of opinions by leading development economists on the theme of culture remain confined to general thoughts. The reason why the economic perspective on this theme has drastically changed is evident: thanks to the progress of game theory, economists acquired an analytic framework within which to conceptualize culture in a way understandable by them. This does not mean that they had found a magical key to comprehend all the main aspects of cultural life in societies, but rather that they had reached a better grasp of why culture is an indispensable component of modern market economies and obtained a more precise idea of what economics can or cannot teach them about the role of culture. In the process, they became acutely aware of where lies the ultimate frontier of their field of knowledge, and where they need to borrow from other disciplines, in particular, history, sociology, anthropology and political science. A direct consequence of the use of game theory to analyse individual behaviour in the context of social groups is that no unique model is available to represent a supposedly universal reality. The profession has thus moved away from the ambitious project of building a unique representation of the market economy based on simple assumptions derived from General Equilibrium theory. Applied to the realm of culture, this means that economists now follow the practice of writing specific models within a contextualized framework. As a consequence, they do not work with a general definition of culture in their minds, but instead use specific concepts of culture that are appropriate to the problem at hand. Let me illustrate this important methodological point. When they try to elucidate the conditions for honest dealing in market exchanges, economists introduce culture under the form of values that influence preferences. For example, they assume that agents are reciprocators rather than purely self-­interested, opportunistic individuals. This may also imply that they are ready to incur costs to punish opportunists even while they do not directly benefit from sanctioning (Platteau 2000). Or, when they address issues of ethnic diversity or discrimination, culture takes on the form of group-­based identity feelings that influence individual behaviour in addition to standard determinants of utility such as material gain or leisure. In studies of the way people relate to each other in the context of small working groups, individual utilities may be posited to depend on self-­esteem considerations that involve social prestige or shame (see, for example, Platteau and Seki 2007).

8   J.-P. Platteau In all these instances, it bears emphasis that values are only one component of individual preferences. A direct implication is the possibility of conflicts between norms (transmitted by socializing agencies) and other motives, hence the existence of a choice problem. To put it another way, a norm is always considered by economists as a behavioural rule rather than as a prescribed action (Greif 2006). It can therefore be subjected to trade-­offs and cost–benefit calculations. Just to illustrate, an individual may be sensitive to identity considerations, yet if satisfying his or her identity feelings is too costly in terms of income, he or she will choose to ignore them in actual choices, such as whether to hire a better qualified worker who is unrelated to him or her, or a less well-­qualified relative. On the other hand, when they deal with issues of coordination in the presence of serious information asymmetries, economists bring social norms into the picture. Such norms perform the necessary function of causing expectations and actions to converge so that an equilibrium position, known as a social convention, can be established. In this new analytical framework, social norms appear as stable patterns of behaviour based upon self-­enforcing expectations, that is, as beliefs that stand confirmed by the observation of the action-­choices of other agents. Since they thus constrain people’s behaviour and coordinate their beliefs once they are established, they are genuine institutions generated by a particular culture. Equally worth stressing is the following characteristic of the economists’ approach to culture: there are typically multiple institutional solutions to a coordination problem arising in societies. An inherent feature of human institutions is that they are only one possible representation among many others. In the words of Masahiko Aoki, ‘If there is only one equilibrium corresponding to the technological specification of the structure of the game, then that equilibrium is little more than a representation of the technological condition, and not an institution’ (2001: 16). Culture then fulfils the function of a selecting device that causes one possible equilibrium to arise and be established. Often, cultural specificity takes on the form of particular equity principles which serve as focal points in a given community. Thus, Roger Myerson defines cultural norms as ‘the rules that a society uses to determine focal equilibria in game situations’, bearing in mind that ‘there may be some situations where people of a given culture might look to equity principles to determine a focal equilibrium’ (1991: 113). In the same vein, Peyton Young writes that ‘equity principles are the means by which the parties coordinate on a particular solution given the constraints imposed by economic rationality’ (1994: 81). Finally, the idea that people have inherent cultural attributes that do not evolve is unacceptable to most economists as it is to most other social scientists. In the framework of economics, social norms, values and conventions are considered to be subject to inertia, but only because they do not change as rapidly as the other variables of concern to economists (prices, quantities of goods and services bought and sold, etc.). To put it another way, it is only in a relative sense that cultural representations are deemed to be slow to change. It is true that a different interpretation seems to follow from the definition of social convention as the equilibrium outcome of a game: a social convention is a behavioural

The role of culture in development   9 regularity that is self-­perpetuating because individual expectations and behaviours are in equilibrium when a convention is well-­established (Young 1996). Nonetheless, economists, like sociologists and cultural anthropologists, are deeply interested in the question as to how a social practices change when an established convention is disrupted, say as a consequence of a change in the economic or technological environment. This is obviously a moot problem, and some economists like to think of culture as providing guidance in such circumstances so that a new equilibrium position can be attained: culture is then conceived as a symbolic device which helps expectations and action/choices to converge. Such a device appears to be needed because, when left to act in a self-­regarding manner on the basis of available information, agents who are strategically rational (they are assumed to have a perfect ability to reason inductively about a feedback mechanism between their own choices and the choices of the others) will not be in a position to find a new equilibrium position after the previous one has been disrupted. If agents are instead assumed to follow rules of limited rationality (they behave myopically), they may be able to find a new equilibrium position, yet there is absolutely no guarantee that they will hit the socially superior one (Bowles 2004: 406–10). The latter result is actually a common finding in economic analyses of social norms and conventions. In the same manner as economists, development economists, in particular, have emphasized pervasive failures in the realm of markets, they are prone to stress the possibility that inefficient social norms come to be established and sustained over time. Economic theory can actually account for those paradoxical situations in which individuals choose to support rules that they do not like or even find oppressive. As shown by Georges Akerlof (1976) and Timur Kuran (1995: chs 6–8), such an outcome can be generated by rational individuals if there exist: 1 an effective network of mutually reinforcing social sanctions against dis­ obedience; and 2 a system of converging expectations that sustain the existing arrangement. The immediate implication of the existence of a web of self-­reinforcing sanctions (think of the ostracization of individuals who have violated caste-­based rules or norms) is that everyone is both a victim and a supporter of a harmful rule or social norm which is not even enforced by any centralized agency. For example, the erstwhile prohibition of remarriage of widows in India works effectively as a social norm although there is no authority that imposes and enforces it, and although many Indian widows acutely suffer from it. A decentralized system of sanctioning is especially effective if meta-­punishment is applied, that is, if a person is considered guilty not only because he or she does not abide by the norm or the rule, but also because he or she does not punish the offender, say by breaking ties with him or her. Note that such an explanatory framework sheds light on how sanctions have been self-­reproducing, not on how they arose (Kuran 1995: 118–36; Basu 2000: 136–47).

10   J.-P. Platteau Even when individual preferences are assumed to be influenced by culturally-­ rooted emotions or feelings (such as the drive to reward people who behave nicely towards us, and to punish those who do not) that have the effect of enlarging the set of possible equilibria, allowing the emergence of new equilibria that are more efficient than those previously existing, there is no certainty that agents will end up actually using these improved opportunities. Thus, the so-­called fairness equilibria existing in the psychological games devised by Rabin (1993) do not necessarily correspond to more efficient outcomes than strict Nash equilibria. On the one hand, fairness may not rule out (strict Nash) equilibria that are inefficient1 and, on the other hand, fairness may rule out (strict Nash) equilibria that are efficient.2 When a dynamic, rather than a static, approach to institutions and social norms is adopted, the conclusion is not markedly different. Thus, the use of evolutionary game theory to model the emergence, diffusion and demise of particular rules or behavioural norms does not lead to the conclusion that efficient rules or norms will necessarily be the only ones to survive. This is despite the fact that they are modelled as the outcome of an organic process of Darwinian natural selection which epitomizes the competitive pressures of the market and the invisible hand: they emerge not as a result of rational, purposeful design by any individual or organization of individuals, but as the result of spontaneous evolution, say, because the people learn from experience that following a given constraint or custom can actually serve their own individual interests (Aoki 2001: 40; Bowles 2004). Path dependence, for example, means that evolutionary trajectories may be produced that are highly dependent on initial conditions and on the order in which players happen to meet. Not only are the paths likely to end up in different equilibrium configurations, but some of these configurations may well be more socially efficient than others (Young 1998: 8; see also North 1990: 92–104). Possible factors behind path dependence are: large set-­up or fixed costs, learning effects, co-­evolution of preferences and norms (for example, rationalizations are constructed to justify past bad choices), or analogues to external economies (spillovers) and increasing returns (‘some institutions may be complementary, each enhancing the functioning of the other, while some institutions may reduce the effectiveness of other institutions’) (North 1990: 93–4; Bowles 2004: 90–1).

About this book As I hope that the above discussion has shown, the time is ripe for a constructive dialogue between economists and other social scientists. There are numerous issues, cultural and institutional, around which they can learn from each other and, if we except extreme positions such as those inspired by the hypermodernist or the cultural critics approaches, there are obvious ways in which different disciplines can come to terms with each other, and acquire genuine respect for contributions made by scholars with a different background. Thanks to what Pranab

The role of culture in development   11 Bardhan and Isha Ray (2008b) have called ‘a thin theory of human action’ based on the three pillars of methodological individualism, utility maximization and exogenous preferences, economics has produced microeconomic models of cultural influences that are precise, parsimonious and endowed with a predictive power. Different models address different issues, as illustrated by Kaushik Basu’s distinction between three categories of norms useful for economic analysis: rationality-­limiting, preference-­changing and equilibrium-­selecting (Basu 2000: 72–3). Sociologists and anthropologists generally feel uncomfortable with the three aforementioned pillars of the economists’ approach. Their analyses tend to be more holistic and more complex than those produced by economists: they spend more time specifying the context, they take into account more aspects of the problem at hand or of the societies/groups under study, and they lend more attention to processes as opposed to outcomes. This largely explains why findings tend to be less sharp and less operationalizable in sociological and anthropological studies than in the economists’ works. It is our contention that the obvious complementarities between the two strands of studies could be exploited to the benefit of all the disciplines involved. What is required, in addition to sufficient agreement around the research questions to be tackled, is that both economists and sociologists/anthropologists make a significant step in the direction of recognizing the worth of other disciplines’ contributions. For economists, such a step amounts to recognizing the need for better contextualization of the problems studied, or at least to realizing that useful insights can be gained from rather holistic studies. As for sociologists/anthropologists, the required effort is to admit that working with a tight analytical framework may be an advantage whenever it is important to assess the logical consistency of the arguments advanced, to specify the conditions under which particular effects may obtain or not, and to derive propositions susceptible of being rigorously tested against empirical evidence. In some instances, the works of sociologists/anthropologists may serve as exploratory advances aimed at clearing up issues that are too complex to probe analytically without a preliminary understanding of what is at stake. The problem with economic models is, indeed, that many different hypotheses are often a priori plausible, so that reliable first-­hand material must be available to vindicate the precise assumptions chosen by the researcher. In other instances, the works of sociologists/anthropologists may be expected to supply additional knowledge about aspects of a problem that the economic methodology is ill-­suited to address (for example, the symbolic dimensions of human actions). Multidisciplinary interactions do not need to take the form of collaboration on common projects. Yet, at the minimum, they require that scholars lend a friendly ear to the discourse of the other social science disciplines. A long way remains to be trodden for this objective to be reached, however. As pointed out by Vijayendra Rao and Michael Walton, indeed, past work has often involved weak interdisciplinary communication. When anthropologists are read and cited by economists, the work is typically

12   J.-P. Platteau several decades old and is used because it fits easily within the hypotheses of the economist, rather than because it represents the best the discipline has to offer. Similarly, many anthropological critiques of economics base their understanding on simplistic ‘Economics 101’ models rather than the more sophisticated thinking in the field. (2004: 368) In reality, whenever they have occurred, exchanges of views and research ideas between social scientists have proven very valuable. This certainly holds true of the pioneering attempts of Pranab Bardhan (and Isha Ray) to bring economists and anthropologists together on two occasions. They resulted in fecund interdisciplinary exchanges that were focused on two well-­delineated topics, the measurement of economic and social change in rural India, on the one hand, and the local commons and collective action, on the other hand. Two books came out of these efforts (Bardhan 1989; Bardhan and Ray 2008a). Likewise, Rao and Walton, again an economist teamed up with a sociologist/anthropologist, organized a transdisciplinary discussion around the theme of inequality and poverty with an interesting focus on the role of public action to mitigate them. Another book, cited above, appeared as a result (2004). Finally, mention must be made of the book edited by Lawrence Harrison and Samuel Huntington (2000). Like the previous ones, it also focus on culture as an independent or explanatory variable, yet a major difference lies in the fact that most of the contributing authors are not economists. The present work is the outcome of yet another attempt made with a view to encouraging development-­oriented cross-­fertilization between different disciplines of the social sciences. It also took the form of a conference in which participants from different disciplines exchanged views on the role of culture in development. This conference, organized by the Agence Française de Développement (AFD) and the European Development Network (EUDN), took place in Paris on 5 December 2007. As in all the above-­mentioned initiatives, the idea followed by the organizers was to bring together economists, sociologists, anthropologists and political scientists to present their views on a number of concrete issues involving cultural aspects and requiring careful analysis before policy orientations can be defined. It was therefore decided from the outset that no time was to be spent on general definitions of culture or on broad conceptualizations of the relationship between culture and development, and between culture and institutions. Conversations between scholars were thus focused on particular issues and central attention was directed to precise findings as well as to possible policy implications. The only exception in the following series of contributions is the contribution by Héla Yousfi (Chapter 2), whose work was commissioned by the AFD to provide, ahead of the conference, a rather detailed review of the sociological literature on the theme of the conference. This was intended to serve as a sort of background material in the light of which the ongoing debates could be read from the vantage point of sociology. All the other contributions, whether written

The role of culture in development   13 by economists or other social scientists, have a clear focus and deal with an issue that lies at the heart of development debates today. They are of unequal length because in the design of the conference some corresponded to full-­fledged contributions while others were more pointed contributions on a quite specific aspect of the role of culture, or were discussions of the main paper presented in the session. Contributions by economists include those by Eliana La Ferrara on the influence of family and kinship ties in development (Chapter 5), by Marcel Fafchamps on social networks, markets and the diffusion of institutional innovations (Chapter 7), by Timur Kuran on the historical roles of culture, institutions and religion to explain the economic underdevelopment of Middle Eastern countries (Chapter 4), and by Rohini Somanathan on the evaluation of positive discrimination policies in India (Chapter 6). Contributions by other social scientists include those by Héla Yousfi (Chapter 2), by Michael Walton on the relevance of cultural processes to the poor (Chapter 10), by Jean-­François Bayart on the role of social struggles (Chapter 3), by Philippe d’Iribarne on the role of culture for the management and functioning of business firms (Chapter 8), and by Jean-­Pierre Warnier on the influence of tradition on the business mindset (Chapter 9). Kuran’s contribution is carefully discussed by Mona Makram-­Ebeid (an Egyptian sociologist), while the chapter written by Michael Walton benefits from two insightful sets of comments, one by Stuti Khemani (an Indian economist), and the other by Stéphanie Mahieu (a Belgian anthropologist). Note that the relationship between culture and institutions receive a lot of attention in the chapters by Timur Kuran, Marcel Fafchamps and Eliana La Ferrara whereas policy aspects are at the centre stage of the chapters written by Rohini Somanathan, Michael Walton and Philippe d’Iribarne. The conference was ended by a synthesis session in the course of which one economist, Jean-­Philippe Platteau, and one anthropologist, Olivier de Sardan, presented their summary views about the dialogue between economists and other social scientists on the subject of the relationship between culture and development as it emerged from the exchanges of ideas that took place during this event. The first synthesis is reported in the present introductory chapter and forms a sort of match to Chapter 2 that follows. The book has been structured in five successive parts that do not correspond tightly to the structure of the conference itself. In deciding the regrouping and succession of chapters, we have ensured a maximum amount of coherence of the subjects addressed by the authors so as to facilitate the reading. This introductory chapter concludes with a short presentation of the content of all chapters. Part I In her survey of various visions of the relationship between culture and development in Chapter 2, Héla Yousfi argues that a common feature that they share beyond their many divergences is their persistent difficulty in grasping the articulation between the role of culture (suspected of bringing back tradition) and

14   J.-P. Platteau individual autonomy (seen as a condition of modernity). There is an urgent need for development thinking to break away from the tradition/modernity dichotomy if more effective policies are to be designed and implemented. In Chapter 3, Jean-­François Bayart expresses strong views in the ongoing debate about the pertinence of culturalist interpretations. A key message emerging from his chapter is that culture is essentially endogenous to more fundamental determinants located in the economic and political spheres. Ethnic, religious and family identities are not socially given but are fabricated by states, whether pre-­colonial, colonial or post-­colonial. Identification is by definition contextual, relational and relative. Ethno-­developmental approaches wrongly attribute a causal status to mystified cultural factors, thus eluding the thickness of historical paths followed by human societies as well as the ideological or political construction of ethnotypes. In the end, the true subject is not culture but history. Part II In Chapter 4, Timur Kuran applies institutional economic analysis to examine whether cultural factors associated with Islam have been conducive or detrimental to economic development in Middle Eastern countries. His central conclusion is that distinct Islamic institutions – the law of partnership, inheritance regulations and the waqf (religious endowment) – lie at the roots of the region’s slip into a state of economic underdevelopment. In other words, upon careful analysis economically harmful attitudes and behaviours that are often treated as manifestations of Islam appear as symptoms of institutional deficiency rather than as immutable traits incapable of institutional evolution. As institutions change, so does the associated culture. Mona Makram-­Ebeid, in her comments, essentially agrees with the approach followed by Timur Kuran. She mostly expands on the types of social movements that the contradictions highlighted by Kuran have nurtured, emphasizing their differentiation along gender and class lines as well as their variations between countries. Chapter 5, by Eliana La Ferrara, discusses the role of family and kinship ties in economic development. In particular, it analyses the functions that families play in transmitting values to younger generations and in establishing social norms that have a profound impact on individual decisions involving fertility, workforce participation and investments in education and health. The role of families as substitutes for markets, with special reference to informal insurance and credit markets, is also considered by the author. A common theme that runs through this contribution is both the ‘comparative advantage’ that families or kin groups may have in performing some of these roles, and the potential inefficiencies that may result from their involvement. In Chapter 6, Rohini Somanathan attempts to assess affirmative action policies as they have been implemented in India since the time of independence. The account provided is highly sceptical about the virtues of quotas set to bring about greater representation of socially marginalized communities in politics, government, and among the educated elite. It is true that greater political repre-

The role of culture in development   15 sentation did actually occur, yet representatives either did not attempt to or did not succeed in providing their constituencies with the opportunities that would bring about a convergence in the welfare of these groups with the rest of the population. A major lesson from the Indian experience is therefore that reservations in higher education or the civil service are unlikely to provide the average individual in marginalized groups or communities with substantial benefits, at least in the near term. Part III In Chapter 7, Marcel Fafchamps starts from the idea that Africa has weak market institutions and unsophisticated market practices. He argues, however, that there is no reason to suspect African cultures to be any more inimical to market development than other cultures. This conclusion arises from comparing market institutions in Africa to those in other parts of the world today and in the past. The problem, according to him, is that many African entrepreneurs are unfamiliar with institutional innovations that have emerged in other parts of the world. It is this lack of familiarity that makes them less competitive and makes African economies less productive. Moreover, private patterns of socialization that influence the formation of business networks often result in the overrepresentation of specific ethnic or religious groups. This fuels political tension and inhibits the spread of institutional innovations leading to a low equilibrium trap in which countries discourage foreign investors for fear of losing control. The other chapters devoted to the problem of culture and entrepreneurship have been contributed by sociologists. In Chapter 8, Philippe d’Iribarne addresses the issue of the suitability of social norms to the requirements of modern market development from the vantage point of business undertakings. His line of approach is different from that adopted by Marcel Fafchamps. As a matter of fact, he points out that ethical values that stress solidarity and loyalty to members of particularistic groups (clan, kinship group, caste, etc.), as commonly found in Africa, are potentially detrimental to efficiency in business activities. Two strategies are available to overcome this contradiction. On the one hand, strong standards can be established at the level of the indigenous firms so as to protect employees against any demand for redistribution emanating from relatives and friends. On the other hand, the firm could be conceived so as to foster new ties of solidarity and loyalty that dominate or evict erstwhile ties and associated obligations. In Chapter 9, and in the same context of firm development in poor countries, Jean-­Pierre Warnier ponders over the meaning of the notion of tradition. His central point is that a paradigmatic shift has recently occurred in the sociological literature. Whereas before tradition was understood as a continuous link with the past, giving rise to analyses of resistance of tradition against the requirements of modernity, now it is increasingly viewed as norms and behavioural patterns that are ‘re-­invented’ or ‘re-­cycled’ depending upon historical circumstances and opportunities. When the latter approach is adopted, both the remarkable

16   J.-P. Platteau r­ esilience of traditions and cultural values in ‘modern’ environments and the diversity of trajectories followed by human societies in the modern era are much easier to understand. Applications to political and business forms are proposed by the author. Another point stressed by Warnier is that actors have available to them several cultural repertoires that are not mixed together. They are ‘bi- or tri-­ cultural’ insofar as they switch from one repertoire to another depending on the context. These views lead him to an optimistic, evolutionary perspective regarding the adaptation of poor societies to modernity. Part IV In Chapter 10, Michael Walton examines the persisting notion that deprived groups stay poor because of their culture, using the case of low caste and tribal groups in India as illustrative material. This notion has shown surprising durability given the lack of support for it amongst both anthropologists and economists. The author strongly argues that the concept is essentially flawed, but that cultural processes are relevant to the poor. Indeed, an analytical prism of ‘inequality traps’ that include cultural as well as economic and political processes in their transmission mechanisms proves extremely useful. These can be considered as self-­enforcing equilibria sustained either by rational behaviour with limited information, or by interactions between unequal power, social status and stigma (which are associated with external exclusion mechanisms). Both normative and policy implications follow from such a diagnosis. In particular, group-­based targeted preferential policies seem to follow from the normative analysis associated with the ‘inequality trap’ story. Yet, as stressed by Somanathan in Chapter 6, the results are not very encouraging if one looks at the situation of the average member of deprived groups. Especially worth emphasizing in Walton’s contribution is the way in which he makes ample use of economic approaches to address an issue that traditionally belonged to the anthropological and sociological terrain. In her comments on Walton’s contribution, Stuti Khemani precisely takes issue with such preferential policies, arguing that they might exacerbate inefficiencies and inequalities in resource allocation if we are in a world of costly bargaining between groups. Moreover, it is very likely that in India and other countries a substantial set of people sharing characteristics of poverty are fragmented along the lines of cultural identity. In the end, political mobilization around social identity might have the perverse effect of increasing within-­group inequality by allowing only some members to escape from poverty through group-­targeted transfers. The alternative, consisting of universal provision of public goods, does not have this disadvantage. The comments offered by Stéphanie Mahieu, coming from an anthropologist, are of a completely different nature. Especially worth noting is her discussion of the risk of cultural relativism (also mentioned by Jean-­François Bayart in Chapter 3) when effort is made to understand odd behaviours of marginalized groups, for instance when the question is raised as to why some groups perpetuate practices that obviously maintain them in a precarious situation. The answer often lies in the cultural sphere, she

The role of culture in development   17 argues. But the problem is that culture is often treated as a black box which can explain apparently irrational behaviour. Part V Finally, Chapter 11, written by the two editors of this book, draws some central lessons coming out of the dialogue that unfolded in the preceding contributions.

Notes 1 Thus, in the Prisoner’s Dilemma, the strict Nash equilibrium in which agents refuse to cooperate is also a fairness equilibrium because in this outcome each player is satisfying both his or her desire to hurt the other and his or her material self-­interest. 2 In a Chicken game, for example, the worst outcome may arise as a fairness equilibrium because all players may want to hurt each other while in the Pareto-­optimal solutions of the game, a player prefers to back down if the other player has chosen to assert his or her interest.

References Akerlof, G. (1976) ‘The Economics of Caste and of the Rat Race and Other Woeful Tales’, Quarterly Journal of Economics, 90 (4), pp. 599–617. Alexander, P. (1982) Sri Lankan Fishermen – Rural Capitalism and Peasant Society, Canberra: Australian National University. Aoki, M. (2001) Toward a Comparative Institutional Analysis, Boston: MIT Press. Badini, A. (1994) Naître et grandir chez les Moosé traditionnels, Paris and Ouagadougou: Sépia-A.D.D.B. Bardhan, P. (ed.) (1989) Conversations Between Economists and Anthropologists, Delhi: Oxford University Press. Bardhan, P. and Ray, I. (eds) (2008a) The Contested Commons – Conversations Between Economists and Anthropologists, Oxford: Blackwell. —— (2008b) ‘Economists, Anthropologists, and the Contested Commons’, in Bardhan, P. and Ray, I. (eds) The Contested Commons – Conversations Between Economists and Anthropologists, Oxford: Blackwell, pp. 1–24. Basu, K. (2000) Prelude to Political Economy – A Study of the Social and Political Foundations of Economics, Oxford: Oxford University Press. Bauer, P.T. and Yamey, B.S. (1957) The Economics of Under-­Developed Countries, Cambridge: Cambridge University Press. Bourdieu, P. (1990) The Logic of Practice, Stanford: Stanford University Press. Bowles, S. (2004) Microeconomics – Institutions, and Evolution, New York: Russell Sage Foundation, and Princeton: Princeton University Press. Breman, J. (1974) Patronage and Exploitation – Changing Agrarian Relations in South Gujarat, Berkeley: University of California Press. Bruton, H.J. (1965) Principles of Development Economics, Englewood Cliffs: Prentice-­ Hall. Cordell, J.C. and McKean, M.A. (1986) ‘Sea Tenure in Bahia, Brazil’, in National Research Council, Proceedings of the Conference on Common Property Resource Management, Washington, DC: National Academy Press, pp. 85–112.

18   J.-P. Platteau Enke, S. (1963) Economics for Development, Englewood Cliffs: Prentice-­Hall. Escobar, A. (1995) Encountering Development: The Making and Unmaking of the Third World, Princeton: Princeton University Press. Greif, A. (2006) Institutions and the Path to the Modern Economy – Lessons from Medieval Trade, Cambridge: Cambridge University Press. Harrison, L.E. (2000) ‘Why Culture Matters?’, in Harrison, L.E. and Huntington, S.P. (eds) Culture Matters – How Values Shape Human Progress, New York: Basic Books, pp. xvii–xxxiv. Harrison, L.E. and Huntington, S.P. (eds) (2000) Culture Matters – How Values Shape Human Progress, New York: Basic Books. Higgins, B. (1968) Economic Development — Problems, Principles and Policies, New York: W.W. Norton [first edition 1959]. Hirschman, A.O. (1958) The Strategy of Economic Development, New Haven and London: Yale University Press. Huntington, S.P. (2000) ‘Foreword: Cultures Count’, in Harrison, L.E. and Huntington, S.P. (eds) Culture Matters – How Values Shape Human Progress, New York: Basic Books, pp. xiii–xvi. Kuran, T. (1995) Private Truths, Public Lies — The Social Consequences of Preference Falsification, Cambridge, MA, and London: Harvard University Press. Kuznets, S. (1966) Modern Economic Growth, New Haven: Yale University Press. —— (1968) Toward a Theory of Economic Growth, New York: W.W. Norton. Landes, D. (2000) ‘Culture Makes Almost All the Difference’, in Harrison, L.E. and Huntington, S.P. (eds) Culture Matters – How Values Shape Human Progress, New York: Basic Books, pp. 2–13. Lewis, W.A. (1955) The Theory of Economic Growth, London: George Allen & Unwin. Meier, G.M. and Baldwin, R.E. (1957) Economic Development, New York: John Wiley & Sons. Myerson, R.B. (1991) Game Theory – Analysis of Conflict, Cambridge, MA and London: Harvard University Press. North, D.C. (1990) Institutions, Institutional Change and Economic Performance, Cambridge: Cambridge University Press. Platteau, J.P. (2000) Institutions, Social Norms, and Economic Development, London: Routledge. Platteau, J.P. and Seki, E. (2007) ‘Heterogeneity, Social Esteem and the Feasibility of Collective Action’, Journal of Development Economics, 83 (2), pp. 302–25. Rabin, M. (1993) ‘Incorporating Fairness into Game Theory and Economics’, American Economic Review, 83 (5), pp. 1281–302. Rao, V. and Walton, M. (2004) ‘Culture and Public Action: Relationality, Equality of Agency, and Development’, in Rao, V. and Walton, M. (eds) Culture and Public Action, Stanford: Stanford University Press, pp. 3–36. Rostow, W.W. (1960) The Stages of Economic Growth, Cambridge: Cambridge University Press. —— (1963) The Economics of Take-­Off into Sustained Growth, New York: St. Martin’s Press. Scott, J.C. (1976) The Moral Economy of the Peasant, New Haven: Yale University Press. —— (1985) Weapons of the Weak: Everyday Forms of Peasant Resistance, New Haven: Yale University Press.

The role of culture in development   19 Young, H.P. (1994) Equity – In Theory and Practice, Princeton: Princeton University Press. —— (1996) ‘The Economics of Convention’, Journal of Economic Perspectives, 10 (2), pp. 105–22. —— (1998) Individual Strategy and Social Structure – An Evolutionary Theory of Institutions, Princeton: Princeton University Press.

2 Culture and development The continuing tension between modern standards and local contexts Hèla Yousfi1

Introduction The effect of culture on development has been a subject of interest for a long time. It is increasingly acknowledged that culture matters as much as economics or politics to the process of development, but some confusion remains about precisely how it matters. The absence of collective agreement on the meaning of the terms ‘culture’ and ‘development’ undermine attempts to shape a clear framework for understanding the effect of culture on development processes.2 The concepts of culture and development may mean different things to different actors, and like the issues they seek to clarify, they have been subject to various controversies. Consequently, over the past 50 years the relationship between ‘culture’ and ‘economic development’ has been, and can be viewed variously as, causal, correlative or relatively autonomous. This chapter does not claim to be exhaustive. It offers some critical considerations on this theme.3 Taking into account the fact that there is no single approach concerning the role of culture on development, we seek to examine the work of some key thinkers in order to shed light on the different ways in which culture has been considered, and the different underlying assumptions about culture that have been taken into consideration by development strategies. Thus, our focus is on whether – and how – culture matters. What are the different connections by which different culture-­defining assumptions can influence diverse aspects of development? We start, in the first part of the chapter, by critically examining how both modernization theory and its critics have shaped the framework within which culture has been deployed and debated in development thinking. We will highlight how the tradition/modernity dichotomy has polarized views, leading to opposing conceptions of culture as either a positive instrument for development or an obstacle to overcome. Then, after considering new ways of thinking about development and culture that have emerged within the context of globalization, we will attempt to analyse the extent to which they have challenged earlier models associated with the tradition/modernity framework. As an illustration of these general considerations, we will emphasize in the second part of the chapter how culture is said to affect economic performance

Culture and development   21 through its ability to create and manage institutions, through the creation of social networks and through its impact on organizations. In fact, as globalization drives rapid changes in the nature of the business and organizational environment, development thinking is urged to provide the insights that can facilitate both an understanding of the role of culture in economic performance and meaningful cross-­cultural comparisons. After tracing the connections that are found between culture and economic performance, we will highlight the persistent difficulty in reconciling what is considered part of tradition and culture and the universalistic assumptions underlying the economic literature’s approach to development issues.

An everlasting controversy: the relationship between culture and development Culture, modernization and development In the 1950s and the 1960s, the role of culture in development received considerable attention within development studies that were dominated by modernization theory. One influential study was Talcott Parsons’ formulation of five sets of pattern variables (1967), which provided a simple binary model distinguishing between modern and traditional societies. The intellectual portrayal of modernization was a political and economic proposition coming to the forefront following the Second World War. It equated the intellectual, cultural and technological advances of the victorious nations as something that needed to be emulated by the ‘poorer, less civilized’ peoples of the world. Samuel Huntington, one of the proponents of modernization theory, pointed out that the concepts of modernity and tradition were central to post-­war modernization theory: ‘These categories were, of course, the latest manifestations of a Great Dichotomy between more primitive and more advanced societies which has been a common feature of Western social thought for the past one hundred years’ (Huntington 1971: 285). The project of ‘modernity’ began with the enlightenment philosophers. By the mid-­nineteenth century, the enlightenment shift from a religious to a secular view of human history had become entrenched in scientific models of human evolution, which fostered a definition of culture as the process of social development. Against a background of European technological and industrial advancement and imperial expansion and aggrandizement, the idea of culture as social development drew on scientific models of human evolution to describe a hierarchy of cultural development across societies and social groups. Bauman (1973: 35) defined culture as ‘a self-­contained system of traits which distinguishes one community from another’. This perspective views culture as a relatively stable, homogenous, internally consistent system of attitudes and values. This understanding of culture assumes that the world consists of separate societies. Each society has its own distinctive culture, which is an integrated totality, radically different from others. It places Western European societies at the pinnacle of cultural achievement and social development, ranking other

22   H. Yousfi s­ ocieties at various ‘stages’ of development down to the lowest level of the ‘primitive’ (Schech and Haggis 2000). The differences between modern and traditional societies were explained in terms of deeply embedded cultural traits. Thus, the traditional traits of Third World societies were thought to dissolve through contact with modernity. The transition process from tradition to modernity was the core theme of nineteenth-­century sociology. The fundamental concepts of sociology formulated by Weber (1922) and Tönnies (1887) invented the analytical distinction between gemeinschaft (community) and gesellschaft (society) as a way of considering different forms of social integration.4 These distinctions have largely been retained. ‘Society’ generally refers to groups held together through anonymous, rule-­bound, more transparent formal contracts and universalistic principles. ‘Community’ is conventionally used to refer to forms of collective life in which people are tied together through tradition, interpersonal contacts and informal relationships. The model of society associated with the modernity project focuses on the autonomy of the individual. Individuals are supposed to defend and maximize their personals interests by being freely involved in contractual relationships and by setting up structures that govern their actions. In contrast, the second model supposes that tradition governs individuals, by ruling their perceptions of the world, their values and their actions. From this perspective, traditional traits were held to be unfavourable to the expansion of the formal, distanced, rule-­ bound, transparent social linkages necessary for achieving a successful market economy and industrial society (Weber 1922). Seen from this second perspective, the difference between developed and so-­ called ‘underdeveloped’ countries is simply that the former have already travelled a historical path that the latter will eventually follow (Rostow 1960). One of the best-­known historical definitions of development is as a succession of stages through which all countries and regions must inevitably pass. ‘Development’ is viewed as a single model defined by the values of ‘Western’ societies, and it is based on the assumption that the political-­economic instruments used to promote economic growth are sufficient for any country to achieve development. This argument tends to assume that culture, viewed as the ‘essence’ of a society, rather than institutions or structural conditions, is responsible for the failure to develop, and that Western cultural values are superior to those of other societies. Consequently, development theorists took for granted that economic development proceeds along a single straight, unambiguous line, from traditional to modern. Modernization theory, which was particularly popular in US social science in the early post-­war period, tended to regard contemporary Western societies as models to which developing countries should aspire. Authors such as W. Arthur Lewis (1954) and David C. McClelland (1964) argued that less-­ developed societies displayed cultural characteristics that constituted obstacles to development. They argued that contact with modern societies would accelerate progress in stagnant, traditional societies. The underlying assumption was that the ‘underdeveloped’ countries had to shed their own traditions and become

Culture and development   23 Westernized. Policy-­makers should promote the modern cultural traits that would enable modernization to take place. A study by McClelland (1964) suggested that stronger achievement motivation could be instilled in individuals both through non-­authoritarian socialization within the family and through Western-­style education. Banfield (1958) focused on questions of cooperation and contrasted cultures that favoured more or less cooperative attitudes. The implication is that modernization must lead to the destruction of the traditions that constitute the essence of developing societies. Along with this framework, economists characterized modern society as consisting of optimistic, self-­interested, calculative individuals, for whom modern culture provides a framework to act freely and respond to market incentives. Thus, the idea of culture as a form of collective thinking, usually seen as fixed and pertaining to a particular group or nation, is presented as a barrier to rational economic calculation. However, many societies resisted the identification of modernity with the West. Economists started encountering poor people in the so-­ called ‘Third World’ who were not interested in greater prosperity. They were surprised that their intended beneficiaries met their offers of economic development with indifference; they met poor people who did not want to have new opportunities and who resisted change. A favourite explanation for economic backwardness used to be an affliction called ‘cultural inertia’. Culture here is the enemy: a voice from the past that inhibits societies from functioning in the modern world. In Huntington’s ‘clash of civilizations’ hypothesis, poverty and low rates of growth are deeply affected by adverse rules and norms that reduce incentives for mobility and investment (Huntington 1996). The challenge for development is then to reform culture by inculcating a more growth- and mobility-­oriented perspective through education or other means of transforming ‘bad cultures’. This perspective was recently exemplified by Francis Fukuyama (1995). In his book, Trust: the Social Virtues and the Creation of Prosperity, he contrasts cultures favourable to trust and cultures favourable to distrust. In the same way, Harrison and Huntington (2000) argue that traditional cultures are unsuited to market-­oriented development and are thus fundamentally hampered in their pursuit of growth. Here again, culture is conceived of as producing mentalities and attitudes that condition individual behaviour in ways that create obstacles to economic growth.5 A variant of this notion, most prominently associated with Max Weber, postulates that it is the content of religious beliefs that is essential to economic development through its impact on behaviour.6 From this perspective, religions could be classified according to their acceptance or rejection of the world: if acceptance, the presence or absence of tension towards the world; if rejection, whether they fostered an orientation of transformation, adaptation or escape from the given world (Eisenstadt 1968). For example, Guizo et al. (2002) characterize Islam as being negatively associated with ‘attitudes that are conducive to growth’, and among adherents to the world’s major religions, Muslims as being the most ‘anti-­market’.7 To summarize, the modernization approach assumes that Third World cultures are a barrier to modernization. In order to experience progress, people in

24   H. Yousfi developing countries were urged to embrace modern culture, which is by definition Western. What is puzzling here is that, on the one hand, many studies tend to relate culture to tradition and argue that the cultural traditions of non-­Western societies must change due to the impact of development, which is conceived in terms of a universal modernity. On the other hand, however, this universal modernity clearly has its cultural roots in the European Enlightenment and therefore easily slips into the concept of the West or Westernization, even if it is not expressly identified in these terms. In addition, we note that the main focus of the modernist perspective is to seek a causal relationship between culture and development – avoiding the difficult task of identifying and analysing the intricate historical and structural interconnections between the many factors that influence development. Some authors argue that economic development brings pervasive cultural change. Others claim that cultural values are an enduring and autonomous influence on society. In both perspectives, culture is seen as a single-­factor explanation of the success or failure of the development process. This leads to the conclusion that we can use a country’s culture when it favours economic development, but ignore or repress it when it is deemed to be an obstacle. This leads to an instrumentalist and essentialist approach to the relationship between culture and development. Culture and the critics of the modernization theory of development The first critiques of the modernization theory of development drew on dependency theory, post-­colonial studies and the deconstructionist school. The main criticism was that modernization theory had seriously neglected factors external to societies, such as colonialism and imperialism, as well as newer forms of economic and political domination. In this section, we will argue that valuable as this work has been in revealing the ideological premises of modernization theory, it leaves little room to think constructively about how culture matters in development. The rise of neo-­Marxist doctrines like dependency theory pointed to the structure of the world economy as the source of underdevelopment. Vigorously rejecting modernization theory, they emphasized the extent to which rich countries exploited poor countries, locking them into positions of powerlessness and structural dependence. They highlighted relations between ‘centre’ and ‘periphery’, arguing that the blocking of development in the latter was the fruit of imperialism in the former. The developed status of these First World countries was, therefore, structurally linked to the underdevelopment of the Third World. This analysis implied an unequal distribution of power between the industrialized centre of the global economy and its underdeveloped periphery (Wallerstein 1976; Amin 1979; Larrain 1989). It follows that the direction and definition of development are seen as objects of political struggle. Culture was conceptually pushed aside as development thinking came to be strongly influenced, if not dominated, by structuralism, Marxism and dependency theory. These analyses focused on economic processes and structure rather

Culture and development   25 than on political, social and cultural processes. Culture is seen as a diversion from the supposedly real processes of ‘domination’ in the contemporary world; or even worse, it is simply a deliberate obfuscation of these processes. In terms of ‘development’, however, the objective remains the same: ‘achieving progress’, even though the interpretation of what is needed to reach this desirable state is different. In fact, regardless of whether it is couched in terms of capitalism or socialism, what is being advocated is the importance of a structural transformation to achieve development. In this regard, Marxist or socialist critiques would not dispute the importance of industrialization and urbanization as manifestations of progress. Another important critique of modernization theory is formulated by post-­ colonial studies. These argue that the dominance of the modernization perspective in development thinking serves colonial domination, rather than merely being a theoretical mistake. Defining culture as ‘a system of control’, post-­ colonial theory revealed the ways in which a politics of representation, formed within European/Western thought in the era of colonialism and empire, continues to inform contemporary Western writing and thinking about the Third World, replicating the stereotypes and power relations of colonialism. For an example, Ferguson (1990), Escobar (1995) and Said (1979) apply techniques of deconstruction, in the tradition of Michel Foucault (1980), to study development as a cultural system, focusing in particular on how economists shaped the modernization perspectives that have dominated development thinking since the 1950s. This dominance is considered to be an aspect of neo-­colonialism, whereby Western ideologies and interests have created a ‘mechanism of control’ that led to the ‘creation’ of the Third World. Economists, and through them the IMF and World Bank, are considered the primary culprits in constructing a development discourse that sustains the distinction between the ‘West’ and the ‘Third Word’, which had emerged during the colonial era. In the words of Homi Bhabha, the strategic function of such narratives is the ‘creation of a space for subject peoples through the production of knowledge in terms of which surveillance is exercised’ (Bhabha 1990: 75). In such a perspective, the idea of culture associated with an ‘ideological system’ can only be indicted for masking the phenomenon of domination behind a false unanimity. Stuart Hall’s 1992 formulation ‘the West and the rest’ is probably the best-­ known articulation of this definition of culture applied to global inequality. Hall’s phrase captures a distinctive way of conceiving of the power imbalance between developing countries and the affluent industrialized societies of Western Europe and North America. He describes this relationship in terms of a pervasive system that represents the non-­West as inferior to the West and underpins the political economy of underdevelopment. The West’s self-­depiction becomes the norm by which ‘the rest’ are identified. In development studies, the West becomes the model to be emulated by Third World societies. This kind of approach to ‘the rest’ is held responsible for creating and extending existing ‘macro’ inequalities between rich and poor countries, and ‘micro’ inequalities

26   H. Yousfi between Westernized and indigenous groups within poorer countries (Mundimbe 1988; Bessis 2001). Demonstrating how development discourse powerfully constructs the ‘Third World’ as the object of development, Escobar, Hall and others argue for a ‘post-­ development’ discourse that would inform a different politics of global inequality – one that builds on and recognizes the diversity and intrinsic worth of the ‘local’ and actively resists the hegemonic Western authority embedded in contemporary notions and the practice of development. However, it is not clear how the development that Third World countries should strive for differs from capitalist development. In fact, by characterizing development as a unitary construct of power/knowledge, this line of thinking fails to explain the alternative models of development that have emerged over the past decades. Correspondingly, how is it possible to conceive of the effect of cultural characteristics on development if culture is associated with an ‘ideological system’ that only expresses the interests of dominant groups?8 To sum up, the critics of modernization have argued that development is not only an amalgam of the process of change, but also a system of knowledge and power that produces and justifies these processes. Furthermore, they have mapped out the extent to which culture is deeply embedded in visions of development and revealed the ideological premises of modernization theory. However, we can argue that modernization’s critics are using the same concept of bounded culture by interpreting development processes not only as the Western imposition of capitalism on the ‘Third World’, but also as cultural imperialism, irrevocably destroying indigenous cultures and identities. Therefore, it appears that little progress has been made in overcoming the dichotomy of tradition/modernity. The binary opposition drawn in development discourse between the global/local and modern/traditional is simply reversed: the local and the traditional become the valued authentic counterweight to a Western modernity seen purely in negative terms. This radical critique of development discourse has the great merit of challenging the superiority of Western values, but it leaves little room for accommodating empirical evidence of the coexistence of ‘modern’ and ‘traditional’ traits in many developing countries, as well as in Western countries. We can conclude that despite constant post-­war efforts to decode the development process, it appears that little progress has been made. Many theories have been proposed (some leading directly to policy), but very few developing countries have succeeded in breaking the bonds of economic backwardness. The ambiguities in the definition of culture and the implicit assumptions about culture and development models led to a cultural blind alley, rather than culturally sensitive development policies and programmes. Globalization and the revival of culture in development thinking The revival of interest in culture and economic development has been a consequence of globalization as well as the observation of the so-­called ‘East Asian

Culture and development   27 Miracle’. East Asia’s economic development has put to the test many of the assumptions underpinning development thinking. This success has challenged beliefs that high rates of economic growth and standards of living could be ‘realisable only within the framework of Occidental civilization’ (Hefner 1998: 1). It questioned the universality of the modernity project. As Rist (1997: 138) points out, the ‘theoretical sequence of modernization is replaced with a multiplicity of new practices that spring forth at the crossroads of history and cultures’. People are reworking their perceptions of other cultures as part of a major shift in relations between nations, regional cultures, and cultural, ethnic and religious groups. The binary models of modern/traditional, or core/periphery, collapse when confronted by the new patterns of interaction generated by globalization. Modernity becomes fluid and multiple rather than a coherent, singular endpoint – whether cast in terms of Westernization or hegemonic capitalism. Thus, globalization as a process and as a model of social change has challenged many of the key concepts of existing development theories, whether of the modernization or dependency schools. We will first seek to trace the way that the globalization process has reshaped the understanding of culture, as well as the building of a new framework for development thinking. Then, we will conclude the section by analysing the extent to which these new ways of thinking have overcome the deterministic view of the relationship between culture and development. Globalization is calling into question conventional ways of viewing culture. As interdependencies become more complex, and simultaneously more diverse between and within nations, as well as between and within organizations, the challenge is to seek the most appropriate way of conceptualizing this new social construct. In fact, the fluidities of global cultural flows undermine the concept of culture as a distinct, discrete and bounded entity – whether conceived of as local, regional, national or global. Instead, ‘culture’ is more accurately conceived as complex and multidirectional cultural interactions and re-­combinations, weaving the local and global together in myriad patterns and configurations. Thus, the most important development in the anthropological understanding of culture is the recognition that the boundaries of cultural systems are leaky, and that traffic and osmosis are the norm, not the exception. This trend of thought underpins some analyses of the cultural dimensions of globalization that have emphasized mixture, heterogeneity, diversity and plurality as critical features of culture in the era of globalization (Robertson 1992). This implies that, for example, national cultures, corporate cultures or professional cultures are seen as symbolic practices that only come into existence in relation to, and in contrast with, other cultural communities. The people whose social interaction makes up these constructs of nation and organization draw on past cultural experience in order to create a new cultural understanding that enables them to make sense of, and live in, a world in which frequent and often wrenching change is commonplace. In other words, people’s constructions of cultural identity and their social organizations of meaning are contextual.9 In this new intellectual context called

28   H. Yousfi ‘post-­modernism’, which pays attention to the fluidity of everything, a perpetual renegotiation of significance is singled out for attention, and the idea of national culture is pronounced ‘dead’. Consequently, it appears that a culture-­based understanding is no longer a static, pre-­existing condition that can be seen as exerting a simple causal influence on action. It is itself a fundamentally constructed phenomenon that arises and is sustained or adjusted through social interaction. More precisely, within an emergent dynamic approach to the conceptualization of culture, culture is seen as being made up of relations, rather than as a stable system of form and substance (Haastrup 1996). This leads to a greater acknowledgement of relational interdependence between social context and the contributions of individuals to social and cultural transformation. Through this conceptualizing of culture, the aim is to balance views that privilege either social determinism or individual autonomy in understanding cultural processes. This suggests that culture is reproduced and transformed not through social determinism but in a constant interaction between the individual’s actions and the social world, as meanings are negotiated and as these meanings change through individuals. As Giddens (1984) points out, the resulting meaning systems are the outcome of individual and patterned social interaction. Such a perspective attempts to reconcile inherited social structures, which at a certain period of time determine the way individuals construct their own perceptions of the social world and react to it, while simultaneously emphasizing the role of individuals who make use of the patterns of thinking and action provided by these structures. According to Giddens, through this implementation individuals guarantee the social reproduction of these structures but, simultaneously, they also create innovations that cause the structures to evolve over time. Along these lines, Mary Douglas (1987, 2004) highlights the role played by the diversity of classifications in social relations, as well as the role of individuals in using this social framework. She suggests that people actively select/ build their identities, as well as their social relations, from available cultural elements: ‘Individuals, when they select from among natural analogies those that they believe in, select at the same time their allies and their enemies as well as the scheme of their future social relations’ (1987: 99). Douglas (2004) quotes Bliss (1993): People are not simply individuals. They live socially and their views, their values, and even their beliefs, as well as their abilities, are formed and sustained within social groupings, families and communities. . . . Perhaps the consideration of life style offers an alternative to the methodological individualism that has been held to be a weakness of orthodox social science. She recognizes that individuals are fundamentally socially embedded beings and are limited in their ability to influence broad trends in behaviour. At the same time, she presents a method by which groups can be categorized and character-

Culture and development   29 ized in order to understand the nature of social interactions that provide for more effective individual agency.10 How does this new approach to culture affect development thinking? A new way of thinking about culture emerged from anthropology. It suggests that culture, as a cement of ‘social organization’, can be harnessed for positive social and economic transformation, particularly through the ways in which power relations and individual agency work within a society. By positioning a group within the social hierarchy, culture provides the means for high status groups to maintain their superior position, whereas for those at the low end, it can limit aspirations, create discrimination and block mobility. Culture is, therefore, fundamentally linked to the perpetuation of inequality. Thus, the new debate in development thinking attempts to respond to the following question: ‘How can interactions between diverse cultures within a society be managed in a manner that allows for more individual agency?’ The challenge is to help the ‘poor’ produce a cultural consensus that best advances their own collective long-­ term interests in matters of wealth, equality and dignity. In fact, the recognition that societies consist of different groups, often structured in hierarchies with unequal social and cultural capital, as conceptualized by Bourdieu (1986), suggests that mechanisms of inter-­group exchange and deliberation need to be set up in a manner that changes the terms of recognition. It implies that development interventions need to be shaped in ways that recognize the relative disempowerment of weaker or subordinate groups in cultural, economic and political terms, especially when they address problems of inequality and empowerment. Enhancing the accumulation of capital in a society and the maximization of material well-­being are no longer the sole preconditions for development. Creating an enabling environment to move towards a culturally equitable form of development through social transformation becomes one of the priorities of the development agenda.11 This approach involves understanding how context matters in ways that are conditioned by such inequalities and the need to design public action that fosters greater ‘equality of agency’ with respect to social hierarchies. Rao and Walton (2004) label this a shift from ‘equality of opportunity’ to ‘equality of agency’.12 As culture is perceived as influencing individual aspirations, as well as the coordination of collective action, the challenge is to redress power inequalities and dominant discriminatory norms in favour of the marginalized. Therefore, it is widely acknowledged that development efforts should no longer focus solely on economic growth and poverty alleviation. Other important goals include democracy, human rights, employment, literacy, health and justice (Rodrik 1999; Sen 1999). These goals are seen as crucial for supporting positive social and economic transformation. In this light, it is worth emphasizing that the linkage between culture and development no longer refers to chance historical processes that emerge over time, but rather to the result of conscious efforts that seek specific social transformations. Along with this type of reasoning, Appadurai (2000) argues that building the ‘capacity to aspire’ in subordinate groups is a direct implication of a cultural

30   H. Yousfi p­ erspective. The capacity to aspire is a forward-­looking cultural capacity that is unequally distributed, with the rich having a greater capacity than the poor. Equalizing the capacity to aspire, and changing the terms of recognition, involves creating an enabling environment to provide the poor with the tools and the voice to navigate their way out of poverty. This may require the development of rituals that help support social agency, such as participatory budgeting, and the identification of key agents that can facilitate the process of connecting the poor to policy-­ makers. In the same vein, other scholars focus on the individual – on understanding people’s ‘true preferences’ in the case of Kuran (2004) and on freeing individuals from the yoke of oppressive elites in the case of Abraham and Platteau (2004) – to ensure ‘equality of agency’. Similarly, Sen (1999, 2000) argues that there is no particular ‘compulsion’ either to preserve disappearing lifestyles or to adopt the newest fashion from abroad, but there is a need for people to be able to take part in these social decisions. Thus, great importance is given to elementary capabilities, such as reading and writing through basic education, being well-­ informed and well-­briefed through a free media, and having realistic chances of participating freely through elections, referenda and the exercise of civil rights. To sum up, enabling development does not require any import of ‘appropriate’ norms and attitudes, as other cultural approaches to development have claimed (Harrison and Huntington 2000). Culture as a cement of social organization calls instead for deliberative political institutions that contain checks and balances, guaranteeing the corporation of all social groups into public decision-­ making. One particular proposal has gained ground: development projects should aim to transform social relations in which lack of trust and cooperation, as well as faltering institutions, bar individuals from leading the lives they deem worthwhile. This, at a minimum, requires full adherence to human rights and democratic procedures (Douglas 2004). However, a question remains: to what extent has this new line of thinking challenged earlier models? As was noted above, the most important shift in thinking on culture and development is the attempt to reconcile the recognition of the existence of social structures with the rehabilitation of the role of individuals in shaping their destiny. This approach to culture consciously downplays the understanding of culture as ‘essence’ in order to focus explicitly on individuals embodying a unique combination of personal, cultural and social experiences. Therefore, if at first sight the tradition/modernity dichotomy is overcome, the emphasis put on the importance of ‘individual agency’ is intimately rooted in the modernity project. Individuals should be ultimately autonomous and free to set up structures that govern their actions. The new emergent framework, by considering culture as an ingredient of social organization, not only allows for social transformation but also seeks implicitly – in accordance with the modernity project – to do away with the supposedly static and negative role associated with tradition. As Sen points out, the ultimate test is the freedom of the citizens to exercise free agency and choose in an informed and participatory way. It is the only way to overcome the conflict between tradition and modernity.

Culture and development   31 The pointer to any real conflict between the preservation of tradition and the advantages of modernity calls for participatory resolution, not for unilateral rejection of modernity in favour of tradition by political rulers, religious authorities or anthropological admirers of the legacy of the past. (Sen 1999) Along these lines, the argument advocating that development strategies should not focus solely on poverty alleviation but also on freeing poor people from traditional hierarchies, informs us on the way modern societies see developing countries, as well as themselves. The hidden assumption remains the same: development should help developing countries get access to modernity. The only difference is that modernity in this new framework is expressed in political, as well as economic, terms. On the other hand, taking into account that the critics of modernization theory radically questioned the superiority of Western values, a question remains: what criteria are to be employed in supporting the ‘empowerment’ of the weak: French ones, British ones or American ones? Similarly, we can ask who has the authority to decide whether a particular social practice is offensive or not, since one person’s offensive practice may be another’s sacred belief. There are always disagreements on such issues, and any disruption of a social equilibrium is likely to result in conflict, hardship and social and economic costs. Besides, if the goal is ‘equality of agency’, who will be able to determine if, and when, this is achieved? The same objective situation can be interpreted by one culture as ‘equal’ and by another as completely ‘alienating’. ‘Equality of agency’ is difficult to achieve, and each culture has its own myths to manage unequal situations. If we have to change the terms of recognition in a given society, what kind of alternatives shall we choose? To conclude, globalization led to a new conceptualization of culture that tends to consider culture as a ‘malleable fact’. From this perspective, any cultural continuity over very long periods of time seems harmful to development because it presupposes an absence of any kind of cultural learning or social transformation. In accordance with the modernity project that prioritizes individual autonomy, any kind of continuity is implicitly compromising the achievement of progress. Thus, although the traditional/modern dichotomy is radically criticized at a national level, it is replicated in the ways development thinking conceives of the challenge of managing diverse cultures and different social hierarchies within a given society. More precisely, this dichotomy is replicated by the prioritization of ‘individual agency’ and ‘social transformation’ over the cultural continuity that supposedly blocks progress. Empirically, however, we find evidence of both massive cultural change and the persistence of distinctive cultural traditions. Even though it is widely acknowledged that culture is malleable, dynamic and adaptable, we still need a better understanding of how the heterogeneity of values or social groups in a given society is articulated with more stable guiding principles and more durable cultural references. Acknowledging and understanding the interdependence between the culture and individual agency cannot be advanced without a clear conception of these relations.

32   H. Yousfi

Culture and economic performance: can we break away from the best practices mainstream discourse? Development thinking is particularly challenged by globalization to seek the most appropriate ways of conceptualizing the role of culture in economic growth. In the following section, we will shed light on some of the ways culture has been taken into account in work that examines the different challenges posed by economic performance. Our interest lies in the nature and form of the connections found between culture and economic performance, and how these illustrate the critical reflections presented above. In particular, we will focus on how culture is said to affect economic performance through the ability to create and manage institutions, through the creation of social networks and through its impact on management. Culture and institutions One of the most important challenges facing development economists is to understand the sources of persistent differences between rich and poor countries. Chronic differences in economic performance have existed throughout history, and they exist across countries today. In the 1990s there was a growing interest in how institutions explain contrasts in economic performance across countries. ‘Governance reform’ has come into prominence in recent years, and the academic literature on institutions and development has exploded. Today, even the World Bank and the IMF, which used to privilege orthodox economic theory tools, have come round to emphasizing the role of ‘institutions’ in economic development. For example, the International Monetary Fund (IMF ) put great emphasis on reforming corporate governance institutions and bankruptcy laws during the 1997 Asian crisis, while two recent annual reports from the World Bank (2002, 2005) focus on institutional development. Nevertheless, development policy has tended to focus on formal institutions, paying much less attention to the informal institutions that shape what North calls the ‘informal constraints’: Institutions are the humanly devised constraints that structure political, economic and social interaction. They consist of both informal constraints (sanctions, taboos, customs, traditions, and codes of conduct), and formal rules (constitutions, laws, property rights). Throughout history, institutions have been devised by human beings to create order and reduce uncertainty in exchange. Together with the standard constraints of economics they define the choice set and therefore determine transaction and production costs and hence the profitability and feasibility of engaging in economic activity. (North 1991: 97) According to North, the sources of contrasting economic performance lie within the institutional structures of society that define incentives for saving,

Culture and development   33 investment, production and trade. Institutions refer not only to the formal private and public sector organizations and rules that influence how agents interact, but also to the relatively stable collective routines, habits or conventions observable in all economies. These kinds of institutions vary greatly among countries and have a significant impact on economic performance. While orthodox economists often seek to give rational, maximizing accounts of the origins of institutions, institutional economists as a group are much more aware of the importance of history, culture, tradition and other so-­called ‘path dependent’ factors in shaping economic behaviour. This section describes the way culture has been integrated into the institutional economics framework to explain economic performance. It also considers the extent to which the conceptualization of culture as part of institutional constraints advances our understanding of the relationship between culture and economic development. Work in institutional economics, associated in particular with Douglass North, has the great merit of integrating the tools of economic, historical and cultural analysis. The analysis operates within the rational choice framework but manages to reconcile anthropological observations with economics (North 1990). North’s special contribution is to map the whole social system and focus on the informal constraints imposed on the choices made by individuals. However, we will argue that North’s interest with institutions and the context within which markets operate still ends up with a dichotomy between ‘modern’ economic relations and the ‘cultures’ that obstruct rational economic activity. North defines culture as ‘communities of common ideologies and a common set of rules that all believe in’ (North 1987: 421). He treats culture as a robustly practical way of life that responds to changes in prices and costs, drawing the economic system along with it (North 1994). Culture mobilizes individuals and turns them into a community with shared norms and values. It does this by setting constraints on behaviour.13 Individuals make choices on the basis of their mental models. These values are acquired through experience that is ‘local’ to a particular environment and which therefore also varies widely across environments. Consequently, there is an immense variation in ‘mental models’ and perceptions of the way the world works (North 1994). This diversity of experiences and learning has produced increasingly differentiated societies and civilizations with very different degrees of success in solving fundamental economic problems and ‘efficient’ institutions. There is increasing recognition in this literature that high-­quality institutions can take a multitude of forms, and that economic convergence need not necessarily entail convergence in institutional forms (North 1994; Freeman 2000; Rodrik 2003; Meisel 2004). There is no ‘one best way’ to secure economic success. As an illustration, Japan, the United States and Europe have managed to generate roughly similar levels of wealth for their citizens, but the prevailing rules of the game are very different in Japanese-­style and US-­style capitalism (Rodrik 2003). Even within Europe, there are large differences in the institutional arrangements (De Rozario 2006). The implication is that transferring the

34   H. Yousfi formal political and economic rules of successful Western market economies to the developing countries is not a sufficient condition for good economic performance (North 1994). Yet, a question remains: if the starting point is that an admixture of rules, norms and enforcement characteristics determines economic performance, how are we to account for the evolution of institutions that create an environment for growth? According to North, the complexity of the institutional environment increased as human beings became increasingly interdependent, and more complex institutional structures were necessary to capture the potential gains from trade. Such evolution required society to develop institutions that allow for anonymous, impersonal exchange across time and space. Thus far, the argument is similar to that of modernization theory. Institutions create the conditions for the impersonal exchanges typical of modern society. For North, the flexibility of Western political and economic institutions have substituted for the traditional role of the family, insured against the new insecurities affecting individuals, and dealt with the externalities, environmental as well as social, that accompany this economic transformation. He postulates that for successful reform, it is essential to change both institutions and belief systems, since it is the mental models of the actors that will shape choices (North 1994). While formal rules can be changed overnight, informal rules change only gradually. Hence, North argues that the evolution of norms of behaviour that will support and legitimize new rules is a lengthy process, and it is by the development of the rule of law and the protection of civil and political freedoms that developing countries can achieve long-­term economic growth. More recently, North et al. (2006) distinguish between modern societies marked by an ‘open social order’ founded on laws and impersonal relations and societies based on a ‘limited-­access social order’ based on interpersonal ties. The key to comprehending modern social development is understanding the transition from limited to open-­access social orders: in other words, the movement towards the formalization of rules. Regarding the issue of the relationship between culture and economic development, the perspective described above calls for two comments: first, these arguments reveal an inherent conflict between culture and development. For North, the coordination problems confronting any poor country can be explained by the fact that the incentives structure works so that community interests outweigh those of the individual (North 1990). The underlying assumption is that communitarian forms of order, represented by the label ‘culture’, often lead to rent-­seeking, non-­transparent behaviour, which is prone to being overwhelmed by principal-­agent problems that frustrate the individual members of groups, while allowing small groups with passionately held views to have undue influence over large, less-­passionate majorities. These properties, economic historians and political economists argue, then lead to sclerosis and lower levels of economic growth than would be possible if market institutions predominated (Olson 1965; North 1981; Moe 1987).

Culture and development   35 So, in order to develop from poverty to wealth an economy needs to obtain the benefits of scale, and in doing so it must weaken culture. Since cultural constraints protect trust and confidence in the future, and since development needs to develop impersonal and anonymous exchange, a sinister trade-­off has appeared. The attempt of the ‘new institutional economics’ to integrate culture within economic analysis arrives at the same old-­fashioned view as the other approaches. It assumes that traditional culture is a dead hand that blocks development, and it ends up reproducing the tradition/modernity dichotomy. It brings us back to the previous view that only two cultures count. One is the entrepreneurial culture – individualist, progressive and modern, almost a non-­culture since it is based strictly on individual rational choice. The other is corporatist and traditional, supposedly an irrational impediment to economic growth (Douglas 2004). This displays once again an unspoken scepticism about any such putative social glue. North’s interest in the context within which markets operate still ends up with a dichotomy between modern economic relations and the ‘cultures’ that obstruct rational economic activity. Second, we agree with Rodrik (2003) and Chang (2003) that this perspective has led to an overemphasis on the role of institutions, such as property rights and the rule of law in enforcing the roles of competition and exit, and as a consequence has confined the role of culture to being a constraint on development. Institutions merely reinforce the market and economic rationality. Similarly, in the world of development agencies, it is argued that because many developing and transition economies lack a clearly defined and secure private property rights system, ‘good’ policies based on ‘correct’ theories recommended by development economists have failed to work. In other words, the institutional argument is being mobilized to suggest that the development policies and theories were never wrong. They failed to work only because the countries implementing them did not have the right institutions for the ‘right’ policies to work. In this regard, d’Iribarne (2003) emphasizes that the importance given to the ‘enforcement of property rights’ as a critical ingredient for building good institutions able to achieve economic development is far from reflecting a ‘universal’ criterion. He shows that this framework is rooted in the specific US conception of a ‘well-­ordered society’, which links the idea of freedom to property.14 He suggests that there are different ways to resolve collective action problems. Each society develops its own ways of collective problem-­solving that both build its governing institutions and affect individual incentives. In spite of increasing recognition in the economic literature that there are different high-­quality institutional models, this literature does not tell us clearly whether diversity merely reflects efficient solutions to different problems derived from different starting points, or whether there are many equally efficient institutions that can be applied to solve the same problem. In addition, the empirical research on national institutions has generally focused only on the protection of property rights and the rule of law, narrowing the spectrum of the comparison and hindering an explanation of what is happening in the real world (Rodrik 2003; Storper 2005).

36   H. Yousfi Correspondingly, Berckowitz et al. (2003) find that countries that developed their formal legal orders internally, adapted imported codes to local conditions or had familiarity with foreign codes, ended up with much better legal systems than those that simply transplanted formal legal orders from abroad. Thus, there is no need to replicate the Western-­style institutional system from scratch; it may be possible to work with such institutions as are available and build on them.15 In this section, we have reviewed some of the key theoretical issues involved in developing a good understanding of the relationship between culture, institutions and economic development. The failure to overcome the dichotomy of tradition/modernity, the excessive focus on property-­rights institutions and the failure to build a sophisticated theory of institutional change have been pointed out as the major problems in the currently dominant literature on institutions and economic development. We can break away from the so-­called best practices prevalent in the mainstream discourse only if we understand the complexity of the relationship between individual agency and cultural continuity in institutional change. Only when we accept the existence of different cultural/institutional ‘traditions’ across societies can we begin to understand that what people believe and do matters in a real sense. An approach that is less focused on the superiority of any particular institutional model and more cognizant of the context-­ specificity of desirable institutional arrangements is needed. Instead of focusing only on ‘big institutions’, work towards gradual improvements in institutions or small institutional practices may help to show exactly how positive institutional arrangements are promoted or blocked. Such an approach may help to open up a new way of looking at institutional reforms. Culture and social networks There is today a debate as to whether large-­scale, rational, bureaucratic principles, along with the individualization, ephemerality and mobility they seem to call forth, have gone too far, weakening forms of community necessary for social order (Etzioni 1996; Putnam 2000). This type of reasoning has also become centrally involved with questions of economic development. It seeks to integrate the potentially positive effects of community with the modern economy. It refers to communities, and hence cultures, as means for improving the functioning of labour markets, generating entrepreneurship and organizing the provision of the public goods that alleviate both private and state burdens in the creation of prosperity and social integration. Similarly, success in small-­firm based industrial clusters or districts, ranging from the most famous case of Italy to examples drawn from Taiwan, Brazil and China, are said to depend critically on the existence of communities that regulate complex inter-­firm and firm–worker relationships through shared norms, reputation effects and mutually aligned expectations. In other words, some constraints on behaviour and some common values can greatly reduce transaction costs and make business more efficient. In the Asian context, certain social networks that span countries are said to have played an important role in facilitating trade and deep transnational integra-

Culture and development   37 tion. In Europe, Dei Ottati (1994) defines the ‘social environment of the ideal-­ type industrial district’ in terms of a common culture, frequent face-­to-face relations and ‘norms of reciprocity accompanied by relevant social sanctions’. Values and social relations in the community are considered functional aspects of the economic performance of local enterprises. Social networks propagate information on job opportunities, relative prices and a host of other economic matters. In the 1990s, the World Bank experimented with so-­called ‘micro-­ lending’ projects that seek to use social networks to extend very small retail loans to poor customers in Africa and other regions. The success of micro-­ lending depends on adequate information about creditworthiness that is best captured through informal rather than formal information channels (Fukuyama 2001). In the same vein, the organizational and management literature has explained the success of inter-­firm relationships in terms of the position of the cooperating partners in a network of relationships (Nohria and Eccles 1992). For example, the economic success of the automobile and the electronics industry in Japan has been explained by the establishment of close and long-­term oriented external relationships (Sako 1992). Inter-­firm networks of actors reduce the necessity of contractual commitments through the emergence of trust. As Granovetter (1985) has argued, the ability of social networks to enhance economic development is strongly related to the building of trust. Actors trust each other because of their common cultural background, shared values and strong reputation effects. These stem from dense interpersonal networks. This notion can also be found in many other empirical studies of regional economic development (Becattini and Sforzi 2002). This literature draws attention to the communities in which enterprises are embedded and to the sociocultural ties that facilitate trust. It challenges the rational choice assumption underlying utilitarian analysis and moves to centre stage the issue of trust and culture in economic development. The question is, to what extent does culture influence the functioning of social networks and their ability to produce trustworthy relationships? The next section outlines the ambiguity concerning the way culture is said to affect the emergence of trust as an output of social networks and an important ingredient in economic performance. Trust as a result of shared values The concept of trust has traditionally been taken to signify and represent a coordinating mechanism based on shared moral values and norms supporting collective cooperation. Casson (1997: 118) defines trust as ‘a warranted belief ’ that someone else will honour his/her obligations, not merely because of material incentives but also out of moral commitment. From this perspective, trust is offered to individuals and/or to organizations through the assumption that they will conform to standards of behaviour that characterize the community: honour, devotion, solidarity, etc. For example, Ouchi (1981) suggests that clan membership is a basis for trust. It determines the members’ behaviour. Fukuyama (1995)

38   H. Yousfi views trust as the expectation of regular, honest and cooperative behaviour based on commonly shared norms and values. Trust is associated with the capacity to cooperate in a spontaneous way on the basis of shared values, rather than on the basis of formal rules. Zucker (1986) defines such a set of shared, ‘taken for granted’ expectations as part of a ‘world known in common’ among certain members of society. Consequently, national differences in values, social context and institutions will have a significant impact on levels of relational quality. Many researchers have mentioned the high trust widely acknowledged to be prevalent in Japanese business society as compared to low trust in the United States (Ouchi 1981). Even though trust is built in different ways, its establishment relies upon the societal norms and values that guide people’s behaviour and beliefs (Hofstede 1980). Since each culture’s ‘collective programming’ results in different norms and values, the processes trustors resort to in order to decide whether, and whom, to trust may be heavily dependent upon a society’s culture.16 This conception of trust is based on a deterministic view of culture understood as those customary beliefs and values that ethnic, religious and social groups transmit fairly unchanged from generation to generation. This view of culture leads to an ambivalence in the understanding of the role of social networks in economic performance. Some scholars suggest that social networks are one of the ways that market failures can be efficiently overcome in developing countries, as well as in modern economies, and they come to the conclusion that networked entrepreneurialism is often superior to bureaucracies (Piore and Sabel 1984; Stiglitz 1994). At the same time, one frequently hears criticism of ‘crony capitalism’ in Asia. Family based production networks – certainly a form of cronyism – work well in Taiwan and are often cited as one aspect of the ‘good communitarian’ structures found in the Third Italy region, but they are deplored when they become clannish, as in the Mezzogiorno (Gambetta 1988; Leonardi 1995). Failure stories in the economic development literature share this ambivalence. For some scholars, social networks can hinder economic development. According to Platteau (2007), strong ethnicity has prevented the modern state from becoming established in Africa during the post-­independence period, thereby creating a perverted path of non-­development. Similarly, Knorringa (1996) argues that in the footwear cluster of Agra (India) the economic divisions between producers and traders were heightened by the distinct social castes to which they belonged. While the producers were largely backward-­caste Hindus and poor Muslims, the traders were higher-­caste Hindus (banias) and rich Muslims. Knorringa suggests that the antagonistic exchange relationship between producers and traders was reinforced by the distrust and social contempt that the two castes had for each other, thereby weakening the prospects for cooperation.

Culture and development   39 Trust as a social construction Most economic theory, whether in mainstream or Marxist variants, regards social ties as obstacles to the building of effective economic relations. Along with this type of reasoning, some authors strain to find the perfectly rational character of participation in networks and governance, even if they acknowledge the role of social networks in economic performance. They deny that these networks could be, at least in part, dependent on group membership or that the participants could be socially ‘embedded’ in ways important to the functioning of these groups. These authors view the development of trust as basically a calculative process (e.g. Buckley and Casson 1988). Individuals are described as opportunistic and seeking to maximize self-­interest. According to these behavioural assumptions, trust is established through a calculative process, whereby one party calculates the costs and/or rewards of another party cheating or cooperating in a relationship. For Dasgupta (1988), ‘Trusting another implicitly means that the probability that he will perform an action that it is beneficial or at least not detrimental to us is high enough for us to consider engaging in some form of coordination with him’. From this perspective, there is no need to make assumptions about a partner’s values and beliefs to understand how he/she behaves. The assumptions of rationality and opportunism are sufficient to explain the essence of this relationship – ‘trust’ is given when the structure of returns to the ‘other’ make defection unattractive. When the structure of returns to both sides to transaction favours cooperation, there is ‘trust’. Williamson (1993) employs a calculative conceptualization of trust. He recognizes that ‘socialization and social approvals and sanctions are also pertinent’, but he goes beyond that calculative process of trust only to the extent that he sets the probabilistic calculus in an undetermined social context. The social norms of trusting behaviour are reduced to a sum of agents who all interact on a purely calculative basis. Social norms are useful, but they are best substituted by economic rationality. In fact, trust as it is conceived of in social science, i.e. the conformity to social standards of behaviour, is difficult to reconcile with a vision of modern society as consisting of self-­interested, calculative individuals. Trust can enhance economic performance only if it is built on a community of ‘interests’, which is distinct from a community of values based on a spontaneous adherence to the same group. Once again, it appears difficult to reconcile between ‘social norms’ implicitly considered as part of ‘tradition’ and the pure coordination of ‘individual interests’ underlying the way that the literature considers the economic development issue. However, another line of thinking suggests a third route based upon trust construction through interaction (Mollering and Stache 2007). Trust is achieved through communication events involving all actors in the negotiation of shared meanings and the development of rituals and practices that establish shared values, norms and beliefs (Das and Teng 1998). Under this alternative view, belonging to the same community is not a prerequisite for the emergence of

40   H. Yousfi trust. Such an approach tries to accommodate the influence of social norms in the building of trust but maintains actors’ autonomy in the shaping of trusting relationships. Actors are freely involved in building trustworthy relationships and jointly set up norms to govern their actions through an interaction process. Firms learn to trust each other over time. They experience the fact that a partner does not take advantage of dependencies or constructively solves small conflicts. As Humphrey and Schmitz argue: This is the distinctive feature of relationships based on trust: the risks taken expose the agent to possible losses which are greater than the advantage being sought. In Luhmann’s words, ‘trust is only possible in a situation where the possible damage is greater than the advantage you seek’. (Luhmann 1988: 98). The reason for accepting this extra risk is the grounded belief that the partner will not behave opportunistically. (Humphrey and Schmitz 1998) This past-­oriented aspect of trust, then allows for a future-­oriented one. A common future creates trust through the overriding consideration that partners are going to meet again (Luhmann 1979). Firms expecting to interact again in the future will place initial trust in the partner, because they assume that the other will not abuse their trust and jeopardize future interactions. These processes are well illustrated by Menkhoff ’s discussion of the development of trust-­based relations among Chinese traders in Singapore (1992). To start a business in Singapore involves establishing credibility. Menkhoff describes how new traders have to slowly build up a track record of trustworthiness through repeat transactions. Nevertheless, such strategies of trust construction have their limitations. Fafchamps (1996) argues that trust, or working with known associates, may be an effective way of operating in a difficult environment, but it is a second-­best solution, narrowing the range of transactions and limiting economic opportunities. Using case studies of manufacturing and trading firms, he shows that compliance with contractual obligations is mostly motivated by the desire to preserve personalized relationships based on mutual trust. However, the absence of reputation mechanisms limits the extent of network construction. Recent research on industrial clusters in developing countries confirms that social ties are an important basis for trust and sanctions but suggests that their importance diminishes as clusters grow (Humphrey and Schmitz 1998; Schmitz 1999). Schmitz (1999) shows that sociocultural bonds facilitate trust, but their influence lessens over time as they are eroded by increasing differentiation within the community and the key role of outsiders. The new ties are based on conscious investment in inter-­firm relationships, and extended trust relies increasingly on economic and technical performance irrespective of social identity. More than that, the ability of clusters to make a shift from a characteristic-­ based trust to a process-­based trust is seen as critical to their ability to compete in a global market.17 Yet, the question of whether culture is a factor in the ability to make this shift remains open.

Culture and development   41 In one of the most explicit efforts along these lines, Fukuyama (1995) argues that low-­trust, highly communitarian societies are less likely to generate successful large enterprises than are high-­trust societies, and low-­trust societies typically have lower long-­term rates of growth than do high-­trust ones. Intriguingly, Fukuyama holds that capacities for direct, spontaneous or informal association of persons facilitate the establishment of large-­scale, transparent and bureaucratic forms of economic life, such as the large corporation. This alternative approach to trust tries to accommodate the influence of ‘social norms’ by suggesting that, rather than the two forms of coordination being mutually incompatible, one is a precondition for the other. The underlying assumption is that economic performance, in conformity with the modernization project, requires less emotional trust in known and familiar persons but more system-­trust predicated on the ways in which modern institutions present themselves.18 This displays the implicit persistence of the utilitarian rational choice assumption as a prerequisite in achieving economic progress. As an illustration, Luhmann (1988) advocates that the conversion towards system-­trust is part of the great ‘civilizing’ processes towards greater social differentiation and societal complexity. This section has highlighted the increasing recognition of the importance of culture as a set of ‘social norms’ that can be used to enhance economic performance, and it describes the different ways in which social networks are said to affect economic development. At the same time, this section emphasizes the difficulty of grasping the articulation between the role of social relations (suspected of bringing back ‘tradition’) and individual autonomy (as a condition of a ‘modern society’) in identifying the role of social networks in economic performance. Thus, the dichotomy of tradition/modernity leads to the perception of the two forms of coordination as juxtaposed and, at best, as substitutes for each other. As a consequence, it becomes difficult to understand the critical association needed to enhance economic performance, as well as to explain the differences seen among economic institutions across countries. Culture and management The work of Harbison and Myers (1959) and Farmer and Richman (1965) stimulated a sizeable body of research centred on the relationship between management and economic development, and comparisons between systems of management (Nath 1986). In these early studies, culture per se was not a research issue. This work was strongly multidisciplinary in orientation, especially that of Farmer and Richman (1965), who included educational/cultural, sociological/cultural, political/legal and economic variables in their model. The research was premised on the convergence hypothesis of socioeconomic development.19 It predicted the ‘convergence of cultures, as well as applicable management principles and practices, throughout the industrial world’ even if ‘this type of universal convergence is likely to take decades, generations and even centuries in some extreme cases’ (Farmer and Richman 1965: 394). In other words, it echoed the approach of modernization theory.

42   H. Yousfi These authors emphasized that the application of scientific thinking to industrial technology would elicit predictable consequences everywhere: a pattern of rational administration to maximize the productivity of machines and personnel; a labour force committed to an industrial way of life; a division of labour based on technical and managerial skills that inevitably arises within functionally related groups; and the development of a rational worldview among people exposed to applied scientific knowledge. The patterns first appear in economic institutions and later spread to related institutions. The application of management principles would not only enhance economic growth but also would help to foster democracy. While traditional cultures may temporarily impede these changes, eventually tradition must yield to the logic of industry. At the same time, academic management research was primarily a Western and, to a large degree, a US enterprise (Boyacigiller and Adler 1991). It was a time of great enthusiasm for the US management model. These factors led to an implicit universalism in much of organization science. The unspoken assumption was that the US model was not the product of a particular culture, but rather the very expression of a universal rationality applicable across cultural boundaries. However, globalization and the relatively rapid growth of transitional economies has, in general, created a burgeoning interest in understanding the management practices of various countries in their own context and within their own frame of reference. As the locus of industrial energy shifted from the United States towards Japan and the newly developing Asian countries, attention focused on the issue of competitiveness. In the United States, as in many other countries, this meant a preoccupation with Japan. The assumption that a unique Japanese culture was contributing to the success of Japanese industrial organizations underlay this preoccupation (Drucker 1971; Ouchi 1981; Pascale and Athos 1981). This trend prompted unprecedented interest among management scholars and practitioners in the influence of national cultures on management practices and the consequences for economic performance. The central question addressed in this section is the following: to what extent does the management literature, in particular that devoted to cross-­national comparative research, challenge the one ‘best practice’ based approach to understanding the relationship between culture, management and economic development? Culture as a set of values and attitudes Hofstede’s seminal book, Culture’s Consequences (1980), with its focus on a readily accessible set of universal dimensions from which measures of culture can be derived, was the precursor to cross-­national comparative research that seeks to understand the link between cultural values and managerial attitudes and behaviours.20 Defining culture as ‘the collective programming of the mind which distinguishes members of one human group from another’ (1980), Hofstede initially found four ‘universal categories of culture’ around which programming occurs. The now well-­known dimensions are individualism-­collectivism, power-­distance, uncertainty avoidance, and femininity-­masculinity. The dimen-

Culture and development   43 sion of Confucian dynamism was added later, developed in a subsequent study with Michael Bond and his Chinese colleagues (Chinese Culture Connection 1987). These universal dimensions of culture ‘describe basic problems of humanity with which every society has to cope; and the variation of country scores along these dimensions shows that different societies do cope with these problems in different ways’ (Hofstede 1980: 313). A vast literature on a wide array of topics based on these dimensions has evolved (Clegg and Redding 1990; Hampden-­Turner and Trompenaars 1993; Whitley 1999). This framework has facilitated the relatively straightforward inclusion of cultural variables into cross-­national research in other management-­ related fields. For example, a dimensions-­based framework informs research in economics with regard to economic growth. Franke et al. (1991) showed that cultural values, measured from Western and Eastern perspectives, are factors in economic performance that explain more than half the cross-­national variance in economic growth over two periods for samples of 18 and 20 nations. Performance seems facilitated by ‘Confucian dynamism’ – stressing thrift, perseverance and hierarchical relatedness, but not traditions impeding innovation. Cultural ‘individualism’ seems a liability, while the propensity for work in cohesive groups is an asset for economic performance. This literature has the great merit of showing that all societies have ‘culture’ and of challenging the one, best-­practice based approach. Nevertheless, the conceptual framework presented in Culture’s Consequences has not escaped criticism. Two critical factors continue to afflict the majority of research from this cross-­national comparative perspective: the use of the nation state as a surrogate for culture, and the assumption that national culture expresses itself as a single, given and immutable characteristic expressed in individuals. This view of culture tends to entail blindness in regard to social variations, diversity and power relations within a nation or an organization, or between nations and organizations. Furthermore, many local case studies within different countries contradict the coherence of cultural dimensions (d’Iribarne 1997; Tayeb 2001). Other critics within the context of globalization move away from the assumption that culture and nation-­state are synonymous (Lenartowicz and Roth 2001). They advocate that the changing nature of the business and organizational environment makes different demands on the type of international cross-­cultural management research that is undertaken. Thus, the study of cultural complexity in the management field becomes contextually sensitive, with qualitative case studies focusing on organizational actors’ interpretations, identity constructions and sense-­making processes (Weick 1995; Soderberg and Holden 2002). As we noted above, national specificities are considered in this new emergent intellectual framework as implicitly traditional, and they are seen simply as hangovers from earlier periods. The question here is: to what extent does this literature have an impact on development thinking? Scholars who focus on management in developing countries are divided into two camps, which can be characterized as the organizational theorists and the

44   H. Yousfi culturalists. The former hold that the theoretical principles underlying and explaining organizational behaviour are universal. The culturalists argue that management practices in developing countries are rooted in local cultural values. However, we can note that when it comes to formulating recommendations for modernizing the functioning of firms in these countries, the gap between these two camps is not as large as it might at first appear (Leonard 1987). Both continue to advocate the best practice, or ‘one best way’, approach. Organizational scholars (Kiggundu 1989; Austin 1990; Jaeger and Kanungo 1990; Womack et al. 1990) continue to take a one best-­practice based approach, and even if some of them acknowledge the importance of culture, they do not consider it of prime interest. They argue that only Western management techniques may be able to improve performance in developing countries.21 The ‘culturalists’ see the character of organizations within these countries as rooted in their larger political and social structures and come very close to saying that the organizations are unreformable. From this perspective, culture is an obstacle to modernization, and it is seen as self-­evident that the persistent ‘traditional’ aspects of life in such countries could only be obstacles to efficient production; it seemed obvious that such efficiency could be based only on the ‘rational’ aspects of the organization. However, many examples show that adopting the ‘best-­management practices’ is not sufficient for improving economic performance in developing countries. The management tools proposed by experts do not have the desired effects; they are often abandoned after the experts leave. Furthermore, when some companies record substantial technical and financial successes,22 the majority of observers, whose concern is seeing the developing countries become part of globalization, see in these cases only the results of universal methods. Few ask themselves about the concrete aspects that explain these singular successes and their effective achievement. This makes it impossible to distinguish what in so-­ called universal management methods is actually universal from what in fact reflects the unique features of the particular context in which they came into being. Culture as a framework of meaning Philippe d’Iribarne (1989) has directly challenged the one best-­practice approach by showing that there are different management models within the so-­called ‘developed countries’, which are largely an expression of different conceptions of interpersonal and social relations with all their political and moral dimensions. Using an ethnographic approach, he showed that when the life of a US factory is observed closely, the organizational model that is a reference point throughout the world can be seen to be a local product. It reflects a political ideal that corresponds to the notion of a society based on contracts freely entered into by equals (de Tocqueville 1835). This contractual ideal is associated with a great mistrust of the arbitrariness of power and a great faith in recourse to objectivity as a means of avoiding this arbitrariness. One finds its traces in the efforts made to

Culture and development   45 organize working relations in the form of contracts that define as precisely as possible the rights and obligations of the parties, whether it be for the relations between a superior and his subordinate, the relations between supplier and client departments or, in unionized companies, the relations between corporation and union (Foner 1998). Similarly, d’Iribarne explains that one finds in France another vision of the right way to live and work together. When the French speak of their work, the omnipresent reference point is the rights and duties associated to a specific position one holds in society, and to the rank associated to this position. When speaking of hierarchical relationships, as well as relationships with customers, the French refer to what seems normal to do in conformity with the customs of the ‘métier’ (profession) that one holds. Without any reference to instructions from superiors or to a ‘contract’, these traditions define how to recognize a ‘good’ way of working, what is ‘normal’ to do, and what an individual cannot stoop to if one is a production engineer, a lathe operator, accountant or in some other calling. This can usefully be regarded as reflecting a society whose functioning is governed by a conception of freedom, quite different from the English or German conceptions, attached to rights specific to a given social position (Montesquieu 1748; de Tocqueville 1856). Such a conception is already evident in medieval France (Bloch 1968). From this perspective, culture is not defined as a set of values and attitudes shared by a group. National cultures do influence economic behaviour, but they do so by serving as a frame of reference that provides a framework of meaning (d’Iribarne and Henry 2007). Culture is essentially seen as a reference system that enables actors to make sense of their own actions and of the world in which they live (Geertz 1973). All cultures provide references to denote, classify, identify, evaluate, connect and order. They establish criteria for distinguishing good from evil – the legitimate from the illegitimate. If the existence of such oppositions seems universal, the form they take differs considerably across societies. In every society, a specific network (constellation) of real or mythical figures and narratives (real or in the form of fables) highlights the principles of classification through which society can be seen to be made up of separate groups. Words are associated to these classifications such as impurity in India, witchcraft in Cameroon, the loss of freedom in the United States, dishonour in Algeria. They provide interpretative systems that give meaning to the problems of existence, presenting them as elements in a given configuration that shapes the relationship between individual autonomy and collective order. In every country that has preserved its unity beyond the vicissitudes of history, one finds largely convergent conceptions of what a well-­ordered society is. These conceptions do not merely influence the political institutions that govern the society as a whole. They even leave their imprint on the functioning of each specific organization. They underlie the way members of organizations, belonging to a specific society, give meaning to what they live daily: the hierarchical functioning of the organization, with its procedures for decentralization, control and evaluation of performances; cooperation among the different departments,

46   H. Yousfi decision-­making and the management of conflicts; the organization of relationship with customers, the quality procedures, the setting up of codes of conduct, etc. They provide references according to which actors interpret situations and events, the taken-­for-granted assumptions that form the basis for their judgement, and the categories they employ to describe their daily reality. Thus, words such as ‘partnership’, ‘cooperation’ or ‘contract’ may appear to be evident and clear but may, in fact, have different meanings from one culture to another. The meanings given to such words influence people’s behaviour and expectations (Yousfi 2006, 2010). In this sense, culture is a stable framework that shapes the way actors or social groups within a given society negotiate their social interaction. One can find consensus or contradictions between actors within the same society, but the negotiation game is framed and legitimated by the same guidelines. Within the so-­called ‘developed world’, one can find evidence of differences in management practices linked to different conceptions of the relations between the individual and society (d’Iribarne et al. 1998; de Bony 2003; de Rozario 2006). These conceptions make the tradition/modernity dichotomy totally obsolete and lead within firms to different ways of making sense of a difficulty, collaboration, a sanction or a failure. Such characteristics are difficult to replicate elsewhere because of their embeddedness in a national culture. However, unlike industrialized countries, the developing countries have not developed indigenous management traditions. Therefore, the challenge is how to balance general organizational theory and the local context in deciding how best to improve management practices. In this regard, d’Iribarne and Henry (2007) show that companies – whether subsidiaries of multinational groups or local companies – that succeeded in implementing imported management tools have sought to benefit from local cultures rather than suppress them. One of the most important success factors of these firms is their ability to develop a high quality of cooperation that is in contrast to common managerial practices in developing countries. They implemented working methods that respected locally accepted behaviours, allowing people to feel comfortable in their work. An illustration of the success of imported management techniques in developing countries is provided by the example of the Cameroon Electric Company (CEC) (d’Iribarne and Henry 2007).23 At the time of the first diagnosis, CEC, like many sub-­Saharan African companies, was suffering from excessive centralization. Highly personalized relationships were pervasive in the company, and everybody tended to constantly look at each other and to care much more about relative than absolute position. Many accounts of everyday life in the CEC thus stressed the constant harassment of successful individuals by envious and jealous members of the kinship group or the community. Behind such a resistance, also encountered in other sub-­Saharan countries, was the fear that assertion of successful individuals would introduce a competitive spirit within the community that would sooner or later undermine its social cohesiveness (Platteau 2007). Consequently, the CEC employees who were involved in a social network outside the enterprise were very used to pursuing informal and personal goals rather than the collective ones that were formally proclaimed.

Culture and development   47 The repeated intervention of experts had changed nothing, reinforcing the ideas held by some about the inability of ‘African cultures’ to adapt to business practices. However, the decision by the director to have all company procedures drawn up in a very detailed way, which was a priori surprising, with respect to the difficulties that needed to be resolved, did produce the long-­awaited improvements. At first sight, one could argue that the introduction of these procedures brings some objectivity into relationships and helps employees escape social pressures. From this perspective, the success of this reform could only confirm the ‘universal’ principle of bureaucratic impersonality as the best way to combat favouritism (Crozier 1964). To explain exactly what happened, d’Iribarne and Henry (2007) highlighted that beyond their universal appearance, the formalization of the procedures led to a greater accountability of the Cameroonian members of the company because the procedures have been reinterpreted positively according to the local frame of meaning. He showed that the relationships between employees within the Company are described in terms of individual avidity and hidden loyalties, or positively in terms of daily ‘mutual aid’ between ‘friends’. ‘Malice’ and ‘jealousy’, as well as ‘nastiness’, are in fact the darker side of ‘mutual aid’ and ‘kindness’. These terms, often used in the account of the Cameroonian interviewees, form oppositions that give meaning to social relationships. Employees at the CEC used this set of interlinked categories to assess and interpret their work relationships. Correspondingly, the detailed written procedures were successfully implemented because they focused not only on elementary tasks but also on how to behave. They implicitly responded to the anxieties related to these rationales: bad faith, malice, hidden envy, sentimentalism. In fact, by assiduously following the instructions, the Cameroonian employees could escape the pressure from friends and avoid making subjective concessions. At the same time, they could provide visible signs of their attachment to standards of ‘good behaviour’ as proof of their ‘good intentions’.24 Moreover, the respect of the detailed procedures, along with their control, did recreate within the company a propensity for ritual that is fairly characteristic of African societies. The sense of ritual leaves room for the expression of personal feelings while limiting their excess. Far from making a trade-­off between ‘social ties’ and ‘rational organization’, the implementation of the procedures at the CEC succeeded because it echoed the way in which Cameroonian employees interpreted social ties. One can find the same procedures everywhere, but the way people give meaning to them may vary from one country to another, leading to different framings of their roles.25 This case challenges those for whom it is self-­evident that ‘traditional aspects’ of culture can be nothing more than obstacles to ‘efficient’ production. It shows that culture is not a backdrop on which universal tools can be placed. It is the ‘terrain’, ‘the grammar’ that should be used for constructing and/or implementing them. Similarly, Yousfi et al.’s (2005) study of Poulina, a successful Tunisian company,26 showed that although it used the US management model, its management practices reveal that local adaptations of these techniques had been

48   H. Yousfi i­ntroduced over the years. These adaptations are, interestingly, not in contrast to local cultural practices, but rather built on them. Indeed, it appears that the essential aspects of the imported management tools (e.g. formalization, regulation, objectives and performance) are reinterpreted within the local cultural context and consequently well accepted. Management principles are not followed as rules, even if they constitute strict procedures. An executive officer explained: There are two things in the procedure, explains an employee. A procedure is a law. First, there is the technical aspect of the law. A law must be respected but one should consider the procedure not only in its disciplinary sense. A procedure has also an educational aspect. We control with a procedure to help employees, to accompany them and to free them from the ‘arbitrary’ power of management. The procedures were perceived as strong guiding principles for managers’ actions, which allow flexibility and enable employees to reconcile them with the Tunisian relational context. An employee commented: ‘Poulina is an organic group which functions like a great family with written rules.’ The management techniques are reinterpreted in connection with cultural metaphors present in the Tunisian culture (e.g. the craftsman as an independent entrepreneur, or the code of honour). By using the metaphor of the craftsman, employees respond to strict regulations, but in autonomous ways. Though they are in a relationship of subordination, they preserve their ‘dignity’ and independence via the strategies they adopt to respond to the strict rules. An executive officer explained: To work for the others is considered degrading in Tunisia.27 As soon as a person starts to succeed in his work, he is pushed by his circle to work for his own account. The term ‘zoufri’, from the word ‘ouvrier’28 means thug in Tunisian dialect . . .. That’s why I think the craftsman is happier than a worker. The procedures help people to preserve their honour: they are working with autonomy like craftsmen, they are engaged, responsible, they do not have a ‘boss’ who controls them and who dictates to them what they have to do. At the same time, they can lose their ‘honour’ if they do not achieve the written objectives. The ways employees talk about the management practices of the company indicate that they make sense of them using the local Tunisian culture, not the meanings usually associated with these management practices in, for instance, Western Europe or the United States. The craftsman example or the metaphor of the family were mobilized by the employees as recognizable social organization models existing in the Tunisian context to make sense of the imported tools.29 The main lesson derived from these examples is that culture can therefore be seen as a resource for providing local and legitimate meanings to management

Culture and development   49 techniques, as long as they can be reinterpreted in conformity with the local framework of meaning. This section has described how culture is said to affect management and consequently economic growth. Most of the contemporary streams of research involved with improving the functioning of organizations in developing countries tend to stress the need for the transfer of the best-­management practices. They look at culture as ‘something to which behaviours can be causally attributed’, basing their investigations on the determination of a scale of attitudes rather than on ethnographic studies (Hofstede 1980). We have considered the limits of both visions. They are based on the assumption that replicating Western management, associated to the idea of modernity, in developing countries is the best means for enhancing firms’ competitiveness. These assumptions are not universally valid in any society, but their applicability is even more limited in Africa than in the West. Furthermore, we have shown how the conceptualization of culture provided by d’Iribarne offers a fruitful way of understanding how general organizational theory can best engage with the local context to improve management practices in developing countries. Once culture is perceived as a producer of meaning, the temptation to attribute mechanical effects to it, or to disregard its influence on development actions, disappears. The goal is not to know how those from a certain culture are supposed to act, whatever the circumstances, in conformance with the attitudes with which they have been inculcated. One the contrary, the focus shifts to the way that attitudes and behaviours change according to circumstances and to the meanings granted to events and situations. It is mostly by giving a specific meaning to management techniques and also to development policies that culture comes into play. It is in such a framework that one goes progressively from what is purely a matter of structure to what is purely a matter of agency. Yet, what are the implications of this discussion on culture and management for the relationship between culture and development in general? Changing the perspective on culture and development In this literature review, we have put the accent on the widely held view in development thinking that culture is associated with attitudes and mentality. This conception of culture has historically polarized views: culture is seen either as a positive instrument, or as an obstacle to development. In reaction to these ‘culturalist’ approaches to development, an opposite current of thought has sharply rejected any kind of cultural explanation, whether in regard to business performance or economic development.30 According to this current, since businesses or economies perform unevenly in the same cultural environment, culture is not a factor: there are simply good and bad universal ways of managing and building institutions that are more or less favourable to economic performance. What is rejected here is a conception of culture that sees culture as directly or automatically responsible for creating cooperative attitudes. Having quite rightly denounced this magical vision of the effects of culture (Womack et al. 1990;

50   H. Yousfi Bayart 2006), the authors ultimately fail to detach themselves from it. As a result, while their own data clearly show that there are considerable variations between the countries that they study, they fail to see how culture intervenes. Culture does intervene, not by producing stereotyped behaviour, but by influencing the way in which the power relationships, management practices and the institutions they promote are received and implemented. In place of this polarized debate between the partisans of culture and universal standards, we have shown that some authors argue that it is the specific nature of the interrelations between the ‘universal’ and the ‘local’ that matters. In fact, with the rise of globalization, there is increasing recognition that effective development is a blend of universal processes and specific local issues. Different forms of connection and interrelation have been investigated in many different fields of development, including the means used to manage social in­equalities within different societies, the influence of social networks and the impact of institutions in enhancing economic performance. Yet, we have emphasized that valuable as this work has been in advancing our way of conceiving how culture matters for development, ambiguities remain with respect to how to conceive of the linkage between cultural continuity and change, between individual autonomy and collective order, and the relative preponderance of the universal and the local. The question that still remains open for debate is: how does one really break away from the framework associated with the tradition/ modernity dichotomy in development thinking? In this respect, the way culture is said to affect management echoes the difficulty of reconciling what is considered part of tradition with the universal assumption underlying the way the literature considers development issues. At the same time, the existence of successful firms in emergent countries, which appear as small islands of modernity, efficiency and good governance, leads us to question the way that linkages between culture and development issues have been considered in development thinking. In fact, looking at the firm as a sub-­ polity, part and parcel of an encompassing society rather than as a profit-­ maximizing ‘black box’, offers a remarkable laboratory for studying the relationship between forms of governance and cultural context. Studying the functioning of the firm can help us to renew our perspective on the relationship between culture and development. The issue then becomes: what lessons can be learned from the existence of successful companies in emergent countries? Do such examples mean that these countries can be modernized despite their culture, by fighting against it or simply by importing the so-­called ‘universal’ standards developed in the West? Or, on the contrary, do they show that there is good in each culture, even in terms of governance, and that these countries can be modernized by leveraging their culture? In this section, drawing upon d’Iribarne’s conceptualization of culture discussed previously, we will seek to analyse how best the universal standards of international institutions can respond to the local context in order to devise more effective development policies. First, we will seek to show how this alternative

Culture and development   51 approach enables an escape from the fatalistic vision embedded in the tradition/ modernity dichotomy. Then, we will analyse how this approach articulates social change and cultural continuity, as well as diversity and homogeneity, within the same society in order to improve our understanding of development challenges. Escaping the tradition/modernity dichotomy As we noted above, the investigations by d’Iribarne and his colleagues show that managerial behaviour in developing countries provides useful insights for development advisors who may be tempted to recommend standard administrative reforms as the remedy for the shortcomings of current approaches to development. The investigations show that while successful firms in ‘developing’ countries were affected and transformed by the acquisition of new ‘universal’ tools, they remained profoundly ‘traditional’ in the way these tools were actually given meaning and interpreted. At a very abstract level, there is barely a difference in different contexts. The same principles are found in Cameroon, Morocco or France. To motivate people, no matter where they are or what their culture might be, one must treat them well, respect them, give them responsibility, listen to and inform them, give just rewards for their efforts, promote the feeling that they belong to a remarkable team and permit them to trust each other. However, the empirical cases discussed above show that these general principles are implemented practically in specific contexts and adapted to them. When deciding how people should be evaluated and rewarded, what emphasis should be given to individual or group contribution, whether to do this formally or informally, reformers had to take into account the specific contexts in which change was being implemented.31 These cases highlight the fact that it was not simply a matter of juxtaposing, or compromising between, the modern and the traditional, but creating an intimate union between them. It was as if the most modern business standards were being given substance through the intermediary of local cultural forms. Because these standards had taken on meaning based on local conceptions of ‘authority’, ‘cooperation’, ‘decision’, etc., employees adhered to them and implemented them efficiently. These ‘modern’ standards thus became an effective vehicle for the organization’s success, because the ‘modern’ was reinterpreted effectively within the local scheme of reference. In each case, the company transformed the way it operated by leveraging elements of the local culture. In development thinking, this way of combining the traditional with the modern and the universal is somewhat foreign to those who normally contrast these two categories. Development theory is currently struggling to find suitable points of articulation between the universal and the local. Debates continue to rage between a majority that defends the universal nature of good development principles and those who emphasize their local character. The two theories are simultaneously correct and inadequate. On a highly abstract level, there are certainly a considerable number of governance principles that have universal scope, whether this involves exercising authority, mobilizing people, business ethics or

52   H. Yousfi reforming institutions. Yet, once we return to the concrete realm of action and implementation, the local level must be taken into account. Certain confusions arise, however, given that development theorists tend to conflate these two levels. Advancing principles with universal scope, experts justify what are no more than expressions of the specific cultures in which these reforms have been conceived. As we discussed above, the importance given to the ‘enforcement of property rights’ as a critical ingredient to build good institutions able to achieve economic development is far from reflecting a ‘universal’ criterion. It is rooted in a specific US conception of a ‘well-­ordered society’ linking the idea of freedom to property. Similarly, the need to assess people’s work in order to reward them is a universal principle, but its implementation may vary. While in the United States performance assessment is ultimately about achieving specific results, in Jordan the criterion would be the extent to which people ‘tried to do their best’ rather than the outcomes of these efforts (Yousfi 2007). In France, the basis for any kind of evaluation refers to the rights and duties associated to a specific position. The issue is not whether people share the same values, or whether they can be expected to act in a manner consistent with a certain ‘universal rationality’. Those who implement development policies in developing countries have to interpret the actions and attitudes of individuals and organizations that operate in contexts quite different from their own. They have to cooperate with groups that have different interpretations of the proposed universal ‘reforms’, different methods of achieving them and different perceptions of their advisors’ behaviour. So, the exercise that experts engage in when they try to implement policies that are more or less evident within their own culture is transformed when it involves a confrontation between two cultures. Thus, knowledge about cultural differences in perceiving the role of the ‘universal principles’ of governance is indispensable. Learning about the interpretation by different cultures of supposedly universal principles may allow development institutions to understand just how much ‘universal’ development recipes owe to their cultural roots, with subsequent policy adjustments. An adaptive process that allows people with very different backgrounds to interpret correctly what others are doing would be very useful. In order to clearly break away from the fruitless tradition/modernity dichotomy, it is important to distinguish the universal from the incidental in such practices. It is necessary to change the way in which development thinking conceives of the relationship between the tools and the people who use them. In addition, some practices are transferable even though they have a different meaning.32 The task for development thinking is not to focus on the tools and disregard the people, but rather to look at how the use of tools is integrated into the context of meanings, and hence the cultural world, of those putting them to work. Indigenous studies as well as cross-­cultural comparisons can thus contribute to an enhanced understanding of both universals and ways in which these universals find locally distinctive expression.

Culture and development   53 Social transformation, diversity and cultural continuity One of the issues raised in this literature review is that the persistence of the ­tradition/modernity framework has undermined our understanding of how social transformation and diversity in a given society are articulated with more stable guidelines and more durable reference points. In fact, within one and the same culture, attitudes vary greatly. It is not just that individuals with a wide variety of attitudes are to be found, which no one would doubt. It is, more radically, that the same individual can have sharply contrasting attitudes according to the circumstances. For example, Americans are commonly described as individualists, as if this were a characteristic of US culture. Yet, this individualism is in fact quite unevenly developed depending on what spheres of life are being considered. For example, the community’s control over each member’s loyalty to its ethical norms is much more accepted in the United States than in Europe. We can find evidence of these differences in both political and economic spheres. It makes no sense to say that Amer­ icans are less individualistic than the French, or that the Chinese are more collectivist than the British. Such broad statements have no more meaning than statements of the exact opposite. In every society, one can find evidence of individualism and collectivism, but the combination of the two aspects varies (d’Iribarne 1997). If culture is seen as a ‘framework of meaning’, then this implies that one would seek to understand the interpretation that culture proposes for particular events and situations. The issue is not to try to find out how individuals from a specific culture are supposed to act in all circumstances in compliance with inculcated attitudes. On the contrary, the focus will be on the fact that they change their attitudes and behaviours according to the meaning they give to particular events and situations. All cultures establish criteria by which individuals denote, classify, identify, evaluate, connect and order.33 They define the principles of classification through which society is viewed as being made up of separate groups. They provide interpretative systems that give meaning to the problems of existence, presenting them as elements that shape the relationship between individual autonomy and collective order. Far from imposing roles on each individual from which he or she cannot escape, culture influences the particular direction of the strategies adopted by each social group as it defends both its interests and its convictions. There are neither purely ‘natural’ interests nor a universal rationality existing independently of the meanings attached to whatever is at stake. In the French case, the importance that French culture has persistently accorded to the distinction between what is noble and what is common (even though the definition of what is noble and what is common has varied considerably in the course of history) has led to a certain structuring of interests. The persistent role of honour is linked to the continuing relevance of this distinction. This cultural continuity does not prevent the social transformations that occur successively in the same country from being different in many respects. It reflects the fact that the frames of

54   H. Yousfi r­ eference within these social changes have taken on meanings that are much more stable than the changes themselves. Therefore, the question is not whether or not developing countries should change their culture to promote progress, but rather: what is the framework of meaning in each society that initiates an effective social transformation, which does not compromise the locally shared conception of a ‘good’ relationship between individual autonomy and the collective order? Everywhere one finds a combination of representations and practices that helps to manage these contradictions. The historical references, the metaphors or the narratives, which give meaning to the instruments enabling members of a group to exercise their role while framing their actions and preventing misuse, do not change radically when passing from one domain to another in societal life or when initiating a reform process. The examples cited above show that successful transformations depend on giving a positive sense to problem-­ridden situations by assimilating them to situations that, in the framework of meaning within the given society, echo positive experiences. It is a matter of organizing practices in such a way that the transformation process authentically echoes the local legitimate conception of a well-­ ordered society. It is this kind of work that makes it possible to see beyond the resistance encountered in developing countries towards imported reforms and to identify the resources offered by each culture for the implementation of effective development strategies.

Conclusion This literature review has identified the many ways in which culture has been considered in development thinking. We have distinguished many perspectives, each with a relatively distinct interpretation of the concept of culture. Throughout, we have stressed how modernization theory, by establishing the tradition/ modernity dichotomy, has had a big impact on the framework in which culture has been deployed and debated in development thinking. The result is that culture has been seen either as a positive instrument or as an obstacle to development. For more than a century, culture has been viewed as a pervasive determinant of psychological traits, mentalities and attitudes, which in turn condition the way people behave and confine them to the realm of tradition. On the other hand, development has always been seen in terms of trade-­off between cultural traditions and access to progress, which is conceived of in terms of a ‘universal modernity’. Thus, it is hardly a surprise that culture was often seen as opposed to development, as tradition is opposed to modernity, and habit to rationality. Following this, we argued that many streams of research have offered different ways of overcoming the deterministic relationship between culture and development. However, we emphasized that these various lines of thought have one thing in common: they have difficulty in grasping the articulation between the role of culture – suspected of bringing back ‘tradition’ – and individual autonomy as a condition of a ‘modern society’. In fact, in these different concep-

Culture and development   55 tualizations of culture, variously referred to as ‘ideology’, ‘informal institutions’, ‘informal rules’, ‘lifestyles’, ‘consensus ideology’, the concept of culture is closely related to the idea of social order (d’Iribarne and Henry 2007). This social order, whether inherited or imposed by the dominant group, is suspect because its stable and normative aspects are thought to downplay the creativity of individuals and to hinder development. When culture is seen as produced essentially by the actors, development occurs as a result of the creation of an enabling environment through which social structure is transformed. Any continuity over very long periods of time seems inimical to development because it presupposes an absence of any kind of cultural learning or social transformation. When culture is seen as a set of values and attitudes, development works by calling into question traditional cultures and enabling the emergence of a modern mentality. It is evident that this perspective offers no basis for long-­term stability. It assumes away phenomena capable of engendering continuities. Any kind of continuity would downplay individual autonomy and compromise the achievement of progress. This review has emphasized the persistence of the tradition/modernity framework, which leads to the view that developing countries must shed their traditional structures to access modernity. However, empirically we find evidence of both massive cultural change and the persistence of distinctive cultural traditions even within ‘developed countries’. Similarly, many scholars have argued that so-­ called ‘modern societies’ are themselves embedded in specific cultures, even though their institutions are following ‘universal’ principles. They show that the existence of such cultural continuity does not prevent a succession of forms of social organization in the same country that are different from each other in many respects. Therefore, the challenge for development thinking is still to improve our understanding of how the heterogeneity of values or the transformation of social organization in a given society is articulated with more durable frames of reference. One possible way to overcome this impasse is to recognize the ‘constructedness’ of culture and to pay more attention to how universal principles are understood and interpreted in local contexts. In the words of LéviStrauss (1958): In all cases there is something that is retained and that historical observation can gradually uncover, through a sort of filtration process that allows what might be called the lexicographical content of institutions and customs to pass through, leaving behind only the structural elements. Having traced the predominance of the dichotomy of tradition/modernity through time, and viewed them from a number of perspectives, what are the lessons that we can drive from this literature review? This chapter leads us to change the way we look at the relationship between culture and development in two important ways. First, the new insights presented in this literature review recognize that the connection between culture and economic development is not causal but ­multiple

56   H. Yousfi and complex. It is widely acknowledged that the importance of culture cannot be instantly translated into ready-­made theories of cultural causation. What is needed is not the privileging of culture as a variable that works on its own, but the integration of culture into a wider picture in which culture is seen as intimately connected to different aspects of development issues. Culture must be seen as an element that shapes the means of the development process, as well as its goals. In fact, culture does not only impact social relations but also the functioning of formal institutions and management techniques. For instance, for scholars studying the issues of inequality and poverty, such as Mary Douglas or Amartya Sen, the question is how interactions between diverse cultures within a society need to be managed democratically and in a manner that allows for free and fair debate. For d’Iribarne, the question is how to shape institutions that take into account in each society legitimate ways of behaving with authority, decision-­ making, managing inequality, etc. This is the way that culture affects the process of building appropriate institutions. Second, the challenge for the development thinking is how to adapt ‘best practices’ to the diversity of cultural contexts. The task proposed is complicated by the fact that the starting points for each society’s conception of a well-­ordered society are different. Meanwhile, the end points, in terms of the precise articulation between what is achievable and what corresponds to the desires of each society, show considerable variation. ‘Universal’ or ‘modern’ standards of aid donors and those who receive the aid are culturally embedded. The objectives of reformers and the programmes of action they adopt do not exist independently of the way in which they see the world. Their actions take on meaning, in their own eyes and in those of the people they have to rally to their cause, within a given structure of meanings. By referring to the fundamental principles of that structure, they make the changes they advocate appear sensible, and in so doing they help to ensure the survival of those principles. This does not lead to complete relativism; it does help place the lessons from certain valuable cultural theories into a more grounded and realistic sociological framework. There is an urgent need to evolve an innovative development discourse underpinned by solid comparative methods and not by simple analogies. We cannot continue to read the history of the South through the history of the North. We can only break away from the so-­called best practices so prevalent in the mainstream discourse if we understand the complexity of the relationship between individual agency and cultural continuity in institutional change. Only when we accept the existence of different cultural ‘traditions’ across societies can we begin to understand that what people believe and do matter in the real sense, and then effectively adapt universal standards to local contexts. An approach that is less focused on the superiority of any particular model and more cognizant of the context-­specificity of desirable reforms is needed. Work on gradual improvements in institutions or in small institutional practices may help towards seeing exactly and concretely how beneficial institutional arrangements are promoted or held back, and how universal standards can be adapted to the

Culture and development   57 local contexts. Such an approach may help to open up a new way of looking more realistically and more effectively at development challenges.

Notes   1 The author would like to thank John Humphrey, the leader of the Globalisation Team at IDS (Institute of Development Studies, UK), for his detailed and insightful comments on earlier drafts. Special thanks are due to Philippe d’Iribarne for his intellectual contribution to this chapter. Thanks also to Aymeric Blanc for his helpful comments.   2 Even when agreement is achieved, the terms are so general that they become tautological or lose all applicability.   3 Surveys of economic development thinking can be found in Ray (1998), Basu (1997) and Schech and Haggis (2000).   4 Durkheim’s notion that there are two different kinds of bonds between people, mechanical solidarity (solidarité mécanique) and organic solidarity (solidarité organique), is a similar line of reasoning (Durkheim 1893).   5 This perspective led to the perception of ‘Confucianism’ as unsuited for a dynamic industrial economy at the beginning of the twentieth century, and to the opposite claim that industrial and economic progress, as currently illustrated by the performance of East Asia, is in large part the result of Confucian ethics.   6 Weber’s thesis on the particular suitability of the Calvinist ethic to capitalism is often evoked, incorrectly, as the distinguished progenitor of this perspective. Weber, in his celebrated Protestant Ethic and the Spirit of Capitalism (1930), was not outlining a causal relationship between Calvinism and capitalism, but merely demonstrating that historically there was an ‘elective affinity’ between them.   7 The basic argument is that Islam preaches fatalism, which is inimical to growth. See Said (1979), Kuran (1997) and Rodinson (1973).   8 For example, today Muslims are considered relatively poor, both in terms of the per capita GDP of Muslim countries compared to the global average and in terms of international differences (Kuran 1997). To explain this phenomenon, there is an active debate between those who ascribe this state of affairs to Islam itself (Lewis 2002) and those who lay the blame at the feet of Western imperialism (Rodinson 1973). Consequently, there is no room to think constructively about how Islamic references may interfere with ‘objective’ constraints, such as Western economic and political domination, in impacting economic performance.   9 Since individuals are seen as simultaneous carriers of several cultural identities, depending on the issue at hand, a different cultural identity may become salient at a given moment. The conceptualization of culture as associated with ‘identity’ helped to emphasize ‘cultural change’ and the role of individuals in this process. 10 The debate surrounding the influence of structure and agency on human thought and behaviour is one of the central issues in sociology. In this context, ‘agency’ refers to the capacity of individual humans to act independently and to make their own free choices. ‘Structure’ refers to factors such as social class, religion, gender, ethnicity, customs, etc., which seem to limit or influence the opportunities that individuals have. 11 See the critique contained in the 2006 World Development Report, Equity and Development. 12 A survey of the new ways of thinking about culture and development can be found in Rao and Walton (2004). 13 To refer to institutional constraints, North sometimes prefers to use the word ‘ideology’, perhaps unwisely, because it suggests too much the fashionable, over-­ intellectualized idea of culture to which he does not subscribe in the least (Kuper 1999).

58   H. Yousfi 14 This point will be developed in the next section. 15 For a discussion on the transferability of legal systems, see Kleinfeld (2006). 16 The differing impact of cultural values on networks of social relations is the basis for the concept of social capital. One of the most important headings under which the issue of culture and development has been discussed in the 1980s and 1990s has been that of ‘social capital’ (Putnam 1993). Social capital consists of the norms or values shared by a group of people that promote cooperation and trust among them. This, like physical and human capital, is a source of wealth. 17 Zucker (1986) makes a useful distinction between characteristic-­based trust and process-­based trust. In the first case, trust is based on the characteristics of the business partner, such as family, ethnic or other attributes. The second kind of trust is derived from the process of doing business directly with X or knowledge of X’s dealings with others (reputation). 18 Luhmann (1979: 23) distinguishes between different forms of everyday trust (including personal trust) and system trust as a more impersonal form. The functioning of all complex political or economic institutions, government bureaucracies or monetary systems, depends directly upon system trust and, at least in part, on trust generated in the more intimate and cognitively accessible contexts of each human being’s everyday life. 19 In its simplest version, the convergence hypothesis states that industry would spread rapidly throughout the world, breaking down traditional loyalties and quickly producing committed industrial workers in the large enterprises, cities and government bureaucracies. Though different elites (dynastic, middle-­class, revolutionary intellectuals, colonial administrators and nationalist leaders) would direct the industrialization of developing countries, the ‘logic’, or the demands, of industrialization would require them increasingly to share decision-­making with workers, managers, scientists and others. A pattern of pluralistic industrialism would arise, and as multiparty rule-­ building replaced revolutionary movements and ideologies, the world would become more homogenous, prosperous and perhaps more democratic (Kerr et al. 1960). 20 Hofstede grounded his research on an extensive literature review, including the values orientation work of Kluckhohn (1951) and Kroeber and Parsons (1958), and on one of the largest databases ever analysed (attitude surveys of 116,000 IBM employees). 21 The strongest exponent of the applicability of Western management techniques cited by Leonard (1987) is the example presented by Cornelius Dzakpasu. Dzakpasu cites the case of an African public company that was near bankruptcy and was rescued through the use of such methods. He states explicitly, though, that this was feasible because the company was an ‘enclave’, i.e. it could be cut off from its environment. He does not tell how such isolation was achieved. 22 L’Afrique des entreprises, by the group of l’Agence française de Développement, La Documentation française, 1998. 23 At the time, the Electric Company employed 3,500 people. It had very high-­level managers and technicians, some of whom had degrees from international colleges and universities. Several missions took place between 1988 and 1994, at the joint request of the company and l’Agence française de Développement. 24 ‘Good intentions’ is the expression used by the Cameroonian interviewees to comment on the role of procedures. 25 It is worth noting that according to Michel Crozier’s own analyses, the real role of rules in the functioning of the ‘monopole’ hardly fits with its description in the French context. In addition, he extended his analysis beyond the universal role of ‘impersonal rules’ in regulating power relationships to the kind of bureaucratic system that can develop within a given culture (1964: 167). His analysis led to his main cultural finding: in the French context, the fear of face-­to-face communication requires impersonal mediation to avoid confrontation with those in authority. 26 This successful company of over 70 subsidiaries (circa 6,000 employees) differs from

Culture and development   59 other Tunisian companies by its use of strict regulations and employee appraisal based on control and regulation systems directly inspired by the US management model. 27 The metaphor of slavery is often used to describe a situation of subordination. 28 The French word for ‘worker’. 29 The interviews were conducted in Arabic and in French. The managers interviewed in this study used French and universal management vocabulary to describe the functioning of their group, but when it came to commentary on how concretely they implemented the ‘American model’, they switched instantly to Arabic and provided us with different insights. 30 Bayart (1996) suggests that the subject to investigate is ‘history’ and not ‘culture’. He criticizes the use of ‘culture’ as a means of hiding the ‘true problems’ that are only, and above all, ‘political’. It is worth pointing out that this line of thinking is particularly strong in France. In this regard, Debray (2006) argues that scepticism towards the idea of ‘culture’ is rooted in the universalism of ‘the Enlightenment philosophy’. As an illustration, there is no equivalent to the field of ‘cultural studies’ in France. 31 A survey of clients’ perceptions of French Development Agency (AFD) programmes in eight countries showed that different meanings given to words such as ‘partnership’, ‘cooperation’, ‘evaluation’ and ‘decision’ led to different expectations from AFD and different interpretations of its actions (d’Iribarne and Henry 2007). 32 The total-­quality management approach, which was originally Japanese, was effectively implemented in many contexts but interpreted and adapted in different ways. 33 One can find a similar approach in Bayart (1996: 115), even though he rejects the concept of culture and speaks of essential ‘repertories’ (repertoires essentiels) of actions that are historically constituted and mixed.

References Abraham, A. and Platteau, J.P. (2004) ‘Participatory Development: Where Culture Creeps In’, in Culture and Public Action, Stanford: Stanford Social Science, pp. 210–33. Amin, S. (1979) Unequal Development, Delhi: Oxford University Press (first published in French in 1973). —— (2000) ‘The Capacity to Aspire: Culture and the Terms of Recognition’, in Culture and Public Action, Stanford: Stanford Social Science. Austin, J.E. (1990) Managing in Developing Countries: Strategic Analysis and Operating Techniques, New York: The Free Press. Banfield, E.C. (1958) The Moral Basis of a Backward Society, Glencoe, IL: The Free Press. Basu, K. (1997) Analytical Development Economics: the Less Developed Economy Revisited, Cambridge, MA: MIT Press. Bauman, Z. (1973) Culture as Praxis, London: Routledge and Kegan Paul. Bayart, J.F. (1996) L’illusion Identitaire, Paris: Fayard. —— (2006) L’Etat en Afrique: La Politique du Ventre, Paris: Fayard. Becattini, G. and Sforzi, F. (2002) Lezioni sullo Sviluppo Locale, Turin: Rosenberg and Sellier. Berckowitz, D., Pistor, K. and Richard, J.F. (2003) ‘Economic Development, Legality, and Transplant Effect’, European Economic Review, 47 (1), pp. 165–95. Bessis, S. (2001) L’occident et les autres, Périphéries. Bhabha, H. (1990) ‘The Other Question: Difference, Discrimination and the Discourse of Colonialism’, in Ferguson, R. (ed.) Out There: Marginalization and Contemporary Cultures, Cambridge, MA: MIT Press. Bliss, C.J. (1993) ‘Life-­style and the Standard of Living’, in Nussbaum, M. and Sen, A.K. (eds) The Quality of Life, Oxford: Oxford University Press.

60   H. Yousfi Bloch, M. (1968) La Société Féodale, Paris: Albin Michel. Bourdieu, P. (1986) ‘The Forms of Capital’, in Richardson, J. (ed.) Handbook of Theory and Research for the Sociology of Education, New York: Greenwood, pp. 241–58. Boyacigiller, N.A. and Adler, N.J. (1991) ‘The Parochial Dinosaur: Organizational Science in a Global Context’, Academy of Management Review, 16 (2), pp. 262–90. Buckley, P.J. and Casson, M. (1988) ‘A Theory of Cooperation in International Business’, in Contractor, F. and Lorange, P. (eds) Co-­operative Strategies in International Business, Lexington and Toronto: Lexington Books. Casson, M. (1997) Information and Organization, Oxford: Clarendon Press. Chang, H.J. (2003) ‘Understanding the Relationship between Institutions and Economic Development–Some Key Theoretical Issues’, paper presented at the WIDER Jubilee Conference, 17–18 June, 2005, Wider, Helsinki. Chinese Culture Connection (1987) ‘Chinese Values and the Search for Culture-­free Dimensions of Culture’, Journal of Cross-­cultural Psychology, 18 (2), pp. 143–64. Clegg, S.R. and Redding, S.G. (eds) (1990) Capitalism in Contrasting Cultures, New York: Walter de Gruyter. Crozier, M. (1964) The Bureaucratic Phenomenon, Chicago: University of Chicago Press. Das, T.K. and Teng, B.S. (1998) ‘Between Trust and Control: Developing Confidence in Partner Cooperation in Alliances’, Academy of Management Review, 23 (3), pp. 491– 512. Dasgupta, P. (1988) ‘Trust as a Commodity’, in Gambetta, D. (ed.) Trust: Making and Breaking Cooperative Relations, Oxford: Blackwell, pp. 49–72. de Bony, J. (2003) Individual Autonomy and Socialization at the Dutch Elementary School: the Relationship between the Individual and the Group, proceedings of the 30th Onderwijs Research Dagen, Open University, Heerlen, the Netherlands. Debray, R. (2006) Aveuglantes Lumières, Paris: Gallimard. Dei Ottati, G. (1994) ‘Trust, Interlinking Transactions and Credit in the Industrial District’, Cambridge Journal of Economics, 18 (6). de Rozario, P. (2006) Le mandat d’intégrer les jeunes en Europe: confrontation de cultures: le FSE en Finlande, en Italie et France. Thèse, Institut d’Etude politiques, Paris. de Tocqueville, A. (1835) De la démocratie en Amérique I, Paris: GF Flammarion. —— (1856 [1952]) L’Ancien Régime et la Révolution, Paris: Gallimard. d’Iribarne, P. (1989) La Logique de L’honneur, Paris: Seuil. —— (1997) ‘The Usefulness of an Ethnographic Approach to the International Comparison of Organisations’, International Studies of Management and Organization, 26 (4). —— (2003) TroisFigures de la Liberté, Annales, 58 (5), pp. 953–80. —— (2006) L’Agence française de développement et ses partenaires, la prise en compte de la dimension culturelle, Working Paper AFD. d’Iribarne, P. and Henry, A. (2007) Successful Companies in the Developing World: Managing the Synergy with Culture, Notes and Documents AFD. d’Iribarne, P., Henry, A., Segal, J.P., Chevrier, S. and Globokar, T. (1998) Cultures et Mondialisation, Paris: Seuil. Douglas, M. (1987) How Institutions Think, London: Routledge and Kegan Paul. —— (2004) ‘Traditional Culture: Let’s Hear No More about It’, Culture and Public Action, Stanford: Stanford Social Science. Drucker, P.F. (1971) Men, Ideas, and Politics: Essays, New York: Harper and Row. Durkheim, E. (1893) The Division of Labour in Society, translated by W.D. Halls (1984), New York: The Free Press.

Culture and development   61 Eisenstadt, S.N. (1968) The Protestant Ethic and Modernization, New York: Basic Books. Escobar, A. (1995) Encountering Development: the Making and Unmaking of the Third World, Princeton: Princeton University Press. Etzioni, A. (1996) The New Golden Rule: Community and Morality in a Democratic Society, New York: Basic Books. Fafchamps, M. (1996) ‘The Enforcement of Commercial Contracts in Ghana’, World Development, 24 (3), pp. 427–48. Farmer, R.N. and Richman, B.N. (1965) Comparative Management and Economic Progress, Homewood: Richard D. Irwin. Ferguson, J. (1990) The Anti-­Politics Machine: ‘Development’, Depoliticization, and Bureaucratic Power in Lesotho, Cambridge: Cambridge University Press. Foner, E. (1998) The Story of American Freedom, New York: Norton. Foucault, M. (1980) Power and Knowledge, New York: Pantheon Books. Franke, R.H., Hofstede, G. and Bond, M.H. (1991) ‘Cultural Roots of Economic Performance: A Research Note’, Strategic Management Journal, 12, pp. 165–73. Freeman, R.B. (2000) Single Peaked vs. Diversified Capitalism: The Relations between Economic Institutions and Capitalism, NBER Working Paper, N W7556, February. Fukuyama, F. (1995) Trust: the Social Virtues and the Creation of Prosperity, London: Hamish Hamilton. —— (2001) in International Encyclopaedia of the Social and Behavioral Sciences, Smelser, N.J. and Baltes, P.B. (eds), pp. 3130–4, Oxford: Pergamon. Gambetta, D. (1988) Trust: Making and Breaking Cooperative Relations, Oxford: Oxford University Press. Geertz, C. (1973) The Interpretation of Cultures, New York: Basic Books. Giddens, A. (1984) The Constitution of Society: Outline of the Theory of Structuration, Berkeley: University of California Press. Granovetter, M. (1985) ‘Economic Action and Social Structure: the Problem of Embeddedness’, American Journal of Sociology, 91 (3), 11/85, pp. 481–510. Guizo, L., Sapienza, P. and Zingales, L. (2002) ‘People’s Opium? Religion and Economic Activities’, NBER Working Paper 9237, Cambridge, MA: National Bureau of Economic Research. Haastrup, K. (1996) A Passage to Anthropology. Between Experience and Theory, London and New York: Routledge. Hall, S. (1992) ‘The West and the Rest: Discourse and Power’, in Hall, S. and Gieben, B. (eds) Formations of Modernity, Cambridge: Polity Press. Hampden-­Turner, C. and Trompenaars, F. (1993) The Seven Cultures of Capitalism: Value Systems for Creating Wealth in the United States, Japan, Germany, France, Britain, Sweden and the Netherlands, Garden City: Doubleday. Harbison, F. and Myers, C.A. (1959) Management in the Industrial World: An International Analysis, New York: McGraw. Harrison, L.E. and Huntington, S. (2000) Culture Matters: How Values Shape Human Progress, New York: Basic Books. Hefner, R.W. (1998) Representation: Cultural Representations and Signifying Practices, London: Sage/The Open University. Hofstede, G. (1980) Culture’s Consequences: International Differences in Work-­Related Values, Beverly Hills: Sage. Humphrey, J. and Schmitz, H. (1998) ‘Trust and Inter-­Firm Relations in Developing and Transition Economies’, Journal of Development Studies, 34 (4), pp. 32–61.

62   H. Yousfi Huntington, S. (1971) ‘The Change to Change: Modernization, Development, and Politics’, Comparative Politics, 3 (3), pp. 283–322. —— (1996) The Clash of Civilization and the Remaking of World Order, New York: Simon and Schuster. Jaeger, A.M. and Kanungo, R. (1990) Management in Developing Countries, London: Routledge. Kerr, C., Dunlap, J.T., Harbison, F.H. and Myers, C.A. (1960) Industrialism and Industrial Man, Cambridge, MA: Harvard University Press. Kiggundu, M.N. (1989) Managing Organizations in Developing Countries: An Operational and Strategic Approach, West Hartford: Kumarian Press. Kleinfeld, R. (2006) ‘Competing Definitions of the Rule of Law’, in Carothers, T. (ed.) Promoting the Rule of Law Abroad: The Search for Knowledge, Washington, DC: Carnegie Endowment for International Peace. Kluckhohn, C. (1951) ‘Values and Value-­orientation in the Theory of Action: An Exploration in Definition and Classification’, in Parsons, T. and Shils, E.A. (eds) Toward a General Theory of Action, pp. 388–433, Cambridge, MA: Harvard University Press. Knorringa, P. (1996) Economics of Collaboration: Indian Shoemakers between Market and Hierarchy, New Delhi and London: Sage. Kroeber, A.L. and Parsons, T. (1958) ‘The Concepts of Culture and of Social System’, American Sociological Review, 23, pp. 582–3. Kuper, Adam (1999) Culture: The Anthropologists’ Account, Cambridge, MA and London: Harvard University Press. Kuran, T. (1997) ‘Islam and Underdevelopment: An Old Puzzle Revisited’, Journal of Institutional and Theoretical Economics, 153, pp. 41–71. —— (2004) ‘Cultural Obstacles to Economic Development: Often Overstated, Usually Transitory’, Culture and Public Action, Stanford: Stanford Social Science. Larrain, J. (1989) Theories of Development, Capitalism, Colonialism and Dependency, Cambridge: Polity. Lenartowicz, T. and Roth, K. (2001) ‘Does Subculture within a Country Matter? A Cross­cultural Study of Motivational Domains and Business Performance in Brazil’, Journal of International Business Studies, 32 (2), pp. 305–26. Leonard, D.K. (1987) ‘The Political Realities of African Management’, World Development, 15 (7), pp. 899–910. Leonardi, R. (1995) ‘Regional Development in Italy, Social Capital, and the Mezzogiorno’, Oxford Review of Economic Policy, 11 (2), pp. 165–79. Lévi-Strauss, C. (1958) Anthropologie structurale, Paris: Plon. Lewis, B. (2002) What Went Wrong? Western Impact and the Middle Eastern Response, New York: Oxford University Press. Lewis, W.A. (1954) ‘Economic Development with Unlimited Supplies of Labour’, Manchester School of Economic and Social Studies, 22, pp. 139–91. Luhmann, N. (1979) Trust and Power, Chichester: Wiley. —— (1988) ‘Familiarity, Confidence, Trust: Problems and Alternatives’, in Gambetta, D. (ed.) Trust: Making and Breaking Co-­operative Relations, Oxford: Basil Blackwell, pp. 94–107. McClelland, D.C. (1964) ‘Business Drive and National Achievement’, in Etzioni, A. and Etizioni, E. (eds) Social Change, pp. 165–78, New York: Basic Books. Meisel, N. (2004) Governance Culture and Development: A Different Perspective on Corporate Governance, OECD. Menkhoff, T. (1992) ‘Xinyong or How to Trust Trust? Chinese Non-­contractual Business

Culture and development   63 Relations and Social Structure: The Singapore Case’, Internationales Asienforum, 23 (1–2), pp. 261–88. Moe, T. (1987) ‘Interests, Institutions and Positive Theory’, Studies in American Political Development, 2, pp. 236–99. Mollering, G. and Stache, F. (2007) German-­Ukrainian Business Relationships: Trust Development in the Face of Institutional Uncertainty and Cultural Differences, MPIfG Discussion Paper. Montesquieu, C. de (1748 [1979]), De l’esprit des lois, Paris: Flammarion. Mundimbe, V.Y. (1988) The Invention of Africa, Bloomington: Indiana University Press. Nath, R. (1986) ‘The Role of Culture in Cross-­cultural and Organizational Research’, in Farmer, R.N. (ed.) Advances in International Comparative Management, Vol. 2, pp. 249–67, Greenwich, CT: JAI Press. Nohria, N. and Eccles, R. (1992) ‘Face-­to-face: Making Network Organisations Work’, in Nohria, N. and Eccles, R. (eds) Networks and Organisations: Structure, Form and Action, Boston: Harvard Business School Press. North, D.C. (1981) Structure and Change in Economic History, New York: Norton. —— (1987) ‘Institutions, Transaction Costs and Economic Growth’, Economic Inquiry, 25 (3), pp. 419–28. —— (1990) Institutions, Institutional Change and Economic Performance, Cambridge: Cambridge University Press. —— (1991) ‘Institutions’, Journal of Economic Perspectives, 5 (1), pp. 97–112. —— (1994) ‘Economic Performance through Time’, American Economic Review, 84 (3), pp. 359–68. North, D.C., Wallis, J.J. and Weingast, B.R. (2006) A Conceptual Framework for Interpreting Recorded Human History, Working Paper 12795, www.nber.org/papers/w12795. Olson, M. (1965) The Logic of Collective Action, Cambridge, MA: Harvard University Press. Ouchi, William G. (1981) Theory Z, Reading, MA: Addison-­Wesley. Parsons, T. (1967) Structure of Social Action, New York: Free Press. Pascale, R.T. and Athos, A.G. (1981) The Art of Japanese Management, New York: Warner Books. Piore, M. and Sabel, C. (1984) The Second Industrial Divide, New York: Basic Books. Platteau, J.P. (2007) ‘Is Culture an Obstacle to African Economic Development?’, IERC Conference, 23–4 February, University of Southern California, Los Angeles. Putnam, R.D. (1993) Making Democracy Work: Civic Traditions in Modern Italy, Princeton: Princeton University Press. —— (2000) Bowling Alone: The Collapse and Revival of American Community, New York: Simon and Schuster. Rao, V. and Walton, M. (2004) Culture and Public Action, Stanford: Stanford Social Science. Ray, D. (1998) Development Economics, Princeton: Princeton University Press. Rist, G. (1997) The History of Development, from Western Origins to Global Faith, London: Zed Books. Robertson, R. (1992) Globalization: Social Theory and Global Culture, London: Sage. Rodinson, M. (1973) Islam and Capitalism, New York: Pantheon. Rodrik, D. (1999) The New Global Economy and Developing Countries: Making Openness Work, Washington, DC: Overseas Development Council. —— (2003) ‘Growth Strategies’, NBER Working Paper No. 10050, October, Cambridge, MA, www.nber.org/papers/w10050.

64   H. Yousfi Rostow, W.W. (1960) The Stages of Economic Growth: A Non Communist Manifesto, Cambridge: Cambridge University Press. Said, E.W. (1979) Orientalism, New York: Vintage Books. Sako, M. (1992) Prices, Quality and Trust, Inter-­firm Relations in Britain and Japan, London and New York: Cambridge University Press. Schech, S. and Haggis, J. (2000) Culture and Development: A Critical Introduction, Oxford: Blackwell. Schmitz, H. (1999) ‘From Ascribed to Earned Trust in Exploring Clusters’, Journal of International Economics, 48, pp. 139–50. Sen, A.K. (1999) Development as Freedom, New York: Knopf. —— (2000) ‘How does Culture Matter’, in Culture and Public Action, pp. 37–58, Stanford: Stanford Social Science. Soderberg, A.M. and Holden, N. (2002) ‘Rethinking Cross Cultural Management in a Globalizing Business World’, International Journal of Cross Cultural Management, 2 (1), pp. 103–21. Stiglitz, J.E. (1994) Whither Socialism?, Cambridge, MA: MIT Press. Storper M. (2005) ‘Society, Community and Economic Development’, Studies in Comparative International Development, 39 (4), pp. 30–57. Tayeb, M. (2001) ‘Conducting Research Across Cultures: Overcoming Drawbacks and Obstacles’, International Journal of Cross Cultural Management, 1 (1), pp. 91–108. Tönnies, F. (1887 [1964]) Gemeinschaft und Gesellschaft [French translation: Communauté et Société: Catégories fondamentales de la sociologie pure] Paris: PUF. Wallerstein, I. (1979) The Capitalist World Economy: Essays, Cambridge: Cambridge University Press. Weber, M. (1922 [1971]) Wirtschaft und Gesellschaft [French translation: Economie et Société], Paris: Plon. —— (1930) The Protestant Ethic and the Spirit of Capitalism, London: Allen and Unwin. Weick, K.E. (1995) Sensemaking in Organizations, California: Sage Publications. Whitley, R.D. (1999) Divergent Capitalisms: The Social Structuring and Change of Business System, Oxford: Oxford University Press. Williamson, O.E. (1993) ‘Calculativeness, Trust and Economic Organisation’, Journal of Law and Economics, 36, pp. 453–86. Womack, P.J., Jones, D.T. and Daniel, R. (1990) The Machine That Changed the World, New York and Toronto: Rawson Associates, Collie Macmillan Canada and Maxwell Macmillan International. World Bank (2002) Building Institutions for Markets, Washington, DC: World Bank. —— (2005) A Better Investment Climate for Everyone, Washington, DC: World Bank. Yousfi, H. (2006) Le contrat dans une coopération internationale; la rencontre des intérêts à l’épreuve de la rencontre des cultures. Thèse, Université Paris X, Nanterre. —— (2007) ‘Gérer en Jordanie: une coexistence problématique entre système hiérarchique et idéal religieux’, Revue Française de Gestion, 171, pp. 157–73. —— (2010) ‘Culture and Trust in International Contractual Relationships, a French Lebanese Cooperation’, in Saunders, M., Skinner, D., Dietz, G., Gillespie N. and Lewicki, R.J. (eds) Organisational Trust: a Cultural Perspective, Cambridge: Cambridge University Press. Yousfi, H., Flipiake, E. and Bougault, H. (2005) Poulina, un management tunisien, Notes and Documents, Agence Française de développement. Zucker, L.G. (1986) ‘Production of Trust: Institutional Sources of Economic Structure’, in Staw, B.M. and Cummings, L.L. (eds) Research in Organizational Behavior, Vol. 8, pp. 53–111, Greenwich, CT: JAI Press.

3 Culture and development Do social struggles make a difference? Jean-­François Bayart

Culture and development: do social struggles make a difference? One of the biggest difficulties faced by the social sciences lies in trying to talk intelligently about ‘culture’. For one thing, we have all become, to a greater or lesser degree, Weberians, and we are all now convinced that human beings are like spiders spinning the webs of their own meanings. Even Marxists with communist leanings, or indeed communists, have finally come to admit as much, thanks to the influence of Gramsci, and the Copernican revolution (for lack of any other revolution) that they were led to bring about under the aegis of E.P. Thompson (1963) and his demonstration that the English working class had created itself as a moral and cultural community, as well as being a community that sprang from the interplay between relations of production. From this point of view, we have all more or less taken the ‘cultural turn’ of the 1980s (Clifford 1988; Hunt 1989; Jameson 1998; Ortner 1999). But, on the other hand, the tidal wave of ‘cultural studies’, a field generally more literary than anthropological or sociological, and the ravages of what I have called, uncharitably (but I don’t take back a single word), the ‘stupidity of identity politics’ – something which arose from the political retraction that accompanied the imperial and capitalist globalization of the nineteenth century – have made our repeated lip-­service to ‘culture’ more or less incomprehensible, on the level both of theory and of political or economic action (Bayart 1996, 2005). Paradoxically, culturalism, i.e. the explanation of social phenomena as the products of a relatively stable and homogeneous ‘culture’, prevents us from grasping the irreducibly ‘cultural’ dimension of social, economic, political and . . . cultural practices. This is not the place to go into the philosophical and religious debate on the ontological specificity of the human race (or its ontological relativity, either). But, whether we are like our cats and dogs or not, we still act socially and individually only through the intermediary of our symbolic and affective production. Over and above the confessional schisms of psychoanalysis, Freud’s great contribution was to make us realize as much – and on this point he in no way contradicts Marx, so long as we read Marx himself rather than his epigones. So our

66   J.-F. Bayart imagination is ‘constitutive’ (Veyne 1988), and the real question is that of the ‘imaginary institution of society’ (Castoriadis 1975) which cannot be dissociated from its relationship with materiality, from what is rather oddly called ‘material culture’ – can we conceive of a culture that is not material, or a materiality that is not cultural? – and the ‘techniques of the body’ induced by it (Mauss 1950: 365–86; Bayart 1996, 2004, 2005; Warnier 1999; Bayart and Warnier 2004).

Optical illusions The biggest methodological error committed by culturalist thinking, however, lies in the way it imagines that a ‘culture’ tends to be a totality and assumes a meaning that allegedly guides the actions of those who are part of this totality. Comparison – which Radcliffe-­Browne used to call ‘the methodological equivalent of experience’ for the social sciences – immediately invalidates this kind of conviction. The post-­Meiji ‘take-­off ’ of Japan and the graceful leaps of the ‘tigers’ of East Asia have often been spoken of in terms of ‘neo-­Confucianism’. But how, then, are we to differentiate between the political and economic paths of the People’s Republic of China and Taiwan? Or between South and North Korea? The objection that springs to the lips of the culturalist – ‘yes, but look at it in perspective, observe the current convergence between the countries on either side of the Strait of Formosa! Isn’t the Greater China of the diaspora a systematic part of the process?’ – confirms that culturalism rests on a short-­ sighted, blinkered approach. It implies downplaying the tens of millions of deaths caused by Maoism, the mute trauma inflicted by the Maoist period on the society of mainland China, and the dead-­end that is a probable result of the one-­ child policy (Liu 2000; Mueggler 2001). It means flattening out the hugely varied landscape of the inversions, inflexions and choices made in industrial, commercial and social policies that have characterized the People’s Republic ever since its foundation (Bing 2003; Wedeman 2003; Kernen 2004; Rocca 2006); however neo-­Confucian they may have been, Mao and Deng, being good Chinese citizens, were not neo-­Confucian in exactly the same way, even though they had both emerged from the same communist matrix – and culturalism does not have much to tell us about this difference. Culturalism is blind to the way the interaction between the People’s Republic and Taiwan rests on contrasting ‘trajectories’, on a political (and more or less militarized) conflict, on complementarities that are constructed rather than arising from any natural affinities (Mengin 1998, 2007). And culturalism also denies that the family-­centred ideology of ‘Asiatic values’ obliterates and legitimates the social relations of power and exploitation at the very heart of the kinship ‘networks’ that establish the ‘solidarity’ of the ‘diaspora’ and make it so ‘competitive’ when measured by global criteria (Ong 1999; Hsu 2000). Similarly, it is rather difficult to explain satisfactorily the completely contrary destinies of Thailand and Burma just by looking the Buddha straight in the eyes: actually, what is happening in the pagodas and the minds of their faithful is less significant than what is happening in the barracks. And things are no clearer if we look at the countries of Allah: from Turkey

Do social struggles make a difference?   67 to Iran, from Malaysia to Saudi Arabia, from Egypt to Libya, to Tunisia, Algeria and Morocco, homo islamicus is not the same homo oeconomicus everywhere. Culturalists tend to highlight certain facts that they judge to be emblematic, interpreting them in a one-­dimensional way that accords with their ideological or theoretical partis pris, and omitting to mention anything that contradicts the latter. The Islamic Republic of Iran, for example, gets reduced to the ritual photographing of a few women wearing a black chador and vociferating against the United States (or, since the election of Mohammed Khatami as president, in 1997, to the more alluring image of young women wearing multi-­coloured headscarves and holding mobile phones): this tells us nothing about the concrete and complex social practices of the hijab, in view of the formation of public space and the participation of the ‘second sex’ in smuggling, the rising power of the liberal professions or service sectors, and the way the process of economic accumulation is becoming more independent from state power (Adelkhah 1991, 1998, 1999, 2003). For the past 20 years, there has also been a considerable amount of chatter about the irresistible rise of the headscarf in the streets of Istanbul and, horresco referens, even as far as the gates of the official residence of the President of the Republic at Ankara – but nobody has much to say about the parallel increase in the number of bars selling beer in Anatolia, or about the appearance of a Turkish vineyard almost worthy of the name and fulfilling the implicit criteria of Copenhagen on membership of the European Union. In fact, another difficulty arises, one on which culturalists observe a discreet silence and which interculturalist approaches evade, in the field of development as in the field of religion or politics: who, exactly, is supposed to be the representative and legitimate interpreter of the ‘culture’ with which we wish to engage, work and converse? Mohamed Khatami, Mahmoud Ahmadinejad or the rebellious Iranian students of 1999? President Ben Ali, the new female President (a Tunisian), of the International Federation of Human Rights, Souhair Belhacen, or the leader in exile of the Islamist movement Rached Ghannouchi? José Bové, Ségolène Royal, Nicolas Sarkozy or Jean-­Marie Le Pen? Brezhnev or Sakharov? But also: the employer or the worker? The civil servant or the peasant? The noble or the captive? The man or the woman? The older or younger child? The believer or the atheist? Culturalist discourse always ratifies a relationship of power and inequality. Worse, it contributes to creating and reproducing this relationship, assigning an identity (of civilization, religion, ethnic group, nation, or some other) to the person whom it thus designates and whom, by this very act, it calls into being as the actor or unit of analysis, attributing (or refusing) material and symbolic resources to him or her. This conjuring trick with identity, and the violence with which it is accompanied over long periods – see the tormented relations between the Turkey of the post-­Kemal Atatürk period and its historical past and human diversity – are made all the easier by the tendency of culturalist thinking to take societies as isolated phenomena and to disconnect them from their regional or international environments, thus concealing the constitutive dynamics of the ‘frontiers’ of every kind through which they have been established (Kopytoff 1987; Rosaldo 1993; Bayart 1996, 2004, 2005, 2007b; Piot 1999).

68   J.-F. Bayart Actually, neither ‘cultures’ nor their constitutive elements (such as systems of kinship or ethnic ‘networks’) represent coherent and unequivocal corpuses. They carry within themselves symbolic and moral repertoires that are heterogeneous and polysemic from the point of view of political, economic and social action. In a country like Congo-­Brazzaville, we can thus observe the coexistence of different repertoires: monarchy, lineage society, prophetism, nationalism, Christianity, the invisible, Marxism, liberal democracy, civil and military bureaucracy, French Jacobinism, etc. This baroque assemblage makes it impossible for us to postulate a ‘Congolese culture’; after all, it has not been stable over the time span of the past 50 years. And above all, this so-­called ‘Congolese culture’ hardly belongs to an ‘African culture’, since this particular assemblage is not identical in the countries next door to it. Furthermore, the way it has guided action has been perfectly contingent, as is demonstrated by the vicissitudes of political life since the 1950s and the path followed by leaders of the stature of Pascal Lissouba, for instance, passing from the Marxist theorization of the ‘tribe class’ in the 1960s to 1970s to the ethno-­nationalist strategy of identity politics and the civil war of the 1990s. If we cross the River Congo, can we view the way such disparate politicians as Lumumba, Mulele, Mobutu, Tschisekedi and Kabila have resorted to the prophetic repertoire as transcending their ideological differences (extending as they do from nationalism to Marxism, via pure and simple villainy)? Does it really enable us to interpret the Sonderweg of Congo-­Zaire? Any sociologist who nourished such hopes would surely be tempted by Messianic expectations! As it happens, prophetism is a mere political ‘style’ (Brown 1992). This in no way affects its content, or lessens the weight of other factors such as the overall strategies of the actors in different sectors of society, including those considered to be ‘cultural’: religious institutions; the galaxy of bars, dance halls and sports clubs frequented by young people – places that have always been a major site of innovation (De Boeck and Plissart 2004); miners’ camps (De Boeck 2004); and the institution of the family in all the complexity of its social relations, whether between people of different status or gender, between allies, etc., including the ways it finds concrete expression, depending on the categories of actors and the contexts involved. More precisely, the social factors that culturalist thinking often highlights, in an attempt to make the idea of ‘culture’ less elusive, are no more tangible than this very idea, whichever overall sense we give it – ‘African’, ‘Arabic’, ‘Chinese’ or other. For several decades, the anthropological critique has completely undermined the consoling illusion of primordial factors – the ‘primordial identities’ of ethnicity, religion and family. The frequently hazy conceptualization of such factors has simply updated the myth of the state of nature, the Noble Savage and his primitive simplicity – or the related myth of the alienation and anomie allegedly fostered by the complexity of the nation, the state, the social differentiation of industrial society and, these days, globalization. Kinship now needs ‘rethinking’ (Needham 1977). It is an artefact, ‘an idiom rather than a system’, by means of which actors constantly negotiate their sense of belonging and their social alliances, as Peter Geschiere (1982) has shown in detail in the

Do social struggles make a difference?   69 case of the lineage society of the Maka people in Cameroon (see also Gulliver 1971; Schildkrout 1978). It was forged in response to the new phenomena of the slave trade or colonization (Piot 1999). It is at the same time a political language in which people are treated as having a certain value – diametrically opposite from their objectification into goods to be traded, as captives or wage-­earners (Kopytoff 1986, 1988). As such, it is first and foremost a field of conflict – for instance, the site par excellence of the practices of witchcraft (Geschiere 1995), of the exploitation of the labour force (Warnier 1993; Ong 1999), of the emancipation of the slaves (Kopytoff 1988), and of the confrontation with religious authorities in Islam, notably the Shiite clergy (Litvak 1998) – though it may then become a field of solidarity and collective action, or else may prop up the social institutionalization of the market (for example, in Iran, that of the bazaar: see Keshavarzian 2007: 95ff., and see also Geschiere 1994). Furthermore, kinship is neither infra-­political nor pre-­political. It is clear that there is a line of generative continuity between the family on the one hand, and ethnicity and state on the other. In ancient Greece, the genos, the phratry and the tribe were not ‘primordial identities’ proper to pre-­civic society; they flourished as institutions of the city, and guaranteed that its members were united by cohesion and philia (Polignac 1984). The same is true of the asabiyya, a term used by certain specialists of the Islamic world to designate the ‘community . . . welded together by blood ties or simply a similar destiny’ (Cheddadi 1980; Carré 1991: 41, for a comparison between asabiyya and Max Weber’s Gemeinsinn). But in the pioneering work of Ibn Khaldun, asabiyya is seen as part of the domain of ‘illusion’ (amr wahmî) and has no real foundation (lâ haqîqata lahu) (Cheddadi 1980). It is ‘in people’s heads’, as Michel Seurat (1989) put it, in his study of a district of Tripoli in Lebanon (see also Roy 1985, 1994, 1996). It is fully political, but it also plays a part in the institutionalization either of the state (for instance of the clergy, in Shiism, where the great ‘sources of imitation’ are organized into ‘households’ (beyt) whose rivalry is constitutive of the religious field both on the level of the different states concerned – Iran, Iraq and Lebanon in particular – and on the transnational level (Litvak 1998; Chehabi 2006). Political anthropology and history have established that the lineage principle is not diametrically opposed to the principle of the formation of state and empire and their social institutions, but, on the contrary, is one of the driving forces behind them. Likewise, empire and state engender kinship, ethnicity and confessionalism – indeed, they ‘manufacture’ them via their public policies, including in the private sphere of the family, thanks to the legal system, the tax system, the public health and education systems, conscription, war and the organization of economic activity. With its drastic policy of birth control, the coercive mobilization of young people, and the control of internal migration, China is, in this respect, a limit-­case (Davis and Harrell 1993). But in France, too, the state has fashioned the institution of the family, sacrificing it on the altar of the defence of the Patrie, placing it under the aegis of compulsory secular education, consummating the separation of Church and state in the socialization of pupils, and subjecting infants and senior citizens to a whole series of preventive measures and

70   J.-F. Bayart controls, while proving incapable of keeping unemployment down and thereby forcing a growing number of young people, including young couples, to continue living with their parents or to fall back on financial assistance from them. Furthermore, de Tocqueville and Foucault, in their different ways, have shown how the state has had a power of individualization with, naturally, effects on kinship relations. We still need to realize that the alchemy of these interactions is decidedly complex. It cannot be reduced to a utilitarian or functionalist vision of society, and even less to an ideological reading of its ‘control’, or its ‘discipline’, as a misinterpretation of Foucault’s work sometimes suggests, or indeed to the conspiracy theories that have always been so attractive in colonial, para-­ colonial (in the cases of China, Thailand, Iran or Turkey) and post-­colonial situations. From this point of view, the formation of ethnic (or ethno-­confessional or ethno-­national) awareness has given rise to a considerable number of misunderstandings. It must be emphasized, to begin with, that we do need to call it ‘ethnic awareness’, not ‘ethnic group’: ethnicity is a matter of awareness, not a structure or an actor – nobody has ever met an ethnic group, just people who identify themselves as belonging to what they call (or what is generally called on their behalf ) an ethnic group, which is quite different. And identification takes place by contextual, relational and relative definition: one defines oneself (or is defined) as belonging to an ethnic group within a given situation, in relation to an Other, without this excluding other lines of identification, whether religious, social or sexual. And these days, it is well known – thanks to the work of historians and anthropologists – that ethnic awareness has emerged in the context of the formation and centralization of the state (whether imperial, colonial or national), instead of being prior to it and constituting a ‘primordial’ resilience or resistance on the part of society to state interference. In Turkey, in Iran, the ethno-­national, ethnic or tribal awareness of Kurds, Sahvesan or Baluchi people is inseparable from this monarchical or imperial process and the piecing together of nations that arose from its ruins after the First World War (van Bruinessen 1978; Tapper 1997; Salzman 2000). In Lebanon, confessionalism took shape at the time of this very same transition from the imperial Ottoman world to the problems posed by the nation-­state and the mandate of the League of Nations. In India, the ‘communalism’ of Hindus and Muslims was an effect of the political triangulation between those two confessions and the colonial state: the Governor General of India from 1772 to 1784, Warren Hastings, constructed the ‘Hindu’ and ‘Muslims’ identities as juridical categories in the domains of civil and penal law, and formalized them by reifying them and rendering them legally performative; bureaucratic institutionalization and the centralization of the sectarian or brotherhood field into clericalized religions contributed, together with electoral practices, to the crystallization and politicization of cultural identifications in a national mode – with Partition – and then in a communitarian mode in which rioting was common (Bayly 1985; Thapar 1989; Jaffrelot 1993). In Central Asia and the Caucasus, the policies of the Russian Empire and, later, the USSR, underlay the configuration of the ethno-­national galaxy that constitutes the

Do social struggles make a difference?   71 special charm of that region of the world and that was broadly reproduced once individual countries became independent: the Muslim socialist ex-­Republics emerged from the Stalinist theory of nationalities (Roy 1997). And, last but not least, ancient Africa was not constituted as a mosaic of ethnic groups (Colson 1969). Ethnogenesis and the ‘creation of tribalism’ (Vail 1989) are directly linked to the history of the colonial state, either because the moment of its formation ethnicized people’s sense of political or social belonging, as in Burundi, or because, in the mode of ethnicity, it gave a fixed meaning – of identity and geography – to fluid ‘chains of society’ (Amselle and M’Bokolo 1985; Dozon 1985) that acted as channels for exchanges of merchandise as well as of religions and cultures. Indeed, ethnicity, as a historical phenomenon, is not an invariant even when it takes a turn that is radically identity-­based: being a Hutu or a Tutsi in Rwanda did not mean the same thing before and during the colonial occupation that racialized ethnic awareness and conferred on it unprecedented material aspects – in landed estate, in monetarization, in bank credit, in commercial techniques, in access to knowledge; nor was it the same before and after the ‘historical traumatisms’ (Vovelle 1982) of the ethnic purification of 1959–61 or the genocide of 1994. And being a Burundi Hutu when you have fled to Tanzania is not the same thing if you live in a camp or in a village among the native population: the refugees in the camps nurse a messianic and twilit nationalism that is frowned upon in the city (Malkki 1995). Being a Tutsi, finally, does not stop one from being at the same time rich or poor, young or old, male or female, educated or uneducated: ethnicity is no substitute for social relations, nor does it necessarily transcend them – pace its organic intellectuals, Western orientalists or the political engineers of identity and its mobilizations. In short, ‘ethnic groups have a history’ (Chrétien and Prunier 1989), and this history is inseparable from that of the expansion of the capitalist mode of production, colonial ‘development’, labour migrations and other kinds of migration, the asymmetries of national economies and ‘economic nationalism’, socialism, statism, the application of programmes of structural adjustment, and policies of economic liberalization (see, for example, Harries 1994 and Piot 1999 on the relationship between the formation of ethnicity and the phenomenon of migration, and Metcalf 2007: ch. 2 on how the ways of engineering identity forged in the British Raj spread across Malaysia). The limit of the utilitarian tradition of the social sciences is that it reduces the historicity of this ethnogenesis to a matter of rational calculation, political or economic in nature, or indeed to a manipulation on the part of the colonizers or their heirs, in accordance with the old logic of ‘divide and conquer’. There certainly is a political economy of ethnic awareness since this latter is linked with the sharing out of, claim on, or appropriation of national resources by social groups, and is historically a manifestation of the universalizing of the capitalist market, notably in lineage societies (Geschiere 1994). At stake in so-­called ethnic conflicts are the most ‘modern’ things: land, banks, education, the voting system, the state. But it is going too far – or, more precisely, not far enough – to assert, as does Robert H. Bates (1983: 152, 164–5), that ‘ethnic groups are, in

72   J.-F. Bayart short, a form of minimum winning coalition, large enough to secure benefits in the competition for spoils but also small enough to maximise the per capita value of these benefits.’ The reality is a little more complex. For one thing, the ‘politics of the belly’ (Bayart 1989) in which ethnicity plays a part is a cultural ‘style’ as much as a rational choice, and it implies other registers – aesthetic, fantasmatic, hedonistic, symbolic – than merely the instrumental quest for profit or interest, whether the latter be symbolic or cultural. Likewise, in India, the political entrepreneurs of the British Raj may have instrumentalized communalist identifications to the point of actually creating them so as to constitute banks of votes for themselves by introducing electoral procedures, according to the heavy-­handed thesis of Paul Brass (1974, 1985, 1997); and yet, it still needs to be pointed out that they were broadly followed by their electors, and that they themselves doubtless believed in their own words: communalism is also a repertoire in the imaginary realm. The fundamental criticism that needs to be made of the ‘rational choice’ approach is not that it overestimates the economic sphere, or the financial or material spheres, but that it overestimates utility. This latter is merely one motor of action among others. The peoples of the world do not think, as they shave every morning, of what they are going to do that day. For another thing, the crystallization and the reproduction of ethnic awareness have occurred in a state of great confusion – that of ambivalence – rather than in the transparency of calculation. They are part of the operatory misunderstanding that skewed the colonial ‘encounter’ or ‘dialogue’ (Irschick 1994; Peel 2000; Bertrand 2005). In John Iliffe’s elegant summary (1979: 324), ‘Europeans believed Africans belonged to tribes; Africans built tribes to belong to.’ Likewise, Hindu nationalists contributed to the crystallization of communalism in accordance with the ‘strategic syncretism’ with which they opposed what they saw as the dual menace of Islam and Christianity (Jaffrelot 1993). Local intellectuals – scholars, lawyers, clerics of the different confessions, Christian catechists, interpreters, informants, teachers – participated in the production of Africanist (or orientalist) and colonial knowledge (Bertrand 2005). Even the idea of ‘race’ – for example, of Aryanism (Ballantyne 2002) – was the object of a native appropriation which resulted for instance in the tragedies of Burundi and Rwanda. In other terms, if there was indeed a colonial hegemony – as I believe (Bayart 1989, 1996, 2004, 2005, 2007a, 2007b; Bayart and Bertrand 2006) – it rested on native ‘organic intellectuals’: Ogotemmêli, the celebrated informant of Marcel Griaule (1966) would have enchanted Gramsci, and so would Hampâté Bâ. It soon starts to become clear that ethnicity (or kinship) cannot be taken for granted and tends to evaporate on closer inspection, and it is highly unlikely that realities as elusive as those of awareness have such an overdetermining influence as culturalists think. Already Weber (1978: 393–5) was warning us: ‘All in all, the notion of “ethnically” determined social action subsumes phenomena that a rigorous social analysis . . . would have to distinguish carefully’ (1978: 394). He saw ethnic belonging as a ‘presumed identity’ (but he curiously stopped halfway,

Do social struggles make a difference?   73 attributing to the kinship group the capacity for ‘concrete social action’ while at the same time emphasizing its historical character) (1978: 395, 357). Furthermore, the culturalists’ orientalist or nativist concern for authenticity turns into a paradox. The categories it highlights stem essentially from the colonial period: from its knowledge, its type of administration, its procedures of legitimation. It is impossible to grasp the ‘tradition’ outside its imperial ‘invention’ in the nineteenth century (Hobsbawm and Ranger 1983). It is impossible to conceive of the family in contemporary Africa in abstraction from colonial hygienist medicine (Hunt 1999) and from Victorian Christian pastoral activity (Comaroff and Comaroff 1991, 1997). It is impossible to understand ethnicity if we ignore the ‘decentralised despotism’ (Mamdani 1996) of British ‘Indirect Rule’ and the compromise with the ‘grands commandements’ (the main coopted kingdoms or chiefdoms) in which the French Army in West Africa indulged, at the cost of a few emancipatory principles, or the ‘politics of race’ of William Ponty which replaced it in 1909. It is impossible to separate contemporary nativist demands in the Americas or in Africa from the Western racialist imaginary of the nineteenth century (Walker 2004). And, as we shall see, it is doubtless impossible to consider the Senegalese Islam of the brotherhoods without also taking into account the colonial and republican character imprinted on it ever since its alliance with the French administration, at the start of the twentieth century. The impasse into which ethno-­development wanders lies in the way it takes at face value the optical illusions of identity, and attributes to various forms of cultural pretence an explanatory virtue or a ‘social capital’ (Putnam 1993), ignoring the specific weight of social relations and of the historical path followed, as well as the ideological or political construction of ethnotypes. Just as ‘martial races’ owed a great deal to the imaginary and to the decisions of colonial armies, the ‘dynamism’ of certain ethnic groups – the Bamiléké, the Kikuyu – is an orchestrated phenomenon and conceals disparities, or indeed inequalities, as has been brought out in detail by Jean-­Pierre Warnier (1993) in connection with the ‘spirit of enterprise’ of the Grassfields of Cameroon: this spirit rests on a radical divergence between the social itineraries of the ‘older’ and the ‘younger’ members (aînés and cadets sociaux), and on the exploitation of the one group by the other. Just like developmentalist discourse (Ferguson 1990), culturalism then becomes a procedure whereby thoroughly political questions are depoliticized – questions such as the primitive accumulation of capital and the manufacture of social poverty or subordination, including the extent to which they have repercussions on the so-­called cultural aspects of society: being a ‘social youngster’ (cadet social) in a Bamiléké country, as in many other African societies, means that one is virtually excluded from the social institution of marriage. The (rather banal) irony of history lies in the fact that this depoliticization occurs by means of noisy political mobilizations on the electoral scene, or by means of ethnic operations, civil wars and even genocides. The contemporary political engineering of culturalism often consists in an alliance between the intellectual and the militia-­man, as was illustrated by the crises of Rwanda in

74   J.-F. Bayart 1994 or Côte d’Ivoire from 2000 onwards, during which academics provided the arguments for violence and the unemployed youngsters provided the strength of their muscles. But this engineering also tends to occur in the shape of a pact between the state, multilateral institutions, multinational enterprises and NGOs or local collectivities which take up the defence of ‘autochthonous’ or ‘minority’ cultures, for example in microcredit, the protection of the environment, and decentralized cooperation (Elyachar 2005). Militarist, jingoistic nationalism in the West or elsewhere has had a certain impact on the social question; likewise, ethnoculturalism, by promoting localist, ‘tribalist’, regionalist and nativist themes, and increasingly also by emphasizing the political problematics of ‘autochthony’ (Bayart et al. 2001; Geschiere 2009), inhibits or defuses the self-­ affirmation of the ‘unimportant’ in the form of an autonomous social movement. We must also remember that the culturalist discourse has been an extraordinary resource of legitimation for authoritarianism to the south of the Sahara (people often recall the well-­known African proverb, probably apocryphal, according to which ‘you can’t have two crocodiles in the same bit of swampland’), in Asia (see the cult of ‘Asiatic values’) or in Arabo-­Muslim countries (in accordance with the theme of the Raïs or the King, of ‘reformism’, and of the Islamic moral order: Tozy 1999; Hibou 2006). It has also provided techniques of control and political submission, such as the use of the family metaphor of ‘Father of the Nation’, neo-­traditional oaths of political allegiance in Kenya and Cameroon, pseudo-­rites of initiation in Tombalbaye’s Chad, and the bogus ‘authenticities’ of the Mobutu and Eyadema regimes in Zaire and Togo. It was thus without any surprise that we watched as this culturalist discourse served the authoritarian restoration that followed the political earthquake of the years 1988–91, for example in the form of the constitutionalization of ‘minority rights’, at least of those useful for dividing the opposition, by the intermediary of a decentralization that appropriated the mechanisms of colonial ‘decentralized despotism’, or indeed by recuperating ‘global’ and multilateral themes in the field of the environment (Bayart et al. 2001; Geschiere 2009). If we maintain the hypothesis of colonial hegemony, culturalism was one of the central elements in its reproduction in the aftermath of independence, and we are obliged to acknowledge that public (and private) aid for development contributed to this continuity. Politically correct naivety, from which post-­colonial studies are not immune, would mean confining the broadened reproduction of colonial culturalist hegemony to the situations of ‘the development of underdevelopment’, and making the former an explanatory factor of the latter. Indeed, neither India nor China, which remain in thrall to the nineteenth-­century imperial and orientalist engineering of identity, seem to have been economically handicapped thereby. Likewise, Malaysia, whose tiger-­like accomplishments are well-­known, has built up its political organization and its system of allocation of educational, economic and financial resources on the basis of a representation of Malay ‘autochthonous identity’ and a communitarizing of the populations of Indian and Chinese origin that is a pure legacy of the colonial government and the intermediary of the British Raj (Metcalf 2007: ch. 2).

Do social struggles make a difference?   75

Let’s have done with the idea of causality: the deus ex machina is dead! In short, culturalist thinking, especially in its ethno-­developmentalist register, continues to believe in causality even though, as we have seen, there is no longer any god, but merely machines which work more or less well. Curiously, the post-­modern anthropology of the 1980s, which granted so much importance to the imaginary and to the dispersion of social facts, in a neo-­Derridean, neo-­ Deleuzian and neo-­Foucauldian perspective, gradually returned to the old scientific superstition of causalism as it degenerated into cultural studies and post-­colonial studies, erecting the colonial moment into the alpha and omega of social action, transforming the lines of continuity of that moment into overdetermination, and granting the ‘postcolony’ a quasi-­ontological status (Bayart and Bertrand 2006; Bayart 2007a). And yet, the fetish author of the culturalists, Max Weber, who is quoted carelessly by people who have misread The Protestant Ethic and the Spirit of Capitalism, rejected the very idea of causality (Troeltsch 1912; Kalberg 2002). He simply investigated the ‘elements that encouraged capitalist expansion’ and ‘the development of a type of human being (Menschentum) created by the conjunction of components of religious origin and components of economic origin’, and was less interested in cultural representations than in ‘modes of life’ (Lebensführung) taken in their ‘daily aspect’ (Alltagsethik, Veralltäglichung) (Weber 1996). He emphasized ‘synchronic interactions’, but also ‘diachronic interactions’, which he defined either as a ‘legacy’, or as ‘antecedent conditions’ (Kalberg 2002: 206–48). He refused to impute a general causal priority to a limited number of factors, and even less to any single factor. In his view, the pluralism of the guiding stars of action had to be at the centre of any causal analysis. Max Weber thought in terms of historical experience or, more precisely, of historical matrix: ‘historical concepts . . . attempt for their methodological purposes not to grasp historical reality in abstract general formulae, but in concrete genetic sets of relations which are inevitably of a specifically unique and individual character’ (Weber 1992: 48). Nonetheless, it is a rule that what has come down from the past becomes everywhere what is considered as valid in the present. This relationship between past and present is fragmentary, latent, evolving, of long duration – in a word, it is contingent. It is contextual in nature, and it is in this sense that the very notion of causality is dangerous if it presupposes that the same causes produce the same effects. For the factors of causality count only in the singular configurations of given historical situations. Action configured by the past also involves a contingent interaction. The question of cultural ‘legacy’ then becomes a question of the context of action configured by the past. And the forms assumed by these ‘survivals’ (in Max Weber’s terms: Reste, Überbleibsel, verbliebener Rest) are easy to misappropriate, and paradoxical, instead of being a one-­to-one ‘causal connexion’ that restores ‘a single leading idea which ex hypothesi itself produces and shapes everything’ (Troeltsch 1912: 88). The ‘legacies’ thus tend to reproduce

76   J.-F. Bayart t­hemselves over time within a single field – for instance the religious or political field – but also tend to shift from one field to another, for example from the religious sphere, or from kinship, to the political or economic spheres. In the very different contexts of Central Asia and the south of Cameroon, Olivier Roy (1994) and Peter Geschiere (1994) have thus demonstrated how kinship relations turn out to be vectors of the market economy and also of political strategies for constituting rents of power. As everyone also knows, they represent classic repertoires of the expression of political relations, of their legitimation, and of obedience to authority (Bayart 1996, 2005; Schatzberg 2001). It is these shifts from one field to another in particular that explain why social change tends so easily to be ‘paradoxical’ and why ‘legacy’ and ‘antecedent conditions’ can never by themselves establish an adequate causality. These apparently highly abstract considerations are nonetheless of immediate use in reaching a better understanding of the problems of development. It has thus been possible to show how the ‘transition to the market economy’, in the variable contexts of structural adjustment in Africa, reforms in China, economic liberalization in Iran and India, or the collapse of the Soviet Empire, drew on representations and practices drawn from the fields of kinship, religion, or native ethical and political repertoires, as much as it drew on neo-­liberal ideology (Bayart 1994). Already, Braudel (1985: 72, 78) wrote that ‘capitalism does not invent hierarchies, it uses them, just as it did not invent the market or consumption’. ‘It is, in the long perspective of history, a belated visitor. It arrives when everything is already in its place.’ This relationship between innovation and the past requires a closer look. Basically, the methodological bias of multiculturalism lies in the way it freezes culture into its patrimonial aspect, seeing it principally as a heritage and forgetting that it is, simultaneously, a ‘combination of operations’ that are possible on the basis of the latter (de Certeau 1980). Furthermore, a culture is inseparable from its relation to the Other and the Elsewhere, from what I call its extraversion (Bayart 1996, 2000, 2005, 2006): an endogamous culture is a dead culture. And at this point we can see more clearly the double trap into which culturalism so easily falls. On the one hand, it thinks in terms of identifying what already exists (what is ‘culture’ in this society?) and in the Malthusian terms of its preservation, rather than in terms of virtualities (what might this culture become?) and ways of defending its capacity for creation, with all the attendant social and political conflicts. On the other hand, it thinks in terms of a zero-­sum game between the global and the local, and thereby cannot understand what any Senegalese pedlar knows as he surveys the world from behind his fake Ray-­Bans: universalization proceeds by appropriation and by the ‘reinvention of difference’, and not by uniformization. If we accept the non-‘leading’ idea that neither ethnicity nor kinship will ever be the ‘golden keys’ that open up development, any more than religions will save it or than the ‘network’ is the demiurge of globalization (Bayart 2004: 180ff. and 2007b), is it still possible to discuss something that was quite correctly emphasized by those who rejected the ethereal and voluntarist economism

Do social struggles make a difference?   77 of the projects and other programmes of structural adjustment? Basically, the proponents of the ‘Cultural Turn’ in development insisted on reminding everyone that every society is a raritas, in the Roman sense of the word, that it needs to be understood as such in view of the economic imperative, and that it can find within itself the motors of its growth without being predestined to stagnation and dependency. True enough. They can also be credited with the irreducibility of the ‘cultural’ or ‘symbolic’ or ‘imaginary’ dimension, call it what one will. Nonetheless, the heavy-­handed pressure of the orientalist and nativist point of view (these are the two sides of the same coin, namely of the colonial moment: Walker 2004) often leads them to transform into an invariant what is first and foremost a place of innovation, of creation, of production, and to subsume the contradictions of interests, conflicts between actors, dissonances of representations or values under a single federative, all-­embracing principle that remains forever ungraspable. They diminish the complexity of the social sphere and take refuge behind a few icons. The true subject is not culture, but history. The real challenge is not the respect for or defence of culture (when will we be finally rid of the Maginot mentality?), but that of its freedom to create. The real, and thoroughly political, stake is the way this freedom of creation of culture, and of the capacity to produce history, are shared out in society.

Conclusion: kinship, ethnicity, status, or institution and social movement? A few months ago, at a meeting, a politician from a West African country, who today occupies a prominent position, said after my presentation, in that sub-­ Saharan French whose redoubtable ambiguity fascinates me: ‘He’s explained things so well to me that now I’m all confused.’ This is the best one can hope for from someone working in the social sciences, and I could not have picked up a better compliment, however ironically it was meant. But the shaking of certainties can open the way to a relatively clear narrative. Resorting to the culturalist explanation has the drawback of ignoring the concrete historical situations that have mediated and constituted the economy of development: for instance, in the case of sub-­Saharan Africa, the slave trade, the ‘industrious revolutions’ (de Vries 1994) of so-­called legitimate trade, the colonial occupation, and the endogenous political, religious or social mobilizations, as well as the more or less ­inegalitarian social relations that have structured those situations, such as (proto) national or trans-­imperial migrations, urbanization, merchandization, salarization, monetarization, relations between different age groups, genders, statuses, and social classes, and their territorialization in precise places such as the market, the sanctuary, the mission, the holy or unholy city, the frontier, the port, the plantation, the mine, and the factory. In its radical form as cultural studies (and in particular post-­colonial studies) it sticks to types of discourse, knowing, of course, that these are performative, but forgetting that this Word is then made flesh – political, administrative, economic, juridical (etc.) flesh: the definition of identities, even if colonial, is not merely part of the imaginary, or a literary

78   J.-F. Bayart fantasy; it has effects (or limitations) that are quite material and social, it develops on the basis of possibilities and, conversely, of impossibilities. In short, the culturalist explanation is resorted to by those with lazy minds, at least insofar as one has not exhausted the other kinds of social, economic, political and demographic explanation. This is what, in conclusion, I would like to show, with the aid of the success story of groundnuts in Senegal. To be sure, this cash-­crop is these days in crisis, and everyone knows why. But the fact remains that it experienced its moment of glory and represented a fine story of ‘development’. How can we do it justice? Was it due to the colonial era that put an end to the ‘immobility’ and ‘repetitiveness’ of African peasant life? But the groundnut cash-­crop and the pioneering development of the land preceded the public policies of the French administration and have remained largely independent of it (Searing 2002), just as happened with the cultivation of palm oil in the Gulf of Benin or of cocoa in the Gold Coast and Côte d’Ivoire. Was it due to culture – in this case Islam – and can we write our The Islamic Ethic and the Spirit of Peanuts? But not all Muslims devote themselves with such zest and endurance to agricultural labour, and the non-­Muslim Serer people also took part in growing groundnuts. By kinship? But the social force, and more precisely the labour force that underlies this economic adventure, was organized precisely on a mode of substitution for established kinship relations, thanks to submission to a non-­kin, the marabout – if you like, it was based on the invention of a fictive kinship of a spiritual nature between the marabout and his talibé, which does not exclude the dynastic principle in the reproduction of the religious hierarchy . . . and of its factional conflicts! The audience for ‘fashionable marabouts’ among young people arises these days, inter alia, from the possibility it gives them of legitimizing matrimonial alliances by short-­circuiting the relations of authority and the conventions observed by families, as well as the splits in status between those belonging to a certain caste, former captives, former free men or former nobles (Audrain 2004; Havard 2005). By ethnicity? But ethnic awareness in Senegal is of ‘low intensity’. The Wolofization of society is neither exclusive nor nativist and proceeds rather by the inclusion and diffusion of a style of life than by the argument of autochtony and rejection (Havard 2005). Finally, the organizational motor of the groundnut industry among the Wolof has been the Murid brotherhood, to which they do not all belong – far from it – rather than ethnicity per se. Here, historical sociology will introduce other parameters into the analysis: social institutions; state formation; historical regions (terroirs); the local, regional or global political economy; and the uncertainties of social and political struggles, i.e. the contingency of events. In itself, none of these explanatory factors is sufficient, nor does it enable us to escape as if by magic from the demons of culturalism. It is, for example, easy to pinpoint a series of social institutions to a greater or lesser degree implicated in ‘development’ – the bazaar, the waqf (the Islamic mortmain), the ‘independent’ or charismatic churches, the camps of miners, etc. – but each of these tends to be interpreted in the hackneyed terms of culturalism and orientalism, as Arang Keshavarzian emphasizes with

Do social struggles make a difference?   79 regard to the bazaar in Iran (2007: 52ff.): in its cohesive heterogeneity, the bazaar is not the fruit of abstract or atemporal cultural beliefs, but of social relations between its merchants, their practices of credit, the political transformations of its governance, etc. – all elements which can, should the need arise, be interrelated with the links of family, ethnic group, region, religion, friendship and so on, but which are not to be confused with the latter any more than they can be reduced to the allegedly natural alliance between business and the mosque that is supposedly the basis for the social and political mobilizations that the country has experienced since the end of the nineteenth century. Likewise, the emphasis on a social movement can easily lead to a sterile dichotomous approach between ‘state’ and ‘society’ that is clearly illustrated by the classic Iranian opposition between dolat and mellat (also criticized by Keshavarzian 2007), and which has recently been systematized by the multilateral and a-­philosophical thematic of ‘civil society’ (Bayart 2004: 96ff.; 2007b): in fact, only the interlinking of the two institutions makes sense (for the paradigmatic case of Iran, see Keshavarzian 2007; Adelkhah 1998), and the concept of ‘civil society’ refers precisely to the relation between society and state, not to its exteriority with relation to the latter. The highly fashionable hold-­all notion of ‘network’, highlighted in particular by the New Economic Sociology (Granovetter 1985), inclines to neglect the social and historical relations of power, inequality and territoriality inherent in the economic actors’ ‘embeddedness’, even if it is able to explain their interaction with the political field (Keshavarzian 2007): for example, the transformations of the sociability of the bazaar, which this author emphasizes, are perhaps less due to the shift in its mode of ‘governance’ and the evolution of its constitutive networks following the change in regime, than to the general upheaval experienced by Iranian society as a whole (Adelkhah 1998 and 1999), a general upheaval which greatly affected the market networks themselves. In addition, this notion of ‘network’ is less a rigorous concept than an ideological figure of neo-­liberalism (Boltanski and Chiapello 1999). It is really and truly the combination of these factors – social institutions, the state, social struggles, historical regions – and the linking together, the concatenation of situations that, in the eyes of historical sociology, opens the door to the multicausal understanding of economic change. The key to the Senegalese enigma resides to a large degree, but not exclusively, in a social institution, that of the brotherhood (and in its innovative institutionalization at the interface of the colonial state), in the political pact that it has drawn up with the latter, and with its legatee, and in the process of emancipation of the servile workforce to which it has given a framework, an economic base and a political representation. In short, the agricultural revolution of groundnuts, its conversion into trade networks and real estate, its transnational and even global developments via smuggling, street trade and emigration, all rest on a social movement, one that is historically situated and has unceasingly renewed itself over the past century (Klein 1998; Searing 2002; Robinson 2004; Bayart 2007a). It corresponds to a complex and contingent sequence of events, it is a raritas. It is the expression of the historicity proper to Senegambia, which,

80   J.-F. Bayart in passing, illustrates the poverty of the dependentist approach that saw it as merely the manifestation of capitalist imperialism. This story can be told without the world ‘culture’ being uttered once, and we do not necessarily lose anything thereby. Try: it’s like giving up smoking – it’s hard to begin with, but it’s possible. Take a smoker’s word for it!

References Adelkhah, F. (1991) La Révolution sous le voile. Femmes islamiques d’Iran, Paris: Karthala. —— (1998) Etre moderne en Iran, Paris: Karthala (revised and augmented edition, 2006). —— (1999) Being Modern in Iran, London: Hurst and New York: Columbia University Press. —— (2003) ‘Iran: femmes en mouvement, mouvement de femmes’, in Bennani-­Chraïbi, Mounia and Fillieule, Olivier (eds) Résistances et protestations dans les sociétés musulmanes, Paris: Presses de Sciences Po, ch. 6. Amselle, J.L. and M’Bokolo, E. (eds) (1985) Au Coeur de l’ethnie. Ethnies, tribalisme et Etat en Afrique, Paris: La Découverte. Audrain, X. (2004) ‘Du ndigël avorté au Parti de la Vérité. Evolution du rapport religion/ politique à travers le parcours de Cheikh Modou Kara (1999–2004)’, Politique africaine 96, December, pp. 99–118. Ballantyne, T. (2002) Orientalism and Race: Aryanism in the British Empire, Basingstoke: Palgrave Macmillan. Bates, R.H. (1983) ‘Modernization, Ethnic Competition and the Rationality of Politics in Contemporary Africa’, in Rothschild, D. and Olorunsola, V.A. (eds) State versus Ethnic Claims: African Policy Dilemmas, Boulder: Westview Press. Bayart, J.F. (1989) L’Etat en Afrique. La politique du ventre, Paris: Fayard. —— (ed.) (1994) La Réinvention du capitalisme, Paris. —— (1996) L’Illusion identitaire, Paris: Fayard. —— (2000) ‘Africa in the World: a History of Extraversion’, African Affairs, 99 (395), April, pp. 217–67. —— (2004) Le Gouvernement du monde. Une critique politique de la globalisation, Paris: Fayard. —— (2005) The Illusion of Cultural Identity, London: Hurst and Chicago: University of Chicago Press. —— (2006) L’Etat en Afrique. La politique du ventre, Paris: Fayard (revised and enlarged edition). —— (2007a) ‘Les chemins de traverse de l’hégémonie coloniale en Afrique de l’Ouest francophone. Anciens esclaves, anciens combattants, nouveaux musulmans’, Politique africaine, 105, March, pp. 201–40. —— (2007b) Global Subjects. A Political Critique of Globalization, Cambridge: Polity Press. Bayart, J.F. and Bertrand, R. (2006) ‘De quel “legs colonial” parle-­t-on?’, Esprit, December, pp. 134–60. Bayart, J.F. and Warnier, J.P. (eds) (2004) Matière à politique. Le pouvoir, les corps et les choses, Paris: Karthala. Bayart, J.F., Geschiere, P. and Nyamnjoh, F. (2001) ‘Autochtonie, démocratie et citoyenneté en Afrique’, Critique internationale, 10, January, pp. 53–70.

Do social struggles make a difference?   81 Bayly, C.A. (1985) ‘The Pre-­history of “Communalism”? Religious Conflict in India, 1700–1860’, Modern Asian Studies, 19 (2), pp. 177–203. Bertrand, R. (2005) Etat colonial, noblesse et nationalisme à Java. La Tradition parfaite, Paris: Karthala. Bing, W. (2003) A l’Ouest des rails, Paris: MK2 Editions. Boltanski, L. and Chiapello, E. (1999) Le Nouvel Esprit du capitalisme, Paris: Gallimard. Brass, P. (1974) Language, Religion and Politics in North India, Cambridge: Cambridge University Press. —— (1985) Ethnic Groups and the State, London: Croom Helm. —— (1997) Theft of an Idol. Text and Context in the Representation of Collective Violence, Princeton: Princeton University Press. Braudel, F. (1985) La Dynamique du capitalisme, Paris: Arthaud. Brown, P. (1992) Power and Persuasion in Late Antiquity: Towards a Christian Empire, Madison: University of Wisconsin Press. Carré, O. (1991) L’Utopie islamique dans l’Orient arabe, Paris: Presses de la Fondation nationale des sciences politiques. Castoriadis, C. (1975) L’Institution imaginaire de la société, Paris: Editions du Seuil. Cheddadi, A. (1980) ‘Le système de pouvoir en islam d’après Ibn Khaldûn’, Annales ESC, 3–4, May–August, pp. 534–50. Chehabi, H.E. (2006) Distant Relations: Iran and Lebanon in the Last 500 Years, London: Tauris. Chrétien, J.P. and Prunier, G. (eds) (1989) Les Ethnies ont une histoire, Paris: Karthala. Clifford, J. (1988) The Predicament of Culture. Twentieth Century Ethnography, Literature and Art, Cambridge, MA: Harvard University Press. Colson, E. (1969) ‘African Society at the Time of the Scramble’, in Gann, L.H. and Duignan, P. (eds) Colonialism in Africa. 1870–1960. Volume I: The History and Politics of Colonialism, 1870–1914, Cambridge: Cambridge University Press. Comaroff, J. and Comaroff, J. (1991, 1997) Of Revelation and Revolution. Volume I: Christianity, Colonialism and Consciousness in South Africa. Volume II: The Dialectics of Modernity on a South African Frontier, Chicago: University of Chicago Press. Davis, D. and Harrell, S. (eds) (1993) Chinese Families in the Post-­Mao Era, Berkeley: University of California Press. De Boeck, F. (2004) ‘La frontière diamantifère angolaise et ses héros mutants’, in Bayart, J.F. and Warnier, J.P. (eds) Matière à politique. Le pouvoir, les corps et les choses, Paris: Karthala, ch. 3. De Boeck, F. and Plissart, M.F. (2004) Kinshasa. Tales of the Invisible City, Tervuren: Ludion. de Certeau, M. (1980) L’Invention du quotidien. Volume I: Les Arts de faire. Paris: Union générale d’édition (collection 10/18). de Vries, J. (1994) ‘The Industrial Revolution and the Industrious Revolution’, Journal of Economic History, 54, pp. 240–70. Dozon, J.P. (1985) La Société bété. Histoire d’une ‘ethnie’ de Côte d’Ivoire, Paris: Karthala and ORSTOM. Elyachar, J. (2005) Markets of Dispossession. NGOs, Economic Development, and the State in Cairo, Durham, NC: Duke University Press. Ferguson, J. (1990) The Anti-­Politics Machine. ‘Development’, Depolitization, and Bureaucratic Power in Lesotho, Cambridge: Cambridge University Press. Geschiere, P. (1982) Village Communities and the State. Changing Relations among the

82   J.-F. Bayart Maka of Southeastern Cameroon since the Colonial Conquest, London: Kegan Paul International. —— (1994) ‘Parenté et argent dans une société lignagère’, in Bayart, J.F. (ed.) La Réinvention du capitalisme, Paris: Karthala, ch. 4. —— (1995) Sorcellerie et politique en Afrique noire. La viande des autres, Paris: Karthala. —— (2009) Political Autochthonies, Chicago: University of Chicago Press. Granovetter, M. (1985) ‘Economic Action and Social Structure. The Problem of Embeddedness’, American Journal of Sociology, 91, November, pp. 481–510. Griaule, M. (1966) Dieu d’eau. Entretien avec Ogotemmêli, Paris: Fayard. Gulliver, P.H. (1971) Neighbours and Networks. The Idiom of Kinship in Social Action among the Ndendeuli of Tanzania, Berkeley: University of California Press. Harries, P. (1994) Work, Culture, and Identity. Migrant Laborers in Mozambique and South Africa, c. 1860–1910, Portsmouth: Heinemann. Havard, J.F. (2005) Bul Faale! Processus d’individualisation de la jeunesse et conditions d’émergence d’une ‘génération politique’ au Sénégal, Lille: Université de Lille-­2, multigr. Hibou, B. (2006) La Force de l’obéissance. L’économie politique de la répression en Tunisie, Paris: La Découverte. Hobsbawm, E. and Ranger, T. (eds) (1983) The Invention of Tradition, Cambridge: Cambridge University Press. Hsu, M.Y. (2000) Dreaming of Gold, Dreaming of Home. Transnationalism and Migration between the United States and South China, 1882–1943, Stanford: Stanford University Press. Hunt, L. (ed.) (1989) The New Cultural History, Berkeley: University of California Press. Hunt, N.R. (1999) A Colonial Lexicon. Of Birth Ritual, Medicalization, and Mobility in the Congo, Durham, NC: Duke University Press. Iliffe, J. (1979) A Modern History of Tanganyika, Cambridge: Cambridge University Press. Irschick, E.F. (1994) Dialogue and History. Constructing South India, 1795–1895, Berkeley: University of California Press. Jaffrelot, C. (1993) Les Nationalistes hindous. Idéologie, implantation et mobilisation des années 1920 aux années 1990, Paris: Presses de la Fondation nationale des sciences politiques. Jameson, F. (1998) The Cultural Turn. Selected Writings on the Postmodern, 1983–1998, London: Verso. Kalberg, S. (2002) La Sociologie historique comparative de Max Weber, Paris: La Découverte. Kernen, A. (2004) La Chine vers l’économie de marché. Les privatisations à Shenyang, Paris: Karthala. Keshavarzian, A. (2007) Bazaar and State in Iran. The Politics of the Tehran Marketplace, Cambridge: Cambridge University Press. Klein, M. (1998) Slavery and Colonial Rule in French West Africa, Cambridge: Cambridge University Press. Kopytoff, Igor (1986) ‘The Cultural Biography of Things: Commoditization as Process’, in Appadurai, Arjun (ed.) The Social Life of Things. Commodities in Cultural Perspective, Cambridge: Cambridge University Press, ch. 2. —— (ed.) (1987) The African Frontier. The Reproduction of Traditional African Societies, Bloomington: Indiana University Press.

Do social struggles make a difference?   83 —— (1988) ‘The Cultural Context of African Abolition’, in Miers, S. and Roberts, R. (eds) The End of Slavery in Africa, Madison: University of Wisconsin Press, ch. 17. Litvak, M. (1998) Shi’i Scholars of Nineteenth-­Century Iraq. The ‘Ulama’ of Najaf and Karbala, Cambridge: Cambridge University Press. Liu, X. (2000) In One’s Own Shadow. An Ethnographic Account of the Condition of Post­Reform Rural China, Berkeley: University of California Press. Malkki, L.H. (1995) Purity and Exile. Violence, Memory, and National Cosmology among Hutu Refugees in Tanzania, Chicago: University of Chicago Press. Mamdani, M. (1996) Citizen and Subject: Contemporary Africa and the Legacy of Late Colonialism, Princeton: Princeton University Press. Mauss, M. (1950) Sociologie et anthropologie, Paris: Presses Universitaires de France. Mengin, F. (1998) Trajectoires chinoises. Taiwan, Hong Kong et Pékin, Paris: Karthala. —— (2007) ‘La présence chinoise au Cambodge. Contribution à une économique politique violente, rentière et inégalitaire’, Les Etudes du CERI, 133. Metcalf, T.R. (2007) Imperial Connections. India in the Indian Ocean Arena, 1860–1920, Berkeley: University of California Press. Mueggler, E. (2001) The Age of Wild Ghosts. Memory, Violence, and Place in Southwest China, Berkeley: University of California Press. Needham, R. (ed.) (1977) La Parenté en question. Onze contributions à la théorie anthropologique, Paris: Editions du Seuil. Ong, A. (1999) Flexible Citizenship. The Cultural Logics of Transnationality, Durham, NC: Duke University Press. Ortner, S.B. (1999) The Fate of ‘Culture’. Geertz and Beyond, Berkeley: University of California Press. Peel, J.D.Y. (2000) Religious Encounter and the Making of the Yoruba, Bloomington: Indiana University Press. Piot, C. (1999) Remotely Global. Village Modernity in West Africa, Chicago: University of Chicago Press. Polignac, F. de (1984) La Naissance de la cité grecque, Paris: La Découverte. Putnam, R.D. (1993) Making Democracy Work. Civic Traditions in Modern Italy, Princeton: Princeton University Press. Robinson, D. (2004) Sociétés musulmanes et pouvoir colonial français au Sénégal et en Mauritanie. 1880–1920. Parcours d’accommodation, Paris: Karthala. Rocca, J.L. (2006) La Condition chinoise. La mise au travail capitaliste à l’âge des réformes (1978–2004), Paris: Karthala. Rosaldo, R. (1993) Culture and Truth. The Remaking of Social Analysis, Boston: Beacon Press. Roy, O. (1985) L’Afghanistan. Islam et modernité politique, Paris: Editions du Seuil. —— (1994) ‘En Asie centrale: kolkhoziens et entreprenants’, in Bayart, J.F. (ed.) La Réinvention du capitalisme, Paris: Karthala, ch. 3. —— (1996) Groupes de solidarité au Moyen-­Orient et en Asie centrale, Paris: Centre d’études et de recherches internationales. —— (1997) La Nouvelle Asie centrale ou la fabrication des nations, Paris: Le Seuil. Salzman, P.C. (2000) Black Tents of Baluchistan, Washington, DC: Smithsonian Institution Press. Schatzberg, M.G. (2001) Political Legitimacy in Middle Africa. Father, Family, Food, Bloomington: Indiana University Press. Schildkrout, E. (1978) People of the Zongo. The Transformation of Ethnic Identities in Ghana, Cambridge: Cambridge University Press.

84   J.-F. Bayart Searing, J.F. (2002) ‘God Alone is King’. Islam and Emancipation in Senegal. The Wolof Kingdoms of Kajoor and Bawol, 1859–1914, Portsmouth: Heinemann. Seurat, M. (1989) L’Etat de barbarie, Paris: Le Seuil. Tapper, R. (1997) Frontiers Nomads of Iran. A Political and Social History of the Shahsevan, Cambridge: Cambridge University Press. Thapar, R. (1989) ‘Imagined Religious Communities? Ancient History and the Modern Search for a Hindu Identity’, Modern Asian Studies, 23 (2). Thompson, E.P. (1963) The Making of the English Working Class, London: Victor Gollancz. Tozy, M. (1999) Monarchie et islam politique au Maroc, Paris: Presses de Sciences Po (revised and enlarged edition). Troeltsch, E. (1912) Protestantism and Progress: the Significance of Protestantism for the Rise of the Modern World, London and New York: Williams & Norgate, J.P. Puttnam’s Sons. Vail, L. (ed.) (1989) The Creation of Tribalism in Southern Africa, London: James Currey. van Bruinessen, M. (1978) Agha, Shaikh and State: On the Social and Political Organization of Kurdistan, Risjwijk: Uroprint. Veyne, P. (1988) Did the Greeks Believe in their Myths? An Essay on the Constitutive Imagination, Chicago: University of Chicago Press. Vovelle, M. (1982) Idéologies et mentalités, Paris: Maspero. Walker, C.E. (2004) L’Impossible retour. A propos de l’afrocentrisme, Paris: Karthala. Warnier, J.P. (1993) L’Esprit d’entreprise au Cameroun, Paris: Karthala. —— (1999) Construire la culture matérielle. L’homme qui pensait avec ses doigts, Paris: Presses universitaires de France. Weber, M. (1978) Economy and Society. An Outline of Interpretive Sociology, edited by Guenther, R. and Claus, W., Berkeley: University of California Press. —— (1992) The Protestant Ethic and the Spirit of Capitalism, trans. Talcott Parsons, London and New York: Routledge. —— (1996) Sociologie des religions, Paris: Gallimard. Wedeman, A.H. (2003) From Mao to Market. Rent-­Seeking, Local Protectionism, and Marketization in China, Cambridge: Cambridge University Press.

Part II

Religion, family and ethnicity

4 Economic underdevelopment in the Middle East The historical roles of culture, institutions and religion Timur Kuran Introduction The belief that Islam inhibits economic development has enjoyed currency at least since the nineteenth century, when rapid economic growth in Western Europe and its cultural offshoots in North America made Muslims appear economically underdeveloped. Although political correctness now discourages the public expression of ‘culturally insensitive’ judgments, in off-­the-record conversations Islam is routinely treated as harmful to economic creativity and progress. Factors invoked as sources of retardation include fatalism, personalism, laziness, lack of curiosity, mistrust of science, superstition, conservatism and traditionalism. In each case, a certain attitude or frame of mind is alleged to be rooted in ‘Islamic culture’. None of these assorted claims can be rejected out of hand. Consider fatalism, which is a tendency to view outcomes as governed by fate rather than human initiative. Fatalism in this sense is integral to Muslim communities, whose members commonly qualify their expressions of intention with the phrase ‘Inshallah’, literally ‘if it is God’s will’. Although the phrase is not always used in a literal sense, there are individuals who act as though a higher being directs their lives and sets their destiny. Those who flaunt traffic rules may be convinced that if their time has arrived no amount of precaution will matter. Yet fatalism is present in every society, and it is correlated with religiosity in general, not just adherence to Islam.1 In any case, fatalism has been present among Muslims from the dawn of Islam, including a long period when the Middle East appeared economically advanced by global standards. Hence, insofar as it has been a source of backwardness, one must explain why its effects on the Middle East’s relative economic standing are now particularly pronounced. That task requires the consideration of factors beyond the attitudinal effects of Islam itself. Not all of the factors invoked as cultural obstacles to Middle Eastern economic advancement can be dismissed so easily. In relation to the economically advanced countries, personal connections play a more significant role in the Middle East’s commercial life. This pattern is consistent with surveys that find Middle Easterners to have unusually low trust in others. In the World Values

88   T. Kuran Survey of 1990, only 10 per cent of the Turks answered ‘yes’ to a question about whether ‘most people can be trusted’. Among the 43 mostly wealthy countries included in the survey, the proportion was lower only in Brazil; the weighted average of all countries was 35 per cent.2 Given that economic modernization has involved a movement from personal to impersonal exchange, this contrast suggests that observed shifts in regional economic standings may well have something to do with differences in attitudes towards strangers. Insofar as people mistrust others, they will exchange primarily within networks of acquaintances, giving up potentially profitable exchanges with strangers. Could the personalism of Middle Eastern economic relations have a basis in religion? The traditional Islamic legal system recognizes only real persons, denying standing to corporations, which are fictitious persons. In the absence of corroborating oral testimony, it also denies independent evidentiary value to documents. So Islamic law may well have fostered personalism among Muslims. In requiring exchanges to be conducted with natural persons, and discouraging the formation of impersonal organizations, it might have shaped the attitudes that surveys now identify as general mistrust. In proposing how Islam might have brought about the observed personalism of Middle Eastern economic life, I am invoking Islamic law as the fount of attitudes harmful to economic growth. Specifically, I am suggesting that the mistrust harmful to trade among strangers is among the consequences of certain institutions central to societies that have lived, historically, under Islamic rule. That, in a nutshell, is this chapter’s thesis. Attitudes commonly treated as economically limiting Islamic cultural traits are dependent, rather than independent, variables. They are among the legacies of mutually reinforcing social institutions that played key roles in past economic successes, but also in subsequent economic disappointments. Our focus will be on two clusters of factors frequently invoked as economically hazardous cultural characteristics. The first entails attitudes that breed immobility directly: conservatism, traditionalism, fatalism, lack of curiosity and low ambition. The second consists of personalism, in other words, the tendency to favour personal exchange over impersonal exchange. Both personalism and immobilizing attitudes are rooted, I shall argue, in Islamic institutions that attained their classical forms around the start of the second millennium. Culture is a concept that commands a bewildering array of meanings, and the same can be said about institutions. To make analytical progress on the controversy under consideration, it is essential to be clear about what these concepts will mean. Here I am defining culture as a pair of distributions that jointly provide a distinct communal identity: a distribution of attitudes and a distribution of behaviours. This focus on distributions accommodates the heterogeneity that every society shows in both attitudes and actual choices.3 An institution is a socially generated regularity of behaviour.4 The institutions that affect a certain class of behaviours form an institutional system. Thus, the laws of a country form an institutional system whose elements include customary legal procedures, rules of promotion within the judiciary, attitudes regarding the role that religion

Economic underdevelopment   89 should play in adjudication, and a hierarchy of courts. By itself, none of these elements captures the social function of a legal system, which is much more than the sum of its parts.5 Both attitudes and behaviours may enter into the nexus of factors forming a particular institutional system. The latter concept thus subsumes what I am treating as culture. Having defined these concepts, let us now critique some influential arguments that ascribe the Middle East’s economic shortcomings to the culture of Islam.

Conservative attitudes During the institutional transformation that turned the West into an economic, political and military powerhouse, and concomitantly the Middle East into a laggard in all these respects, certain knowledgeable observers of the Islamic world inferred that the culprit is the latter’s dominant religion. Islam stands for unchangeability, they proposed. Lord Cromer, the British governor of Egypt from 1883 to 1907, argued that Islam is un-­reformable, because it promotes a mindless attachment to tradition. Practicing Muslims cannot pursue economic modernization, he claimed, as they are bound to consider the requisite innovations sinful.6 In the late nineteenth century the French scholar Ernest Renan advanced a complementary view. Islam and its Arab propagators have always been hostile to science and philosophy, he claimed. Knowledge advanced under Arab domination only until Persian and Hellenic influences faded. Once Islamic culture freed itself of foreign influences, it quashed open inquiry in an effort to suppress findings incompatible with Islamic beliefs. Renan understood that the essence of all religion is acceptance of certain claims on faith. But Islam stands in a class by itself, he thought, because in refusing to compartmentalize life, ‘it extends the rule of dogma to the temporal realm’.7 In principle, Islam seeks to control not just faith, worship and morality but the whole of human civilization, including governance, science and economics, realms that Christianity eventually recognized as secular. As a matter of practice, however, Islam’s span of control has always been limited. Arab scientists who made huge advances in optics and metallurgy were not expected to ground their findings in the Qur’an. Likewise, Muslim merchants of the early Islamic centuries hardly relied on the Qur’an, or on other religious teachings, in deciding when, where, how or what to trade. Even if Islam rejected a formal separation between ‘mosque and state’ or ‘religion and science’ or ‘faith and economic reality’ there were always many realms that it effectively treated as secular. Insofar as religious control hindered the Middle East’s economic advancement, understanding its record requires, then, specifying why the freedom to make changes was more limited in certain areas than in others. Neither Cromer nor Renan explored why Muslim practices proved more flexible in the early centuries than later, in military technology than in philosophy, and in Ottoman Turkey than in the holy cities of Mecca and Medina. Vastly oversimplifying the

90   T. Kuran historical record, they considered ‘Islamic culture’ monolithic and eternally fixed, when in fact it has harboured much diversity and movement. Cromer and Renan lived well before intellectuals began to shy away from criticizing foreign ways of life openly. But their views have endured. Scholarly debates carry arguments that treat Islam as a fount of economically hazardous attitudes.8 None of the commonly expressed arguments explains why the effectiveness of the alleged attitudes, if in fact they have been identified correctly, has varied across time, places and contexts. As we shall see, the variations are related to the trajectory of Middle Eastern institutions. But before we turn to institutions, let us take up other popular arguments centred on attitudes.

Lack of curiosity In his best seller What Went Wrong? (2002), Bernard Lewis argues that the Middle East fell behind because Islam discouraged curiosity about civilizations it treated as inferior. Early on, Lewis observes, Muslims learned to look down on Europeans on the grounds that they followed a superseded religion. Medieval Arabic writings about Europe depict it as a remote and unexplored wilderness inhabited by exotic and essentially primitive people from whom there was nothing to fear and little to learn. The attitude of disdain was reinforced by Arab military victories in Spain and, later, Turkish conquests in the Balkans. Among the attitude’s manifestations is that few Muslims considered it useful to learn European languages. The perception of European inferiority gradually hardened, making Muslims oblivious to various European innovations that were ushering in the capabilities and comforts of modern life. People cannot adapt to developments that they cannot observe. If Muslims failed to embrace European scientific, intellectual, political and economic advances, a major reason, says Lewis, is that they lacked an awareness of them. In sharp contrast, European Christians took an interest in Muslim practices early on, for they saw Islam as a formidable rival. They wanted to understand Muslims so as to determine how best to eliminate a danger, just as centuries later Americans poured resources into the study of communism in order to deal with the Soviet threat. European studies of the Middle East were spearheaded by monks, who wanted to save Christian souls from conversion to Islam, keep territories from falling under Muslim rule and, in the long run, reclaim Jerusalem and convert Muslims to Christianity. The asymmetry between learning about the ‘other’ is evident in the production of foreign-­language dictionaries and grammar books. By the late eighteenth century Europeans had produced dozens of books on the major languages of Muslims. Yet, Arabs, Persians and Turks had not produced a single book on any Western language. What makes the asymmetry particularly puzzling is that the Middle Eastern disinterest in learning about Europe continued in the face of mounting evidence of economic slippage vis-­à-vis the West. Prior to the nineteenth century, practically no Middle Eastern ruler showed awareness of the new economic values and institutions that enabled Europe’s growing might, and none

Economic underdevelopment   91 commissioned the translation of an economic treatise.9 Middle Eastern rulers took an interest mainly in the weaponry that helped to defeat Muslim armies, in evolving techniques of shipbuilding, and in new medicines. When a civilization that has enjoyed dominance starts falling behind, one expects it to produce diagnoses. Indeed, in the seventeenth and eighteenth centuries increasing numbers of Muslims took to speculating about the ongoing erosion of their global power. On the economic front, they noticed budget deficits, markets taken over by foreigners, and inflation. Typically they attributed such economic difficulties to internal factors such as corruption, oppression, greed and high taxes. Things had gone wrong at home, not right elsewhere. And domestic problems had arisen because institutions supportive of past glories had eroded. The appropriate remedy, most thought, was to rediscover and restore institutions that worked effectively in the past. Meanwhile, incumbent European institutions were being challenged in the name of progress. However, until well into the nineteenth century the Middle East spawned no major movements pursuing radical social change. No one thought of examining European institutions for clues to solving domestic problems. Lewis’ observations raise the question of why the imperviousness to European developments persisted. There were two barriers, he suggests, both involving the practice of Islam. The first was the system of education. Run by religious functionaries, it taught people a finite set of information, rather than how to use their own judgment, evaluate claims critically, and make decisions for themselves. More so than in Europe, schools made students memorize facts and venerate classic texts, so their graduates saw knowledge as something to be acquired rather than discovered and expanded. The scientific method, which relies on observation and experiment, failed to take hold. The second barrier to recognizing the advantages of reforms was the so-­called ‘closure of the gate of independent judgment’. In early Islam, scholars and theologians more or less freely developed answers to problems that scripture and tradition left unresolved. Their innovations helped sustain an internal engine of growth and contributed to the fluidity of Muslim worldviews. However, between the ninth and eleventh centuries, freedom of innovation ended when it was declared that independent judgment was no longer permissible – in the traditional formulation, ‘the gate of ijtihad ’ was closed. This closure meant that all answers were already available: thenceforth, one needed only to follow and obey. In treating Islamic learning as having attained perfection and the Islamic world as self-­sufficient, it gave legitimacy to values, attitudes and practices that discouraged inquisitiveness. Fostering an educational system based on rote learning and memorization, it also fuelled a culture that limited curiosity about the outside world. Historians have not identified a time when Muslims assembled and declared all problems solved. Nor does Lewis claim otherwise. He knows that the decline in creativity occurred through a decentralized process rather than by decree. Yet he does not elucidate the mechanism that drove the decline. If the end of

92   T. Kuran i­ndependent judgment was widely accepted, what accounts for the shift? Why did the malcontents within Muslim communities – as in any society, they existed – not see the spread of conservatism as a source of deepening problems? What, precisely, restrained the impulses of people who stood to gain from introducing one innovation or another? Even if the prevailing educational system limited inquisitiveness, it could not have extinguished the desire for change. So the Lewis thesis consists of observations rather than a complete causal explanation. A satisfactory explanation of obstacles to the Islamic world’s quick economic recovery must make sense of why they could not be overcome through individual or collective action. It must also put forth the mechanism that kept Muslims who came in contact with Western Europe from appreciating the huge transformations underway. Intellectuals were not the only group in contact with foreigners. Many merchants had dealings with their counterparts across the Mediterranean. What Went Wrong? does not explore why groups outside of intellectual circles accepted the freezing of Islamic institutions. There is a methodological bias at work here, which is common in scholarship focused on ideas. It entails the idealist fallacy, whose most prevalent form treats all history as the history of thought.10 In the Lewis variant, innovations emerge exclusively from producers and guardians of knowledge, including scholars, teachers and religious functionaries. In reality, people of all walks of life are capable of identifying and demanding useful innovations. Credit-­constrained merchants will pressure officials to create ways to circumvent restrictions on interest. Peasants unable to pay their taxes will communicate their frustrations by petitioning state officials, revolting or migrating; such responses may stimulate fresh thinking on the part of intellectuals, but also policy-­makers and landowners. In brief, the starting point of an innovation need not be the mind of an intellectual. The demand for change may emerge from any part of a social system, just as anyone can contribute to the supply of an innovation. Hence, a satisfactory explanation of why the Middle East fell behind economically must go beyond the identification of attitudes originating from intellectuals and religious leaders. A full explanation must deal with the understandings, incentives and opportunities of every social group. As we shall see, broadening the analysis in that direction will require the consideration of institutions.

Personal exchange The weaknesses of the attitudinal arguments discussed thus far are present also in arguments that treat personal exchange as the main cultural factor responsible for the Middle East’s observed economic failures. This point may be developed with reference to the works of Abraham Udovitch, a leading economic historian of the Middle East. Among the puzzles that Udovitch addresses, one pertains to the stagnation of the Middle East’s financial system. In the Middle Ages the region had a sophisticated financial system by the standards of the day. Yet this system did not

Economic underdevelopment   93 develop further, and by the middle of the second millennium it looked underdeveloped vis-­à-vis its counterparts in Western Europe. A major obstacle to the Middle East’s financial development, suggests Udovitch, lay in the ‘social context of the region’s economic life’.11 In economic relations, he says, personal relations and reputation played central roles. The peoples of the region operated within networks in which they were known, rarely doing business with strangers. It is precisely because personal relations mattered so much, Udovitch goes on, that banks did not emerge prior to the nineteenth century. By construction, banks are impersonal entities that intermediate among borrowers and depositors unknown to one another. For the region’s financiers to develop banks, they would have had to extricate themselves from the ‘personalism’ of Islamic culture. But the culture proved far too powerful. In the medieval era, Middle Easterners received financial services from other individuals rather than from firms. They borrowed from money lenders working on their own account or as members of small and temporary partnerships enjoying no legal standing. The parties to financial disputes were always individuals. Because borrowers and lenders operated within networks, this atomistic financial system worked remarkably well for the time. But the very factor that facilitated financial relations also prevented the transformation of atomistic credit suppliers into banks. With business restricted to small and intimate circles, large organizations specializing in credit, deposit taking, and lending simply could not form. In the Middle Ages there were no banks anywhere, not even in Western Europe, which was to spearhead financial modernization. Credit relations were as personalistic in Florence and London as they were in Aleppo and Tunis. So the personalism that Udovitch presents as a key ingredient of Islamic economic culture is not unique to the Middle East. Where the regions differ is that exchanges became progressively less personal in Western Europe, even as they remained highly personal in the Middle East. Thus, personalism need not be a fixed attribute of an economic system. The types of relations that prevail in a region’s financial markets can become less personal over time. Indeed, the move from personal to impersonal exchange is among of the most conspicuous manifestations of the institutional transformation that generated the modern global economy.12 Regional differences in the ‘social context of exchange’ do not explain fully, then, the divergence that took place between the Middle Eastern economy and that of the West. Although the ‘social context’ was critical, this is not for the reasons Udovitch offers. The cultural elements that he includes in the ‘social context’ proved adaptable in certain respects but not in others. Hence, to link Middle Eastern cultural features, or Islamic principles, with the region’s observed economic trajectory, one must identify why as a matter of practice the degree of flexibility was highly variable. Consider the large body of Islamic law on the regulation of commerce. Called muamalat (literally, ‘commercial relations’ or ‘transactions’) and usually ­translated as the ‘law of sale’, it provides intricate rules for conducting voluntary transactions. According to the law of sale a transaction must be free of arrangements construable as unjustified enrichment, defined as obtaining an advantage

94   T. Kuran without yielding something equivalent in value.13 To ensure that everyone gives up an equivalent amount, all parties must know everything knowable about the goods being exchanged. Transparency is vital to commercial legitimacy, for ignorance in any respect can lead to unjustified enrichment.14 Moreover, no exchange is to take place until the goods are identified and measured precisely. Thus, the fruits growing on an orange tree cannot be sold legitimately, because the number and quality are unknown. They can be sold only after they have been picked, counted, weighed and individually examined. The logic of this ban is to prevent disappointment. No one pays too much or receives too little, thus minimizing the possibility of discord. To prevent unjustified enrichment, it is generally necessary also to have the buyer and seller hold a face-­to-face meeting. The two sides must look each other in the eye and indicate through a gesture or handshake that the deal is satisfactory. In practice, of course, not all transactions involved a formal face-­to-face meeting. Nor did people take the time to inspect fully and carefully every good being traded. Exhaustive inspections would have raised transaction costs to the point of eliminating the gains from trade. Exchanges often took place within networks of acquaintances. The members of these networks considered it a waste of time to conduct the detailed inspections demanded by law. Trusting their trading partners, they did not worry about getting cheated. This is not to say that they knew all the knowable specifics of the exchanged commodities. However well shielded against cheating, they traded in a condition of ignorance about the goods changing hands. Since surprises remained possible, they broke the Islamic law of sale routinely. One might have expected the legal community to take measures aimed at improving enforcement of the law of sale. To the contrary, it tried to adapt interpretations of the law to realities of the marketplace. An instructive example lies in a treatise of the renowned legal scholar Shamseddin al-­Sarakhsi (d. 1090), the Book of Sale. Sarakhsi notes that it is illegal to order a pair of shoes from a shoemaker. This is because you would be negotiatiating over something nonexistent, and the sale of a nonexistent object is not permissible. The seller cannot sell what is not in his possession, and the buyer cannot inspect an object that remains to be produced. Does it follow that made-­to-order shoes should be prohibited? Sarakhsi answers in the negative, on the ground that people have been buying made-­to-order shoes happily from time immemorial. The underlying criterion is that a practice is legitimate if it is common and causes no problems. To justify this criterion he invokes a saying attributed to the Prophet: ‘Anything which Muslims consider appropriate is also appropriate in the eyes of God.’15 Sarakhsi thus finesses the stringent requirements of the law of sale by appealing to social custom. If something is commonly done, he holds, then it is legitimate. Here is another example of this criterion at work. In the medieval Middle East people bathed not at home but at public bathhouses (hamams), paying the fee up front. In principle, the practice violated the Islamic law of sale, because at the time of payment, the prospective bather did not know the quality of service. Would the water be sufficiently warm? Would the bathhouse be too crowded?

Economic underdevelopment   95 Under a strict interpretation of the law of sale, all bathhouses would have to close. The only way to legitimize their operations would be to seek refuge in customary practice. In fact, Sarakhsi justifies the practice of charging for a bath on the ground that it has endured for centuries. A practical thinker, Sarakhsi understood that many forms of economic exchange would cease if one insisted on having both sides to a contract get themselves fully informed at the outset. If bathers do not mind paying a fee up front, they are content with the service, and they seem to be getting clean, it makes sense to leave the bathhouse sector alone. Presumably, if the services of a bathhouse decline, the number of bathers will decline; people cannot be cheated indefinitely. The key lesson is that Islamic law proved quite adaptable in certain contexts. The law of sale was applied pragmatically, allowing exceptions in certain contexts where the benefits of imperfect transparency appeared to outweigh the costs. So Islamic law was not necessarily applied rigidly, without regard for practical trade-­offs or for what we now call transaction costs. Likewise, attitudes regarding economic relations were not fixed in stone. By the same token, Islamic law failed to undergo various transformations that we associate with economic modernization. Most relevant here, it failed to generate from within the institutions necessary to achieve the transition from personal to impersonal exchange. To attribute this failure of the Middle East’s social context is to confuse explanans with explanandum, in other words, to treat as an explanation precisely what needs to be explained. The persistence of personalistic attitudes was not a manifestation of generalized inertia. It occurred even as other attitudes underwent refinements and adjustments. Islamic culture offers abundant examples of both change and rigidity. To sum up thus far, neither alone nor in combination do the suggested links between Middle Eastern economic failures and ostensibly fixed Islamic characteristics, such as fatalism, conservatism, lack of curiosity and proclivity to favour personal exchange, explain observed variations in the region’s economic performance. Each is a symptom of outcomes whose explanation must involve deeper social characteristics.

Enduring effects of classical Islamic law These deeper characteristics are the region’s institutions. Of particular relevance here are certain components of classical Islamic law, which was largely in place by the end of the first millennium. One key component is the law of Islamic commercial partnerships. Although Islamic partnerships assumed several variants, often they involved a contract between one or more sedentary investors with one or more travelling merchants, with profits meant to be split according to a predetermined ratio. Inspired by various pre-­Islamic institutions, an Islamic partnership resembled also certain partnerships that European merchants used throughout the Middle Ages. Islamic partnerships stand out as unusually flexible, in that they allowed the profit shares of the partners to depend on such factors as

96   T. Kuran the riskiness of the intended venture and the merchant’s reputation. These flexibilities were so advantageous that non-­Muslim traders and investors of the Islamic world, though under no compulsion to do business under Islamic law, commonly opted to do so anyway.16 Right up to the nineteenth century these partnerships consisted of temporary commercial enterprises with few members, often just two. Moreover, their forms barely changed even as European partnerships expanded in size and gained complexity.17 Another component of Islamic law, the religion’s distinct inheritance system, contributed to the stagnation of Islamic partnerships. Islamic inheritance rules require two-­thirds of any estate to be divided among members of the deceased person’s extended family. Although there were ways to circumvent the rules, on balance they raised the cost of maintaining enterprises across generations. Because adding members to a partnership increased the likelihood of its premature dissolution, the rules also discouraged the formation of large enterprises. The consequent failure to make the transition to large commercial partnerships critically influenced the region’s economic evolution. In particular, the local merchant community saw no reason to pressure local courts to create fundamentally new laws or to re-­interpret the existing legal code. Consequently, the region produced nothing akin to the auto-­catalytic process that in the West led, first to increasingly complex partnerships, and then, in the nineteenth century, to commercial and financial corporations. Among the manifestations of this divergence is that Europe began the transition from personal to impersonal exchange earlier than the Middle East. In the Middle East substantial transitional moves came in the nineteenth century, with the transplant of pertinent European institutions in the context of top-­down reforms. In the intervening years, exchange has become generally less personal in the Middle East; today, a much greater portion of the region’s exchanges are among mutual strangers who are matched by impersonal intermediaries such as banks and stock markets. Evidently, Islam is not incompatible with impersonal exchange.18 Yet, Islam’s medieval commercial infrastructure failed to bring about the transition from personal to impersonal exchange organically. It is this failure that necessitated fundamental reforms involving the transplant of institutions developed elsewhere. The region’s inability to generate modern commercial institutions from within also had a political consequence, with economic effects of its own. A persistently atomistic commercial sector failed to advance civil society, further limiting pressures for institutional reforms. The resulting inertia helped to make the region’s peoples seem apathetic, conservative and disinterested in outside developments. It also limited the feasibility of democratic rule, allowing authoritarian regimes to form and gain stability. Had the private economic sector developed large and durable organizations, political leaders would have found it harder to rule dictatorially. At a minimum, they would have sought to share power with merchants and financiers, thus stimulating the expansion of democratic liberties, including free speech and the right to dissent. Most critical here, Islam may never have been linked to inertia and resistance to change.

Economic underdevelopment   97 The weaknesses of civil society in the region are rooted also in another historically prevalent institution: the waqf, which is a distinctly Islamic form of trust.19 Founded through the personal act of an individual property owner, a waqf consists of property immobilized in order to finance a designated religious or charitable mission in perpetuity.20 Within the limits of Islamic law, the choice of a waqf ’s mission was up to its founder. Accordingly, from the early Islamic centuries to modern times, a vast array of social services were provided through the waqf system. To name a few of varying economic significance, waqfs were formed to build and maintain mosques, support inns on caravan routes, pay the taxes of designated neighbourhoods, finance schools, and maintain parks. In the great cities of the medieval Middle East waqfs supplied most of the urban services now provided largely by municipalities. They also contributed to poverty alleviation through public shelters and soup kitchens. Waqfs proliferated partly because they served also as wealth shelters. A founder could make himself his waqf ’s trustee for life, compensate himself for his services, choose the waqf ’s employees from among his relatives, and bequeath self-­conferred privileges to his descendants. Waqf assets were considered sacred, which gave them considerable security against confiscation.21 Rewarding waqf founders with material security thus turned out to be an ingenious way to promote philanthropy and reduce the state’s burden as a provider of social services. With some success states endeavoured to influence the most significant investments made through the waqf system. A dignitary planning to establish a major waqf would be advised to take account of the state’s strategic priorities. Nevertheless, the Middle East’s huge network of waqfs took shape largely through the uncoordinated decisions of individual property holders. One might have expected this network to stimulate the development of civil society, for waqfs did not depend on state handouts. By virtue of their financial independence, waqfs might have given private groups bases from which to assert democratic rights and impose checks on arbitrary government rule. Yet a waqf is not a corporation; once chartered, its mission cannot be amended, except through means of questionable legality. The consequent rigidity limited the waqf ’s contributions to the growth of civil society. Insofar as modern capitalism requires strong ‘intermediate’ organizations situated between the individual and the state, this rigidity ultimately retarded economic modernization by keeping waqfs inherently conservative and unresponsive to evolving social needs. If the Western experience offers any guidance, waqfs capable of genuine self-­ governance might have stimulated experimentation as well as reformist movements from below. In Europe, the emergence of greater and lesser corporate bodies in the early Middle Ages gave rise, over subsequent centuries, to new laws that gradually strengthened civil society. Lack of incorporation is also a striking feature of other social structures of the pre-­modern Middle East. Not even the region’s world-­class cities were chartered as corporations, and not until the 1850s did any boast a municipality. Although certain urban craft guilds exercised considerable autonomy in their internal affairs, the state reserved the right to meddle in their activities. Colleges were .

98   T. Kuran not free to alter their curricula in response to changing educational needs. Generally chartered as waqfs, they were required, in principle but to a degree also in practice, to teach what their founders had stipulated. Just as Islam’s inheritance and partnership rules jointly hampered the advancement of civil society by keeping commercial enterprises small and short-­ lived, its waqf rules and lack of incorporation opportunities limited civil society by making its philanthropic organizations rigid as well as politically impotent. A persistent legacy of these weaknesses consists of the region’s limited capacity to solve social problems privately, without direction from the state. Another is widespread apathy and complacence, which are defensive responses to feelings of powerlessness in the face of vexing problems. Attitudes commonly viewed as elements of a fixed Islamic culture are thus among the effects of specific institutional handicaps. When the problematic institutions were discarded, the associated attitudes would have started to fade away. If the transformation has been slow, the primary reason is that political structures accustomed to filling the void left by weak civil societies have had a stake in limiting the role of non-­state actors. Significantly, although Arab businesses and philanthropic organizations have now been able to incorporate for decades, Arab civil societies remain strikingly weak. There is still little in between Arab dictators and the masses they rule – no real labour unions, no influential human rights groups, no free press, and no parliaments with political power. That is why an Arab nation will descend into chaos when its dictator is removed from power – as the aftermath of the American occupation of Iraq so clearly shows. A complementary reason why the attitudinal transformation in question could drag on is that in a system of interconnected institutions the functioning of each particular institution depends on the others. The efficiency of large organizations requires, among other things, a competent and honest cadre of accountants, a legal system equipped to handle disputes within and among such organizations, and an impartial judiciary. By itself, then, the introduction of a law of corporations need not lead to a proliferation of corporations, or keep established corporations efficient. In the absence of a legal system capable of adjudicating corporate cases, people will be reluctant to form corporations, preferring to conduct commerce and finance through smaller, shorter-­lived and less complex forms of organization. The movement from personal to impersonal exchange will thus be delayed. Moreover, technologies whose efficient exploitation requires large-­scale organization will remain unexploited, strengthening the impression of a lethargic society. Such a scenario played itself out, in fact, from the mid-­nineteenth century onward, when Middle Eastern governments initiated chains of institutional reforms aimed at overcoming economic handicaps. Each reform transformed one element of the system within which the region’s merchants, producers, financiers and investors must operate. Until all the reforms ran their course, people did not start forming modern enterprises, except as legal protégés of a foreign power.

Economic underdevelopment   99

The primacy of Islamic institutions over ‘Islamic culture’ Having asked whether the Middle East’s economic shortcomings are somehow attributable to attitudes associated with Islam, I have ended up suggesting that the problems are rooted in distinctly Islamic institutions. Economically harmful attitudes and behaviours that are often treated as manifestations of Islam are not immutable traits independent of the region’s institutional evolution. Rather, they are lingering symptoms of past institutional deficiencies. Observers who ascribe the Middle East’s economic deficiencies to ‘Islamic culture’ maintain that conservatism, fatalism, personalism and incuriosity are more prevalent among Muslims than among non-­Muslims. They claim, moreover, that such growth-­inhibiting attitudes stem from basic Islamic teachings, making them essentially fixed. Therein lies the problem. Islamic civilization has had periods of remarkable dynamism. Besides, although personal exchange remains common in the Middle East by the standards of advanced countries, its share among all exchanges has fallen greatly since the Middle Ages. Today, the region’s banks, stock markets and credit card companies intermediate myriads of exchanges among parties unknown to one another. An institution, as defined here, consists of social factors that produce interconnected social regularities. Being part of a system, its operation may depend on other institutions. Islamic law, which for many centuries gave the Middle East a persistent identity, formed a system of institutions. Its elements directly relevant to the region’s economic development include a law of partnerships, inheritance regulations and the waqf. Individually and jointly, these elements of Islamic law delayed the transition from personal to impersonal exchange, discouraged mass production, kept civil society weak, and set the stage for sustained authoritarian rule. Their consequences have included the very traits that are commonly considered integral to ‘Islamic culture’. Institutions are not immutable, and neither are the associated cultural features. As institutions change, so does the associated culture. This is not to posit a monotonic relationship between the trajectory of any given institution and associated attitudes. Because the effects of each institution depend on many others, a law can be reformed, or a new law introduced, without any immediate impact on economic attitudes. At the same time, certain reforms may yield results in short order. Witness the proliferation of local Middle Eastern banks in the early twentieth century and the quick acceptance of financial relations with strangers. Once various institutions supportive of banking, including corporate law and modern accounting, fell into place, banking rapidly became an indispensable element of Middle Eastern economic life. Today, no one in the region objects to banking as an innovation incompatible with traditional Muslim life. Although Islamists would like banking to be at least cosmetically ‘Islamic’ it does not bother them that it violates the Islamic ‘law of sale’ fundamentally. Even Islamists are accepting of essentially impersonal relations between lenders and borrowers. Their attitudes towards credit relations bear no resemblance to those that a millennium ago Sarakhsi identified as the Islamic ideal.

100   T. Kuran

Notes   1 Inglehart et al. (1998: Table V151).   2 It may well be more common in certain economically advanced countries characterized by high religiosity than in certain predominantly Muslim countries. The World Values Survey of 1990 found Turks to be less religious than either Italians, Austrians or Americans, all nations that are economically more advanced. Participants were asked, ‘Generally speaking, would you say that most people can be trusted or that you can’t be too careful in dealing with people?’ (Inglehart et al. 1998: Table V94). If fatalism is not demonstrably more prevalent among Turks than among much richer Americans, it alone cannot explain why the Middle East is economically underdeveloped. For a range of studies that place Muslim attitudes within a global context, see Moaddel (2007).   3 It accommodates, for instance, the mix in religiosity found within every society, as well as the commonness of tensions between attitudes and behaviours. A society that is relatively religious on the whole may contain atheists. Its atheists may feel compelled to practice the dominant religion, whether to avoid isolation or to gain acceptance. The two distributions will never be independent of one another. Although attitudes are not the sole determinant of behaviours, they do exert an influence. The definition draws on Kuran and Sandholm (2008), where further details may be found.   4 This definition is drawn from Greif (2006: especially 35–45).   5 Institutions may be nested within broader, more comprehensive institutions. A country’s legal system is part of its system of government which belongs, in turn, to a system of global governance. Higher institutions constrain lower ones, but they are also shaped by them.   6 Cromer (1909, Vol. 2: especially 228–9).   7 Renan (1883).   8 After their time, Max Weber (1963: especially 263–5), and the modernization theorists who followed, such as McClelland (1961: 340), treated Islam as a source of attitudes harmful to economic development. More recently, Guiso et al. (2003) have offered partial empirical support for this view. Scholarly critiques of the Weberian thesis include Rodinson (1973), Turner (1974) and Noland (2005).   9 Lewis (1982: 196). Even Adam Smith’s Wealth of Nations went unnoticed. 10 Fischer (1970: 195–200). 11 Udovitch (1985: 272–3); see also Udovitch (1979). 12 This is a central theme of institutional economics. See, in particular, Greif (2006), North (2005) and Platteau (2000: chs. 6–7). 13 Udovitch (1985: 448). 14 Medieval legal treatises carried this insistence on transparency very far. For an illustration, suppose that a store owner sells a customer a water jug. After consummation of the sale, the buyer discovers a crack in the jar, unnoticed during inspection. The law gives the buyer a ‘right of rescission’: because the jug is now worth less to him, he may cancel the exchange unilaterally. Discovery of the crack is taken to imply that the store owner took money from the buyer without giving something of equivalent value in return. See Udovitch (1985: 449). 15 As cited by Udovitch (1985: 455–6). 16 Kuran (2004: sects. 5–7). See also Goitein (1967: especially 169–79). 17 Kuran (2003). 18 Significantly, this transition has taken place even as the share of Muslims has risen as a result of emigration and population exchanges. 19 The waqf has no basis in the Qur’an. It achieved identification with Islamic civilization about a century after the advent of Islam, at a time when the Arab-­Islamic Empire was governed from Damascus. 20 Kuran (2001), Powers (1999) and Lev (2005).

Economic underdevelopment   101 21 Waqf assets were sometimes confiscated, usually under the pretext that their founders had acquired them illegally. But at least until the eighteenth century such confiscations were far less common than those of private property, because officials were reluctant to appear impious. So the waqf ’s wealth-­sheltering function contributed enormously to its popularity, all the more so in times of political instability, when private property generally became less secure.

References Cromer, E.B. (1909) Modern Egypt, 2 vols. New York: Macmillan. Fischer, D.H. (1970) Historians’ Fallacies: Toward a Logic of Historical Thought. London: Routledge & Kegan Paul. Goitein, S.D. (1967) A Mediterranean Society, I: Economic Foundations. Berkeley: University of California Press. Greif, A. (2006) Institutions and the Path to the Modern Economy: Lessons from Medieval Trade. New York: Cambridge University Press. Guiso, L., Sapienza, P. and Zingales, L. (2003) ‘People’s Opium? Religion and Economic Attitudes’, Journal of Monetary Economics, 50, pp. 225–82. Inglehart, R., Basañez, M. and Moreno, A. (1998) Human Beliefs and Values: A Cross-­ Cultural Sourcebook. Ann Arbor: University of Michigan Press. Kuran, T. (2001) ‘The Provision of Public Goods under Islamic Law: Origins, Impact, and Limitations of the Waqf System’, Law and Society Review, 35, pp. 841–97. —— (2003) ‘The Islamic Commercial Crisis: Institutional Roots of Economic Underdevelopment in the Middle East’, Journal of Economic History, 63, pp. 414–46. —— (2004) ‘The Economic Ascent of the Middle East’s Religious Minorities: The Role of Islamic Legal Pluralism’, Journal of Legal Studies, 33, pp. 475–515. Kuran, T. and Sandholm, W.H. (2008) ‘Cultural Integration and Its Discontents’, Review of Economic Studies, 75, pp. 201–28. Lev, Y. (2005) Charity, Endowments, and Charitable Institutions in Medieval Islam. Gainesville: University Press of Florida. Lewis, B. (1982) The Muslim Discovery of Europe. New York: W.W. Norton. —— (2002) What Went Wrong? Western Impact and the Middle Eastern Response. New York: Oxford University Press. McClelland, D.C. (1961) The Achieving Society. Princeton: Van Nostrand. Moaddel, M. (ed.) (2007) Values and Perceptions of the Islamic and Middle Eastern Publics. New York: Palgrave Macmillan. Noland, M. (2005) ‘Religion and Economic Performance’, World Development, 33, pp. 1215–32. North, D.C. (2005) Understanding the Process of Economic Change. Princeton: Princeton University Press. Platteau, J.P. (2000) Institutions, Social Norms, and Economic Development. Amsterdam: Harwood. Powers, D.S. (1999) ‘The Islamic Family Endowment (Waqf )’, Vanderbilt Journal of Transnational Law, 32, pp. 1167–90. Renan, E. (1883) L’Islamisme et la Science (transcript of Sorbonne lecture, March 29, 1883). Paris: Calmann Lévy. Rodinson, M. (1973) Islam and Capitalism, trans. Brian Pearce. New York: Pantheon. Turner, B.S. (1974) Weber and Islam: A Critical Study. London: Routledge and Kegan Paul.

102   T. Kuran Udovitch, A.L. (1979) ‘Bankers without Banks: Commerce, Banking, and Society in the Islamic World in the Middle Ages’, in The Dawn of Modern Banking, edited by the Center for Medieval and Renaissance Studies, UCLA. New Haven: Yale University Press, pp. 255–73. —— (1985) ‘Islamic Law and the Social Context of Exchange in the Medieval Middle East’, History and Anthropology, 1, pp. 445–65. Weber, M. (1963) Sociology of Religion, trans. Ephraim Fischoff. Boston: Beacon Press.

Comments on Timur Kuran’s chapter Mona Makram-­Ebeid

The twentieth century witnessed the popularization of the notion of ‘Islamic economics’, which is the idea that economics should be structured according to the timeless principles of Islam. This ascendancy of Islamic economics in the Muslim world was triggered by the desire to restructure society according to so-­ called Islamic injunctions. However, discussions of contemporary Islam in the West tend, in general, to be framed in black-­and-white terms, with the religion depicted either as a backward, dangerous and hateful force or as a misunderstood and moderate foundation for peaceful living. Having said that, there is no doubt that discourses involving Islam are welcome in every hue, either because of their informative content or as manifestations of growing ideological controversies. Nevertheless, there are a few studies of exceptional value that stand out from the mass of publications and deserve to be read by every serious observer, even if one does not fully agree with their content. Timur Kuran’s chapter on economic underdevelopment in the Middle East is one of these few. In his chapter, he provides prominent examples of proposed linkages between culture and economic development, arguing that the suggested cultural obstacles do not elucidate the Middle East’s observed trajectory, at least not by themselves. He argues that attitudes inimical to economic progress were also present in periods when the region was remarkably creative and adaptable. This is unfortunately not the case with the present enforcement of the Islamic law of sale, which would amount, he argues, to replacing the institutions of modern economic conditions with ones that reflect the needs of a millennium ago, and would also harm productivity. The problem with these campaigns that aim at building a distinctly Islamic economic system is that they are inspired by cultural and specifically religious goals, rather than by efforts to stimulate economic life, whereas, he argued elsewhere, the inspiration for economic development must come primarily from outside Islam and Islamic precedents. Here, I would like to underline what to my mind is one of the major problems of economic underdevelopment: the weakness of civil society.

104   M. Makram-Ebeid As Kuran incisively argues, because the commercial sector remained atomistic, civil society failed to develop, limiting pressure from below for institutional reform, limiting the feasibility of democratic rule, which allowed authoritarian regimes to form and gain stability, and, most relevant here, limiting the expansion of democratic liberties including free speech and the right to dissent. In such a context, how can you expect any creativity, innovation or far-­ reaching re-­interpretation of Islamic laws? Kuran is strongest in his argument about the primacy of institutions suggestions that distinct Islamic institutions lie at the roots of the region’s slip into a state of economic underdevelopment – in other words, the historical mechanisms of institutional stagnations, as he once called it. His thesis is that the economic institutions of Islam, though well suited to the era in which they emerged, were poorly suited to a dynamic industrial economy. These institutions have fostered social equilibria that reduced the likelihood of modern development emerging from within Islamic civilization. The obstacles he has identified involve inheritance practices, contract law, court procedures, the absence of corporations, the financial system, and the delivery of social services. Islamic law, which for many centuries gave the Middle East a distinct and persistent identity, formed a system of institutions, including a law of partnerships, inheritance regulation and the waqf, all of which have delayed the transition from personal to impersonal exchange, discouraged the use of the technologies of mass production, kept civil society weak, and set the stage for sustained authoritarian rule. Kuran’s analysis of the waqf and the rigidity of its system, which limited its contribution to the growth of civil society, remaining conservative and unresponsive to changing social needs, is immensely insightful although he fails to mention that this particular institution is used by non-­Muslims as well. However, I believe that Kuran should have distinguished between Islamist thought and what is currently termed post-­Islamism, which in general refers to political and social conditions where (like the failed Saudi and Pakistani experiences he refers to), following a phase of experimentation, Islamism’s appeal, energy and sources of legitimacy are exhausted even among its once ardent supporters. Many Islamists become aware of their system’s anomalies and inadequacies. Eventually, pragmatic attempts to maintain the system reinforce an abandon of its underlying principles. Islamism becomes compelled, both by its own internal contradictions and by societal pressure, to reinvent itself, but it does so at the cost of a quantitative shift. Of course, this definition applies particularly to Iran, but new stirrings can be seen in many other Muslim countries. Moreover, the term ‘Islamist’ is no longer a functional unit of understanding because it encompasses too wide a variety of political forces. It is interesting to note that in Saudi Arabia, for example, Islamist groups have acted as both supporters and opponents of the process that produces a generation that does not trust institutions and is beginning to adopt an ideological and militant attitude.

Comments   105 Unfortunately, there has not been much debate about the extent to which Islamic political thought in many countries has moved over the past decades from unequivocal endorsement of Islamic rule to thoughtful consideration of how Islamic thought can dovetail with the practices, attitudes and social expectations that democracy and economic development entail. While Kuran makes clear his critique of the modern-­day implementation of Islamic economics, he fails to account for the unprecedented intellectual and theological ferment that is roiling Islam around the world for a new reading of the Qur’an. The craving for new thinking has led to an explosion of ijtihad, the Islamic practice of exerting one’s utmost to understand. Moreover, the interest has allowed a thousand ijtihads to bloom. Disillusionment with government and religious authorities is helping fuel a re-­examination of Islam. Restoring this interpretative relationship with the sacred scriptures is a key to many doors. Individuals who revitalize the faith through ijtihad unlock their own spiritual powers and take responsibility for their lives. When this occurs on a large scale, which is not yet the case, new political and cultural attitudes emerge and intellectual creativity blossoms. These long term effects are why the current devotion to ijitihad (mainly pursued by Muslims in Western countries) is likely to be the most important and enduring aspect of Islam’s contemporary revival. Closer security is therefore required to see the seeds of new thinking emerging. Beyond its implications for Islamic intellectual life, the new interest in Qur’anic criticism also harbours a long-­term potential for revolutionary social and political change. As Muslims begin to read the Qur’an with critical intelligence they see through the myths, the inconsistency with principles, and the cultural prejudices and literary devices imposed by humans on interpretations of the text. A new understanding of religion is the prerequisite of any social change. Muslims today are faced with a three-­sided ‘prison’ so to speak: an archaic Islamic past, a seductive Western future, and the problematic present. How to shed an ossified legalistic interpretation of Islamic teaching (as it is still taught at Al Azhar University, the main bastion of Islamic teaching) is one of today’s greatest challenges. In other words, Islam is the subject of intense conflict between different segments of the faithful. Women, youths, the middle classes, the poor, the powerful, the modern, the traditional clerics, laymen are all engaged in redefining the truth of their creed through either ordinary daily practice or deliberate campaigns. In doing so, they render religion a plural reality with multiple meanings. The agents of historical change are not theories but the decisions taken by Muslims at every level of society as they strive for a better life for themselves and their children, and as millions of young Muslims crave to participate as equals in the global economy. The pertinent question is how and under what conditions can Muslims legitimize and popularize an inclusive reading of their doctrine in the same way that democrats in the West have been struggling to broaden narrow notions of democracy when it applied only to white, male, propertied people.

106   M. Makram-Ebeid In my opinion, the theoretical envelope that Kuran constructs to contain and elucidate his observations is broad enough to challenge all serious readers. His insights are relevant to many different situations, from the parochial to the transnational and from the quotidian to the philosophical. I hope it will challenge other authors to use a similar combination of real-­world observation, analytical sophistication, and constructive purpose.

5 Family and kinship ties in development An economist’s perspective Eliana La Ferrara

Introduction The role of family and ethnic ties varies considerably across societies. Some countries have relatively small and cohesive nuclear families (e.g. Italy, Spain); others have a comparably small but more ‘independent’ organization (e.g. Scandinavian countries), other still have very big extended families with a pervasive role (e.g. most of Africa, Asia and Latin America). Why family structures have evolved in such a different way is a fascinating question which goes beyond the scope of the present chapter. This chapter will focus on the economic effects of different family structures, and in particular on the extent to which societies with more extended (or stronger) family ties benefit or lose from these ties in a strictly economic sense. I will adopt a microeconomic perspective, starting from the smallest unit of analysis – the nuclear family and the parent/child relationship – and then move to the broader notions of kin and ethnic group.1 Before proceeding, a terminological clarification is in order. The notions of ‘family’, ‘kin’, ‘clan’ and ‘ethnic group’ are quite complex and much debated upon in the anthropological literature. In this chapter, I will use the word ‘family’ to broadly refer to any form of blood relationship (hence encompassing very proximate and very distant ones), and use the terms ‘nuclear family’, ‘extended family’, ‘clan’, ‘kin group’ and ‘ethnic group’ when I need to employ narrower definitions. In that case, ‘nuclear family’ will identify a group of two parents and their children; and ‘extended family’ will include close relatives (such as grandparents, aunts/uncles and cousins) often living in close proximity. I will use the term ‘clan’ to identify a unilineal group of relatives living in one locality, and the term ‘kin group’ to denote a collection of various clans that comprises ‘socially recognized relationships based on supposed as well as actual genealogical ties’ (Winick 1956:  302). To define ‘ethnic groups’, on the other hand, I will refer to Max Weber’s definition: groups that entertain a subjective belief in their common descent because of similarities of physical type or of customs or both, or because of memories of colonization and migration. This belief must be important for group

108   E. La Ferrara f­ ormation; furthermore it does not matter whether an objective blood relationship exists. (Weber 1978: 389) The chapter will start by analysing families as the origin and vehicle for the transmission of values and social norms. I will discuss the economic benefits as well as the potential distortions and inefficiencies generated by different types of family structures and norms. I will then move to analyse families as substitutes for markets, in particular their role for information transmission and enforcement of informal agreements, and discuss again the pros and cons of their involvement from an economic point of view. The last section will briefly conclude with some open issues and policy implications.

Families, values and social norms Cultural transmission One of the fundamental roles played by parents is that of transmitting distinct cultural traits to their children. These traits include elements of preferences such as the degree of altruism, risk aversion, attitudes towards fertility and labour force participation, religious traits, etc. Bisin and Verdier (2001) model the transmission of cultural traits from parents to children and derive predictions on the equilibrium composition of the population. They assume that parents are altruistic and choose the cultural trait they transmit by anticipating the effects that different traits will have on their children’s welfare. However, parents evaluate their children’s welfare with their own preferences, i.e. they derive higher utility from the fact that their children share their own culture (imperfect empathy). In equilibrium, socialization choices differ depending on whether family and society are substitutes or complements. If they are substitutes, parents belonging to a minority group will socialize their children more intensely, leading to a preservation of heterogeneous cultural traits in the population. On the contrary, when they are complements, majority groups opt for more intense socialization, leading to homogeneity in the long run. Several empirical papers have tried to assess the extent and the determinants of cultural transmission among generations. Bisin et al. (2004) analyse religious identification and intermarriage patterns in the United States and find that Protestants, Catholics and Jews exhibit a strong preference for children who share their own religion. Fernandez et al. (2004) find that preferences on female labour force participation and educational investment are transmitted from mothers to sons. Giuliano (2007) finds that parents tend to reproduce the pattern of living arrangements common in their country of origin when they move to a different country. Looking at second-­generation immigrants in the United States, she finds that children tend to live with their parents longer if their culture of origin in Europe exhibits that pattern.

Family and kinship ties in development   109 The socialization choices that occur within the family are extremely important for economic development, insofar as crucial decisions such as fertility, investment in education and labour force participation depend on these choices. But they also dramatically affect the ability of a society to solve collective action problems and sustain cooperative norms. Tabellini (2008) extends the model of Bisin and Verdier (2001) to study situations in which individuals are randomly matched with others to play a prisoner’s dilemma game. The scope for cooperation, i.e. the set of matchers in which cooperation can be sustained, depends on economic incentives and on the values that individuals have been taught by their parents. Parents choose whether to socialize their children to cooperative behaviour based on environmental conditions (e.g. the quality of legal enforcement) and on the expected pattern of future transactions. In a similar vein, Guiso et al. (2008) model the intergenerational transmission of beliefs regarding the trustworthiness of other individuals. They propose an overlapping generations model in which children receive from their parents a prior on trustworthiness of others, they participate in exchanges where they can be cheated or realize gains from trade, and then transmit an updated prior to their own children. The authors show that in equilibrium, parents transmit conservative priors in order to protect their children from future losses. However, this behaviour is suboptimal from the social point of view, leading to the possibility that societies are trapped in a low trust equilibrium. Families as constraints While the transmission of cooperative norms may be univocally seen as a benefit deriving from the stronger involvement of the family in an individual’s life, a recent literature has identified a number of side effects, or constraints, generated in the same way. I will start by addressing constraints that affect both industrialized and developing countries, and then move to some issues that are specific of the latter group of countries. In a recent paper, Alesina and Giuliano (2007) investigate the relationship between the strength of family ties and participation in market economic activities. They quantify differences across countries in the strength of family ties by using a set of questions contained in the World Value Survey. The first asks ‘how important the family is in one’s life’ on a scale of 1 to 4. The second question ask respondents if they agree with the following statements: 1 ‘Regardless of what the qualities and faults of one’s parents are, one must always love and respect them’; or 2 ‘One does not have the duty to respect and love parents who have not earned it.’ The third question asks if the respondents agree with one of the following statements:

110   E. La Ferrara 1 ‘It is the parents’ duty to do their best for their children even at the expense of their own well-­being’; or 2 ‘Parents have a life of their own and should not be asked to sacrifice their own well being for the sake of their children.’ The authors aggregate all these variables through principal component analysis into one measure of ‘weakness of family ties’, the values of which are displayed in Figure 5.1. As the index is decreasing in the strength (increasing in the

Ireland Uruguay France Slovenia Bulgaria Hungary New Zealand Australia Slovakia Azerbaijan Ukraine Croatia Belgium Great Britain China Latvia Republic of Czech Republic Luxenbourg Japan Russian Fed. Iceland Greece Switzerland Austria Norway Finland Estonia Sweden Belarus Netherlands Germany Lithuania

Figure 5.1  Weakness of family ties (source: Alesina and Giuliano (2007: 41)).

1.2

Algeria

1

0.8

0.6

0.4

0.2

0

0

–0.1

–0.2

–0.3

–0.4

–0.5

–0.6

–0.7

Nigeria Zimbabwe Egypt Venezuala Philippines Malta Bosnia and Herz. Indonesia Pakistan El Salvador Morocco Vietnam Jordan Macedonia Singapore Algeria Tanzania Chile Mexico Uganda South Africa Montenegro Brazil Serbia Albania India Colombia Poland Argentina Georgia Peru Italy USA Dominican Rep. Spain Turkey Canada Iran Taiwan Moldova Bangladesh Portugal

Family and kinship ties in development   111 weakness) of family ties, smaller values correspond to societies in which family links are perceived as more important. It can be seen from Figure 5.1 that most developing countries have relatively strong family ties (left panel), while OECD countries tend to have weaker ones (right panel), with weakest ties in Lithuania, Germany, the Netherlands and Scandinavia. Alesina and Giuliano argue that maintaining strong family ties requires spending a lot of time at home (e.g. caring for children or elderly people, cooking meals at home, etc.) and this takes time away from market activities. Panel A of Figure 5.2 shows that women are more likely to participate in the labour force in countries with weaker family ties. Panel B shows that a similar pattern holds for youth, namely, young people are less likely to seek employment when they live in societies with stronger family ties. The latter result is consistent with a literature finding that in Mediterranean countries (Italy, Greece, Spain and Portugal) more than half of the people aged 18–33 live with their parents, while this figure is less than 30 per cent in France, Germany, the UK, the United States and Scandinavian countries (Giuliano 2007). This has ambiguous effects on aggregate welfare. On the one hand, increased reliance on the family for support may allow a greater number of people to achieve higher education. Manacorda and Moretti (2006) argue that increased income allowed Italian parents to offer higher household consumption to their children in exchange for their living at home. On the other hand, residence with one’s parents is associated with delayed entry into ‘adult life’, including delayed entry in the labour force, delayed age at marriage and lower fertility patterns, with obvious implications for pay-­as-you-­go pension systems. An issue that emerges in some of the above papers is that the strength of family ties may endogenously respond to different institutional settings or economic conditions. Alesina and Giuliano (2007) address this point in two ways. First, they look at second-­generation immigrants in the United States and how they respond to the strength of family ties in their country of origin, similarly to what other studies on cultural transmission have done. Second, they instrument family ties with a variable that captures whether the first-­person singular pronoun can be dropped or not. The intuition is that societies in which this cannot be dropped are more individualistic. Their results on the negative relationship between family ties and market participation are robust to these strategies. To the extent that the choice of living with one’s parents or not entering the labour force is voluntary, one may invoke a revealed preference argument and claim that despite the social inefficiencies that these choices may induce, they are still ‘optimal’ from the individual’s point of view. When looking at some developing countries, however, one gets a sense that there are instances in which the effects of family ties go beyond voluntary compliance by individuals, and are closer to the notion of ‘coercion’ than of ‘choice’. A well known example is the practice of dowries in South Asia. Dowry payments are transfers of wealth from the family of the bride to the groom and his family at the time of marriage. These transfers were a common feature of medieval

60 50 40 30 20 10 –0.5

0

0.5

1

Weak family ties Women LFP

Fitted values

100

90

80

70

60 –0.5

0

0.5 Weak family ties

Youth LFP

Fitted values

Figure 5.2  Family ties and labour force participation.

1

Family and kinship ties in development   113 Europe but declined with modernization, while they are on the rise in contemporary India, leading to the well-­known phenomenon of ‘dowry inflation’ (Anderson 2003). There is widespread concern about this phenomenon among policy-­makers and society as a whole, as the extent of payments is often so large as to lead to the impoverishment of families with one or more daughters to wed. An extreme consequence is that forward-­looking parents may resort to selective abortion or female infanticide in order to avoid dowry payments in the future. Other consequences that have been documented include bride-­burning or wife-­ beating, in case payments from the bridal family are considered insufficient (Bloch and Rao 2002). Dowries – and more generally marriage payments – also have distortionary effects on household investment decisions. In fact, to the extent that resources are transferred not to the couple but to the parents of the bride or groom, if the older generation has a lower propensity to save or to invest in the education of the couple’s children, these marriage practices may slow down the accumulation of physical and/or human capital. Edlund and Lagerlof (2004) suggest that the switch from arranged to love marriage in Europe starting in the eighth century may have been one of the causes of Europe’s economic success in the following centuries.2 Despite the fact that policy-­makers have outlawed marriage payments in several countries (see for example the Dowry Prohibition Act passed in India in 1961), the reason these norms persist is that they are part of a ‘market equilibrium’ in which families are the players and strategies of ‘conformity’ dominate unilateral deviations towards no payments. How to induce an equilibrium switch from the ‘everyone pays’ to the ‘nobody pays’ equilibrium is a dramatically important, and still open, question. Another important role played by families in marriage decisions is that of establishing at what age a girl should get married. Adolescent and child marriage is a common phenomenon in many developing countries, and there is widespread concern that women who marry too early tend to attain lower education levels, have worse health outcomes (including higher maternal mortality and domestic violence), and have lower bargaining power within their household (Jensen and Thornton 2003). In a very interesting paper on Bangladesh, Field and Ambrus (2005) estimate the loss of education caused by adolescent marriage in Bandgladesh. Their identification strategy relies on the variation in the timing of menarche as an instrument for age at first marriage (given that girls are typically withheld from marriage until they reach puberty). Their estimate suggest that postponing marriage by one year in the age range 11–16 would increase girls’ schooling by an average of 0.3 years, and increase adult literacy by 6.5 per cent. At the aggregate level, they estimate that if the government could enforce a ban on marriages below age 17, female schooling would increase by 9 per cent. These are sizeable improvements that call for policy action in this setting. Marriage payments are not the only type of social norm ‘imposed’ by family structure in some societies. Another important class of norms relates to inheritance rules. Broadly speaking, inheritance systems may be classified under two

114   E. La Ferrara dimensions. The first is the blood line along which property is transmitted. Historically, most societies have featured patrilineal inheritance, where property passes from fathers to sons. Some cultures, however, maintain matrilineal inheritance systems, where property is transmitted along the female line.3 Another distinction relates to the division of the property among heirs. At one extreme there is primogeniture, where the entire property goes to the eldest child (typically, son) of the deceased. At the other extreme, there is equal division, where each child inherits the same share. Intermediate forms of partible inheritance are also common. The evolution of inheritance rules over time is a fascinating topic. Platteau and Baland (2001) review inheritance practices in medieval Europe and in contemporary developing countries, contrasting primogeniture with partible inheritance. They argue that while some regions (typically, Asia) have resorted to primogeniture as a way of maintaining the integrity of the family real estate, in sub-­Saharan Africa there is a widespread attachment to the equal division norm. They anticipate that as land scarcity becomes more severe in the face of growing population pressure, traditional land tenure arrangements are likely to give way to market transactions. While it is not possible in the present context to analyse the evolution of inheritance rules as an endogenous response to economic conditions, it can be interesting to examine some of the implications of these norms for the allocation of resources within and between households. In particular, I will focus on a small number of studies that have analysed the consequences of matrilineal inheritance rules in contemporary developing countries. The matrilineal inheritance system can be illustrated with reference to Figure 5.3. In this kinship diagram, following the convention in social anthropology, triangles indicate males, circles females, vertical links indicate a descent bond, horizontal ones a codescent bond, and the sign ‘=’ stands for a marriage relationship. The shaded triangle is the (male) household head. The numbers inside the

=

=

1

=

2 Male Female Dark area refers to household head. Numbers indicate order of inheritance

Figure 5.3  Matrilineal inheritance (source: La Ferrara (2007: 285)).

3

Family and kinship ties in development   115 circles or triangles indicate the order in which a given relative should inherit the head’s property. According to the matrilineal rule, the man’s heir will be his sister’s son. However, in most matrilineal society a seniority principle requires that property be passed first to any living male codescendant of the head (i.e. any living brother) and then to the younger generation. If no brother or nephew exists, the closest male relative to inherit the head’s property will be the maternal aunt’s son. This inheritance structure is bound to have profound effects on the investments that parents make in their natural offspring, but also on the children’s efforts in caring for their natural parents. As far as the former effect is concerned, an optimizing parent who cannot bequeath land to his natural son will have an incentive to over-­invest in his education (La Ferrara and Redaelli 2005). A similar effect has been documented by Quisumbing and Otsuka (2001) for Western Sumatra. In that context, the customary rule requested that land be passed on from mothers to daughters. Quisumbing and Otsuka consider the allocation of land and schooling between sons and daughters and test whether the evolution of inheritance patterns from a strict matrilineal rule to a system where both sons and daughters inherit has an effect on the pattern of schooling investments between boys and girls. They find that indeed the traditional rule had generated a suboptimal allocation of land and schooling, with daughters receiving much less education than sons. In their data, the shift to a bilateral inheritance system is associated with a smaller gender gap in education. But this system also distorts children’s choices of how much (and in what form) to take care of their parents. In a study on Ghana, La Ferrara (2007) formalizes the choice of a father who faces an inheritance rule like the one depicted in Figure 5.3.4 The father is not unconstrained in his choice about the allocation of inheritance, but rather must obey a customary norm that requires a strictly positive fraction of his land to remain with the matrikin. In particular, the default allocation is for the entire land endowment to be bequeathed to the nephew, unless the father makes a donation to his own children during his lifetime (a common practice among the matrilineal Akans in Ghana). The model shows that in equilibrium children will send monetary transfers to their parents to induce the latter to donate the maximum socially acceptable amount of land to them. This prediction is confirmed in the data: ceteris paribus, Akan men are significantly more likely to receive transfers from their children compared to other ethnic groups, the more so the greater the strength of customary matrilineal norms. This has important implications for the pattern of inter-­vivos transfers, hence for saving decisions of young individuals who are ‘constrained’ by customary norms.5 Finally, the structure of the family itself may generate inefficiencies. Given the widespread variation in the size and notion of families across countries, one could ask if children’s outcomes are better if they live in a nuclear family or in an extended family. Edlund and Rahman (2005) provide an answer to this question both from a theoretical and from an empirical point of view. They start from the premise that, viewed from the point of view of a child, the household head is the father in a system based on the nuclear family, while it is

116   E. La Ferrara the grandfather in one based on the extended family. Because the father has a greater likelihood to survive until his children reach adult age, he will have greater incentives to invest in the children’s human capital. The counterpart of this is that extended families can provide better insurance to their members and more public goods. Edlund and Rahman use data from Bangladesh – a country with a traditional family structure based on extended families – and exploit the transition from extended to nuclear family that occurs when the grandfather dies. They find that children in nuclear families receive more education, but they have no advantage in terms of height-­for-age.

Families as substitutes for markets The interaction between family values and markets brings us to a second set of considerations. A key role played by families, kin and ethnic groups in poor countries is that of facilitating exchange in the absence of markets. There is a wide variety of context in which markets are either nonexistent, or poorly functioning. This is due largely to the imperfect information available to individuals or to the absence of legal enforcement means, or both.6 As useful example on the role of information and enforcement is the well known work of Greif (1993). Greif documents the practices of a medieval group of traders, the Maghribi traders, who operated in the Mediterranean through a large number of commercial agents. There was an obvious problem of asymmetric information between agents and their principals, as the latter could not observe the costs incurred by the agents. However, the Maghribi traders put in place a ‘monitoring’ mechanism based on reputation, so that no merchant would employ an agent who had cheated on another merchant. The Maghribi traders were descendants of Jewish merchants who had emigrated to North Africa one century before and, although they were not all related by blood, their common ancestral links allowed them to easily spread information among members of the coalition and give incentives to merchants to carry out punishment even when this was individually costly, for fear of being ostracized. Indeed, because a key component of enforcement strategies in games of repeated interaction is the ability to exclude deviators from future transactions, the costlier the exclusion the greater the set of profitable arrangements that can be sustained. Kinship embeds a variety of relationships – social, emotional, economic – and thus constitutes a very powerful venue to enforce informal transactions. Insurance An important role played by families and kin groups all over the world is that of providing insurance and credit to their members. This is particularly crucial in societies that are poor and prone to economic shocks (e.g. agricultural societies in developing countries), with severe credit market imperfections, and where the state provides few or no social safety nets.

Family and kinship ties in development   117 For industrialized countries, Bentolila and Ichino (2006) study how families cope with unemployment shocks as a function of the generosity of unemployment benefits. They find that family networks tend to respond more (i.e. send more transfers) to those hit by a shock the smaller the welfare state in the country of residence. For developing countries, empirical studies have shown that there is a considerable degree of risk sharing between members of the same family (Fafchamps 1992; Murgai et al. 2002; Fafchamps and Gubert 2007) or of the same ethnic group (Grimard 1997). The rationale for having extended families or kin groups act as risk pool, rather than a generic network of friends, comes from two factors. One is the degree of altruism, the other is that mutual insurance arrangements critically rely on repeated interaction, and the long-­term nature of kin relations makes them an ideal risk-­sharing group. In what follows we analyse these two motives in isolation. As argued – among others – by Foster and Rosenzweig (2001), altruism can be instrumental to sustaining self-­enforcing mutual insurance arrangements. In fact when households care about each other, they stand to gain more from insuring each other than they would otherwise, and this increases the scope for cooperation. The authors find evidence in support of their argument using data from rural South Asia. When transfers do not take the form of monetary payments, but rather of in-­ kind payments, there may be further reasons for relying on family networks rather than on networks of friends. Consider for example the institution of child fostering widespread in sub-­Saharan Africa. This occurs when parents send a child to live with another family for a more or less extended period of time. In a recent study on Burkina Faso, Akresh (2007a, 2007b) shows that fostering serves a number of purposes, including that of adjusting to idiosyncratic income shocks (hence, risk sharing), that of balancing the demographic composition of a household’s children (e.g. to care for a newborn), and that of helping a child get an education, in case the host family lives nearer to the school. Because of the higher familiarity and degree of altruism among members of the same family, child fostering most often occurs between close relatives. A fundamental prerequisite for risk-­sharing agreements to work is reciprocity. The reason why people share their crops or livestock is that they expect to become recipients in the future, although the exact time and extent of the reciprocation may not be known at the date of the transaction. Repeated interaction then becomes a key element of any informal enforcement strategy (see, among others, Kimball 1988; Coate and Ravallion 1993; Ligon et al. 2001). Effective implementation of insurance arrangements also requires information on the shocks received by different individuals or households, information which is not readily available. As I will argue below in the context of credit, family ties give a fundamental advantage both in terms of information transmission and of reciprocity.

118   E. La Ferrara Credit The role of close-­knit communities (though not kin groups) in enforcing repayments of informal loans has been first analysed by Udry (1994) and Besley and Coate (1995). The latter refer to the sanctions imposed by community members as ‘social collateral’, and show that these sanctions can lead to higher repayment. These papers take social pressure as a given, modelling it as a direct utility loss, and do not model the way in which social sanctions are carried out, nor the role of families in this process. La Ferrara (2003) studies informal credit transactions in Ghana and addresses precisely the question of what makes relatives (or kinsmen) different from a generic set of individuals who interact on a regular basis and possibly have low transaction costs. She identifies two key characteristics. The first is that kin membership is exogenous (one is born with a given set of blood relations) and cannot be freely changed. This makes it costlier for individuals to opt out of a kin group in case of defection than it would be for a generic social group. The second characteristic is that kinsmen are dynastically linked in such a way that the actions of parents can fall upon their children, for good or bad. The author proposes a model with overlapping generations in which people are born rich or poor, an endogenous matching process between rich and poor determines who borrows from whom, and parents depend on their children for support in old age. She shows that the non-­anonymity of the dynastic link discourages unilateral deviations even in short-­term interactions. Younger generations can in fact be punished for deviations by their parents, and to the extent that parents internalize this punishment, cooperation (repayment) is easier to enforce. This implies that even when ‘bilateral’ reciprocity cannot be enforced because the two partners do not expect to interact in the future, ‘generalized’ reciprocity can. In other words, within a family network players can expect reciprocation from someone who is not a direct beneficiary today. The model also shows that dynastic links encourage preferential dealings between some sets of partners (i.e. members of the same kin group) and change the terms of economic transactions (specifically, lower the interest rate on transactions between kinsmen). Evidence from rural Ghana is consistent with the model’s predictions. Why should families enforce contracts? The above discussion suggests that families (intended in a broad sense that includes kin groups) have a comparative advantage in the enforcement of informal agreements along several dimensions. First, to the extent that family members display a greater degree of altruism towards each other compared to generic sets of players, the temptation to play non-­cooperatively will be lower. In fact, the gains from cooperation that accrue to the other player(s) will be at least in part internalized by each individual. This means that, other things being equal, a greater set of cooperative arrangements can be sustained.

Family and kinship ties in development   119 Second, even in the absence of altruism the dynastic nature of family ties increases the possibility of enforcing informal agreements by allowing for ‘generalized reciprocity’ (La Ferrara 2003). In fact, while in bilateral transactions reciprocation can only be carried out by the original beneficiary (‘bilateral reciprocity’), among members of the same family or kin group it is possible for reciprocation to involve the offspring of the original benefactors (‘generalized reciprocity’). In many developing countries, for example, it is common for young people who receive support from older relatives to finance their studies to reciprocate by helping their younger relatives once they start earning money. The power of dynastic links helps not only to reciprocate favours but also to carry out punishments. Kinsmen often obey the principle of collective responsibility, whereby members of the same clan are held jointly responsible for each other’s actions (Posner 1980). Following a defection, retaliation can thus fall on the defectors as well as on other members of their clan, increasing the cost of breaching the contract. Finally, family ties facilitate enforcement because they limit the possibility of opting out in case of defection. In a society when most economic transactions (e.g. insurance) occur within the boundaries of the extended family or clan, the ascriptive nature of these ties make it impossible for someone to enjoy those benefits if they unilaterally leave their original group. A similar argument is made by Caselli and Coleman (2006) in the context of the organization of political competition along ethnic lines. They argue that ethnic identity can be used as a ‘marker’ to recognize members of the opposing coalition and prevent their infiltration. When would it be better for families not to enforce contracts? The fact that families enjoy a comparative advantage in enforcement does not mean that the outcomes of family enforced arrangements are always welfare improving. First, they may not be welfare improving for society. As early as 1958, Banfield argued that the reason why the South of Italy was relatively less developed from an economic point of view was due to its ‘amoral familism’, i.e. the fact that its inhabitants were willing to cooperate with members of their own families but not with anyone else (Banfield 1958). This idea was later incorporated in the distinction between ‘bonding social capital’, which involves links among members of a relatively homogeneous and closed group, and ‘bridging social capital’, where the links cut across diverse social groups (Putnam 2000). The Sicilian ‘mafia’ can be seen as an extreme form of social organization based on family ties, which enjoys considerable enforcement power but implements agreements that are detrimental to social welfare (Gambetta 1993). Second, they may impose welfare losses on family members themselves. Membership in family groups is sometimes cemented by rituals or practices that involve substantial costs for the participants. A useful example is provided by Berman (2000) in his analysis of Ultra-­Orthodox Jews. Men in this group typically remain

120   E. La Ferrara in religious seminaries (yeshiva) and do not work until the age of 40 despite the poverty of their families to ‘signal’ their commitment to the community. A number of social practices and traditional rituals performed in developing countries nowadays may be interpreted along the same lines.

Concluding remarks This chapter has examined the role of family and kinship ties in economic development. It has started with an overview of the functions that families play in transmitting values to younger generations and establishing social norms. These values and norms have a profound impact on individual decisions involving fertility, labour force participation, investments in education and health. It has then proceeded to analyse families as substitutes for markets, with particular reference to informal insurance and credit markets. A common theme in the chapter is to highlight the comparative advantage that families have in performing some of these roles, but also the potential inefficiencies that may result. The chapter neglects, for reasons of space, two important strands of literature. The first looks at families in their role as ‘employers’. This includes the literature on ethnic networks and their effects on job search, employment and wages (see among others, Granovetter 1995 and Munshi 2003), and the recent literature on family firms, which looks at the economic performance of family controlled firms and at the potentially distortionary effects on the allocation of capital (Fisman 2001; Bertrand and Schoar 2006; Perez-­Gonzalez 2006). The second is a recent literature on endogenous family structure. Many of the arguments above related to the economic effects of certain patterns of family links are subject to the criticism that such patterns are not set in stone, and may themselves evolve in response to changing economic conditions. This is true at the micro level, in the sense of endogenous household formation (see, among others, Pezzin and Schone 1999; Konrad et al. 2002 and Fafchamps and Quisumbing 2007). But it also holds at a more macro level, in the sense of endogenous kinship groups. For example, Bates (1990) reports evidence that in many parts of East and Central Africa varying degrees of kinship ties reflect different needs to cope with risk. Despite these limitations, some of the contributions examined in this chapter offer helpful insights into the design of public policies that take into account the involvement of families in economic decisions. Some of these policies should explicitly address inefficient social norms that are perpetuated by family decisions, such as the payment of dowries or the practice of early marriage. Action has been taken in this direction, for example by institutions like the Grameen Bank in India, which does not lend to parents who pay a dowry or marry their daughters before 17, or the programme of national education vouchers in Bangladesh, for which married girls are not eligible. Other policies may be targeted to other goals, but take into account the structure of family relations. For example, the design of public safety nets should be responsive to the existing informal insurance networks and be sensitive to the

Family and kinship ties in development   121 fact that certain groups (e.g. migrants or members of the minority local population) may not enjoy the same degree of social connectedness, hence rely more on the public system. Also, to the extent that learning and information transmission in certain areas occur within family boundaries, public policies and information campaigns may be targeted to parents as well as to children (e.g. for the preventions of HIV-­AIDS). To conclude, sociologists and anthropologists have always taken family structure as a starting point for understanding society. Not so development economists, at least until recently. It is time to use the tools of economics to enrich our understanding of how individuals, families and groups decide in developing countries, and improve our ability to fight poverty effectively.

Notes 1 For a recent survey on the role of family and kinship networks, see Cox and Fafchamps (2007). For a review of the literature on ethnic diversity and economic performance, see Alesina and La Ferrara (2005). 2 For a different rationale of dowries in the context of land market imperfections, see Botticini and Siow (2003). 3 This does not mean that only women inherit. As will be described below, property typically passes from a man to his sister’s sons. 4 The model builds on and extends the literature on strategic bequests initiated by Bernheim et al. (1985). 5 Although the above analysis treats customary norms as given, it is likely that they endogenously adapt to the economic environment. Using data from Western Ghana, Quisumbing et al. (2001) argue that traditional matrilineal institutions are evolving towards individualized rights in circumstances where it is important to provide incentives for cocoa tree planting. 6 For an early contribution to this line of research, see Ben-­Porath (1980).

References Akresh, R. (2007a) ‘Risk, Network Quality, and Family Structure: Child Fostering Decisions in Burkina Faso’, BREAD Working Paper, 65. —— (2007b) ‘Adjusting Household Structure: School Enrollment Impacts of Child Fostering in Burkina Faso’, BREAD Working Paper, 89. Alesina, A. and Giuliano, P. (2007) ‘The Power of the Family’, NBER Working Paper, 13051. Alesina, A. and La Ferrara, E. (2005) ‘Ethnic Diversity and Economic Performance’, Journal of Economic Literature, 43 (3), pp. 762–800. Anderson, S. (2003) ‘Why Dowry Payments Declined with Modernization in Europe but Are Rising in India’, Journal of Political Economy, 111, pp. 269–310. Banfield, E. (1958) The Moral Basis of a Backward Society, New York: The Free Press. Bates, R. (1990) ‘Capital, Kinship and Conflict: The Structuring Influence of Capital in Kinship Societies’, Canadian Journal of African Studies, 24 (2), pp. 151–64. Ben-­Porath, Y. (1980) ‘The F-­Connection: Families, Friends, and Firms and the Organization of Exchange’, Population and Development Review, 6 (1), pp. 1–30. Bentolila, S. and Ichino, A. (2006) ‘Unemployment and Consumption Near and Far Away from the Mediterranean’, mimeo, CEMFI.

122   E. La Ferrara Berman, E. (2000) ‘Sect, Subsidy, and Sacrifice: An Economist’s View of Ultra-­Orthodox Jews’, The Quarterly Journal of Economics, 115 (3), pp. 905–53. Bernheim, D., Shleifer, A. and Summers, L. (1985) ‘The Strategic Bequest Motive’, Journal of Political Economy, 93 (6), pp. 1045–76. Bertrand, M. and Schoar, A. (2006) ‘The Role of Family in Family Firms’, Journal of Economic Perspectives, 20, pp. 73–96. Besley, T. and Coate, S. (1995) ‘Group Lending, Repayment Incentives and Social Collateral’, Journal of Development Economics, 46 (1), pp. 1–18. Bisin, A. and Verdier, T. (2001) ‘The Economics of Cultural Transmission and the Dynamics of Preferences’, Journal of Economic Theory, 97, pp. 298–319. Bisin, A., Topa, G. and Verdier, T. (2004) ‘Religious Intermarriage and Socialization in the United States’, Journal of Political Economy, 112, pp. 615–64. Bloch, F. and Rao, V. (2002) ‘Terror as a Bargaining Instrument: A Case Study of Dowry Violence in Rural India’, American Economic Review, 92 (4), pp. 1029–43. Botticini, M. and Siow, A. (2003) ‘Why Dowries?’, American Economic Review, 93 (4), pp. 1385–98. Caselli, F. and Coleman, W.J. (2006) ‘On the Theory of Ethnic Conflict’, mimeo, LSE. Coate, S. and Ravallion, M. (1993) ‘Reciprocity without Commitment: Characterization and Performance of Informal Insurance Arrangements’, Journal of Development Economics, 40, pp. 1–24. Cox, D. and Fafchamps, M. (2008) ‘Extended Family and Kinship Networks: Economic Insights and Evolutionary Directions’, in Schultz, T. and Strauss, J. (eds). Handbook of Development Economics, Vol. 4, North Holland. Edlund, L. and Lagerlof, N. (2004) ‘Implications of Marriage Institutions for Redistribution and Growth’, mimeo, Columbia University. Edlund, L. and Rahman, M. (2005) ‘Household Structure and Child Outcomes: Nuclear vs. Extended Families. Evidence from Bangladesh’, mimeo, Columbia University. Fafchamps, M. (1992) ‘Solidarity Networks in Preindustrial Societies: Rational Peasants with a Moral Economy’, Economic Development and Cultural Change, 41 (1), pp. 147–74. Fafchamps, M. and Gubert, F. (2007) ‘The Formation of Risk Sharing Networks’, Journal of Development Economics, 83 (2), pp. 326–50. Fafchamps, M. and Quisumbing, A. (2007) ‘Household Formation and Marriage Markets in Rural Areas’, in Schultz, T. and Strauss, J. (eds). Handbook of Development Economics, Vol. 4, North Holland. Fernandez, R., Fogli, A. and Olivetti, C. (2004) ‘Mothers and Sons: Preference Transmission and Female Labor Force Dynamics?’, Quarterly Journal of Economics, 119 (4). Field, E. and Ambrus, A. (2005) ‘Early Marriage and Female Schooling in Bangladesh’, mimeo, Harvard University. Fisman, R. (2001) ‘Estimating the Value of Political Connections’, American Economic Review, 91 (4), pp. 1095–102. Foster, A. and Rosenzweig, M. (2001) ‘Imperfect Commitment, Altruism, and the Family: Evidence from Transfer Behavior in Low-­Income Rural Areas’, Review of Economics and Statistics, 83 (3), pp. 389–407. Gambetta, D. (1993) The Sicilian Mafia: The Business of Private Protection, Cambridge: Harvard University Press. Giuliano, P. (2007) ‘Living Arrangements in Western Europe: Does Cultural Origin Matter?’, Journal of the European Economic Association, 5 (5), pp. 927–52.

Family and kinship ties in development   123 Granovetter, M. (1995) Getting a Job: A Study of Contacts and Careers, Chicago: University of Chicago Press. Greif, A. (1993) ‘Contract Enforceability and Economic Institutions in Early Trade: The Maghribi Traders’ Coalition’, American Economic Review, June, 83 (3), pp. 525–48. Grimard, F. (1997) ‘Household Consumption Smoothing through Ethnic Ties: Evidence from Cote d’Ivoire’, Journal of Development Economics, 53 (2), pp. 391–422. Guiso, L., Sapienza, P. and Zingales, L. (2008) ‘Social Capital as Good Culture’, Journal of the European Economic Association, 6 (2–3), pp. 295–320. Jensen, R. and Thornton, R. (2003) ‘Early Female Marriage in the Developing World’, in Sweetman, C. (ed.) Gender, Development and Marriage, Oxford: Oxfam GB. Kimball, M. (1988) ‘Farmers’ Cooperatives as Behavior Toward Risk’, American Economic Review, 78 (1), pp. 224–32. Konrad, K., Junemund, Lommerud K. and Robledo, J. (2002) ‘Geography of the Family’, American Economic Review, 92 (4), pp. 981–98. La Ferrara, E. and Redaelli, S. (2003) ‘Kin Groups and Reciprocity: A Model of Credit Transactions in Ghana’, American Economic Review, 93 (5), pp. 1730–51. —— (2005) ‘Social Norms, Inheritance and Human Capital’, mimeo, Bocconi University. —— (2007) ‘Descent Rules and Strategic Transfers. Evidence from Matrilineal Groups in Ghana’, Journal of Development Economics, 83 (2), pp. 280–301. Ligon, E., Thomas, J. and Worrall, T. (2001) ‘Informal Iinsurance Arrangements with Limited Commitment: Theory and Evidence in Village Economies’, Review of Economic Studies, 69, pp. 209–44. Manacorda, M. and Moretti, E. (2006) ‘Why do Most Italian Youths Live with Their Parents? Intergenerational Transfers and Household Structure’, Journal of the European Economic Association, 4 (4), pp. 800–29. Munshi, K. (2003) ‘Networks in the Modern Economy: Mexican Migrants in the US Labor Market’, Quarterly Journal of Economics, 118, pp. 549–99. Murgai, R., Winters, P., Sadoulet, E. and de Janvry, A. (2002) ‘Localized and Incomplete Mutual Insurance’, Journal of Development Economics, 67, pp. 245–74. Perez-­Gonzalez, F. (2006) ‘Inherited Control and Firm Performance’, American Economic Review, 96 (5), pp. 1559–88. Pezzin, L. and Schone, B. (1999) ‘Intergenerational Household Formation, Female Labor Supply, and Informal Caregiving: A Bargaining Approach’, Journal of Human Resources, 34 (3), pp. 475–503. Platteau, J.P. and Baland, J.M. (2001) ‘Impartible Inheritance versus Equal Division: A Comparative Perspective Centered on Europe and Sub-­Saharan Africa’, in De Janvry, A., Gordillo, G., Sadoulet, E. and Platteau, J.P. (eds). Access to Land, Rural Poverty, and Public Action, Oxford: Oxford University Press. Posner, R.A. (1980) ‘A Theory of Primitive Society with Special Reference to Law’, Journal of Law and Economics, 23, pp. 1–53. Putnam, R. (2000) Bowling Alone: the Collapse and Revival of American Community, New York: Simon and Schuster. Quisumbing, A. and Otsuka, K. (2001) ‘Land Inheritance and Schooling in Matrilineal Societies: Evidence from Sumatra’, World Development, 29 (12), p. 2093. Quisumbing, A., Panyongayong, E., Aidoo, J.B. and Otsuka, K. (2001) ‘Women’s Land Rights in the Transition to Individualized Ownership: Implications for the Management of Tree Resources in Western Ghana’, Economic Development and Cultural Change, 50 (1), pp. 157–81.

124   E. La Ferrara Tabellini, G. (2008) ‘The Scope of Cooperation: Norms and Incentives’, Quarterly Journal of Economics, 123 (3), pp. 905–50. Udry, C. (1994) ‘Risk and Insurance in a Rural Credit Market: An Empirical Investigation in Northern Nigeria’, Review of Economic Studies, 61 (3), pp. 495–526. Weber, Max (1978) Economy and Society, edited by Guenther Roth and Claus Wittich, Berkeley: University of California Press. Winick, C. (1956) Dictionary of Anthropology, New York: Philosophical Library.

6 The demand for disadvantage1 Rohini Somanathan

Introduction Disadvantage is a popular and controversial word in India these days. In October 2007, half a million Gujars, members of a traditionally pastoral community of north and central India, filled the streets of several towns in the Indian state of Rajasthan demanding they be classified by their state government as disadvantaged. The Gujars wish to be listed as Scheduled Tribes, and thereby receive greater parliamentary representation, preferential treatment in public employment and lower admissions standards in many educational institutions.2 Yet, ethnographers have cast doubt on their aboriginal descent, they share customs with other groups in the middle of the social ladder,3 and a current web site hosted by members of the Gujar community refers to the group as ‘a proud people’ with ‘the desire and ability to rule the world’.4 The case of the Gujars illustrates, oddly but powerfully, the ways in which culture and politics mingle to shape acceptable notions of social justice and government policy in democracies. In a poor, growing economy with academic costs well below the market value of educational training, the tag of disadvantage has come to acquire value and, ironically, the desire for mobility has brought about a demand to be classified as disadvantaged. It is this demand that I would like to reflect upon here – its cultural roots, its social rationale, the political mechanisms through which it is expressed and some of the economic implications of the policies that it has generated. Group-­based policies of preferential treatment began under British rule in the first half of the twentieth century. After political independence in 1947, the Indian constitution converted some of these policies into rights and thereby facilitated the expansion of state-­led affirmative action. The constitution was unusual in that it juxtaposed provisions for the equality of all citizens before the law with those that mandated the proportional political representation of specific groups and allowed the state to make special concessions for their advancement. In the decades that followed, these provisions did dilute the dominance of the elite in political and social life but also generated caste-­based contests for the rents from public office and the gains from spending on public goods.

126   R. Somanathan Mandated political representation and other types of affirmative action changed the balance of power but also created new types of inequalities within the set of targeted communities. Demographic data from the census, public employment and college admission records, and studies of electoral outcomes all suggest that minimally disadvantaged and the numerically strong communities benefited more than the others. The constitutional space given to affirmative action was initially valuable because it encouraged the state to acknowledge its responsibility towards the socially marginalized. Over time, however, it has created a peculiar discourse of social justice and development in India in which individual advancement is linked to group mobility and groups move forward by claiming that they have been left behind. In the process, the state has neglected less controversial and more fundamental rights such as access to primary and secondary education that may have done more for larger numbers of truly disadvantaged communities. The next section describes the constitutional basis for affirmative action policies in India and provides a brief history of these policies. The third section presents secondary evidence on the characteristics of beneficiaries and the distribution of benefits. It also documents the inequality in educational attainment that emerged within the set of communities that were targeted as recipients of affirmative action over the 1931–91 period. I conclude with reflections on the divergence between the intended and actual effects of affirmative action in India.

An unusual constitution The social classification that is used as a basis for affirmative action policies in India appeared during colonial rule in the first half of the twentieth century, when British politicians and nationalist leaders deliberated on the manner in which Indians were to receive greater political representation. In 1906 separate electorates were granted to Muslims as a disadvantaged religious minority, and similar claims were advanced by other communities (Dushkin 1967: 626–9). Job reservations in public employment first appeared in the Madras Presidency in south India and in some of the independently ruled areas of southern and western India. These were the result of struggles to limit the power of the Brahmin elite in favour of the middle classes. The reservations were far-­reaching and elaborate in that they assigned specific shares for each of the major caste groups. In the state of Mysore in south India, only one-­fifth of government posts and seats in colleges continued to be assigned through open competition. These movements did little however to integrate groups at the bottom of the social hierarchy who were largely illiterate with no prospects for either college degrees or public employment. During the inter-­war years debates on compensatory preference came to be centred around the Untouchables, a culturally and occupationally diverse group of castes regarded by other Hindus as ritually impure. Their disadvantage could be linked to a long history of discrimination; they were excluded from temples, marketplaces, water sources and most types of social interactions. The abolition

The demand for disadvantage   127 of untouchability became part of a vision for independent India. It was central to Mahatma Gandhi’s agenda for social change and B.R. Ambedkar, later to become one of the architects of the Indian Constitution, was himself from one of the Untouchable castes of western India and became their most prominent spokesman.5 During census operations in 1931, enumerators were asked to create lists of these exterior castes to facilitate their ‘representation in the body politic’ and to design appropriate policies for ‘raising them from their present backward position’.6 The term Scheduled Castes first appeared in the Government of India Act of 1935 when this diverse group of castes were placed in a single category and a fraction of seats in provincial and state legislatures were reserved for them. After political independence, historical disadvantage and compensatory state policies were extensively discussed by the Constituent Assembly that was created to draft the Indian constitution. Untouchability was widely viewed as incompatible with a modern society and preferential treatment to these groups was advocated in 1947 in the first report on minority rights. At the time one of the Assembly members pointed to the social isolation and poverty of tribal groups in India and, after a committee investigated their condition more fully, they too were accepted as worthy recipients of preferential policies (Guha 2007: ch. 6). Lists of Scheduled Castes and Scheduled Tribes were created separately for each state and the constitution in 1950 required that seats in parliament and in the state assemblies be reserved for candidates from these two groups. In addition, they were entitled to jobs in public employment and seats in higher education, all in proportion to their share in the population.7 Preference policies took three principal forms. The reservations described above were the most coveted and the most controversial. These included seats in legislatures, posts in government and quotas in institutions of higher education.8 In addition, there were programmes which provided these groups with scholarships, loans and land grants. There were also laws, such as those that limited transfers of land between them and the general population, that were designed to protect them from exploitation by other classes (Galanter 1984: ch. 3). Béteille (2005) makes an important distinction between mandatory provisions and enabling provisions of the constitution. Proportional representation in the parliament and in state legislative bodies was the only mandatory provision of the constitution. All other policies were based on the constitutional directive of paying special attention to the conditions of these groups and did not require specific forms or levels of representation. Difficulties in interpreting the constitution surfaced early. An upper caste girl was denied admission to a medical college in 1950 and she claimed this violated her constitutional right to be treated at par with other citizens.9 Prompted by this case, the constitution was amended in 1951 to explicitly allow the state to favour backward classes of citizens. A commission was appointed in 1955 to identify and enumerate such classes and although a number of criteria were established to identify backwardness, the commission concluded that social backwardness ‘is mainly based on racial, tribal, caste and denominational

128   R. Somanathan differences’ (Government of India 1955: Vol. I, p. 42). Backward classes at this point came to mean backward castes; various groups made representations to the commission claiming backwardness and lists of these castes were generated for each of the Indian states. The final report listed 2,399 communities as backward and recommended that 70 per cent of seats in higher education be reserved for this category. Those not classified as Scheduled Castes and Tribes were known as Other Backward Classes (OBC) and in some states these three categories together accounted for more than three-­quarters of the population. The central government did not implement these recommendations and there was disagreement even among members of the commission on the objectivity of the lists.10 State governments in southern and western India with large numbers of politically active Backward Castes did continue to implement quotas for them in public employment and higher education.11 In 1978 the Mandal Commission was appointed to investigate the condition of the Backward Classes and it recommended preferential treatment for them in politics, employment and education, once again identifying such classes through their caste affiliations. The methodology and the recommendations of the commission were widely criticized and its report in 1980 was met with violent student demonstrations in many parts of the country.12 Eventually, however, many of its recommendations were adopted; starting in 1990, 27 per cent of public service jobs were reserved for OBCs and in 2005 the constitution was amended to explicitly allow quotas and lower evaluation standards for them in higher education. In 2006 the legislature passed an act requiring all federally funded universities to implement these quotas.13 The Supreme Court of India subsequently stayed the implementation of this act and questioned its constitutional validity. Although it might appear that the relevant constitutional provisions simply led to the expansion of a culture of differential treatment that had begun under colonial rule, they were a significant break from the past because they converted benefits from government policies into rights. This makes the Indian constitution an ambiguous document because statements for the equality of all citizens before the law, that are standard in the constitutions of democracies across the world, are qualified by those that grant particular communities special status. The contrast with the United States is instructive in this regard.14 It is another large democracy where inequality has well-­established racial boundaries and historical roots. Yet, while affirmative action in India vastly expanded, support for race-­based admissions dwindled in the United States in spite of large and persistent racial gaps in academic achievement. The Constitution of the United States does not allow for exceptions to the Equal Protection Clause and many affirmative action programmes in US universities have been declared unconstitutional during the same period that Indian programmes have expanded. In a series of cases since the late 1970s, the United States Supreme Court has not upheld any admissions policy that insulates minority applications to educational institutions from competition with a broader applicant pool. Affirmative action policies have survived only if the judiciary has been convinced that the policies ‘achieve that diversity which has the potential to enrich everyone’s education’.15

The demand for disadvantage   129 I believe, as do many others, that not enough has been done to address racial gaps in the United States and that, to use Glenn Loury’s phrase, racial stigma16 and its consequences continue to shape social interactions. Yet, a comparison of the two countries is useful because it illustrates the difficulties in achieving social equality through a route that recognizes group-­based identities. In spite of the wide-­ranging and somewhat problematic constitutional provisions for social equality in India, the stigma attached to caste remains, albeit diluted, and, as the following section will illustrate, the gains from these provisions have not touched the majority within these communities.

The beneficiaries and the benefits As outlined in the preceding section, affirmative action policies in India have relied almost entirely on caste classifications. This approach has been justified on the grounds that the origins of backwardness lie primarily in the historic discrimination faced by certain castes and, once identified, caste membership provides the most feasible means of reaching the disadvantaged. It is also argued that economic measures of backwardness do not fully capture the social condition of these groups since discrimination and social stigma denies them opportunities available to others in their economic situation. These claims have been uncontroversial for some of the communities that were classified as Untouchables, Depressed Classes or Exterior Castes during the colonial period and are now listed as Scheduled Castes. There is abundant evidence of their exclusion from schools, roads, wells and other public places, they were often prohibited from adopting the social customs, attire and ornaments of the upper castes and after political reforms by the British government in the 1920s, only a small fraction of the population within these castes was entitled to vote.17 Once the association of backwardness with caste was established, it became the principal strategy through which the state executed its mandate of social justice. Several castes without any history of untouchability came to be viewed as worthy of preferential treatment and no individual could enter the ambit of affirmative action without membership of one of these groups. Our assessment of these policies therefore depends on: 1 the extent to which disadvantage was captured by the identities of included groups; 2 the nature of benefits conferred on those entitled to them; 3 their effectiveness in excluding socially mobile groups from the purview; and 4 their implications for the excluded population. This section deals with available evidence on these issues. There is a wealth of information on the Indian caste system, but there are serious limits to a systematic empirical investigation of the relationship between caste and disadvantage. A major obstacle is the sheer size and complexity of the

130   R. Somanathan caste structure. Several thousand different communities have appeared in the ethnographic literature on caste and the social standing of a particular caste may vary from one region to another. The Anthropological Survey of India recently embarked on ambitious project aimed at generating anthropological profiles of all major communities in India and listed 4,635 different caste groups (Singh 2002). Even if the required data on these groups were available, a multidimensional regional ranking of the type that was attempted by the two Backward Classes Commissions is necessarily somewhat arbitrary. A related point is that caste counts rely on self-­reported data. Over time, sections within some castes have broken away and used new names as passports to new untarnished identities. The process by which communities changed social practices to facilitate their movement up the social ladder has been carefully studied by the well-­known Indian sociologist M.N. Srinivas. He comments on how the decennial census, introduced during colonial rule, provided an instrument for social mobility: ‘Prosperous low castes, and even those which were not prosperous, sought to call themselves by new and high-­sounding Sanskrit names.’ The recording of these names by census enumerators validated these new identities (Srinivas 1957: 531). This was recognized by census authorities and was a source of considerable frustration to them, but caste was after all a social construction, new social identities had to be acknowledged, and there was therefore no process that could ensure consistent caste-­wise data across census years.18 Finally, there is a lack of nationally representative data on caste. The last complete caste enumeration was done by the colonial administration in 1931. After independence, detailed caste affiliations were only recorded for the Scheduled Castes and Tribes. Census data cannot therefore be used to make caste-­wise comparisons of educational attainment or employment over time except for castes in these two categories. A standard method of evaluating government policies is to measure the difference in relevant outcomes for a set of affected individuals with an otherwise similar group that was excluded. Historical census records suggest that there were castes with demographic characteristics similar to some of the Scheduled Castes and these could have functioned as a control group for this type of analysis had data on them been systematically collected. In the face of these constraints, studies of caste and mobility have usually adopted one of two alternative routes. The first is to use available secondary data to compare the entire group of Scheduled Castes or Scheduled Tribes with the rest of the population without making any distinction within these categories. Census data can be used for comparisons of educational attainment, household demographics and occupational structure and several large nationally representative surveys are available with data on the health and household expenditures for these broad categories. Mendelsohn and Vicziany (2002) document some of the atrocities that have continued against the Scheduled Castes after independence. Drèze and Sen (2002) use various official data sources to illustrate how region, gender and caste combine to depress educational attainment in some of the Indian states. In 1991, the literacy rate for Indian males was 64 per cent, for females it was 39 per cent, in some backward districts of Rajasthan it was less than 5 per cent for

The demand for disadvantage   131 Scheduled Caste women and less than 1 per cent for women from the Scheduled Tribes (chapter 5, pp. 146–9). Banerjee and Somanathan (2007) show that over the period 1971–91, fewer education, health and transport facilities were located in parliamentary constituencies with Scheduled-­Tribe concentrations. Pande (2003) combines budgetary and electoral data at the state level and finds that greater parliamentary representation of Scheduled Castes and Tribes led to increases in job quotas but did nothing to improve education spending. An alternative approach to the study of caste mobility has been based on more specialized datasets based on, for example, the membership of major political parties, electoral outcomes, parliamentary debates, government reports and public employment. Christophe Jaffrelot has painstakingly collected data on the caste identities of members of parliament and legislative assemblies since the 1950s to decipher trends in the caste composition of elected representatives. He finds that most Scheduled Castes in the 1950s and 1960s were loyal to the Congress Party and the increase in their share in the legislature brought about by reservations did not result in their taking over positions of leadership within the party. The most significant change that he documents is the silent revolution by which the Other Backward Castes have come to dominate politics in north India. In 1952, less than 5 per cent of MPs elected in the Hindi Speaking Belt came from this category. In 1999, their share had risen to over 22 per cent and these seats were held primarily by Yadavs and Kurmis, the more prosperous groups among the backward castes. Much of the increased representation of these castes was through the formation of new political parties that promoted their interests.19 Marc Galanter, in a remarkable book, compiles material from a variety of administrative and judicial sources to explore the effects of policies of preferential treatment. He finds that disparities in school attrition rates between upper and lower castes widen considerably as students move to higher grades. School enrolment rates for grade 9 and above for Scheduled Caste students in the early 1970s were about half those for the unreserved groups and rates for the Scheduled Tribes were one-­third below those for the Scheduled Castes. He also finds that in higher education, Scheduled Caste and Scheduled Tribe students tend to be concentrated in the less prestigious courses. Data from the reports of the Commission on Scheduled Castes and Tribes shows that only 6.4 per cent of post matriculate students from these groups were studying medicine or engineering while close to 40 per cent of students from Other Backward Classes were in these fields in the early 1970s (pp. 60–3). He also finds that in the two preceding decades, reservations in public employment did lead to a rise in the numbers of both Scheduled Castes and Scheduled Tribes in higher levels of government, but the change was much more substantial for the castes than for the tribes. The share of Scheduled Castes in the central administrative services went from 1 per cent to 8 per cent between 1953 and 1975 while that of Scheduled Tribes was just a little over 1 per cent at the end of this period. Harry Blair’s study of assembly elections in Bihar is worth mentioning here because it examines the distribution of political power within the group of non-­ Scheduled Castes. He constructs Lorenz curves that relate the share of seats held

132   R. Somanathan in the Bihar Legislative Assembly to the population shares of different castes and finds that the poorer communities among the Hindu Backward Castes (collectively called Shudras in this case) are grossly under-­represented while ‘upper Shudras have more or less managed to hold their own over the period, with just under 24% of the non-Scheduled population and an average of 25.7% of unreserved seats in the three elections’. This body of research reveals several interesting patterns. There is an asymmetry in the gains of the Scheduled Castes relative to the Scheduled Tribes in spite of very similar mandated entitlements. The Backward Castes have appeared as major political players in spite of no mandated representation in the legislature and, in states with OBC reservations, they have occupied much larger fractions of seats in coveted educational institutions than the other two groups. Within the Backward Castes, the wealthier groups are the ones who have acquired political influence. Finally, greater legislative control by the Scheduled Castes and Tribes seems to have encouraged job reservations rather than spending on education in spite of very high levels of illiteracy among these populations. Elected representatives for these groups do not appear especially committed to the overall advancement of their communities. I would like to end this section by suggesting an alternative approach to examining the distribution of benefits from affirmative action among targeted communities. For the purpose of monitoring the demographic and social outcomes of the Scheduled Castes and Tribes, the census has been enumerating each community within these categories. These data have been little used and can potentially tell us a great deal about the distribution of benefits across castes within these categories. The 1931 census data, because it provides a detailed caste enumeration of the entire hierarchy, can be used to measure the extent to which these policies did in fact target disadvantaged groups. One can then focus on the distribution of gains within the Scheduled Castes and Tribes in the post-­independence period to see how entitlements within the targeted population were translated into real gains for the groups who most needed them. I proceed along these lines to examine disadvantage and mobility for major castes in the area that used to form the British province of Bihar and Orissa. About 11 per cent of the Indian population lived in this area in 1931 and roughly 14 per cent lived in the corresponding states in 2001.20 This region is relatively well-­suited for a historical study of preferential treatment and caste mobility because Scheduled Castes, Scheduled Tribes and Other Backward Castes are all sizeable fractions of the population and the region remained relatively undisturbed during periods of political upheaval in 1947 and 1971, when changes in national boundaries resulted in large-­scale movements of families in and out of many parts of the country. Literacy levels and rates of educational attainment for this region for the period 1931–91 are shown in Tables 6.1 and 6.2. Table 6.1 lists caste-­wise literacy rates for the period 1931–91. Included are all Hindu and tribal castes that numbered more than one per thousand of the province population in 1931 and whose literacy was tabulated for that year.

The demand for disadvantage   133 Starting in the 1950s, caste-­wise figures on educational status are available only for the Scheduled Castes and Tribes, the targeted beneficiaries of affirmative action policies. The current classification of both these groups as well as Other Backward Castes vary by state and have changed over time. The table has the current classification of these groups for the states of Bihar, Orissa and Jharkhand, which roughly covers the area of the former British province of Bihar and Orissa.21 The literacy rates in the table reveal a striking divide between the upper castes and the rest of the population. Brahmins, the traditional priestly class, and Kayasthas, often referred to as the caste of scribes (Crooke 1974: Vol. 3, p. 185), had literacy rates far above the rest of the population. Literacy among Kayasthas was seven times the provincial average in 1931 and 24 per cent of Kayastha males were literate in English. These groups also dominated the jobs held by Indians in the colonial administration and seats in provincial legislatures after political decentralization in the 1930s (Hutton 1933: ch. 9). In contrast, literacy rates among many of the tribal and lower caste communities were less than 1 per cent. Other Backward Classes lay typically between these two extremes, although there are groups such as the Mallah caste (traditionally boatsmen and fishermen), who fared distinctly worse than some of the Scheduled Castes. The Mallah caste is one of the groups that was found to have almost no political representation in Bihar politics (Blair 1972). Over the 60-year period following the 1931 census, literacy rates and rates of primary and secondary school completion for the Scheduled Castes and Tribes remained well below the rest of the population. The higher literacy rates for Scheduled Tribes at the beginning of this period and the slighter higher rates for the Scheduled Castes at the end suggest that the latter group gained relative to the former, but the differences between these groups appear small in relation to those between them and the rest of the population. By 1991, educational attainment rates for the population of this area, taken as a whole, were roughly double those for the Scheduled Castes and Scheduled Tribes. Within each of these groups, caste ranks in educational attainment are mostly preserved over time. The Dhobis and Pasis were the most literate among the Scheduled Castes in 1931 and they both had significantly higher rates of secondary school completion and college graduation in 1991 than other castes in this category. The same is true for the Scheduled Tribes, with the Mundas and Oraons starting and remaining on the top of the major tribes in this area. One important difference between the castes and the tribes reflected in these data is that large groups were the most mobile among the lower castes but there appears no such pattern among the tribes. The Chamars (the caste of skinners and tanners that was regarded as ritually impure even by many of the other Scheduled Castes) and the Dusadhs each constituted about one-­fifth of the population of these castes in the province and were the most illiterate among them in 1931. By 1991, their rates of secondary school completion and college graduation were not very different from those of the Dhobis. The Doms and Bhuiyas were much smaller groups that started a little ahead of the Chamars and Dusadhs and rapidly fell behind. In contrast, the Santhals, who formed nearly one-­quarter

134   R. Somanathan Table 6.1  Literacy rates by caste in Bihar and Orissa, 1931–91 Category

Caste

Literacy (per cent) 1931

Forward Castes

Other Backward Castes

Scheduled Castes

Scheduled Tribes

All Scheduled Castes All Scheduled Tribes All groups

Babhan Brahman Kayastha Rajput Gaura Kurmi Mallah Teli Bauri Bhuiya Chamar Dhobi Dom Dusadh Nat Pasi Munda Oraon Santhal Savar

13.6 19.6 37.4 12.0 2.5 4.9 0.8 5.9 0.7 0.7 0.5 1.5 0.4 0.6 1.3 1.4 2.8 1.1 0.5 0.9 5.3

1961

1991

11.6 1.1 6.6 12.6 6.1 7.0 5.7 8.9 10.8 9.9 6.0 8.6 7.9 7.9 21.8

29.9 13.7 21.1 34.8 21.2 21.0 10.9 25.6 28.9 32.8 20.7 25.1 22.8 23.0 41.4

Sources: literacy rates are taken from Census of India volumes for each of the three census years. Figures for 1961 and 1991 are population-weighted averages of literacy rates for the states of Bihar and Orissa. The 1961 rates are based on the population above five years and 1991 rates on the population above seven years.

of the population of this region, started disadvantaged, and experienced little change in their relative position. These differences are consistent with evidence on effective political mobilization among the Scheduled Castes and their much greater electoral success. In the mid-­1990s, the Bahujan Samaj Party came to power in north India and the Chamars were its principal supporters (Jaffrelot 2003: 399–404). The gains in educational attainment among the Scheduled Castes are also reflected in the changing composition of public employment. Although government jobs for the castes and the tribes were reserved in proportion to their population shares, Scheduled Castes filled a much larger fraction of these seats than the Scheduled Tribes (Galanter 1984: Table 7, p. 96). Overall, these data illustrate both the substantial heterogeneity within each of the official caste categories and the overlap in their distributions of literacy. This is to some extent unavoidable whenever an intricate structure is mapped into a small number of coarse categories. In the Indian case, however, it resulted

3 4 21 6 5 19 0.1 3 100 10 10 23 4 100

Bauri Bhuiya Chamar Dhobi Dom Dusadh Nat Pasi All Scheduled Castes Munda Oraon Santhal Savar All Scheduled Tribes All groups

9.02 4.02 6.47 9.7 5.39 6.3 3.62 7.79 6.72 8.73 9.35 6.68 6.81 6.54 12.51

1991

1961 0.51 0.23 1.64 2.5 0.86 1.87 1.55 2.74 1.44 4.87 4.77 3 0.89 2.31 4.78

Secondary (% completed)

Primary (% completed)

0.09 0.01 0.17 0.33 0.07 0.18 0.2 0.41 0.14 0.31 0.43 0.01 0.02 0.17 1.13

1961

2.81 1.58 4.98 6.64 2.62 5.35 1.78 6.65 3.92 4.71 6.64 3.13 1.46 3.13 8.8

1991

0.44 0.29 1.26 1.77 0.47 1.32 0.44 2.41 0.96 1.13 1.91 0.59 0.29 0.71 4.3

1991

College (% completed)

Sources: rates of educational attainment have been computed using the total number at each level, divided by the population over 15 for primary school and over 19 for secondary school. These rates may therefore be inflated if sizeable numbers complete these levels of schooling before these ages.

1961 (% population)

Caste

Table 6.2  Educational attainment by caste in Bihar and Orissa, 1961–91

136   R. Somanathan mainly from the initial association of backwardness with ritual purity and the difficulties of excluding socially mobile groups from the benefits of preferential treatment once these had been granted to them. In 1965 the Lokur Committee was appointed by the government to suggest changes to the lists of Scheduled Castes and Tribes. The committee recommended the exclusion of about half of the Scheduled Caste population of North India and one-­fifth of the Scheduled Tribes. Resistance from Scheduled Caste members of parliament followed and when an act addressing the coverage of these policies was finally passed in 1976, it retained all the initial beneficiaries and in fact added about five million persons to the list by eliminating intra-­state restrictions which had previously limited the benefits of particular castes to specific regions within each state. In the state of Bihar alone, the population of Scheduled Castes increased by over 400,000 (Galanter 1984: 135–40).

Conclusions The most salient feature of affirmative action policies in India was a set of quotas intended to bring about greater representation of socially marginalized communities in politics, government and among the educated elite. Greater political representation did occur, but representatives either did not attempt to or did not succeed in providing their constituencies the opportunities that allowed a convergence in the welfare of these groups with the rest of the population. The extremely low educational attainment among most Scheduled Castes and Scheduled Tribes today makes it unlikely that reservations in higher education or the civil service can provide the average individual in these groups with substantial benefits in the near future. In fact, a recent study using nationally representative survey data finds that most of the gap in college completion rates across the official caste categories can be attributed to differences in the eligible population – the fraction of the population that has completed high school – and the sources of gaps in achievement are therefore at the bottom, and not the top, of the education pyramid (Sundaram 2006). The finding that large and politically influential groups were the most mobile is hardly surprising. After all, is it not numbers and influence that dictate state policies in democratic systems? One is left wondering why a group of statesmen with a genuine desire for development, modernity and the removal of caste-­ based social inequalities introduced a constitution that supported these reservations. What were they thinking? An interview with B.R. Ambedkar, the chair of the drafting committee of the constitution, by the Backward Classes Commission in 1955 provides some sort of an answer. Consider the following excerpt from his response to questions on the causes of backwardness and the role of reservations:36 Different status should disappear. It can disappear only by the advancement of education, when all the communities are brought to the same level in the matter of education not everybody but the community as such. If there are

The demand for disadvantage   137 10 barristers, 20 doctors, 30 engineers etc. in a community, I regard that community as rich although everyone of them is not educated. Take for instance, Chamars, you look upon this community with hatred, but if there are some lawyers, doctors and educated persons among them, you cannot put your hand upon them . . . no body will look down upon them. My idea is something fantastic: collect the best boys from the primary schools and give them food, shelter and education, send them in a college and give them tuition fee. After that select a number of students from amongst them and send them to foreign countries where they can get the best of education . . . Thus you will create a few people with high qualification and place them in high posts. (Government of India 1955: Vol. III: 73–4) Ambedkar viewed reservations as helping create an elite within the Scheduled Castes that, given their increased political mobilization, would help shed the stigma that accompanied their caste names within a period of ten years (ibid.: 74). He believed that the appropriate leadership, together with other constitutional provisions, such as article 45 which guarantees free and compulsory primary education, could ensure progress among the majority within these communities. The quotas in the legislature were designed to ensure that the talented among the historically disadvantaged groups could assume positions of leadership. There is nothing to suggest that these were viewed as instruments that would directly redistribute opportunities from one group to another. This vision, of developing an elite within a group that has faced discrimination, was similar to that of W.E.B. Du Bois (1903), when he wrote that ‘The Negro race, like all races, is going to be saved by its exceptional men.’ Many of those who supported special constitutional provisions for backward classes in the 1950s would not have foreseen these groups developing a ‘deep attachment to these measures’ (Deliège 1999: 195). In the state of Uttar Pradesh, home to the largest number of Scheduled Castes in the country, one-­third of all villages are without a primary school and 96 per cent without a high school. As M.N. Srinivas wrote 50 years ago, ‘it is understandable that groups which are classed as backward show reluctance to give up the privileges of backwardness’ (Srinivas 1957:  547). It may however be time to devote scarce judicial and administrative energy to more fundamental, less controversial rights.

Notes   1 Thanks to Aditya Bhattacharjea, Hemanshu Kumar, Rajiv Sethi and E. Somanathan for useful discussions.   2 See The Hindu: ‘Gujjars take to the streets demanding ST status, October 2, 2007’.   3 See Crooke (1974: Vol. II), for comparisons of the Gujars with other castes and tribes during the colonial period. Here, and in the 1931 Census of India, they are classified together with other pastoral castes such as the Yadavas who have recently become prominent in north Indian politics (Hutton 1933: Part II, Table XVII).   4 www.gujarsonline.com.

138   R. Somanathan   5 Ambedkar’s father was an officer in the army and worked for the State of Bombay after his retirement. Ambedkar recalls the pain from the ostracism he and his elder brother faced when travelling to visit their father during their summer holidays. With the innocence of a nine-­year-old, he revealed his caste to the railway station-­master and norms of pollution made both transport and water inaccessible. The following excerpt from his account of this trip reveals the deep impression this incident made on him: There was plenty of food with us. There was hunger burning within us; with all this we were to sleep without food; that was because we could get no water and we could get no water because we were untouchables. Taken from the chapter ‘On the way to Goregaon’ in Rodrigues (2002: 52).   6 These lists were to contain ‘castes who suffered disability on account of their low social position and on account of being debarred from temples, schools or wells’. No more specific criteria were framed since it was recognized that ‘conditions varied so much from province to province and from district to district, that it would be unwise to tie down the Superintendents of Census Operations with too meticulous instructions’ (Hutton 1933: 471–2).   7 Article 15 deals with equality before the law, article 16 with public employment, 330 and 332 in Part XVI deal with political representation, and article 46 with state policy.   8 The reservation of seats in the legislature was originally for ten years but has been consistently extended.   9 The Supreme Court in this case held that the Communal Government Order of the Madras government which explicitly allowed for caste-­based admissions violated the principle of equality before the law (Vijayan 2006: 92). 10 One of the members, P.G. Shah, notes in his minute of dissent (Government of India 1955: Vol. III, p. 9): In several cases there was great difficulty in arriving at a decision as to whether a community should be treated as backward or not. Generally, this decision was taken after free, full and frequent discussions with an anxiety not to omit the name of any community which was entitled to social or educational relief. But, in several cases, in the absence of any information the decision had to be taken on the strength of the name of the community only, on the principle of giving the benefit of doubt. . . . While it is correct to give, in a welfare state, the benefit of social relief to as large a part of the community as possible, it is unscientific to prepare these lists in this manner. 11 Radhakrishnan (1996) provides a historical account of the Backward Class Movement in Tamil Nadu and details of quotas for major communities in that state. Table 18 in Galanter (1984), lists state-­wise concessions for these classes in 1951–2. 12 Eloquent criticisms of the report are found in Das (1990) and Kumar (1992). 13 In 2005, Article 15 of the constitution was modified to read: (5) Nothing in this article or in sub-­clause (g) of clause (1) of article 19 shall prevent the State from making any special provision, by law, for the advancement of any socially and educationally backward classes of citizens or for the Scheduled Castes or the Scheduled Tribes in so far as such special provisions relate to their admission to educational institutions including private educational institutions, whether aided or unaided by the State, other than the minority educational institutions referred to in clause (1) of article 30. (Constitution (Ninety-­third Amendment) Act 2005) This amendment does not explicitly mention evaluation standards. These were addressed by the Eighty-­second Act, 2000:

The demand for disadvantage   139 nothing in this article shall prevent the State in making of any provision in favour of the members of the Scheduled Castes and the Scheduled Tribes for relaxation in qualifying marks in any examination or lowering the standards of evaluation, for reservation in matters of promotion to any class or classes of services or posts in connection with the affairs of the Union or of a State. These are implicitly extended to Backward Classes by the 2005 Act which permits the state to offer this group the same treatment as the Scheduled Castes and Tribes. 14 Part 1 of Thomas Weisskopf ’s book (2004) is devoted to this comparison. Also see Béteille (2005). 15 Quoted from the opinion of the US Supreme Court in the Grutter v. Bollinger case. The court upheld the use of race as a factor determining admission to the University of Michigan Law School. At the same time, the Court held that undergraduate admissions in the same university were unconstitutional because they explicitly allowed African-­American students to enter with lower scores. Both opinions can be found in the Supreme Court collection of the Legal Information Institute at Cornell University Law School (www.law.cornell.edu). 16 The role of stigma in perpetuating racial inequality is a central theme in Loury (2002). 17 Hutton (1933: Appendix 1) is devoted to the exterior castes under colonial rule and Deliége (1999) contains numerous contemporary examples of their exclusion from mainstream society. 18 J.H. Hutton, the Census Commissioner in 1931 also recognized that while ‘Caste is still of vital consideration in the structure of Indian society. . . . Experience at this census has shown very clearly the difficulty of getting a correct return of caste.’ He records the Superintendent of Census Operations in the south Indian province of Madras lamenting the ‘fluidity of present appellations’ among the traditional barber castes (Hutton 1933:  432): ‘Had caste terminology the stability of religious returns caste sorting might be worthwhile . . . 227,000 Ambattans have become 10,000 Navithan, Nai, Nai Brahman, Navutiyan, Pariyari claim about 140,000 – all terms unrecorded in 1921. Individual fancy apparently has some part in caste nomenclature.’ 19 Table 2.5, p. 68 has data on Congress cadres in the state of Uttar Pradesh. Tables 2.14 and 10.4 contain shares of different communities in the set of all elected members of parliament between 1952 and 1999. 20 These figures are not strictly comparable because of some boundary changes over this period. The 11 per cent figure refers to the British province of Bihar and Orissa and is from the 1931 census. For 1961 and 1991, the figures refer to the states of Bihar and Orissa which include some small regions that were not directly under British rule during the colonial era. By the census of 2001, the state of Jharkhand had been carved out of Bihar, and any census figures for this year are population-­weighted averages of all three states. 21 Lists of Other Backward Castes for each state are available with the National Backward Classes Commission (www.ncbc.nic.in).

References Banerjee, A. and Somanathan, R. (2007) ‘The Political Economy of Public Goods: Some Evidence from India’, Journal of Development Economics, 82 (2), pp. 287–314. Béteille, A. (2005) ‘Matters of Right and Policy’, Seminar 549. Blair, H.W. (1972) ‘Ethnicity and Democratic Politics in India: Caste as a Differential Mobilizer in Bihar’, Comparative Politics, 5 (1), pp. 107–27. Crooke, W. (1974) The Tribes and Castes of the North Western Provinces and Oudh, New Delhi: Reprint by Cosmo Publications (originally published in 1896).

140   R. Somanathan Das, V. (1990) The Mandal Commission Report: A Critical Appraisal, New Delhi: The Institute for Studies in Industrial Development. Deliège, R. (1999) The Untouchables of India, Guildford: Berg. Drèze, J. and Sen, A. (2002) India: Development and Participation, New Delhi: Oxford. Du Bois, W.E.B. (1903) ‘The Talented Tenth’, in The Negro Problem: A Series of Articles by Representative Negroes of To-­day, New York. Dushkin, L. (1967) ‘Scheduled Caste Policy in India: History, Problems, Prospects’, Asian Survey, 7 (9), pp. 626–36. Galanter, M. (1984) Competing Equalities: Law and the Backward Classes in India, Delhi: Oxford University Press. Government of India (1955) Report of the Backward Classes Commission, Volumes I–III, Simla: Government of India Press. Guha, R. (2007) India After Gandhi: The History of the World’s Largest Democracy, New Delhi: Picador. Hutton, J.H. (1933) Census of India 1931: Volumes I and II, Delhi: Gian Publishing House. Jaffrelot, C. (2003) India’s Silent Revolution, Delhi: Permanent Black. Kumar, D. (1992) ‘The Affirmative Action Debate in India’, Asian Survey, 32 (3), pp. 290–302. Loury, G.C. (2002) The Anatomy of Racial Inequality, Cambridge, MA: Harvard University Press. Mendelsohn, O. and Vicziany, M. (2002) The Untouchables: Subordination, Poverty and the State in Modern India, New Delhi: Cambridge University Press. Pande, R. (2003) ‘Can Mandated Political Representation Increase Policy Influence for Disadvantaged Minorities? Theory and Evidence from India’, American Economic Review, 93 (4), pp. 1132–51. Radhakrishnan, P. (1996) ‘Backward Class Movements in Tamil Nadu’, in Srinivas, M.N. (ed.). Caste: Its Twentieth Century Avatar, Delhi: Viking, pp. 110–34. Rodrigues, V. (2002) The Essential Writings of B.R. Ambedkar, New Delhi: Oxford University Press. Singh, K.S. (2002) People of India: Introduction, New Delhi: Oxford University Press. Srinivas, M.N. (1957) ‘Caste in Modern India’, Journal of Asian Studies, 16 (4), pp. 529–48. Sundaram, K. (2006) ‘On Backwardness and Fair Access to Higher Education’, Economic and Political Weekly December, pp. 5173–82. Vijayan, P.P. (2006) Reservation Policy and Judicial Activism, New Delhi: Kalpaaz. Weisskopf, T.E. (2004) Affirmative Action in the United States and India: A Comparative Perspective, London and New York: Routledge.

Part III

Culture and entrepreneurship

7 Markets and the diffusion of institutional innovations Marcel Fafchamps

Introduction Institutions are important for development (e.g. Keefer and Knack 1997; Acemoglu et al. 2005). This is particularly true of market institutions which play a key role in directing valuable scarce resources to their best possible use. Alternative allocation mechanisms exist, such as command and control within large organizations. But most firms in developing countries are quite small and governments often lack the resources to manage large and efficient public services. This means that markets in poor economies are even more central to the efficient allocation of resources than in developed economies (Fafchamps 1997). This is certainly true in sub-­Saharan Africa. Yet the evidence suggests that Africa has weak market institutions and unsophisticated market practices (Fafchamps 2004). This raises two questions: is the lack of market sophistication the result of an inherently market-­unfriendly culture; and what can be done to remedy the current situation. The purpose of this chapter is to offer elements of answer to both questions. We argue that there is no reason to suspect African cultures to be any more inimical to market development than other cultures. This conclusion arises from comparing market institutions in Africa to those in other parts of the world today and in the past. The problem is that many African entrepreneurs are unfamiliar with institutional innovations that have emerged in other parts of the world. This lack of familiarity makes them less competitive and makes African economies less productive. Social and economic isolation is the most likely explanation for the lack of familiarity with modern market institutions. This is suggested by observing that non-­indigenous entrepreneurs in Africa are more likely to use modern market institutions (Fafchamps 2004). Non-­indigenous groups have managed to establish themselves in a number of African countries. Their historical and geographical origin is quite varied, but they seem to share a better familiarity with a small number of key innovations, not just in terms of production technology and internal organization of the firm, but also in the way they interact with others through the market. This familiarity has often enabled non-­indigenous entrepreneurs to outcompete native Africans in business. Unfortunately, the institutional

144   M. Fafchamps innovations introduced by non-­indigenous groups have not always spread beyond the confines of their own social group. The unanswered question is why? We suspect it has to do with political conditions that do not favour social and economic integration.

Market institutions To understand market institutions in Africa, it is essential to recognize that an efficient organization of market exchange results from the combination of three essential components: formal rules of conduct enforced by courts; social norms; and expectations of behaviour. Laws and courts The policy focus to date has been on formal rules of conduct, namely laws and regulations. Researchers, for instance, have divided countries according to their legal tradition and investigated whether countries that inherited a common law tradition from their British colonizers have fared better than countries that inherited a Roman law tradition from the French and other European colonizers (e.g. Beck et al. 2003a, 2003b). The literature has also focused on the legal environment required for specific markets to operate smoothly. Calls for land titling are a case in point, given the role that land can play as loan collateral (Deininger 2003). Much attention has also been paid to the organizations in charge of promulgating and sanctioning laws and regulations. The process by which laws are passed and regulations enacted by governments is the object of an ever expanding political economy literature. Some attention has also been paid to the organizations that sanction these formal rules, namely courts and, to a lesser extent, prisons (e.g. Messick 1999; Ministère de la Justice 1999). Figure 7.1, for instance, shows that countries differ in the extent to which they regard courts as fair. The comparison suggests that African entrepreneurs are less inclined to believe that courts would treat them fairly, possibly because of corruption, possibly because the law itself is regarded as unfair. With a few well-­noted exceptions (e.g. North 1990; Acemoglu et al. 2005), economists to date have paid less attention to the role that social norms and rules of behaviour play in market performance. Yet laws and courts can accomplish little if societies choose to ignore them. As Table 7.1 suggests, for instance, only a minority of African entrepreneurs actually use courts to resolve commercial disputes. Laws can only affect market efficiency if they are internalized in social norms and expectations of behaviour. This takes us into the realm of business culture (e.g. Greif 1994; Fukuyama 1995). Laws can influence behaviour in two ways: directly by serving as a deterrent through the threat of court action; and indirectly by influencing social norms. For the threat of court action to work as deterrent against opportunistic breach of contract, this threat must be credible. This is seldom the case for small transactions because the magnitude of the loss is not commensurate with the direct and indirect cost of court proceedings. Even for

Markets and the diffusion of innovations   145 1 0.9 Cumulative frequency

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1

0

2

3

4

5

6

From 0 = very unfair to 6 = quite fair India

Kenya

Morocco

Saudi Arabia

Vietnam

Ethiopia

Tanzania

Uganda

Figure 7.1 Belief in fairness of courts (source: Author’s calculations from World Bank ICA Survey data). Table 7.1  Recourse to legal institutions Ghana (%)

Kenya (%)

Following a dispute with a supplier Ever saw a lawyer Ever went to court Ever used arbitration Threatened to call the police Ever called the police

13 2 4 5 5

6 0 0 2 0

Following a dispute with a client Ever saw a lawyer Ever went to court Ever used arbitration Threatened to call the police Ever called the police

8 6 4 14 5

38 21 6 4 4

Source: Fafchamps (2004: Table 4.5).

large transactions, civil action is not worth initiating if the debtor has no assets to foreclose upon. This immediately puts the overwhelming majority of market transactions in Africa de facto beyond the reach of courts. This is especially true in the so-­called informal sector where microenterprises dominate, entrepreneurs are poor, and transaction sizes are small.

146   M. Fafchamps Relationships and reputation It is important to recognize that, even in developed economies, the threat of court action is not credible for most market transactions. Think of all the transactions we engage in as consumers. For large transactions, such as the purchase of a home or unfair dismissal, the amounts at stake are typically large enough for the threat of court action to be credible. For most everyday transactions, however, the amount involved is so small that we would never realistically engage court action. Yet we expect contract compliance with respect to quality, warranty, delivery date, etc. If the threat of court action is not what disciplines the seller, what is? For many transactions, the fear of losing a valuable relationship serves as a deterrent to opportunistic behaviour. This implies that expectations of behaviour depend on expectations of continued interest in the business relationship: only those who wish to continue the relationship fear losing it. An immediate corollary is that transactions between firms or individuals unknown to each other are risky. This serves as a disincentive to experiment with new clients and new sources of supply – and reduces market flexibility. One important solution is reliance of reputation mechanisms. What distinguishes modern economies is the widespread use of a multiplicity of reputation mechanisms. There are old ones – like publicizing the names of those who fail to pay a bill of change. There are new ones – like eBay feedback. Other examples include credit rating agencies, specialized press, and consumer reports. These reputation mechanisms typically rely on laws and courts to punish misreporting. This support role of laws and courts may in practice be more important than their direct role in the enforcement of contracts themselves. For reputation mechanisms to work, economic agents must be unambiguously identifiable. There are formal institutions devoted to this purpose alone, such as business registration. The terms used in contracts must be given an unambiguous meaning. There are formal institutions devoted to this purpose as well, such as quality control agencies, industry standards, weights and measures. In some cases, the formal institutions are private (e.g. ISO certification), in others the government intervenes directly (e.g. weights and measures). As soon as a name acquires a reputation for quality and reliability, this name becomes valuable and the fear of losing the reputation associated with the name itself becomes a deterrent against breach of contract (Tadelis 1999). Brand name recognition is an example of reputation value attached to a name or identifier. Franchising can be seen as a way to capitalize on such a reputation by guaranteeing a certain level of quality to consumers. Franchising is common in hotels and restaurants that cater to the needs of travellers. By giving value to a name, reputation processes generate incentives for undue appropriation: it is in the interest of fly-­by-night and cheap imitators to posture as the real firm or brand name in order to reap a profit from defrauding customers. For formal reputation mechanisms to work, business names must therefore be protected. This again calls for protection from laws and courts – trademark protection, copyright, etc.

Markets and the diffusion of innovations   147 What this discussion shows is that laws and courts are important institutions for supporting markets, but their role goes much beyond the simple enforcement of private contracts. In fact, one could even argue that, with the possible exception of large transactions, courts play little or no role in deterring opportunistic breach directly. The most important role of courts is indirect, that is, in supporting a large set of private and public institutions that activate various reputation mechanisms. It is the fear of losing a valuable reputation that is the real deterrent to opportunistic breach of contract. Once this is understood, it becomes readily apparent that multiple equilibria are likely to arise. If all economic agents act in a reliable fashion, breach of contract is interpreted as a sign of fraud or impending bankruptcy, and results in an immediate loss of reputation and, consequently, of business. Breach of contract results in an immediate fall in profit. However, if all economic agents interpret contract terms in a flexible manner, breach of contract provides little information regarding the state of one’s business. As a result, loss of reputation is not immediate (Fafchamps 2002). This argument can be generalized because all enforcement mechanism are ultimately vulnerable to massive defection. We have just seen that this is the case for reputation mechanisms. It is also true for courts since they could not cope with the case volume. It is even true for situations in which the fear of losing a relationship is what disciplines parties: once a pattern of behaviour is established, e.g. paying late, it is very difficult to credibly threaten to sever the relationship when this pattern is repeated. It follows that for a norm to be enforceable, it must be followed by the majority of the intended population. The equilibrium in which an economy is can be taken as one dimension of ‘business culture’: in some business environments, contracts are seen as hard constraints, in others they are nothing but a general guide from which parties can depart at will. These patterns of behaviour generate expectations of behaviour which, over time, become enshrined in social norms, i.e. what is regarded as ‘normal’ behaviour. Of course what is normal in one location or sector may not be normal elsewhere. Multiple equilibria need not be equivalent in terms of economic efficiency. Some equilibria may dominate others. The question is: how can we move an economy from an inferior equilibrium to a superior one? Since this requires changing social norms, the answer to this question is unclear. But it probably has something to do with the social context within which market transactions take place. To this we now turn. Embedded markets Norms and expectations are themselves embedded within a social structure consisting of associations and overlapping social networks (e.g. Granovetter 1985; Putnam et al. 1993). This is true anywhere, even in the most developed economies. But it is particularly true when formal market institutions are weak or nonexistent and the enforcement of contracts relies almost exclusively on informal enforcement mechanisms (e.g. Bernstein 1992, 1996). In such economies, the

148   M. Fafchamps social context becomes essential because it provides the information sharing needed for reputation mechanisms as well as the environment in which norms of acceptable behaviour get formed. The shape of this social context is itself affected by culture because business patterns of socialization at least partly reproduce private ones. Belonging to the same church, temple or golf club predisposes individuals to exchange business information and emulate each other’s business practices (e.g. Greif 1993; Fafchamps and Minten 1999; McMillan and Woodruff 1999a, 1999b). Socialization within the extended family can perform the same function. Social contacts can assist the functioning of markets in many different ways. They can circulate valuable information about trade opportunities, prices and market conditions. They can convey information about hidden characteristics of products and economic agents. They can also provide the context for collective action, for instance to lobby political authorities for support, or to exclude individuals whose have violated norms of conduct. Table 7.2 illustrates this for information on market conditions. In my research I have shown that social contacts are valuable for business: individuals who are better connected have higher profits, are better able to obtain supplier credit, and grow their business faster (e.g. Fafchamps and Minten 2001, 2002; Fafchamps 2003). Table 7.3 shows results from regression analysis that illustrates that this is indeed the case. Table 7.2  Sources of information on market conditions Small

Medium

Large

Total

A. Prices: Other traders Suppliers and clients Messengers Public sources

81% 15% 4% 0%

61% 31% 8% 0%

40% 37% 23% 0%

60% 28% 12% 0%

B. Supply conditions: Other traders Suppliers and clients Messengers Public sources

32% 65% 2% 1%

20% 76% 4% 0%

19% 68% 12% 0%

23% 70% 6% 1%

C. Demand conditions: Other traders Suppliers and clients Messengers Public sources

30% 68% 1% 1%

11% 86% 2% 2%

10% 79% 9% 2%

16% 78% 4% 2%

Number of observations

227

254

243

729

Source: Fafchamps and Minten (1999). Notes Table reports the main source of information on the following: firm size: size categories are based on total sales.

Coeff. 0.460 n.a. 0.298 0.840 −0.245 0.040 0.143 −0.248 Included but not shown 3.052 0.446 636

Regressors

Number of traders known (log) Member of trader association (yes = 1) Working capital (log) Manpower, in man-months (log) Female trader (yes = 1) Years of education Years of experience in trade (log) Number of languages spoken Regional dummies Intercept R-squared Number of observations

Source: Fafchamps and Minten (2001).

Madagascar OLS

(Dependent variable in log)

Table 7.3  Determinants of value added among agricultural traders

1.88

8.38 8.14 −2.18 2.01 1.56 −2.17 4.25 0.882 0.425 0.438 535 585

t-stat. 6.19 6.44 10.21 −1.93 −1.01 −1.52 0.60 1.65

Coeff. 0.328 1.000 0.628 −0.200 −0.189 −0.034 0.066 0.089

5.74

t-stat.

Benin OLS

3.695

0.301 −0.018 0.591 0.361 −0.353 −0.026 0.065 −0.044

Coeff.

Malawi OLS

10.97

5.42 −0.06 12.56 3.06 −2.85 −1.39 0.80 −0.70

t-stat.

150   M. Fafchamps If markets are embedded in a social context and social links are economically beneficial, the social context will affect how markets operate. Perhaps the most visible manifestation of this simple reality is the role that ethnicity and religion play in the development of markets. In most countries, people tend to intermarry with members of their caste, ethnicity, religion and social class. Since many interpersonal contacts follow – or are influenced by – extended family lines, socialization tends to be segmented along the same lines. It is common for business socialization to reproduce, at least in part, whatever segmentation is present in the private realm. This means that business communities are often segmented along ethnic, religious and social lines – even though there may be a lot of variation from country to country as to which factor of differentiation is most prominent in segmenting social interaction (Fafchamps 2000). This variation is illustrated, for instance, in Table 7.4. Segmentation along ethnic or religious lines is often reinforced by other processes. Signalling, for instance, arises whenever a costly signal – e.g. participation in religious activities – is used by individuals to show that they are reliable business partners (Ensminger 1992). Statistical discrimination – and other forms of discrimination – arises whenever able but less well connected individuals find it difficult to establish their credentials and, as a result, are unable to gain a foothold in the dominant business community (Fafchamps 2000). Fortunately, such reinforcing processes are not universal. Using survey data from three African countries, I have shown that agricultural markets are not subject to ethnic bias even though network effects are quite strong. When multiple reinforcing processes are at work, however, once a group becomes dominant in a given activity, for whatever reason (e.g. historical or accidental), this group tends to remain dominant as long as there are benefits from network proximity. This has been documented, for instance, in the manufacturing sector of certain African countries (e.g. Himbara 1994; Fafchamps 2000).

Institutions and development The segmentation of market exchange along ethnic or religious lines can be extremely damaging for development because it acts as a barrier to the diffusion of technological and institutional innovations. To this we now turn. Growth and innovation Over the last 200 years the world has witnessed unprecedented growth in standards of living. During the pre-­industrial era, the prosperity that empires brought for their citizens was always limited and much of it was achieved by impoverishing someone else. The Industrial Revolution brought an application of science to technology. This dramatically increased the productivity of the invention process, making it possible for societies to increase their standards of living

Source: Fafchamps (2004).

African 82 Asian 3 European 6 Mid-Eastern 0 Other 1

81 2 14 1 3

Burundi (%) Cameroon (%) 60 0 23 7 10

83 0 1 0 15

91 0 1 8 0

Côte Ethiopia (%) Ghana (%) d’Ivoire (%)

Table 7.4  Ethnic composition of the ownership of African manufacturing firms

42 51 4 2 2

Kenya (%) 73 24 0 2 1

Tanzania (%)

59 26 13 2 0

Zambia (%)

33 13 47 0 7

Zimbabwe (%)

152   M. Fafchamps through innovation. Before the Industrial Revolution, science had little or no practical intent. Applying scientific discoveries to technology is what has made it possible to produce more with less. Without this, the massive increase in world prosperity would not have been possible. In the initial phases of the Industrial Revolution, science was applied primarily (or at least most visibly) to the physical production process itself – e.g. the steam engine, the cotton gin. Hence the nickname ‘Industrial Revolution’. But this obscures the fact that science can be applied to all sectors of the economy, not just industry. The application of science to agricultural technology has led to the green revolution. The service sector has also benefited from the application of science to technology, for instance through the use of office machinery, transport equipment, computers, telephones, and tall buildings to facilitate agglomeration externalities. All these examples focus only on material things – equipment, machinery, agricultural inputs. But science can also be applied to immaterial things, like organizations and institutions. This is where economics comes into the picture. Thanks to science in general and economics in particular, we now know that hyperinflation is a scourge, that auctions are a good way of allocating goods, that managers work more if provided with strong incentives, that markets are a cost-­ effective way of decentralizing the allocation of goods and services. A continued increase in prosperity requires not only that the physical production process be continuously adapted to keep up with innovations. It also requires that institutions and organizations be continuously adapted as well. For instance, the physical infrastructure for providing telephone services has changed dramatically in recent years. The older technology had massive returns to scale: there was no point in duplicating wires installed in the ground or on telephone poles. This kind of technology generated a natural monopoly – hence the need for regulation, hence the tendency to have government providing telephone services. Cellular phones have changed all that. The technology is different, it does not have such large returns to scale, there is room for multiple providers and hence room for competition. With sufficient competition, regulating prices becomes unnecessary. This calls for a change in institutions – a change in the law, an abolition of state monopoly – and a change in organization, e.g. auctioning off the airwaves. Maintaining old institutions and organization becomes counter-­ productive, and ultimately hurts consumers and businesses. Change in institutions and organizations is needed if the full benefits of innovation in physical production are to be achieved, and this change must itself build on the application of science to technology – in this case, the economics of competition, regulation and auctions. Improving market institutions While social networks and business communities play an absolutely essential role in Africa today, nearly exclusive reliance on them is fraught with problems. One problem is that social networks are not equitable, they do not provide a

Markets and the diffusion of innovations   153 level playing field, they do not provide equality of opportunity. Informal market institutions typically result in a system that is inequitable as well as inefficient: jobs do not go to the most qualified person, orders do not go to the best suppliers, and loans do not go to the most dynamic entrepreneur (e.g. Barr and Oduro 2002; Fafchamps 2004). Moreover, because social networks and business communities are often organized around ethnic lines, the day-­to-day operation of social networks results in ethnic bias. With time, this generates rancour and political tension which serve as disincentives to investment. The solution is better market institutions, institutions that are open to all, regardless of race, gender, religion, language or ethnicity. This means formal institutions. Formal institutions include the police and courts that protect property rights against thieves and crooks. But, as we have seen, they also include a large number of specialized institutions, some private, some public, some hybrids. Examples include: credit reference and credit rating agencies such as Dun and Bradstreet and Moodys; private and public agencies that set grades and standards; quality certification agencies such as ISO certification; bonded warehousing services that are essential to organized commodity exchanges; auctions and auction floors for grain, coffee, copper, fish, foreign exchange, futures, stocks, airwaves; many kinds of financial institutions and the innovations they bring, such as the letter of credit or hire-­purchase contracts, secondary markets for mortgages, junk bonds; bankruptcy liquidators; customs and port services; secondary market in machinery and equipment; formal business associations with open membership. The list goes on. These formal institutions are innovations. They result from the application of science to technology, in this case the technology of market exchange. Like other forms of technological innovation, they only happen if people know about them, become familiar with them, and make them part of their social norms. Business practices To survive in a given environment, familiarity with local practices is essential. To most people this familiarity comes with upbringing and imitation. But when people are thrown into unfamiliar circumstances, they have to adjust to their new environment. This means changing their behaviour. Most African entrepreneurs are unfamiliar with the conditions prevailing elsewhere, with modern institutions for economic exchange. This means they find it difficult to survive when faced with competition from entrepreneurs from other parts of the world, who are familiar not only with more productive physical technology but also with more efficient modes of organization, both within the firm and in market exchange with other firms. This lack of competitiveness is visible not only in poor export performance but also in the displacement of domestic firms by foreign firms in African domestic markets. The success of Asian and Lebanese businessmen in a few African countries is striking in this respect. Perhaps even more striking is the

154   M. Fafchamps success of natives from Central and West Africa as small entrepreneurs in South Africa. How is it possible that these individuals perform better than native entrepreneurs in an alien environment? The reason is that they bring familiarity with relatively simple innovations, such as networking, invoicing and supplier credit, that outperform local institutions. These institutional innovations are simple enough that they can survive in a weak institutional context – mostly in combination with strong social networks. African entrepreneurs can learn how to survive in the global village. The question is: who will they learn from? Ultimately the new knowledge comes from scientists and innovators. But few people learn directly from the innovators themselves. For most people, learning comes from imitation. To imitate, there must be someone nearby to imitate from. This is where things get complicated. In their overwhelming majority, African entrepreneurs operate very small businesses and do not have much knowledge regarding technological and institutional innovations used elsewhere. With sufficient funding, they can acquire the new technology by importing the necessary equipment and learning how to use it. Funding is not enough, however. African entrepreneurs are also largely unfamiliar with institutional innovations in widespread use elsewhere. Unlike physical technology, an individual entrepreneur cannot simply import the institution and learn how to use it. An institution is not something that can physically be moved from one place to another. It is a complex combination of formal rules, expectations of mutual behaviour and social norms of conduct. Importing one component only – say the formal rules – will not work without the other two. This is because formal rules can only be enforced if a sufficiently large proportion of the target audience follows them. While it is possible for a single entrepreneur to adopt a new technology independently from others, institutional change is subject to coordination failure: a large enough group of people have to change the way they do things at the same time for institutional change to be sustainable. This makes decentralized institutional change slow and unpredictable.

Diffusion of institutional innovations So how can we improve the institutional environment in which African firms operate? One possibility is to give them time. With sufficient time, they will eventually come up with innovations that improve upon the current situation. For this process to be reasonably fast, it must be based on the application of science to technology, that is, the application of social sciences to institutional design. Economists thus have a crucial role to play in this transformation process since their special skill is precisely the application of science to institutional design. Even if we were willing to assume that policy-­makers listen to evidence-­ based advice, homegrown institutional innovation would take a very long time –

Markets and the diffusion of innovations   155 probably of the same order of magnitude as it took Europe to come up with its market institutions and export them to other parts of the planet, that is, several hundred years (North 1973). And even this would require developmental states, that is, states that have their own development in mind and are willing to make the decisions required to foster it. There have been plenty of counter-­examples in Africa. So, left to its own devices, a homegrown process would take a very long time. There is no need to reinvent the wheel, however. Africa can just imitate others by adapting and adopting institutional innovations introduced elsewhere. The adaptation and introduction of institutional innovations requires government intervention. But, as we have noted earlier, it also requires changes in social norms and patterns of behaviour. Social norms and expectations of behaviour that help shape business practices must spread through the business community for adoption to be complete. This diffusion process can be understood as a network diffusion process. Our understanding of such processes has greatly improved over recent years thanks to epidemiology models on networks (Vega-­Redondo 2006). The main intuition of this literature can be illustrated with a simple example. Take a contagious disease that spreads through close contact. This disease will spread across the social network, first reaching individuals who are closely connected to the originator of the disease. This means that social segmentation will slow the spread of the disease across ethnic lines. Individuals or groups who are socially isolated from the infected part of the network may never contract the disease. By analogy, a useful business practice will spread through the social network, reaching first those individuals most closely connected with the source of the innovation, and then spreading slowly to other parts of the network. Here too social segmentation will slow the spread of adoption of the innovation, and may even result in some groups being bypassed altogether. As a result, we may observe different groups following different business practices at the same time. In this sense it becomes possible to talk about the business culture of a specific community within a given country at a specific point in time. The analogy with the spread of a disease is imperfect, however. In the case of a disease, contact with a single infected person is sufficient to catch the ailment. For institutions, this is not the case. For someone to switch behaviour, a sufficient number of his or her immediate neighbours must have already changed their own behaviour. This is because market interaction based on social norms and expectations of behaviour results in local network externalities. The upshot of this observation is that the transfer of institutional innovations across social networks requires dense connections. A single bridge is not enough. This means that social segmentation is much more likely to hinder the spread of institutional innovations than it is to hinder the spread of disease. The successful spread of institutional innovation throughout society requires a high level of social integration along ethnic and religious lines (e.g. Vega-­Redondo 2004; Galeotti 2007; Jackson 2007; Young 2007).

156   M. Fafchamps Foreign investors Armed with this understanding, we can now consider the most likely source of information on institutional innovations, namely, foreign investors. Letting foreign investors into Africa brings them close enough that they can in principle be observed and imitated by African entrepreneurs. This avenue offers the fastest growth potential for Africa. But is it sufficient to ensure that institutional innovations diffuse widely? This ultimately depends on the level of social integration between foreign investors and local entrepreneurs. We have seen that when market institutions are weak, social networks acquire a critical role as a substitute for formal institutions. Network externalities naturally lead to a selection bias. This bias is typically of an ethnic or religious nature, simply because socialization anywhere primarily takes place along ethnic and religious lines. Once a group becomes overrepresented among the modern business class, entry becomes more difficult for those who do not have the right background. If foreign investors are much more sophisticated in their technology and business practices, they are very likely to outperform local entrepreneurs. If the gap between foreign investors and local entrepreneurs is too large, the latter will be forced to exit before they have had the time to learn about the new technology and business practices. Furthermore, once a foreign business class has gained a foothold in a specific industry or country, network externalities will favour further investment and entry by members of the same class. This process in turn may thwart the very imitation by African entrepreneurs that it was intended to foster. The histories of South America and South Africa are not reassuring in this respect: in both cases white settlers gained long-­lasting economic supremacy over indigenous people. Although in both cases political factors played a key role in this process, the knowledge gap between indigenous and foreign entrepreneurs was wide at the outset. As a result, the comparative advantage of foreign investors may have been so large that imitation became very difficult. Over time this often means that foreign investors – and their descendents – face political resentment. This further limits opportunities for imitation if imitation is seen as politically alienating. There is also the possibility that a wealthy entrepreneurial class may use its financial might to seek political control, thereby furthering its economic advantage and further complicating social integration and imitation by native entrepreneurs. The fear of losing control to a foreign entrepreneurial class may explain the mixed feelings that many Africans have towards foreign interference in general, and foreign investors in particular. As a result, foreign entrepreneurs in Africa often find themselves in a politically subordinate situation, not dissimilar to those of the Jews in medieval Europe or of Armenians in the Ottoman empire – i.e. a business class without political rights subject to the whims of the prince and readily usable as scapegoats to feed public resentment. In this, Africa is not very different from other parts of the world. China and India, for instance, remained closed to foreign investment for a long time. They

Markets and the diffusion of innovations   157 only began to enjoy serious growth when they opened up to foreign investors. But they did so in a position of strength, when they were confident it would not entail a loss of control over their own affairs. I suspect that Africa will similarly open up to foreign investment once it is ready for the challenge, that is, once African elites are confident they will be able to imitate foreign investors and will not lose control to foreign interests. When this condition is satisfied, average productivity and competitiveness will rise across the board – not just in small foreign enclaves.

Conclusion The application of science to technology is the ultimate driving force behind the massive increase in standards of living that the world has witnessed over the past 200 years. This includes technological innovation in the production process, but also innovations in institutional and organization design. In many instances, the two can only work together. An immediate corollary is that an upgrading of market institutions is essential for development and growth. Modern market institutions are a complex mix of formal rules and social norms, the main purpose of which is to support reputation mechanisms. When these modern institutions are absent, business networks serve as an imperfect substitute. Private patterns of socialization influence the formation of business networks and often result in the overrepresentation of specific ethnic or religious groups. This fuels politicial tension and inhibits the spread of institutional innovations. The end result is a low equilibrium trap in which countries discourage foreign investors for fear of losing control. The solution is to promote social integration, especially in business circles. Integrating socially with foreign investors is the best short-­term strategy for getting Africa out of a vicious circle in which its entrepreneurs are not sufficiently cognizant of up-­to-date institutional practices to form an attractive business environment for modern production, but local elites are reluctant to let foreign entrepreneurs fill this need for fear of losing political control.

References Acemoglu, D., Johnson, S. and Robinson, J. (2005) ‘The Rise of Europe: Atlantic Trade, Institutional Change and Economic Growth’, American Economic Review, 95 (3), pp. 546–79. Barr, A. and Oduro, A. (2002) ‘Ethnic Fractionalization in an African Labour Market’, Journal of Development Economics, 68 (2), pp. 355–79. Beck, T., Demirguc-­Kunt, A. and Levine, R. (2003a) ‘Law and Finance: Why Does Legal Origin Matter?’, Journal of Comparative Economics, 31 (4), pp. 653–75. —— (2003b) ‘Law, Endowments, and Finance’, Journal of Financial Economics, 70 (2), pp. 137–81. Bernstein, L. (1992) ‘Opting Out of the Legal System: Extralegal Contractual Relations in the Diamond Industry’, Journal of Legal Studies, XXI, pp. 115–57. —— (1996) ‘Merchant Law in a Merchant Court: Rethinking the Code’s Search for

158   M. Fafchamps Immanent Business Norms’, University of Pennsylvania Law Review, 144 (5), pp. 1765–821. Deininger, K. (2003) Land Policies for Growth and Poverty Reduction, Cambridge: Cambridge University Press. Ensminger, J. (1992) Making a Market: The Institutional Transformation of an African Society, New York: Cambridge University Press. Fafchamps, M. (1997) ‘Introduction: Markets in Sub-­Saharan Africa’, World Development, 25 (5), pp. 733–4. —— (2000) ‘Ethnicity and Credit in African Manufacturing’, Journal of Development Economics, 61 (1), pp. 205–35. —— (2002) Spontaneous Market Emergence, Topics in Theoretical Economics, Vol. 2 (1), Article 2, Berkeley Electronic Press at www.bepress.com. —— (2003) Ethnicity and Networks in African Trade, Contributions to Economic Ana­ lysis and Policy, Vol. 2 (1), Article 14, Berkeley Electronic Press at www.bepress.com. —— (2004) Market Institutions in Sub-­Saharan Africa, Cambridge, MA: MIT Press. Fafchamps, M. and Chiu, R. (2007) Making Markets Work, unpublished report submitted to the World Bank. Fafchamps, M. and Minten, B. (1999) ‘Relationships and Traders in Madagascar’, Journal of Development Studies, 35 (6), pp. 1–35. —— (2001) ‘Social Capital and Agricultural Trade’, American Journal of Agricultural Economics, 83 (3), pp. 680–5. —— (2002) ‘Returns to Social Network Capital Among Traders’, Oxford Economic Papers, 54, pp. 173–206. Fukuyama, F. (1995) Trust: The Social Virtues and the Creation of Prosperity, New York: The Free Press Paperbacks. Galeotti, A. (2007) ‘A Theory of Strategic Diffusion’, mimeograph. Granovetter, M. (1985) ‘Economic Action and Social Structure: The Problem of Embeddedness’, Amer. J. Sociology, 91 (3), pp. 481–510. Greif, A. (1993) ‘Contract Enforceability and Economic Institutions in Early Trade: The Maghribi Traders’ Coalition’, Amer. Econ. Rev., 83 (3), pp. 525–48. —— (1994) ‘Cultural Beliefs and the Organization of Society: A Historical and Theoretical Reflection on Collectivist and Individualist Societies’, J. Polit. Econ., 102 (5), pp. 912–50. Himbara, D. (1994) ‘The Failed Africanization of Commerce and Industry in Kenya’, World Development, 22 (3), pp. 469–82. Jackson, M. (2007) ‘Naive Learning in Social Networks: Convergence, Influence and the Wisdom of Crowds’, mimeograph. Keefer, P. and Knack, S. (1997) ‘Why Don’t Poor Countries Catch Up? A Cross-­National Test of Institutional Explanation’, Economic Enquiry, 35 (3), pp. 590–602. McMillan, J. and Woodruff, C. (1999a) ‘Dispute Prevention Without Courts in Vietnam’, Journal of Law, Economics and Organization, 15 (3), pp. 637–58. McMillan, J. and Woodruff, C. (1999b) ‘Interfirm Relationships and Informal Credit in Vietnam’, Quarterly Journal of Economics, 114 (4), pp. 1285–320. Messick, R.E. (1999) ‘Judicial Reform and Economic Development: A Survey of the Issues’, World Bank Research Observer, 14 (1), pp. 117–36. Ministère de la Justice (1999) Justice selon les justiciables: Une enquête auprès des usagers du système judiciaire [Justice According to Citizens: A Survey Among Users of the Judicial System], Gouvernement de Madagascar, Antananarivo. North, D.C. (1973) The Rise of the Western World, Cambridge: Cambridge University Press.

Markets and the diffusion of innovations   159 —— (1990) Institutions, Institutional Change, and Economic Performance, Cambridge: Cambridge University Press. Putnam, R.D., Leonardi, R. and Nanetti, R.Y. (1993) Making Democracy Work: Civic Institutions in Modern Italy, Princeton: Princeton University Press. Tadelis, S. (1999) ‘What’s in a Name? Reputation as a Tradable Asset’, American Economic Review, 89 (3), pp. 548–63. Vega-­Redondo, F. (2004) ‘Diffusion, Search, and Play in Complex Social Networks’, mimeograph. —— (2006) Complex Social Networks, Cambridge: Cambridge University Press. Young, P. (2007) ‘Innovation Diffusion and Population Heterogeneity’, mimeograph.

8 Culture, management and development1 Philippe d’Iribarne

Multiple research currents offer widely contrasted views of the relationship between culture and development. Some attribute development difficulties to traditional cultures and the resistance to change that they generate, which leads them to advocate modernization policies (McClelland 1961; North 1990). Others, on the contrary, emphasize the effects of situations of dependency that generate a culture of apathy, which leads them to advocate empowerment policies (Douglas 2004). Some take an interest in entrepreneurship, and others look more to the issues of coordination and transaction costs. Nevertheless, these diverse research currents share one thing: they see culture as a set of customs, beliefs and values. In these conditions, they lead to the affirmation that cultural change is necessary to transform practices so as to make it possible to break out of underdevelopment. A set of empirical work comparing how companies operate in diverse places around the world has led to another vision of culture in which culture is seen as forming a framework of meaning associated with a general concept of mankind and society. This framework is very stable across the centuries, and it cannot be modified to allow development. It is, however, compatible with very diverse practices. A lot can be done, in the ways companies and institutions operate, to build practices that are both meaningful in this framework and conducive to development. Management issues within companies are therefore central. The same is true for governance issues on a larger scale.

The standard view of culture: a set of practices In the sociological tradition, the notion of culture was long linked to the notion of community, in the sense given to this term where one compares community (Gemeinschaft) – in which each individual’s existence is governed by a set of inherited beliefs and standards that impose themselves on the individual – to society (Gesellschaft) – a free association of individuals attached to their interests and masters of their values (Tönnies 1887; Weber 1922). Culture is then seen as being on the side of community. It is assumed to be associated with customs, traditions, and the inherited values that legitimate them. Even when social sciences place the emphasis on the role of actors (agency), they continue to associate the notion of culture with clearly defined behaviours.

Culture, management and development   161 Thus, A. Giddens’ structuration theory (1984) is oriented in this direction. The existence of habitual ways of acting, of routines, is at the heart of the theory. Actors are seen as generally following these routines and likely to modify them constantly. Hence, culture is seen as largely created by actors. The classic approaches to the relationship between culture and development have been inspired by such a vision. Thus, for North (1987: 421), culture is linked to ‘communities of common ideologies and of a common set of rules that all believe in’. And, for Douglas (2004), culture is ‘a way of thinking that justifies a way of living’. Yet, the observation of how companies operate, in both developed and developing countries, reveals another dimension of culture: a framework of meaning that is both inherited and compatible with very diverse practices.2 This plays a large role in questions of economic efficiency, and thus of development.

Cooperation concepts that resist modernization while being compatible with wide-­scale transformations in ideas and behaviours When comparing how organizations operate in ‘modern’ societies – the United States, European countries – one sees the extent to which these modes of operation differ from one country to the next. This is because they are marked, in each country, by concepts of cooperation that are rooted in a unique vision of mankind and society. Let us look at the form of social organization based on strict enforcement of property rights in which North sees the only possible path to development. It is part of a very specific tradition, unique to the Anglo-­Saxon world and more specifically to American society. It consists in organizing society in such a way that each individual has, as much as possible, the feeling that no one else controls his or her destiny. Contracts, by which each individual commits to the extent that they conform with his or her will, thereby play a central role. In companies, the utmost is done to bring day-­to-day operation into line with the mythic image of a society of property owners that cooperate through freely negotiated contracts (d’Iribarne 1989). Individuals are seen as owning themselves, negotiating the use of their labour as they would negotiate goods that belong to them. In this way, even when placed in a subordinate position, they appear to remain in control of their destiny (Foner 1998). This contract-­based logic concerns the relations between a superior and a subordinate, the latter seen as one of his or her superior’s suppliers, and the former assimilated to a client. In the same way, it concerns the relationships between a ‘supplier’ service and a ‘client’ service, or, in unionized businesses, the relationships between the company and the union. So-­called universal management practices (goal-­based management, budgetary control, development of in-­house company contracts with internal transfer prices) represent as many ways of implementing, within companies, a form of cooperation that corresponds to this American ideal of contractual relations

162   P. d’Iribarne between a supplier and a client. Thus, the careful separation of responsibilities – which leads to setting freely negotiated objectives supposed to faithfully represent each party’s contribution to the common work – goes in this direction, even if this faithfulness is, in fact, highly relative (Ouchi 1981). We are dealing with a vision of mankind and of society bequeathed by history. Thus, for Locke, the characteristic of a free man’s work is that his obligations are strictly defined by the contract concluded between him and his employer (Locke 1690: 322). In France, the ideal vision of mankind and society is very different. What ­everyone seeks to avoid above all is their being reduced to a servile position. References to the specific position each individual occupies in society, to the rank associated with this position, and to the specific rights and duties it imposes are omnipresent. With regard to hierarchical relationships as well as relations with clients, it is constantly a question of complying with the standards of one’s profession, of what the profession demands that one do or forbids one from doing lest one should ‘déchoir’ (i.e. position oneself below one’s due status). France’s ancien régime was already marked by the concept of a society based on such a fear and on the reference to the rights attached to a specific social position as the means of protecting against that fear (de Tocqueville 1856). The contemporary reference to one’s profession, to professionals, to the profession’s importance, and the full set of mental images and practices it carries, ensures the link between what one goes through in organizations’ day-­to-day life and the collective imagination that marks French society. In this way, this reference provides a way of connecting the work one does in a subordinate position, actually subject to a boss’s authority, to a vision of independence, honour and nobility. When individuals working for a company see themselves as guided by the vision they have of their profession, seeking to live up to the duties associated with it allows them not to think too much about the constraint and interest elements implied, despite it all, by their situations. In addition, the way authority is exercised in French companies truly leaves room for the form of autonomy that professionals have (d’Iribarne 1989, 2006). The equivalent is to be found in client relations. Generally speaking, management practices implementing forms of coordination that make up local references have progressively taken shape in the various developed countries. Of course, they have hardly been turned into theories, and are not taught in management schools, but in practice they play a decisive role. Yet, this inheritance hardly resembles what one habitually calls culture, i.e. well-­defined ways of behaving. What is at stake here rather consists in a general interpretive framework. This framework is marked by the opposition between certain feared-­above-all types of peril, different from one society to the next (having one’s destiny controlled by others in the United States, being reduced to servitude in France, etc.), and the paths to salvation that allow these perils to be avoided (d’Iribarne 2008). The vision of an ideal society – a society in which one would have effectively warded off what is perceived as the utmost peril –

Culture, management and development   163 bears the mark of this framework of meaning. Such an inheritance is not only seen in the proponents of traditional social orders. It contributes just as much to giving shape to the way in which a society modernizes. It influences the form that democratic institutions take (d’Iribarne 2003b, 2008) and similarly conditions the type of labour law established to protect employees.

Successful companies in developing countries In developing countries, the implementation of management practices in accordance with the view of mankind and society that prevails locally is still in its infancy. This is why the methods that experts, management consultants, representatives of international organizations, and foreign-­trained executives generally seek to impose are the so-­called ‘universal’ methods (that is to say, American methods more or less adapted by those transmitting them according to their own visions of life in society). When the graft takes poorly and the results are mediocre, directors are generally encouraged to be firm in effectively applying ‘best practices’. And, the less successful the approaches one seeks to impose are, the sharper the insistence on applying them. In the face of such an observation, three hypotheses compete with one another: •

• •

Hypothesis 1: developing countries will only manage to supersede their current poor conditions if they finally manage to establish the ‘best practices’ that have proven themselves in developed countries. Given the cultural resistance to such an evolution, large-­scale cultural change is needed in these countries. Hypothesis 2 (which is currently tending to gain ground with regard to sub-­ Saharan Africa): such cultural change is so difficult to obtain that some areas of the planet are condemned to remain underdeveloped. Hypothesis 3: development demands that one design and implement, in each cultural context, a management approach suited to the concept of mankind and society that prevails locally. In such areas, vast innovations in management are necessary for this purpose.

Our research in developing countries in Africa, Latin America and Asia contributes to validating this last hypothesis. The research carried out in particularly successful companies was especially enlightening on that matter. For instance, let us look at the transformation of the Société d’électricité du Cameroun’s management in the 1990s (Henry 2003). When we started examining this company’s operations, multiple international consultants had prescribed implementing a classic decision decentralization system, with the establishment of target-­setting, evaluation and control systems. But it had no effect, despite calls for ‘voluntaristic’ action by the management. A few years later, the company had managed to set up a form of decentralization that indeed worked. Very surprisingly for foreign observers, that was a somewhat unorthodox

164   P. d’Iribarne system. The writing of an imposing handbook (made up of approximately ten binders, each containing several hundred pages) describing in a wealth of detail the procedures to be followed when carrying out all technical, commercial, administrative and financial tasks played a central role. Such a form of decentralization suited the cultural context. In that society, suspicion weighs heavily. Above all else, one fears that others (supervisors, subordinates, equals, clients, suppliers), guided by their interests, may be plotting against one without one’s knowledge (Platteau 2009). This fear is activated whenever anything is not ‘clear’ or ‘transparent’ in the actions of those with whom one works, whether they be one’s superiors, subordinates or equals. In particular, difficulties arise as soon as delegation and verification procedures leave room for interpretation when evaluating subordinates’ actions. After-­the-fact judgments of a subordinate for whom one had set objectives and whom one had allowed to choose the means of attaining them are strongly suspected of being based on reasons that have nothing to do with the quality of the subordinate’s work. If one succeeded, was it not because a complicit superior had set undemanding objectives? If the other failed, was it not because a superior who had it in for him had set impossible objectives? If the person in question had agreed to those objectives, perhaps the superior knew of, but deliberately did not disclose, aspects of the situation that made it inevitable for the mission to fail. In such conditions, there is a good chance that an agent who gets criticized or punished will react negatively and counter-­attack, either overtly (false outages, small-­scale sabotage) or even covertly, calling on witchcraft. For superiors, it is prudent to avoid exposing themselves to such reactions. It is better to pass decisions along to those higher up in the chain of command. To run a company well, these suspicions need to be silenced. To do so, one must remove the uncertainty as to the real intentions of those who assign work and as to the degree to which workers intend to accomplish the work well. In small companies, one good way to do this is to agree in advance on the tasks to be accomplished. For instance, during a meeting with a local manager, we observed a constant parade of his staff: ‘Boss, I’m going to do this’, ‘Boss, I’m going to do that.’ The more specific the agreement, the less the parties involved risk suspecting each other of lacking good faith, or even of laying ‘traps’. In large structures, one cannot operate in such a way without falling into paralysing centralization. However, the case-­by-case, on-­the-spot agreements can be replaced by an extremely detailed definition of the tasks to be accomplished in all circumstances. The procedure manual used plays this role. It defines the tasks in such detail and thereby leaves so little room for interpretation that those who follow its instructions find themselves in the same position as they would if they had obtained formal agreement for each task. The use of such manuals is not only a simple preservation of traditional practices; such an approach is partially inspired by management tools coming from elsewhere. But neither is it a blind importation of foreign practices. We are dealing with an innovative way of managing.

Culture, management and development   165 We have found the same type of situations in companies located in very diverse countries: Mexico, Morocco, Argentina (d’Iribarne 2003a) and Tunisia (Yousfi 2005). These companies rely on a modern reinterpretation of traditional forms of cooperation that are present in their societies but generally absent from the business world. In Mexico, for instance, a unique management method led to the emergence of a sort of ‘family of brothers’ who continuously provide each other with mutual help and support; and in Morocco, the same thing can be said of a religious brotherhood brought together by a shared ideal and run by a holy man. In each of the countries concerned, there are local forms of cooperation compatible with a good level of efficiency. And these forms of cooperation can be mobilized by companies if they use appropriate forms of management – which is still only rarely done.

Effective management in particularistic societies The conditions under which good cooperation is possible within a group of human beings who are not reduced to small, close-­knit groups is an important question with regard to the debates on the relationship between culture and development. Are societies that are marked by particularistic cultures condemned to nepotism and corruption? At the very least, is their economic development not irremediably hindered by the fact that transaction costs become very high as soon as one seeks to operate outside of small communities? Would it then be necessary, in line with the theory developed by North, for these societies to undergo a sort of cultural revolution leading to the establishment of a system of contractual relations associated with strict enforcement of property rights? In fact, this theory does not fit with our observed data. One first point is that, within the most prosperous societies, acceptable forms of cooperation between individuals who are not linked by any particular ties (i.e. forms that do not have prohibitive transaction costs), exist but that these forms are very diverse. The development of contract-­based relations is only one specific example of such forms. For instance, in France one frequently finds a form of coordination based on each individual’s attachment to the duties of his or her profession, with what these duties imply vis-­à-vis other professional groups. This attachment makes it possible to establish cooperation between individuals who have no particular ties to each other that can be quite reasonably considered as satisfactory, both within organizations and in client–supplier relations. Whatever one thinks of one’s boss, clients or the neighbouring department, the duty to do one’s work ‘in a proper way’ remains. The existence of such a sense of duty generates confidence. It makes an ensemble of opportunistic behaviours sufficiently unlikely that it does not seem necessary to protect oneself from them. This cooperation mode is not merely an imperfect substitute for coordination based on strict enforcement of property rights. Indeed, when contracts are signed, it generally makes it possible for them to be simpler and less subject to dispute than in the United States, which reduces transaction costs by as much.

166   P. d’Iribarne Similarly, the form of coordination that dominates in Northern Europe – based on the search for pragmatic compromises between people who, having exchanged their arguments, are far from having reached a communion of thought – does not seem less efficient than strictly contract-­based coordination. It is true that one shared characteristic of European societies (including the United States, Canada, and a few others) makes it relatively easy for large-­scale cooperation to function efficiently: in these countries, one finds forms of duty that come along with obligations to anyone, independently of the specific relationships one has with them. The obligation for contractual honesty concerns dealings with everyone, not only with one’s close relations. The same is true of the obligation to be worthy of one’s social position and of the obligation to manage one’s relations with one’s peers in a spirit of compromise. It is in compliance with such forms of a sense of duty that it seems reasonable to fulfil one’s professional obligations even when they imply being insensitive to the particular relationships that one has with a parent, friend, or member of a mutual-­help network. It is true that, outside these societies endowed with a ‘universal’ sense of duty, it is more difficult to establish effective forms of cooperation on scales larger than small communities. However, the cases that we have analysed show that it is far from impossible. Indeed, these cases concern sets of several thousand people. Of course, in each case, deviations linked to the strength of particularistic relations have marked the company’s past, entailing very harmful consequences on its operations. But these deviations could be fought effectively thanks to appropriate management. Going back to the Cameroonian example: before the company’s management reform, there were already some efficient forms of coordination in small local units where the person in charge could define on a case-­by-case basis what ­everyone had to do, if only by granting his/her acceptance when informed in advance by the subordinate. The management reform consisted in finding an equivalent to this form of coordination to make it compatible with a larger structure operation. Ethical questions provide a good example of what can be done in particularistic societies provided one uses the appropriate forms of management. In such societies, placing one’s professional duty above the favours that one can do for one’s relations (family, clan, brotherhood, network) risks being seen as betraying them in favour of one’s company. In addition, such types of ‘traitors’ are easily suspected of having hidden, unspeakable reasons for doing so: wanting to keep for themselves everything they can take from the company, not really having the power (over hiring, for example) that they claim to have, etc. Accepting various forms of ‘corruption’ often appears to be a much more ethical behaviour than refusing them. Should one conclude that, when such ethics prevail, changing behaviours requires a major cultural transformation leading to the advent of universalistic ethics? Of course not, but one needs to use methods suitable to this type of context. Two paths are possible. A first path is to help those who are put in positions where temptations exist. This means allowing them to show the other members of the solidarity networks

Culture, management and development   167 they belong to that, if they refuse their requests, it is not because of a lack of loyalty to them but because they cannot do otherwise. Establishing rigorous procedures, the application of which is subject to very tight control, that govern the accomplishment of all acts likely to be occasions for opportunism, works in this direction. The rigorous control exercised over those that run the risk of being subject to outside pressure is not perceived as an insulting mark of mistrust, but as the normal control exercised by a group over its members to ensure that they are loyal to the group. One can see such control perceived as ‘help’, ‘support’ or a source of ‘comfort’ (d’Iribarne 2003a). The second way is to make the company itself, according to modalities suited to the local culture, into a group of belonging that carries sufficiently strong loyalty duties to be a source of obligations taking precedence over one’s obligations to family or friends. Let us add that, in many countries, the fact that particularistic ethics govern (or in any case predominate in) economic life does not mean that universalistic ethics are not simultaneously present in society, whether they are of religious inspiration or a matter of honour. We have encountered this situation in Argentina and Morocco (d’Iribarne 2003a). One of the roles of management, then, is to ensure that these ethics do not remain a reference in principle but take life within the company. This does not require one to change national culture, but to set up appropriate management methods. One can wonder to what extent what is valid for companies could also be valid for the institutions that regulate the entire economy. The experience of European countries, or European cultures, shows that the cultural logics that mark company operations are also found in the operation of institutions that govern how the entire economy is run. For instance, in the United States, the institutions that govern labour relations – labour law in the first place – were designed to make the relationships between employers and employees conform as fully as possible with the ideal image of the freely negotiated contract. This image inspires current legislation as for the importance given to the notion of a level playing field in which employers and employees negotiate from positions of equal power (Foner 1998). On the contrary, in France, the institutions that govern labour relations were built through endowing workers with a status derogating from general law (Castel 1995). The same contrast can be seen with regard to general-­interest service policies. Whereas, in the United States, one seeks to organize how the market operates, in France one seeks to give a status derogating from general law to operators entrusted with public service missions (Dobbin 1994). Not only does this concern the functioning of economy, one can, along the same lines, cite the role played in France but not in the United States by administrative jurisdictions which, due to the status of those under their jurisdiction, derogate from general law. The United States and France are not the only countries concerned. One could similarly evoke the place given to a consensus-­seeking logic between the concerned actors in the Netherlands, either within companies or in overall economic regulation, or even in the way democratic institutions operate.

168   P. d’Iribarne In each of the countries concerned, adapting institutions and practices to the local cultural framework is always the outcome of a long process of trial and error. Ideas circulate. The attempts – sometimes enthusiastic – to assimilate, in a given country, institutions and practices from elsewhere have always been given a lot of importance. But, in practice, those institutions and practices are, when it comes to their concrete application, judged according to the indigenous vision and conserved, abandoned or modified in consequence. Why would what is true in countries where the modernization movement is already ancient not be true in places where it is much more recent and where one is still in a period of institutional trial and error? A priority task for development research should be to help such trial and error processes converge more rapidly.

Conclusion When one seeks to understand the relationship between culture and development, one usually relies on a vision of culture that assimilates it with traditional practices. Such practices are suspected of hindering development. In this view, transforming practices implies changing the underlying culture. This vision of culture is incorrect. In fact, within the most ‘advanced’ societies, one encounters cultural traits that have remained while societies were modernizing. The institutions and practices to which this movement led over the last centuries have remained marked by these cultural traits. This is equally true for company management and for the institutions that govern the overall functioning of society. Observation of particularly successful companies in developing countries suggests that, here too, a modernization movement compatible with the local culture (culture being understood as a general interpretive framework associated with an ideal vision of society) is underway, even if this movement is still more or less embryonic. These are modern companies in which the major abstract principles of management (managers’ integrity, listening to staff, transparency, mutual trust, etc.) are applied. But, when one looks at the concrete manifestations of this implementation, one sees that the major principles are incarnated in local forms, and that it is thanks to such incarnations that ‘modern’ management receives the support of those that implement it. These companies are still happy exceptions, but the way in which they operate is meaningful. The path that they show is not an illusionary cultural revolution, but the advent of a form of modernity that will be no more or less local than those that can already be seen in the countries – the United States, France, Germany and others – that blazed the trail. Much remains to be done to learn from the examples they provide, not only for company management but also for country governance.

Notes 1 Paper presented at the 5th AFD/EUDN Conference, 5 December 2007. 2 We shall rely on the lessons from a research programme examining the diversity in the forms of coordination encountered around the world and their influence on how organizations operate. The goal is to reveal how these forms of operation (hierarchical rela-

Culture, management and development   169 tionships, cooperation between divisions, how client relations are organized) take on meaning for those involved, and how this influences the way in which cooperation between individuals and functions is organized. Begun in 1982, the fieldwork has, to date, been conducted in more than 40 countries located on various continents and having very diverse degrees of development. Among the publications resulting from this research, we can cite: Henry et al. 1991; d’Iribarne et al. 1998; d’Iribarne 1989, 2003a, 2006, 2008, 2009; Yousfi 2005.

References Castel, R. (1995) Les métamorphoses de la question sociale, une chronique du salariat, Paris: Fayard. de Tocqueville, A. (1856, 1952) L’Ancien Régime et la Révolution, Paris: Gallimard. d’Iribarne, P. (1989) La logique de l’honneur, Paris: Seuil. —— (1990) ‘Face à l’impossible décentralisation des entreprises africaines’, Revue française de gestion, 80, pp. 28–39. —— (2003a) Le tiers monde qui réussit, Paris: Odile Jacob; trans. Gill Gladstone, Jon Graham and Eleanor O’Keeffe (2007) Successful Companies in the Developing World: Managing in Synergy with Cultures, Paris: Agence Française de Développement. —— (2003b) ‘Trois figures de la liberté’, Annales, 58 (5), pp. 953–80. —— (2006) L’étrangeté française, Paris: Seuil. —— (2008) Penser la diversité du monde, Paris: Seuil. —— (2009) L’épreuve des différences; l’expérience d’une entreprise mondiale, Paris: Seuil. d’Iribarne, P., Henry, A., Segal, J.P., Chevrier, S. and Globokar, T. (1998) Cultures et mondialisation, Paris: Seuil. Dobbin, F. (1994) Forging Industrial Policy: The United States, Britain and France in the Railway Age, Cambridge: Cambridge University Press. Douglas, M. (2004) ‘Traditional Culture – Let’s hear no more about it’, in Vijayendra R. and Walton, M. (eds) Culture and Public Action: Relationality, Equality of Agency and Development, Stanford: Stanford University Press. Foner, E. (1998) The Story of American Freedom, New York: Norton. Giddens, A. (1984) The Constitution of Society: Outline of the Theory of Structuration, Berkeley: University of California Press. Henry, A. (2003) ‘Une révolution des procédures au Cameroun’, in d’Iribarne, P. (ed.) Le tiers monde qui réussit, Paris: Odile Jacob. Henry, A., Tchenté, G.H. and Guillerme, P. (1991) Tontines et banques au Cameroun, les principes de la société des amis, Paris: Karthala. Locke, J. (1690, 1960) Two Treatises of Government, edited by Peter Laslett, Cambridge: Cambridge University Press. McClelland, C.D. (1961) The Achieving Society, Princeton: Van Nostrand. North, D.C. (1987) ‘Institutions, Transaction Costs and Economic Growth’, Economic Inquiry, vol. XXV, July, pp. 419–28. —— (1990) Institutions, Institutional Change and Economic Performance, Cambridge: Cambridge University Press. Ouchi, W.G. (1981) Theory Z, New York: Avon. Platteau, J.P. (2009) ‘Institutional Obstacles to African Economic Development: State, Ethnicity, and Custom’, Journal of Economic Behaviour and Organization, Vol. 71, Issue 3, pp. 669–89.

170   P. d’Iribarne Tönnies, F. (1887, 1964) Gemeinschaft und Gesellschaft; trans. Charles P. Loomis (1957) Community and Society, East Lansing: The Michigan State University Press. Weber, M. (1922, 1971) Wirtschaft und Gesellschaft; trans. Guenther Roth and Claus Wittich (eds) (1978) Economy and Society: An Outline of Interpretative Sociology, Berkeley: University of California Press. Yousfi, H. (2005) Poulina, un management tunisien, Paris: Agence Française de Développement.

9 The invention of traditions and entrepreneurship A critical perspective Jean-­Pierre Warnier

The invention of traditions and entrepreneurship: a critical perspective Tradition exists. We have all encountered it. Certainly, no action – individual or collective – can be invented afresh at each new occurrence. Were this the case, the actors would exhaust themselves weighing up the reasons for their choices, agreeing on a melodic line and making sure they sing the same tune, instead of acting. The end result would be powerlessness and cacophony. Even when action is innovative, it nonetheless brings into play widely tested and recognized repertoires as well as the experience of past action. There is a huge gamut of terms to denote the sediments of experience underlying present actions: norms, ethos, procedures, customs, habitus, habits, repertoires, behaviour, routines, culture, traditions, civilization, spirit, values, etc. Yet the contours of these notions are hazy. They attempt to apprehend a moving reality, at the limit of common sense and critical and scholarly undertakings. The different traditions, be they intellectual (philosophy, sociology, ethnology), national (American, German, French), or theoretical (from Montesquieu, Marx, Weber, Durkheim, etc.), all have their own preferences and viewpoints. This singularly complicates any rigorous attempt to address the cultural phenomenon. The myriad viewpoints impose on all alike the constant gymnastic of translating, transposing and clarifying, which I will not attempt in this chapter for want of space. In the 1960s, the first theories on the subject set tradition and modernity at odds as being mutually incompatible. Subsequently, the manifest resilience of traditions and cultural values in ‘modern’ environments threw into question those theories predicting that civilizations were to converge under the influence of Western modernization. An extensive survey carried out by IBM between 1966 and 1973 in 72 countries, and published by Bollinger and Hofstede (1987), was one of the first to combine cultural values and modernity. This has stood as a landmark for ‘cross-­cultural management’ and the many strands of related research. At the time it was launched (1966), the survey was an innovation. Its theoretical and methodological framework was underpinned by social psychology and American cultural anthropology. Its perspective was macrosocial

172   J.-P. Warnier i­nasmuch as it attempted to capture the average attitudes of IBM employees with respect to four factors, on the basis of country or nationality. With the 1970s, and even more so the 1980s, came an impressive diversification of theoretical perspectives, objects and methods. For instance, with respect to theoretical paradigms, d’Iribarne’s research (1989) into corporate management made the linkage between a macrosocial approach to national traditions and company monographs (at a more microsocial level) with strong reference to the work of Montesquieu. J.-F. Bayart (1994), on the other hand, was less concerned with corporate management than capital formation and modes of accumulation within ‘peripheral capitalism’ with reference to a historical sociology of Weberian inspiration. Monographic studies mushroomed. Large collective works like Ellis and Fauré’s essays (1995) on African enterprises proved somewhat eclectic in theoretical and methodological terms, reflecting the extraordinary diverseness of the social sciences at the end of the twentieth century. This was visible, for instance, in Janelli’s work (1993) on a Korean conglomerate, in the study of an Ivoirian firm by Bazin (1998), whose approach falls within the scope of ethnology, or in the L’Afrique des entreprises (1998), a collective work reflecting the concerns of management experts. In 2007, the landscape is extremely complex. Theories and analyses have been reformulated several times over. In 2000, enterprises were no longer operating in the same environment as in the 1960s. For the sake of simplicity, I shall be focusing on a single object (or, in other words, one single entry point into the subject), provided by the notion of tradition.1 Certainly, it seems high time to make a critical appreciation of what is meant by ‘tradition’ and to highlight what effect this may have on the analysis of the relationship between enterprises and traditions. Finally, to be perfectly clear, I shall be doing so in my capacity of ethnologist, with the intellectual and methodological biases of my profession, which is more inclined than others to show interest in what is local, over protracted periods of time.

Inventing traditions Being part of history, social relationships are subject to change. This said, insofar as they are of a relatively regular nature, they also produce traditions. Certain procedures, ways of doing, habitus and norms become traditions to the extent that they are transmitted by the group to individuals and from one generation to another, until history takes it upon itself to reshape them. Traditions stabilize norms and behaviour. They allow them to become institutionalized. They safeguard social action against improvisation. They crystallize rules of conduct. In the 1960s, this trait of permanence and inscription in the long term seemed incompatible with a modernity chiefly typified by change and the systematic search for innovation, both of which had ostensibly sprung up in the Europe of the Enlightenment: elective and participatory democracy, the incipient processes of individuation, the Industrial Revolution. Since the turn of the nineteenth century, these modernizing currents, underpinned by the second Industrial

The invention of traditions   173 Revolution (that of electricity) and the imperial designs of countries bordering the north Atlantic had supposedly spread to the rest of the globe, upsetting tradition-­dominated societal structures. The opposition between tradition and modernity has been progressively clarified by a critical analysis of these two notions, of their empirical content and their differences. On the side of ‘modernity’ (which had its critics as well as eulogists – Leiris referred to ‘merdonité’2), the analyses have underscored its contradictory traits, the fact that it was in no way a Western monopoly, the crises affecting it and its failures. Analysts have sought to overstep the limits of the notion by calling on other notions such as ‘post-­modernity’, ‘second modernity’ or ‘hypermodernity’. On the side of ‘tradition’, the ideal type of archaic society, for which Louis Dumont (1967, 1977) gave a canonical description, was supposedly hierarchical in structure on account of the individual’s subordination to the group encompassing him, the frailty of individuation processes, and the resulting absence of certain factors, such as entrepreneurs or political operators, and of the space for individual freedom that made the democratic and industrial revolutions possible. Contemporary ethnologists and historians now almost unanimously agree that L. Dumont was mistaken, as much from the empirical as the theoretical point of view.3 Human beings are – and have always been – individualists in the current sense of the term. Individuation processes exist to some degree in all societies. Name-­giving, religious practices and power relationships all produce individuation effects. Hierarchies and subordination to the group, like the emancipation of the individual from group ties, are historical and political constructs; they depend on historical circumstance and not on the intrinsic nature of a certain type of ancient/historical society as opposed to a contemporary/egalitarian one. Western ‘modernity’ places value on the dynamics present in ‘traditions’. The latter are subject to change, and ‘modernity’ produces new traditions by stabilizing certain norms. It was E. Hobsbawm and T. Ranger’s book, The Invention of Tradition (1983), that brought this line of reasoning to the attention of the analysts. The term ‘invention’ has given rise to controversy and misunderstanding insofar as it seemingly suggests that traditions are created ex nihilo and in an arbitrary fashion. Hobsbawm and Ranger’s argumentation, however, is quite different. They employ the word ‘invention’ in the archaeological sense of identifying an ancient site. The ‘Chauvet Cave’ has existed for some 30,000 years with its rock art. Yet Mr Chauvet, the ‘inventor’ of the site, discovered it and recognized it as a major find whose patrimonial value, uses and importance for twentieth-­century Europe have taken over those of the hunter-­gatherers who decorated it, while at the same time completely modifying them. Likewise, in the specific case of Europe and its nationalities, which are the focus of Hobsbawm and Ranger’s book, local traditions (in Scotland, Wales, the British monarchy, Victorian India, Europe between 1870 and 1914) existed prior to the nation states, but are ‘invented’ and re-­ appropriated within the context of a Europe of Nations, underpinned by provincial peoples with distinctive cultures, regional customs, their own ‘dialects’, etc. This ‘invention’ set in motion processes of sorting, elimination, re-­appropriation,

174   J.-P. Warnier innovation and bricolage that made it possible to present practices that prolonged old customs as constituting ‘traditions’. These practices, however, were substantially redefined and reforged within a political project that went beyond the limits of the customs themselves. Hobsbawm and Ranger’s book had consider­ able impact. It was a landmark work, separating the ‘before’ and ‘after’ in the sociological and historical study of tradition. This theoretical detour will help me to clarify what can be understood by the ‘traditional values’, norms and institutions that shape the entrepreneurial environment in the developing world. In 2007, within what context are these traditions ‘invented’, in the above-­defined sense? For, unquestionably, traditions exist (in business, having Jains or Kikuyus as partners can make a difference). Yet equally unquestionably, the ‘traditions’ of a Jain stock market operator from Bombay or a Kikuyu engineer from Nairobi are markedly different in 2007 from those of their grandparents 50 or 100 years ago. Along what lines have these traditions been configured, and by what processes of ‘invention’?

‘Modern’, ‘urban’ and ‘global’ traditions Nineteenth-­century European nationalism (which I take as a given) furnished the framework and motives for the invention of provincial traditions. What is it then that provides the framework and motives for inventing enterprise-­related traditions in each part of the developing world in the twenty-­first century? In the first place, the minimum narrative of modernity has effectively been adopted worldwide. It is even highly likely that this happened simultaneously in several regions of the globe (China, Japan, the Arab world) in the nineteenth century, as ‘modernizing’ currents independent of any Western reference (as shown by Beaujard 2005) are visible. To make myself clearer: if the figure is in any way meaningful, it can be said that 95 per cent of human beings reaching adolescence today know that medicine and hospitals are available to treat them with a good chance of success; that money exists to enable them to obtain quantities of desirable goods on the market; that paid employment allows them to earn money; that having a job, a house, a car, founding a family and sending their children to school is an enviable life project; that participation in regulated public debate is desirable; that the whole world is potentially open to plans to migrate and that global exchanges affect each human being. This narrative of modernity makes it possible to envisage and strive for a life programme underpinned by a number of instruments, the most important being money and the market, but also business, the state, international trade, etc. This minimum programme is matched with narratives about suitable means for collectively implementing it (through socialism, progressive Islam, international aid, ‘development’, etc.) and about its successes and failures. I describe this programme as a ‘minimum’ programme since the actors who really measure its implications (regional and international peace, sustainability, environmental constraints, democratic requirements, a sound banking system, global governance, etc.) are very few. These implications define what could be a ‘maximum’

The invention of traditions   175 programme of modernity, which would ipso facto include elements of its own critique. This minimum programme has the merit of being attached to a material culture and symbolic corporeal practices, which are beyond discussion: the Shiites and Sunnis in Baghdad, the Jains in Mumbai and the Navajo in the western United States have directly appropriated the materialities of family housing, cars, running water, electricity, cell phones, together with the accom­ panying practices such as school, employment, money, the market and a better life for future generations. In 2007, depending on the country, between 20 and 100 per cent of humans are urban dwellers (18 per cent in Burkina Faso, one of the planet’s poorest countries; industrialized countries have a 70 to 90 per cent urbanization rate, and emerging countries around 50 per cent). This proportion is steadily increasing. And enterprises, particularly industrial ones, are an integral part of the urban fabric. Finally, after the minimum narrative of modernity and of the urban setting, the third element that helps to structure the invention of traditions at the outset of the twenty-­first century is the globalization of commercial, financial and cultural flows. Of these three factors (modernity, urbanization and globalization), it seems to me that globalization of flows implies and subsumes the other two (modernity and urbanization). These three factors are operative the world over, and I know of no notable exception. I hope that I will be excused for this detour via the invention of traditions, modernity, urbanization and the globalization of flows, because this allows me to point out that none of these three historical phenomena toll the bell for either traditions or cultures. These remain – sometimes reacting with renewed vigour. The three factors do however provide the historical context of invention, and enable me to elaborate my questioning of how ‘traditional values’ are ‘invented’ in a historical context that is no longer one of nineteenth-­century nationalism, but one of modernity, urbanization and globalization. How do these values, and the norms and institutions attached to them, affect the way in which a company is set up and managed? To put it differently, two avenues of investigation seem to be equally impossible in 2007. The first would have ‘traditions’ linked to a rural, village-­based, primordial authenticity. Those traditions have disappeared in whole swathes and what remains of them has been profoundly reworked. The second posits that particularizing traditions are vanishing under the effect of modernization and the standardization/globalization of practices. The avenue that I am proposing is that humanity is a machine that produces difference amidst change, conflict and incompleteness by inventing traditions. Today, this machine functions in a ‘modern’, urban, globalized context.

The ‘return of the kings’ and the invention of the African monarchies A case drawn from Africa will allow me to enrich the debate with empirical data and address the core of the fourth topic of the AFD/EUND conference. What is

176   J.-P. Warnier more ‘traditional’ than the African monarchies? These are so linked to tradition that in the 1960s, at the time of independence, they were deemed to be a dying phenomenon, doomed to disappear. Africa was aspiring to economic and political modernity. Rituals were neglected. Palaces were crumbling. The kings, almost all of whom were illiterate, were rejected by the new elites. Now, at the start of the twenty-­first century, palaces have been renovated and rituals are staged. Kings are solicited. This remarkable ‘return of the kings’ to the centre stage of African life since 1980 was highlighted at a conference, and published as a book edited by C.-H. Perrot and F.-X. Fauvelle-­Aymar (2003). Not all of the kingdoms have managed to play their cards right. The smallest ones have disappeared, as have those that suffered from excessively violent dissension. However, most of the relatively large kingdoms have now emerged as a political resource in the various crises of African states and the formation of hegemonic alliances likely to bring stability to a given regime. Who are these kings? They have all completed at least secondary education, and more often university. Almost all of them are professionally qualified and work as engineers, doctors, lawyers, teachers and senior civil servants. Almost all have a foot in business, including plantations, trade, services, manufacturing (breweries, construction). Finally, as the invention of these kingdoms largely depends on the political context, all of the kings have a foothold in a political party, the state and international relations. For example, C.-H. Perrot (2003: 9) mentions that: ‘In Niamey, in 1996, eleven kings from West and Central Africa met together to offer their mediation in various countries torn by civil war.’ ‘In 1999’, adds C.-H. Perrot, ‘a further meeting was held in Benin to discuss the problems of development and ecology’. Everywhere, these kings have integrated the urban phenomena. Sometimes they reside in a city, in which case they have a lieutenant to represent them at the palace and supervise daily ritual and ceremonial life – the king only attending important occasions. Alternatively, they take up residence in their kingdom, making frequent stays in the country’s capital or in the largest city near to their kingdom (the city has sometimes developed out of the latter, as is the case with Ife, Benin, Zaria, etc.). They practice what experts on Africa have identified as ‘straddling’, between business, politics and an important position in local social structures. The kingdoms are enmeshed in the globalization of cultural and commercial flows, both locally through the impact of these flows, and also thanks to the elites and the diaspora of former kingdom residents who form associations. A case in point is the Mankon Cultural Development Association (MACUDA), whose members are scattered between Yaounde, Douala, London, New York, Minneapolis and Texas. The kingdom (Mankon – Cameroon) has its Web page linked to the palace museum (www.museumcam.org/mankon/kingdom) promoting Mankon cultural heritage and ‘traditions’.4 The US branch of MACUDA has its own website (macudaamerica.org). The king of the Mankon people, like all his peers, is contributing to the invention of African modernity with respect to urbanization and globalized flows by inventing monarchical traditions. These kingdoms could be suspected of no longer having anything ‘authentic’ or ‘traditional’ about them. This, however, is the hub of my argument: these

The invention of traditions   177 kingdoms were formerly born out of the ‘high seas’ of continental exchanges by fabricating locality and capturing resources of extraversion. In the latter half of the twentieth century, they proved highly reactive to historical circumstances by continuing, in a changing environment, to do what they had always done: take part in exchanges, produce locality, capture the resources of extraversion, and concentrate wealth and power. In one sense, they are more authentic and traditional then ever before. African traditions of royalty are thus the product of ‘invention’ in Hobsbawm and Ranger’s sense of the term. They existed before an African modernity was produced in an urban and globalized context. Yet it is in this context that they have been rediscovered, like old objects stored away in the attic. They have been recovered, reworked and recycled for a new lease of life. These reinvented monarchical traditions have an obvious relevance to the questions asked within the fourth theme of today’s conference. These kingdoms are instruments of capital formation. For example, when the Mankon king succeeded his father in 1959, he founded the Mankon State Union, to which each citizen was ‘invited’ to subscribe. The capital thus collected enabled the king to invest in several undertakings in the provincial capital (breweries, film theatre) as well as in plantations. This foundation has also permitted grants to be awarded to students from the kingdom. Kingdoms are, by definition, composite, pluriethnic entities, united by a monarch, one of whose roles is to capture and absorb what can further the kingdom’s interests and reject useless or waste elements. The king systematically attracts foreign elements: NGOs, Peace Corps volunteers, researchers, entrepreneurs. In the 1970s, he encouraged the setting up of a French bakery that employed as many as 40 people. When the baker retired, the king granted him the title to a plot of land on which he built a house and settled down for good.5 The king, a trained agronomist and public official in the Ministry of Agriculture until he succeeded his father, is on the lookout for all kinds of innovations (fish-­farming, zootechnical research, transport and communications, public health, water adduction, Internet, museum) and welcomes into his kingdom outside vectors for these innovations. He has a politically conservative approach to economic and technical modernization. I know of various contracts that he has negotiated and signed, for example a multipartner contract between the Cameroonian Institute of Zootechnical Research, an American NGO (Heifer Project), a French institute and the state of Cameroon. The king’s contribution was 250 hectares of agricultural land that he had appropriated in the name if his ‘traditional’ land rights. I was unable to consult the terms of the contract, which was signed in the early 1980s. The research station was built, the employees recruited and it operated normally into the late 1980s, until it was badly affected by the political and economic crisis experienced by Cameroon. Let us now return to the formation of capital. In at least two cases (the Mankon State Union foundation and the Zootechnical Research Institute), the king, without consultation, appropriated the financial and land resources at his subjects’ expense. This behaviour is directly in line with age-­old royal powers, but it can hardly be said that these powers are attuned to the functioning of a

178   J.-P. Warnier state of law. The king is reproached for this by many of his subjects, especially as a land market, based on land registry services and a system of private property, came into being even before the country’s accession to independence in 1960. The king, however, overrode all objections. This was only possible with support from the ruling party (President Ahidjo’s Cameroon National Movement [UNC], and subsequently the Cameroon People’s Democratic Movement [RDPC] led by P. Biya), and from the political elites to which the king is affiliated as National Vice-­President of the ruling party, president of the provincial section first of the UNC then the RDPC, and member of parliament for two successive terms. What is more, the king is not content simply to capture his kingdom’s resources. Another of his functions is to capture state resources – where they can be found, which means in the country’s capital city – and to bring them back to his palace and kingdom. This blurring between the state, the kingdom as a moral entity and the king’s and his family’s own patrimony has been created by the sequence of historical circumstances. Prior to the modernization process I am describing, the king’s physical body, his palace and his kingdom were all identified as composing a single ‘corporation’, made up of material and physical envelopes with orifices through which elements transited, being absorbed as food or rejected as the kingdom’s excrement. The ‘modern’ state, whose legislation introduces a second legal norm alongside the (‘traditional’) norm of the kingdom, produces a situation conducive to trials of strength or even violence,6 ‘corruption’ or illegality, which is not contingent on, but rather essential to, the invention of African kingships and the social and economic divergence they engender.

Hobsbawm and Ranger ‘at large’? Hobsbawm and Ranger’s line of reasoning may thus be usefully transposed from a nineteenth-­century nationalist Europe that ‘invented’ its provinces to a twentieth-­century Africa experiencing urbanization, modernization and globalization and ‘inventing’ its kingdoms. Can this pattern of ‘invention’ be generalized to all traditions worldwide? At the outset, the two historians’ objective was much more limited. It involved explaining why the second half of the nineteenth century in Europe had seen the combined explosion of invented regional provinces and national narratives. Their reasoning, however, resonated so forcefully in the analyses of other historical situations that most experts in social matters now unhesitatingly assert that what actors and observers call ‘tradition’ is always ‘invented’ in the sense of our two historians. A conference on E. Hobsbawm at Nice University (proceedings edited by D. Dimitrievic, 2004) shifted closer towards such generalization with its evocative title Fabrication de traditions, invention de modernités (Producing Traditions, Inventing Modernities). As a result, a new conception of traditions has recently overlaid an earlier conception. The contours of this earlier conception are clearly outlined by the ethnologist Jean Pouillon (1991: 710) for whom ‘tradition is defined . . . as what from the past persists into the present, where it is transmitted and remains active and is

The invention of traditions   179 accepted by those who have received it and who, in turn, over the generations, pass it on.’ Pouillon adds that tradition is transmitted orally (which implies that it is of an essentially verbal nature and thus accessible to consciousness). Yet nothing is less certain. M. Bloch (1995) and myself (2007), drawing on advances in the cognitive sciences, have emphasized the fact that tradition is very largely transmitted by way of material and bodily cultures, in an unconscious and non-­ verbalized way. Pouillon’s conception is not invalidated by the work of Hobsbawm and Ranger. In fact, they have taken up the earlier definition and added an original, often neglected parameter: the historical context in which traditions are transmitted, and thus the constraints governing the selection processes that determine what is brought out of oblivion from among the old practices and representations, then transmitted or reformulated depending on the circumstances. I would propose the following definition of tradition to replace that of J. Pouillon: ‘That which, from among the practices and representations of the past, is taken up, reinvented and completed in line with a new historical context, and which remains active and accepted by those who appropriate it.’ This tradition is in part verbalized and orally transmitted, and in part incorporated into material cultures and bodily techniques that are transmitted without any verbalization or explicit consciousness. This second approach would seem more pertinent whatever the historical situation, and more particularly so in the case of the relationship between entrepreneurship and tradition in modern times. I will thus take it as a given that the traditions impacting on enterprises are practices and representations that are more or less ancient. They have stabilized and crystallized in such a way as to make the actors’ behaviours more or less predictable and regulated, following a process of selection and reformulation within a historical context characterized by the taking on board of a minimum programme of modernity, in a mainly urban and globalized environment, and the implementation of money and market tools. The earlier approach to traditions authorized social analysts to trace back their original, vital sources as far as possible, and to even reconstitute what they had been before the shock of colonialism or the ravages of modernity. Traditions had homelands, their own temporality, and supporters. These were traditions from village society, from the bush and from a past predating the ‘contact’ and shock of colonialism. They were the traditions, for example, of the Dogon, the Lao and the Nambikwara as they used to be. They provided a yardstick for measuring the degree to which individuals were ‘detribalized’, traditions had been ‘forgotten’ or ‘disappeared’, and rural populations and migrants were ‘detraditionalized’. To my mind, we should now turn the page and abandon this perspective once and for all. Traditions have no reference point other than that of the society that keeps them alive in the present. It is thus in cities and organizations (kingdoms, enterprises, political parties, states) at grips with globalized modernity that they can be apprehended. This now means shifting from such generalities and the particular case of the ‘return of the kings’ to an empirical study of traditions throughout the world. Yet

180   J.-P. Warnier this objective poses a problem of documentation. Ethnologists have collected a wealth of information on the traditions of the past, and their body of work is in no way obsolete. Certainly, provided that the traditions they have recorded are restituted in their historical context, their investigations are precious.7 This said, the traditions that interest us here are those that can be collected mainly in cities and in enterprises. However, documentation of this subject is scarce and changing rapidly. Longitudinal follow-­up tools are lacking. Certainly, the first prerequisite for satisfactory treatment of the issues raised by the fifth AFD/EUND conference would be to possess a state of cultures in the world updated each year. In 1998 and 2000, UNESCO published a World Report on Culture (‘culture’ in the singular), which proved a short-­lived project. The challenge has now been taken up by two scholars, Helmut Anheier and Raj Yudhishthir Isar, who edit an annual publication entitled The Cultures and Globalization Series (with the word ‘cultures’ in the plural). The first issue, on the theme of Conflicts and Tensions, came out in 2007. The 2008 issue explores the economy of cultures and, in 2009, the focus will be on the relationship between art and cultures. Given the large array of contributors, the size of the work (639 pages), the 27 regional reports presented, the abundance of statistical data, the diagrams and index, this work is already an invaluable tool, and will prove even more so over time. Entrepreneurship and traditions There is undeniably a substantive lack of monographs on companies and ‘invented traditions’ that could enrich thinking on how these interrelate.8 In the following pages, for want of more extensive material, I will attempt to analyse several studies and publications that furnish empirical data relating to the questions above and can thus further a better understanding. I will refer in turn to the case of an Ivoirian firm, the question of the ‘compartmentalization principle’, a South Korean example, ‘licit illegal’ practices and the ‘tradition of invention’ in Africa. An Ivoirian firm Bazin’s (1998) monograph on an Ivoirian firm has the merit of being based on a lengthy in-­company survey. Bazin gives this industrial firm, with 200 employees and based in urban Abidjan, the pseudonym Subsi. It produces consumer food products for the domestic market in Côte d’Ivoire. I will pick up the parameters of my analysis one by one (modernity, urbanization, globalization – and finally: money and market), and try to relate them to what can be grasped of the ‘traditions’ within the firm itself and in wider society. The modernity project has been wholly taken on board by company employees, both inside the firm and in their personal lifestyles. The factory (described on pp. 17–18) was built in the late 1980s and its facilities are among the most sophisticated in the world for this type of activity. The factory, the ­laboratory and the administrative building stand in a landscaped setting. The offices are computerized,

The invention of traditions   181 functional and well equipped. The firm is a public-­limited company formed by Ivoirian, Senegalese and French capital, and with the support of two investment banks. It is not a family business of the kind that would have prospered over time and retained some traces of its origins. The technical and financial package was engineered at the end of the 1980s, out of a Senegalese company and with support from a French firm specializing in the same sector. The director is Ivoirian. The mostly Ivoirian managers were recruited from the Senegalese and French companies to create the Abidjan company. Supervisors and workers were hired in Abidjan, trained on the spot or in Senegal and, in a few cases, enticed away from a competing firm run by a Syro-­Lebanese. This explains the ‘modernity’ of the company itself. As for the lifestyle of its employees, Bazin’s study remains relatively discreet. Yet, the remarks of the workers and supervisors taken from interviews, and sometimes condensed into short life histories (cf. pp. 180–98 and 218–24), clearly reflect their overriding concerns and ambitions, namely their ranking on the ladder of responsibilities and salaries, levels of remuneration, and thus their access to money, the market, consumption, housing and related facilities (running water, electricity), as well as food, health, children’s schooling and transport. ‘Modernity’, money and the market represent more than a mere aspiration. What is involved here is a will and plan for life and action. The city only appears as a backdrop in Bazin’s book, his main focus being power relations within the firm. The city is depicted as a cosmopolitan and composite space bringing together Ivoirians from all backgrounds and ‘expatriates’ from West Africa (Senegalese as far as the company is concerned) and the rest of the world (French, Syro-­Lebanese, etc.). For the Ivoirians, the city contrasts with the more or less distant rural and ethnic origins of some of them. According to Bazin, these origins, like kinship, provide nothing more than the idiom or window-­dressing for conflictual relationships within the company. On the contrary, it can be argued that some ‘traditions’ have crystallized during Africa’s urban migrations, all ‘ethnic’ groups combined. The case of Maurice (pp. 218–19) shows that migrants who have settled in the city serve as a foothold for the subsequent migration of family members: Maurice lives in a three-­roomed dwelling where he is lodging thirteen other people in 1993: his wife and four children, as well as a younger brother, an elder brother’s two sons, four brothers and one sister (younger) of his wife. This phenomenon is more typically urban than African. The same description would fit the case of a Breton migrant in the Montparnasse district of Paris in 1925 or an Italian migrant in New York in 1910. Finally, the globalization of commercial, financial and cultural flows (and to a lesser degree, migratory flows) is prevalent in the case of this firm, along with its corollaries (internationalization of capital and know-­how, trips made by executives to negotiate contracts or industrial and financial packages, staff recruitment and training, etc.).

182   J.-P. Warnier The analysis of kinship ties within the firm occupies a central place in Bazin’s study. There are rare, but they exist. The Ivoirian director mistrusts this type of relationship and has always refused to recruit family members. On the other hand, real or imagined kinship ties are invoked in order to discredit an employee caught up in a conflictual relationship, and forgotten or put aside when this is not the case. The idiom of kinship, clientelism and ethnic preference appears as the stereotype and metaphor of Africans wielding power over Africans. Given that this power has been configured via the authority of (white) expatriate managers over the blacks, it is ill received. Bazin (1998: 244) even claims that it helps to reproduce a colonial relationship within Africa in the context of conservative decolonization. All in all, the harvest of ‘traditions’ at Sebsi is a meagre one. ‘Ethnic’ and primordial traditions are markedly absent. There is a colonial tradition, which deserves further analysis, and a specific and unquestionably urban tradition that correlates with the migratory phenomenon and its massive, enduring nature. This urban tradition, partially revealed in the quotes from the life histories and interviews, is linked to lifestyles and consumption, to a material culture and bodily techniques typical of most African villages, but which have no direct impact on the firm, albeit that these practices accommodate comparatively low levels of remuneration without, however, endangering the employees’ active participation in the functioning of the firm. Certainly, these ‘traditions’ are linked to the countless number of typically African practices and representations, analysed by J. Guyer (1984, 1987, 2004), which manage to make ‘things work’ whatever the circumstances, including most importantly feeding the African cities. Finally, there is a tradition of ‘modernity’ that has unquestionably been induced since the Africans have shown the desire to appropriate the resources of extraversion, including those of the colonizers. It can be said that these different traditions have now been in existence for a good century. Bazin has a very detailed knowledge of the existing literature on the relationship between cultures and entreprises, including Bollinger and Hofstede (1987), d’Iribarne (1989), Henry (1988, 1995), Henry et al. (1991), among others. He is fully aware of the questions explored by authors such as these, one example being the issue of establishing trust in African companies. He shows that the idiom of kinship helps to verbalize the problems of trust and mutual mistrust. To what extent are these traditions ‘invented’? I would like to highlight the fact that the practices involved in what could be termed a mass artisanal food supply system, studied by Guyer, are little more than extremely old economic practices reconverted within a modern urban context. Likewise, the modernity promoted by the Ivoirian firm recycles and reinterprets the appropriation of the resources of extraversion so characteristic of the ‘African frontier’, as analysed by Kopytoff (1987). Bastide and the compartmentalization principle At this point in the argumentation, it is important to clear up a possible misunderstanding. Subsi was set up after the ‘Ivoirization’ of the country’s

The invention of traditions   183 economy. Out of 200 job positions, only between two and four are held by Europeans. The company’s African employees are present with their own subjectivity, their family solidarity networks, their ethnic affiliation and identifications, or in short, their culture and traditions. The observer is thus tempted to posit that their repertoires of action are the same both inside and outside the company. This tacit assumption appears to inform the literature on relationships between culture and companies to varying degrees. Most analysts fail to envisage that an employee can leave his culture and subjectivity in the firm’s locker rooms when he dons his dungarees or his lab coat or when he sits down in the director’s armchair. Yet, Roger Bastide, an ethnologist specializing in Afro-­Brazilian society, has observed that this is in fact the case. He has assimilated this phenomenon to what he calls ‘the principle of compartmentalization’. Afro-­Brazilians, he says, have several cultural repertoires that do not mix. They activate a sort of switch enabling them to change from one culture to another depending on the context – either that of dominant white society, or that of the Afro-­Brazilian community. They are bi- or tri-­cultural, much in the way that people who are bilingual, trilingual or multilingual can switch between languages to suit the context with equal skill in all of them. Humans are sufficiently gifted to be able to act in several scenarios, making a clear separation between each of them. Bastide and a good many ethnologists have subsequently extended the compartmentalization principle to all multicultural situations, as opposed to the essentialism of common sense which upholds that ‘a black is a black whatever you do, my dear sir’. It is conceivable – but has still to be empirically proven on a case-­by-case basis ­– that a company, a business environment or a city construct norms and values that are sufficiently autonomous as to discredit ethnic traditions within its own sphere of influence. It is then up to the actors to negotiate the compartmentalization and compromises they wish or are able to make with respect to this shift. It would therefore seem impossible to posit a single culture (ethnic or national) for one individual, category or social group, or a homology between mainstream culture and corporate culture. If such homology exists, it has yet to be proven. A South Korean example Janelli (1993) highlights this type of compartmentalization in the case of a South Korean conglomerate, in which he conducted a lengthy survey and whose values, he says, break away from the dominant values of Korean culture that frame the employees’ practices outside the company. For example, the ‘family traditions’ that seem to prevail within the firm are radically different from those of society as a whole. Thus: Most Teasong (an assumed name for the conglomerate under study) workers did not explicitly challenge the family metaphor but rarely did they invoke it. Despite its pervasiveness in the company magazines and training

184   J.-P. Warnier s­ essions, the persons who drew analogies between the company and the family were older managers. (p. 223) And he remarks earlier (p. 215) that: ‘Not all South-­Korean organizations are characterized by a topdown authoritarian style. Academic societies, alumni associations, villages, and even lineages can conduct their affairs quite differently.’ After an intricate discussion, Janelli concludes that a Japanese colonial tradition has been re-­appropriated within Korea’s post-­colonization industrial development, involving the practices of setting the Korean population to work and instituting disciplinary control. This re-­appropriation has not however given rise to a critical elucidation of how this relates to Japanese colonization. On the contrary, these practices have been associated with ‘family traditions’ invented exclusively within companies and not in Korean society. There are thus two ‘cultures’, with this duality being felt (but assumed and put into practice) by the actors in two different contexts. The conglomerate’s employees do not have to choose between two cultural registers. They alternate between the two, in different contexts. This Korean case raises similar questions with respect to the African context, where the theme of kinship ties seems to constitute the common ground for studies on entrepreneurship. The idiom of kinship in Africa Even reinvented, the tradition of African monarchies is undoubtedly a product of local origin and of the continent’s specific history. The fact that it has been recycled in a fully modern context supports the argument that, in some cases, neo-­ traditional values, norms and institutions have a noticeable impact on capital formation, innovative capacity, trust, etc. Conversely, the Ivoirian firm studied by Bazin appears to escape the influence of ethnic traditions, or at least traditions other than those developed within an urban context and industrial environment. The contrast between these two cases merits discussion. Is it possible to imagine a kind of magic table with the lines listing the largest possible number of enterprises in emerging countries, and with each column representing the different traditions that impact (or not) on these enterprises’ performance indicators (this impact would be signed as either + or –)? For purposes of symmetry, the columns for urban, entrepreneurial and modernizing traditions could be placed after the ‘ethnic’ or ‘national’ traditions. To my mind, this is an envisageable project. Carrying it out would of course be hindered by the scarcity of company monographs to inform the boxes in the table. Data are scant, incomplete, difficult to compare and, more importantly, they age very quickly: in 2007, the profiles and dynamics of enterprises in emerging countries were already substantially different from what they were ten years previously. Markets, including financial markets, are now globalized. China is henceforth a salient player in the emerging countries of Africa, and Central and East Asia. A great many enterprises have disappeared over this

The invention of traditions   185 period. Others have been created or changed hands through succession or acquisition. Managers and directors are more mobile and more highly professionalized, etc. Overall, it can be assumed that the impact of ethnic and national traditions is less forceful than industrial and entrepreneurial traditions due to the ‘sequence of circumstances’ at urban and global level. This is evidenced by a notable change in tone with respect to kinship since the 1990s. Kinship relationships provide one of the commonplace themes of Africanist literature on business enterprise. I shall begin by giving a brief critical review of these. On the whole, firms are located in cities or in the urban sprawl. Cities are by definition composite and cosmopolitan. They are spaces in which the threads of kinship are severed and then rewoven through urban migration. The urban labour market is also inevitably composite. Any enterprise with more than 50 job positions will find it difficult to source personnel entirely from either kinship networks or ethnic groups. This will have repercussions not only on personnel recruitment, but also on capital formation. Having scoured the existing literature on this subject, Clignet (1995: 200) concludes that kinship ties are ineffective. The most common processes for capital formation result from windfall effects (which Roitman [2005] terms ‘fiscal disobedience’), trafficking, privatizations, political employment and positions, individual saving backed by the tontines that short-­circuit and neutralize kinship ties by replacing them with ‘friendship’ ties (cf. Henry et al. 1991). On the other hand, Clignet underlines that transfer of entreprises is far more sensitive to the different traditions of succession imposed by filiation systems, the varying degrees of hierarchized status in political organizations or ancient lineages, or by particular succession rights, as under Islamic law. When the head of a business dies, this may even result in the enterprise being liquidated and its capital dispersed. ‘Licit illegal’ practices Another commonplace observation involves the ‘corruption’ and illicit activities that are regularly denounced and fought in the name of good governance. The work of MacGaffey (1987, 1991, 1995) on the Congo (Zaire) and Roitman’s work (2005) on the Lake Chad region highlight not only the potential for illegal activities and the almost imperative need to resort to illegality in historical circumstances characterized by lasting crises and complexity, but also the extraordinary creativity that it produces: Zaire since independence in 1960 has moved from one political and economic crisis to another, but the extraordinary dynamism of Zairian entrepreneurs is nevertheless constantly evident. Obbo’s comment illuminates the reason: these crises produce extraordinary opportunities for some of these men and women. (MacGaffey 1995: 187) Some years earlier, MacGaffey (1991) had already tempered the demarcation between formal and informal economies drawn by Hart (who himself takes a

186   J.-P. Warnier very distanced view of the unexpected fate of his proposals). She points out that these two ‘sectors’ are vitally interdependent and so thoroughly intertwined that the frontier between the two is in many cases hard to distinguish. It often exists quite simply due to the devices of authorship or the discretion of those concerned. In a recent book, Roitman (2005) develops a line of argumentation similar to and explicitly referencing MacGaffey’s own. The following is a summary of her investigations: They [the entrepreneurs around lake Chad] say that their own activities are illegal and yet licit because they are simply modes of economic accumulation and of governing the economic, or ways in participating in forms of reasoning that constitute a particular political economy. Therefore, because evaluations of what constitutes licit practice and licit self-­conduct are not derived out of a set of ethical standards that are distinct from those that make power legible, such resistance is necessarily generated out of states of domination. (Roitman 2005: 189) Yet Roitman does not attribute the regular occurrence of these practices to transitory crisis conditions. She shows that they are embedded in the region’s long and medium-­term history. Ancient practices of conquest, plunder and enslavement were reshaped under colonial rule around taxation, pricing policy and the movements of a ‘floating population’ shifting in line with regional opportunities. What we have here, therefore, are traditions that are gradually reworked within changing historic contexts. Today, these traditions focus on what Roitman calls the ‘fiscal subject’, who is both the target of taxation and the agent of wealth creation. On page 204, she concludes that the ‘licit-­illegal’ practices do not imply the bankruptcy of the state (whose legitimacy and reasons for being are not, despite everything, put into question by those concerned), nor even perhaps a weak economy. They constitute a ‘liberal’ type of functioning that places debt at the heart of exchange and partnership relations, and which is market-­based and in no way market-­unfriendly. According to her, the state is indeed involved in a form of regulation, but this is an exception to the official declarations of politicians – particularly Paul Biya – which conform to ‘political correctness’ in words rather than acts. Through these ‘licit-­illegal’ practices, the economic agents of north Cameroon are themselves able to take on the project of modernity, not because this might be jeopardized by the state, but because such practices are an extension of those of the state, targeting the same objectives. There is no doubt that these ‘traditions’, which are observable throughout Central Africa, are instrumental in gaining contracts and establishing trust within a moral economy of debt (which I had already noted in the same region; cf. Warnier 1994). The arguments of Macgaffey, Roitman (like Hart) and other observers of Central and West Africa pose a broader problem, already raised by Braudel and,

The invention of traditions   187 before him, Marx and Proudhon: historically and concretely speaking, does capital formation occur – as Marx assumed – when an entrepreneur normally and legally reaps surplus value, or mainly through other means deriving from abusive or even illegal activities, predatory practices, the privileges of power, positions of strength, etc. Bayart (1994: 35) comes to this clear conclusion: ‘The “angelism” of the social sciences would like the relationship between capitalism and crime and violence to be contingent. But may we be permitted our doubts’, and he goes on to cite many historical studies along the same lines. Should this be the case, there is call for debate on the possible relationship between such ‘traditional’ practices and the ‘good practices’ laid down by the requirement of ‘good governance’. The tradition of invention A book by Jane Guyer published in 1987 on feeding the large African cities clearly poses the problem of what she calls (in 2000) The Tradition of Invention in Africa, inverting Hobsbawm and Ranger’s title to better fit her own purposes. She observes that, from the early years of the twentieth century, colonial authorities were greatly concerned by the issue of feeding urban populations and tried to encourage, organize and manage this issue. Their policies involved initiatives that, although praiseworthy, were of little use inasmuch as food supply had always been reliably ensured by extremely reactive artisanal networks, even and especially during periods of high urban growth in the big cities. Drawing on these observations, Guyer investigates what, in African economies, functions independently of government policy and rather in what could be termed ‘African-­fashion’ – in other words, by relying on Africa’s economic past, complex commercial practices, a profusion of old currencies that are mutually non-­ convertible, ordinal scales of values, asymmetrical exchanges, and practices of gain ‘at the margin’. These traditions of invention are transmitted in part orally (for example, through the computation systems or verbal interactions of traders), but also through corporeal, material and non-­verbal ways, which are far less accessible to reflexive consciousness, and present in all the habits and materialities of packaging, transport, the marketplace, handling different currency types, etc.

Questions of method and objects of research Anyone involved in international matters in any sphere whatsoever (business, universities, large religious organizations, diplomacy) constantly comes up against persistant and stubborn differences in practices and representations that are described as ‘cultural’ and associated with various ‘traditions’. There is no doubt that traditions exist, in the sense of stabilized practices transmitted within institutions or social groups. Once these traditions have been identified in a company, group or city, they help actors to predict certain individual or collective behaviours within a known range of variations. They help to reduce uncertainty, even in cases where they are conflictual and subject to dispute.

188   J.-P. Warnier This said, the issue at stake in the debate on traditions and the relationship between entrepreneurship and traditions seems to have shifted to such an extent since the period of the 1960s to the 1990s that I felt it important to highlight and measure this shift. At the close of the 1980s, traditions were usually associated with the past of a country, nation, social or ethnic group that persisted into the present. The main paradigm shift of the last 15 years concerns the question of how traditions relate to the past and present. Thanks to the thrust of certain historians, it is becoming heuristically profitable, in each case study, to ask the question whether these ‘traditions’ are not the product of an invention process that selects, rejects, retrieves, innovates and reworks traditions according to specific historical contexts. At the beginning of the twenty-­first century, the context relevant to this invention would seem to comprise the programme of modernity, the urban phenomenon and the globalization of flows. It could be thought that these three factors might combine in such a way that traditions/civilizations would converge. This, however, does not match empirical observation, since the diversity of political dynamics the world over is such that inventions are diffracted in myriad ways, as is shown by the collective publication edited by Bayart (1994) on the reinvention of capitalism. This paradigm shift goes hand-­in-hand, as is always the case, with methodological readjustments to suit new objects of investigation. As long as traditions were considered as the persistence of age-­old practices into the present along more or less the same lines as the past, the research method generally applied involved referring back to a former state of society X or Y, sometimes deemed to be a ‘primordial’ or ‘primary’ state, and then reconstituting the ‘traditional’ functioning at the date ‘T0’. Only then, as in a second step, did one search the present (at date ‘T1’) for signs of continued age-­old practices. Most often and by construction, a twofold conclusion was reached: on the one hand, the finding that there has been an impressive erosion of traditions due to the mechanical action of modernity and, on the other, the long-­term persistence of certain more resistant elements of tradition. Today, the perspective has shifted: it now involves analysing, in the present, practices that are relatively stable, relatively predictable and hence institutionalized as ‘traditions’. The word ‘relatively’ points up all the ambiguity of this kind of exercise as, in each case study, the temporality specific to the traditions under study must be suitably identified. This means ipso facto extracting them from a primordialist essentialization and integrating them into a historical trajectory. Then, in a second stage, current traditions can be linked both to the existing historical context (modernity, urbanization, globalization of flows) and, using a regressive approach, to the institutional and historical configurations of the past. Thus, the pessimistic, backward-­looking account of the ‘loss of traditions’ (cultural ‘reference points’, ‘roots’, etc.) will lose sway, leaving room for a different account that is more open to current dynamics, including their conflicting aspects. This regressive approach is illustrated by the works of Guyer, Roitman, Janelli, Bayart, and others cited in this chapter. With respect to the relationship

The invention of traditions   189 between entrepreneurship and traditions, this paradigm shift therefore proposes new lines of questioning, research objects that are changing place, and additional to the existing set of methodological tools.

Notes 1 It should be emphasized that ‘culture’ or ‘repertoire of actions’ as entry points into the subject would not necessarily have the same heuristic effects, as MacLuhan’s saying ‘medium is message’ affirms, which suggests that content depends partly on the container provided by the notional perspective one chooses. 2 This wordplay is derived from the French slang for excrement: ‘merde’. 3 Drawing on ethnographic field studies conducted in Indian villages, Deliège (2004) convincingly showed that the intellectual tradition to which Louis Dumont belongs – that of Tönnies and Durkheim who made the opposition between traditional and modern societies – relied on incomplete and fragile empirical knowledge and on weighty cultural, even ideological, presuppositions. In addition, Dumont’s structuralist approach, being more sensitive to verbalized norms than to actors’ practices and social historicity, did not allow him to avoid the trap of stereotyping with respect to the so-­ called ‘traditional’ and ‘hierarchized’ societies as opposed to ‘individualist’ and ‘modern’ societies. 4 I propose a monograph of this kingdom and its contribution to the invention of African modernity in a work published in two versions: The Pot-­King. The Body and Technologies of Power (2007), and Régner au Cameroun. Le Roi-­pot (2009). For more on the Mankon museum, see the splendid catalogue by Jean-­Paul Notué and Bianca Triaca (2005). 5 The Mankon kingdom recognizes the citizenship of a category of assimilated foreigners and grants them a specific status. They are called ankyeni, a word designating a species of migratory grasshopper that appears at the end of the rainy season and is edible and much appreciated. 6 In 1992, the political unrest in Cameroon had repercussions on the kingdom. The king remained loyal to his alliances and campaigned for Paul Biya, who stood as candidate to his own succession as President of the Republic, against John Fru Ndi who then lived in Mankon and was supported by all of the king’s subjects. On the periphery of the ‘Dead Cities Campaign’, an angry crowd of Mankon citizens set fire to the king’s offices in the provincial capital of Bamenda and marched on the palace, burning the village of a leading member of the royal clan on its way. 7 The ethnology and history duo have had a stormy and complicated relationship. The account that is most relevant to my present topic was written by Thomas (1989), who underlines the success of Radcliffe-­Brown in founding a professionalized ethnology, according to which only data collected in the field by qualified ethnologists using appropriate methodology were to be taken into account. Radcliffe-­Brown thus rejected the inclusion of data gathered or elaborated by merchants, missionaries, state officials, doctors, colonial administrators, the military or adventurers – in other words, all data that evidences and reflects the social historicity. Consequently, ethnologized societies and their ‘traditions’ have been placed out of time, out of history. Since the 1990s, and with great effort, this bias has been surpassed, but it still weighs heavily on the analysis of traditions. 8 The ongoing ambitious research programme headed by d’Iribarne will provide documentation and answers to these questions.

190   J.-P. Warnier

References L’Afrique des entreprises (1998) Groupe de l’Agence Française de Développement et La Documentation Française. Anheier, H. and Isar, Y.R. (eds) (2007) The Cultures and Globalization Series 1. Conflicts and Tensions, Los Angeles and London: Sage. Bayart, J.F. (ed.) (1994) La Réinvention du capitalisme, Paris: Karthala. Bazin, L. (1998) Entreprise, politique, parenté. Une perspective anthropologique sur la Côte d’Ivoire dans le monde actuel, Paris: L’Harmattan, coll. Connaissance des homes. Beaujard, P.H. (2005) ‘The Indian Ocean in Eurasian and African World-­Systems before the Sixteenth Century’, Journal of World History, 16 (4), pp. 411–65. Bloch, M. (1995) ‘Le cognitif et l’ethnographique’, Gradhiva, 17, pp. 45–54. Bollinger, D. and Hofstede, G. (1987) Les différences culturelles dans le management. Comment chaque pays gère-t-­il ses hommes?, Paris: Les Editions d’Organisation. Clignet, R. (1995) ‘Les entreprises et la question de l’héritage’, in Ellis, S. and Fauré, Y.A. (eds) (1995), pp. 195–206. Deliège, R. (2004) Les castes en Inde aujourd’hui, Paris: P.U.F., Coll. Sociologie d’aujourd’hui. Dimitrievic, D. (ed.) (2004) Fabrication de traditions, invention de modernités, Paris: Ed. de la M.S.H. d’Iribarne, P.H. (1989) La logique de l’honneur. Gestion des entreprises et traditions nationales, Paris: Le Seuil. Dumont, L. (1967) Homo hierarchicus: le système des castes et ses implications, Paris: Gallimard. —— (1977) Homo aequalis, genèse et épanouissement de l’idéologie économique, Paris: Gallimard. Ellis, S. and Fauré, Y.A. (eds) (1995) Entreprises et entrepreneurs africains, Paris: Karthala-­Orstom. Guyer, J. (ed.) (1984) Family and Farm in Southern Cameroon, Boston University, African Studies Centre. —— (1987) Feeding African Cities. Studies in Regional Social History, London: Manchester University Press for IAI. —— (2004) Marginal Gains. Monetary Transactions in Atlantic Africa, Chicago and London: University of Chicago Press. Henry, A. (1988) ‘Peut-­on redresser une entreprise africaine en respectant la parole des ancêtres?’, Annales des Mines, gérer et comprendre, 15, pp. 86–94. —— (1995) ‘Entreprises modernes, entreprises métisses’, in Ellis, S., and Fauré, Y.A. (eds) (1995), pp. 175–86. Henry, A., Tchente, G.H. and Guillerme-­Dieuregard, P.H. (1991) Tontines et banques au Cameroun, les principes de la société des amis, Paris: Karthala. Hobsbawm, E. and Ranger, T. (eds) (1983) The Invention of Tradition, Cambridge: Cambridge University Press. Janelli, R.L. (1993) Making Capitalism. The Social and Cultural Construction of a South Korean Conglomerate, Stanford: Stanford University Press. Kopytoff, I. (ed.) (1987) The African Frontier. The Reproduction of Traditional African Societies, Bloomington/Indianapolis: Indiana University Press. Macgaffey, J. (1987) Entrepreneurs and Parasites. The Struggle for Indigenous Capitalism in Zaire, Cambridge: Cambridge University Press. —— (1991) (ed.) The Real Economy of Zaire. The Contribution of Smuggling and Other

The invention of traditions   191 Unofficial Activities to National Wealth, London: James Currey, Philadelphia: The University of Pennsylvania Press. —— (1995) ‘Chaotic Social Change in Zaire. Its Opportunities and Problems for Entrepreneurs’, in Ellis, S. and Fauré, Y.A. (eds) (1995), pp. 187–94. Notué, J.P. and Triaca, B. (2005) Mankon. Arts, Heritage and Culture from the Mankon Kingdom (Western Cameroon), Milan: 5 Continents Editions. Perrot, C.H. (2003) ‘Introduction’, in Perrot, C.H. and Fauvelle-­Aymar, F.X. (eds) (2003), pp. 7–23. Perrot, C.H. and Fauvelle-­Aymar, F.X. (eds) (2003) Le Retour des rois. Les autorités traditionnelles et l’Etat en Afrique contemporaine, Paris: Karthala. Pouillon, J. (1991) ‘Tradition’, in Bonte, P. and Izard, M. (eds) Dictionnaire de l’ethnologie et de l’anthropologie, Paris: P.U.F., pp. 710–12. Roitman, J. (2005) Fiscal Disobedience. An Anthropology of Economic Regulation in Central Africa, Princeton and Oxford: Princeton University Press. Thomas, N. (1989) Out of Time. History and Evolution in Anthropological Discourse, Cambridge: Cambridge University Press. Warnier, J.P. (1994) ‘La bigarrure des patrons camerounais’, in Bayart, J.F. (ed.) (1994) pp. 175–202. —— (2007) The Pot-­King. The Body and Technologies of Power, Amsterdam: Brill. —— (2009) Régner au Cameroun. Le Roi-­pot, Paris: CERI-­Karthala.

Part IV

Culture and poverty reduction

10 Culture matters for poverty, but not because of a culture of poverty1 Notes on analytics and policy Michael Walton

Do poor people stay poor because of their culture? Does culture shape the dynamism or stagnation of groups, even of nations? The phrase ‘a culture of poverty’ was coined by the anthropologist Oscar Lewis. But the idea that there exists such a thing as a culture of poverty is generally derided by the anthropology profession. Most economists tend to just ignore questions of culture. Yet the culture of poverty view keeps coming back, sometimes implicitly in attitudes to the poor and approaches to policy, sometimes explicitly, as, for example, in contributions of the recent book by Harrison and Huntington (2000), entitled Culture Matters. This chapter takes on the question of the relationship between culture and poverty. It provides a review of the analytics of the issues in four areas: micro (or group-­based) interpretations of links between culture and poverty; potential links to growth processes; implications for normative assessments; and consequences for policy. We first state a version of the culture of poverty view based on the characteristics of poor groups. We argue that this formulation of the relationship between cultural influences and poverty is incorrect, and potentially dangerously misleading for both normative assessment and policy design. Culture of poverty perspectives have three problems: they view cultural features of the poor in isolation from their relations with other groups; they underestimate the adaptability of economic behaviour to distinct cultural histories; and they give insufficient attention to the pliability of culture, in response to economic, political and social conditions. The chapter then argues that culture does matter, and offers an alternative approach: cultural factors can play a role in sustaining inter-­group differences in wealth, status and power. Such inequalities can shape beliefs, aspirations, stigma, social narratives and symbolic and information structures, with manifestations in collective organization, economic behaviour and individual psychology. If the mechanisms involved are self-­enforcing and persistent, this can be considered to be an ‘inequality trap’. Where such an inequality trap exists, it implies that subordinate groups are maintained at least in relative poverty, and that these are associated with culturally shaped behaviours and stigmatized identities. These can include endogenous preferences that limit the prospects of

196   M. Walton poorer, subordinate, groups. The sustained condition of (relative) deprivation may look like a ‘poverty trap’, but such a view constitutes a misdiagnosis of the underlying causes of the condition.2 Group-­based differences in wealth and status are sustained over time by interlocking (formal and informal) economic institutions, power structures and cultural relations between groups. It is important then to understand the dynamics of the system, including both cultural dynamics and how these interact with both economic and political processes. We then turn to the question of whether culture can influence economic growth. The national version of the culture of poverty view is reviewed and rejected. But there is again a case for cultural factors influencing growth-­related processes, though here the chapter is more speculative. There is some micro support for the view that there are interactions between culturally shaped inequality traps and accumulation. At the macro level, the evidence is inconclusive, but there is a case for exploring how group-­based identities can interact with political processes in ways that tend to be growth-­dampening or promoting. The chapter then shifts from positive to normative analysis – since normative considerations are an important aspect of the ‘culture of poverty’ view. It first relates the analysis to both aggregative and equity dimensions of welfare – in Sen’s language the level and equality of capabilities. We argue that equality of agency, that is culturally shaped, is (often) an ingredient of equality of opportunity, and this concept goes beyond many existing formulations of this notion. For normative evaluations, we are particularly interested in transitions that lead to either greater equality or increased growth-­enhancing processes. Either is potentially superior, though we have a special interest in shifts to a ‘better’ and more efficient inequality equilibrium, that also has a positive causative relationship with growth-­related processes. Such an equilibrium will have superior aggregate outcomes and equity characteristics, but is not necessarily Pareto-­superior. Breaking inequality traps will often intrinsically involve losses to those who benefited from initial structures of inequality. Finally we review policy implications, in terms of categories of policies that may be relevant to breaking culturally shaped inequality traps. We suggest that change may occur from a number of sources, including exogenous economic and political shifts, and endogenous processes within the system, including social mobilization, or from elite groups deciding to effect change from above. Where there is a self-­enforcing inequality trap, there may still be some scope for agency within the system; it is important to understand this for the design of policy. The chapter is an analytical review of these issues, with a primarily non-­ formal presentation. In many parts we use the Indian case as an example, including in particular the position of scheduled castes (dalits) and tribal groups (adivasis). This is not intended as a contribution to the extensive literature on these groups, but rather as an illustration of the analytical arguments. Since the focus is on poverty and inequality, the chapter does not review an important set of other domains in which economic processes interact with cultural processes and social norms, such as the role of culturally shaped norms in managing the commons.3

Culture and poverty   197

From ‘cultures of poverty’ to cultural dimensions of inequality Culture has many meanings. In this chapter we do not confine ourselves to a specific definition.4 We are broadly concerned with ‘culture as identity’, rather than culture as artistic production, even though the two are related. Culture as identity concerns how groups define themselves in relation to other groups, through a broad range of social practices, symbolic meanings, aspirations, norms of social interaction and belief systems. As a fundamentally relational concept, it is both dynamic and contested, changing in the process of intra- and inter-­group inter­ actions. Cultural dimensions of living can both influence and be influenced by economic and political processes. This approach implies that we are concerned with issues relating to socially meaningful groups. People are ‘members’ of multiple groups – an individual may be of a particular caste, gender, community, nation, geographic region, sports group and so on. Each of these identities plays a role in his or her life, and the mix of identities that matter to a person can change over time (Sen 1998). Which of a person’s many identities matter is a function of context. It depends on social and political processes that can lead to more or less ‘groupness’ of a particular category, as much as individual choice. As Brubaker argues, ‘if we treat groupness as a variable and distinguish between groups and categories, we can attend to the dynamics of group-­making as a social, cultural and political project, aimed at transforming categories into groups or increasing levels of groupness’ (Brubaker 2002: 170–1)) Take caste and tribal status in India. Caste is a complex social structure that involves the hierarchical allocation of individuals to distinct social categories, traditionally associated with rigid occupational allocations and a variety of ritualistic and symbolic restrictions. The primary unit is the jati, that is typically an occupationally and geographically specific grouping, endogamous and heritable.5 At the bottom of the hierarchy are groups, into which a set of jatis fell, that were formally outside the caste structure, and known in the past as untouchables, because of ritualistic restrictions on contact between higher castes and individuals in these groups. Their occupations were restricted to low status and menial jobs, such as leather tanning, sweeping, or clearing human waste. Since independence the term dalits, meaning ‘oppressed’, was adopted by politically mobilized members from these groups. As with (almost) any conception of human identity, dalit is a constructed and dynamic concept, as we discuss further below. By virtually any dimension of well-­being ­– including life expectancy, material wealth, income, education, dignity, insecurity – individuals in this group traditionally suffered from severe deprivations associated with their ascribed social position. Tribal groups – also known as adivasis, or ‘original’ occupants of India – are more akin to what in other regions are often referred to as ‘indigenous’ groups. They fall outside the caste hierarchy, live primarily in relatively remote forest and hilly areas, have distinct beliefs, practices and customs, and are systematically

198   M. Walton more deprived in terms of access to public goods, educational status and wealth than most Indians. Dalits account for some 15 per cent and adivasis for around 8 per cent of India’s population. Both groups have been the subject of explicit government policy since independence, with the banning of untouchability and the introduction of affirmative provisions for so-­called ‘scheduled’ castes and tribes (essentially corresponding to dalits and adivasis) in the Indian Constitution, backed by a series of subsequent policy initiatives to promote their position in society.6 The continued relative deprivation of these groups is evidenced in measures of human development and expenditure poverty (Tables 10.1–10.3). With Table 10.1  Infant mortality rate by social category, 1980 and 2000 (in number per 1,000)

Scheduled castes Scheduled tribes Others All

1980

2000

Difference

127 101   99 109

83 84 68 73

−44 −17 −31 −36

Source: Thorat (2007).

Table 10.2  Literacy rate by social category, 1980 and 2000 (in percent)

Scheduled castes Scheduled tribes Others All

1980

2000

Difference

24 19 47 41

55 47 69 65

30 28 22 23

Source: Thorat (2007).

Table 10.3  Poverty incidence by social category, 1993/4 and 2004/5 (in percent)

Rural Scheduled castes Scheduled tribes Others All rural Urban Scheduled castes Scheduled tribes Others

1993/94

2004/05

Difference (percentage points)

48.6 51.7 31.4 37.2

37.4 47.0 22.9 28.4

−11.2 −4.7 −8.5 −8.9

51.5 47.8 30.5 33.9

41.2 40.3 23.6 26.7

−10.3 −7.5 −6.9 −7.2

Source: National Sample Survey, Ministry of Statistics and Programme Implementation, Government of India.

Culture and poverty   199 respect to human development indicators, both scheduled castes and tribes were substantially more deprived in terms of infant mortality and literacy in both 1980 and 2000, with a mixed picture on changes. In terms of simple percentage point differences, scheduled castes experienced somewhat faster progress than ‘others’ on both measures; scheduled tribes experienced slower gains with respect to infant mortality, faster gains on literacy.7 With respect to expenditure poverty, both social groups remain substantially more deprived in 1993/4 and 2004/5, but with larger reductions than for ‘other groups’ for scheduled castes and significantly lower reductions for scheduled tribes in rural areas (where most live). Note that these results are already inconsistent with most theories of a poverty ‘trap’, since the latter would imply groups stuck in absolute poverty. An important feature of these group-­based differences is their persistence, at least in ranking. It is said that the caste system has persisted for thousands of years, though this undoubtedly overstates its rigidity. As with ‘tribal’ categories in Africa, there is evidence that caste was made more rigid, and possibly more salient, by the practices of the British to codify and then measure the hierarchy they believed they had found.8 But what is more immediately relevant here is the persistence of difference in the post-­independence period, despite democratization, major economic change and concerted policy effort. As with blacks in the United States, indigenous and Afro-­Latino groups in Latin America, the hypothesis is that severe stigmas can outlive the abolition of formal barriers to equality of opportunity, or even proactive measures to redress them. So we are interested in explaining this absolute and relative structure of deprivation. We treat such outcomes as exemplars of deprivations associated with particular social groups in many societies, and explore whether there is anything that can be characterized as cultural about such profound, persistent, group-­ based differences. If so, we want to know what it implies for the design of policy. The culture of poverty concept – groups Let’s start with the original coining of the phrase – that lies in the characterization of poor families. Oscar Lewis used the phrase a ‘culture of poverty’ in his 1959 ethnographic study of Mexican families (Lewis 1959). He was concerned with describing the features of the poor, in terms of a series of social, economic and psychological traits. These included, in Lourdes Arizpe’s summary: chronic unemployment and underemployment leading to low income, lack of property ownership, absence of savings . . . lack of effective participation and integration in the major institutions of the larger society . . . high illiteracy rates . . . tend not to participate in national welfare agencies, labour unions, or political parties . . . a mistrust of government, and those in high position . . . experience gregariousness with a minimum of organization beyond the nuclear and extended family. (Arizpe 2004: 168–9)

200   M. Walton What is striking about this account is how similar it looks to typical recent poverty studies undertaken by development agencies, such as the World Bank.9 Yet this notion quickly became linked to the view that such a range of attributes were a consequence of ‘cultural’ aspects of the poor – socialized norms, behaviours and practices that kept them in poverty rather than fostering whatever it would take to lift themselves out of deprivation. The idea of a culture of poverty spread fast. It became popular in some circles in the United States, especially in relation to an explanation of the causes of African-­American deprivation. Such views remain current today. As Glenn Loury argues, racial disparities in outcomes, such as imprisonment, incomes, jobs and so on, are often treated as if they ‘can be accounted for by a narrative line attributing the outcome to the inadequacies of the persons who suffer the condition, not to any as yet undiscovered problems with our own social organization’ (Loury 2002: 81). What of economists? We come below to some of the more sophisticated economic treatments of cultural aspects of economic behaviour. Here we note what a culture of poverty would mean in terms of standard economic theory of household behaviour. Unlike most anthropological accounts, basic economic theory is relentlessly individualistic, with outcomes determined by the interaction between the preferences and constraints of individuals (usually bundled into households in economists’ models). Households seek to maximize their welfare, in the form of present and future consumption of goods, subject to a budget constraint, the latter determined by initial wealth and work opportunities, and with decisions also shaped by preferences between leisure and work. Investments, including in the education of their children, is determined by their assessment of future returns to this education, the foregone cost of schooling (including in reduced work by the kids), their willingness to wait, access to capital markets and any intrinsic value given to schooling. Households also are concerned about risks, and are most commonly ‘risk averse’ (in part because the value of increments to income at low levels of spending is generally expected to be greater than the same increment to income at higher levels of income). Within this perspective, a common view (at least implicitly) is that the difference between being poor and non-­poor, is that the non-­poor face less stringent constraints – whether in terms of their initial endowments of wealth, the returns to work, the cost of borrowing, the risks they face, or the possibilities of insuring. So a simple way of bringing cultural factors into this perspective would be to say that the poor have different, culturally shaped, preferences – in terms of leisure, time preferences (and so savings), the intrinsic value of education, attitudes to risk and so on. While behaviour is individualistic, preferences can be shared across other members in the group. And since standard economic perspectives treat preferences as given, that’s the end of the story. A simple illustration of this is given in attitude surveys on why the general public think the poor are poor: in the United States 60 per cent of those surveyed in the World Values survey believe the poor are ‘lazy’ (compared with ‘only’ 26 per cent of Europeans – though in reality there is no relationship between work hours across countries or in relation to those of the rich). ‘Laziness’ is a manifestation of a

Culture and poverty   201 preference. In similar spirit, an initial class of economic models to explain discriminatory behaviours against groups – defined by race, ethnicity, gender, etc. – relied on a ‘taste’ for discrimination, that meant that individuals were willing to pay more to avoid a transaction or contact with a person from the group that they did not like (Becker 1971). But the preference, or taste, is primitive. A second strand of work in economics concerns the idea of a ‘poverty trap’, in which the poor stay poor because they are poor. While this has superficial parallels with the culture of poverty view, the types of models that explain poverty traps are almost always driven not by fundamental differences in preferences, but by other features of the environment of the poor. One story, for example, is that the very poor have such low food intake that they cannot work hard enough to gain the income necessary to lift them out of their low-­nutrition condition.10 Another would be that they can only save the bare minimum for the reproduction of the household system (e.g. to cover subsistence consumption plus enough to keep the bullocks alive in a farm-­household) with none left over for human or physical capital accumulation. These stories have nothing to do with culture, and are not consistent with the brute facts of significant trends in conditions of dalit and adivasis at least for the averages of the groups reported in Tables 10.1 to 10.3. But we flag this class of model here, since we contrast poverty traps with inequality traps below. Critique The main critiques we are interested in here come from within anthropology. (We come back to economics next, when we look at alternative characterizations.) There are positive and normative dimensions to these. With respect to positive analysis, the essence of the critiques is that a culture of poverty view fails to account for the fundamentally relational and dynamic character of cultural features of life. Yes, there are distinctive cultural features of the lives of the poor, but these are not primordial, essential aspects of the group that cause the state of poverty. They are rather products of interactions within the overall society, and with their economic and physical environment. As Appadurai says: ‘There may not be anything which can usefully be called a “culture of poverty” (anthropologists have rightly ceased to use this conceptualization), but the poor certainly have understandings of themselves and the world that have cultural dimensions and expressions’ (Appadurai 2004: 65). There are two aspects of the critiques that it is useful to highlight: power and the future. A culture of poverty view is detached from overall structures of power. Yet even mainstream accounts have come to recognize that ‘disempowerment’ is both the cause and consequence of poverty. In the World Bank and other development agencies ‘empowerment’ of the poor is typically part of what is advocated for a poverty reduction strategy. Empowerment is often portrayed as something that can occur without disturbing overall structures of power, but this was nevertheless an important step for organizations that had typically given

202   M. Walton limited attention to power.11 Within anthropology, the work of Bourdieu is central to the exploration of the relationship between power and culture. A key concept is that of habitus, or the implicit, taken-­for-granted structures that shape the full range of daily, ritualistic and symbolic behaviours, in ways that perpetuate existing structures of social difference, within the home, between genders, within the community and in broader interactions within and between groups. Where dominant cultural structures lead to the internalization of a restricted, demeaning and lower status view of the world for subordinate groups this is said to be a form of ‘symbolic violence’. As Swartz says: Habitus tends to shape individual action so that existing opportunity structures are perpetuated. Chances of success or failure are internalized and then transformed into individual aspirations or expectations; these are then in turn externalized in action that tends to reproduce the objective structure of life chances. (Swartz 2000: 103) Thus habitus provides a bridge between the past and the future. The second aspect of critiques we highlight explicitly concerns the future. Appadurai (2004) argues that there has been a tendency to see ‘culture’ as primarily concerned with the past, as reflecting the current social deposits of a group’s history and traditions. This has fed the view of many development practitioners that culture is the enemy of development, embodying resistance to change, excessive spending on ritual and festival, restrictive behaviours and inertia. All this needs to be swept aside if the positive forces of economic and technological change are to tackle problems of societal and individual poverty. More subtle is the implicit allocation of the past to anthropology and the future to economics, the latter bringing its set of tools on forward-­looking agents, making decisions in relation to the prospect of future returns, risks, opportunities to borrow and accumulate human and physical capital. But these views are inadequate as accounts of how culture interacts with social behaviour, that is as much about how people engage with and prepare for the future as the past. Appadurai formulates the concept of the ‘capacity to aspire’ as a way of organizing cultural aspects of future-­oriented behaviour, a concept developed from ethnographic work with poor slum-­dwellers and social activist groups in Mumbai. This is conceived as a capacity to conceive of, plan, organize and take action with respect to future social and economic paths, formed as a product of group-­based interactions. These may occur in the home, in processes of social mobilization and in the implicit or explicit expectations or barriers met in daily living. His thesis is that this capacity is unequally distributed across different groups in a society, a product of unequal ‘terms of recognition’ across different groups, and associated internalized patterns of imagining of future alternatives, or ‘navigational’ paths. The normative dimension of the critiques is a direct corollary of the positive analysis, and was already reflected in the above quotation from Loury. It is a small step from arguing that the poor are poor because of ‘their’ culture, to

Culture and poverty   203 taking the ethical position that it is their fault. Yet if cultural dimensions of the lives of the poor are a product of relations within the context of unequal social and power structures, this amounts to blaming the victim. This matters for policy design. A culture of poverty view can form the basis, implicitly or explicitly, of justifications of a policy of neglect or punishment for being poor, or, at best, a paternalistic stance. These critiques can be related to the Indian examples. The caste system is intrinsically a society-­wide, hierarchical system, in which dalits are sustained in their poverty because of the range of economic, social and symbolic restrictions in the caste structure. It is thus intrinsically about unequal, relational features of social life. The position of adivasi groups is different, since these groups were not formally part of a hierarchical system, but rather experienced histories of exclusion, low status, weak provision of public goods and lower opportunities relative to the dominant social and political structures of the society. This is somewhat closer to the position of ‘indigenous’ groups in contemporary Latin American societies, that have lived outside formal hierarchical structures for a long time. But the mingling of cultural forces with political influences is still salient, and we could again examine the unequal terms of recognition that they face. This is certainly the position of scholars and activists of adivasi groups, who see them as historically marginalized, neglected and demeaned, with a consequential restriction in their horizons. An alternative: culturally shaped inequality traps So far the argument has been negative, that the culture of poverty view is inadequate as an account of the role of culture for the poor, let alone as a causal theory of poverty. In this section we turn to a more constructive approach, describing how cultural factors can play a significant role in the creation and persistence of poverty. This builds on the view that culture is essentially relational, and argues that we need to focus on the persistent effects of social and economic inequality, with effects on the poor reflective of overall relationships between groups. An intuitive definition is provided by Rao: Inequality traps . . . describe situations where the entire distribution is stable because the various dimensions of inequality (in wealth, power, and social status) interact to protect the rich from downward mobility, and to prevent the poor from being upwardly mobile. (Rao 2006: 11) The argument involves two moves. First, we argue that the concept of an ‘inequality trap’ is a useful way of looking at the condition of many societies, that is analytically distinct from a poverty trap. A society can get stuck in a position of a high level of inequality, with cultural factors playing a role in the manifestation and persistence of inequalities (interacting with economic and political

204   M. Walton structures). A formal account is given in the Appendix. Second, we outline the different ways in which cultural factors can lead to a high-­inequality equilibrium. These have been approached in quite different ways, depending on whether the underlying source of a high inequality trap flows from issues around information, or more directly from group-­based differences in power, status and stigma. The precise definition of an inequality trap formalized in the Appendix confines this concept to a long-­run equilibrium, in which one group enjoys a distribution of advantages that are permanently superior to another group. Advantage is used broadly to refer to positions of value in a society, and can include dignity and social status, political influence, as well as health status, income and monetary wealth. While this chapter is focusing on implications for the bottom of the distribution, exemplified by dalit and adivasi groups, this can apply to different parts of the distribution. There is often keen interest in which group is on top: for example, the traditional dominance of Brahmins relative to middle groups was an issue of major social and political mobilization in Tamil Nadu and other parts of southern India.12 And while the formal definition refers to a steady state, we may more loosely classify as an inequality trap a situation where the process of moving to a less unequal equilibrium is glacial. What do the data suggest? As already seen in Tables 10.1–10.3, there are signs of both absolute, and possibly relative, progress of scheduled castes with respect to poverty incidence and human indicators in the past couple of decades. Table 10.4 reorganizes the expenditure data to present the ratio of group averages of expenditure per capita for the three aggregated groups of scheduled castes, scheduled tribes and ‘others’. Even at this level of aggregation there is a striking result: there is no evidence of any trend towards convergence, let along ranking changes, between these three groups. If anything, there is mild divergence between scheduled castes and ‘others’, and large divergence between scheduled tribes and ‘others’. These aggregate results undoubtedly hide more complex distributional shifts at finer levels of aggregation, with possible a small minority of these two subordinate groups – and more so for scheduled castes – gaining substantially in the post-­independence period (see Somanathan 2010). So far the definition, and the empirical illustration in Table 10.4, is about the distribution of advantage. Yet we are also interested in levels, especially given Table 10.4 Average expenditure per capita by group in India as a ratio of the population average, 1993/4 and 2004/5 Rural (%)

Scheduled castes Scheduled tribes Other

Urban (%)

All (%)

1993/4

2004/5

1993/4

2004/5

1993/4

2004/5

85 81 107

85 74 108

76 83 105

73 77 106

80 75 108

78 65 110

Source: National Sample Survey, Ministry of Statistics and Programme Implementation, Government of India.

Culture and poverty   205 the initial motivation of this chapter in terms of poverty. With an absolute concept, poverty is determined by both the average (or aggregate) level of advantage and its distribution across individuals and groups. Thus we are keenly interested in the efficiency properties of alternative equilibria, that determine the aggregate level of advantage. We return to this below, but note an analytical point: even the cursory empirical exploration of Tables 10.3 and 10.4 suggests that India is in a situation not of persistent absolute poverty, but of persistent inequalities between groups. An inequality trap has been defined in terms of group-­based differences, and often groups will correspond to apparently cultural divisions. But this has so far nothing on cultural processes. A theme of this chapter is that treating cultural groups as given is one of the major weaknesses of simplistic, but often influential, approaches to culture. An inequality trap could be caused by many factors, flowing from economic, political, environmental or other influences, as much as cultural ones. These will generally, indeed intrinsically, interact. We focus on cases where cultural factors play a role, in keeping with our focus. There are a variety of approaches to bringing culture in. We categorize these into three (Figure 10.1): approaches driven primarily by issues of information structures; and two categories of approach where the culture–power interaction is central, dividing between those in which subordination is an essentially rational response to incentives, and those in which unequal terms of recognition seep more profoundly into internalized structures of identity.13 In a given society it is quite likely that all three processes are relevant, feeding off each other. But it is useful to characterize them separately. We emphasize in particular work within economics, in keeping with one of the objectives of the chapter, to illustrate how economics has something to offer in this traditionally anthropological and sociological terrain. Information, coordination and stereotypes. In the first category of approach, a fundamental driver is limited information, often embedded within strategic interactions. Individuals have to navigate an uncertain world, and use their beliefs on Inequality traps

Information and stereotypes

Unequal power and culture

‘Rational submission’

Figure 10.1  Cultural factors and inequality traps.

‘Internalized hierarchy’

206   M. Walton types of people to make decisions on their behaviour, whether this concerns education, employment, a contract or a more subtle informal social contact. Views on types may be supported by histories of discrimination and inequality and sustained by social custom. But the driver of behaviour in the specific, contemporary processes is cognitive – working out how best to make a decision given expectations on the behaviours of others. Take the results of recent surveys of employer behaviour in India from an ongoing study on discrimination against dalits in India. In a study involving in-­ depth interviews with the human resource departments of 25 companies, virtually all asserted discrimination by caste or other category was a thing of the past, in what they characterize as the modern, forward-­looking part of the economy (Jodhika and Newman 2007). Yet many did take other considerations into account that they considered to be good indicators of productivity for the jobs for which they were recruiting: this included family background, education of parents, geographic origin, and sometimes positive stereotypes (e.g. an airline company favouring sardar [Sikh] girls that are well spoken, and from ‘good families’). What is interesting about these responses is that they effectively amount to a set of alternative mechanisms for screening out potential dalit workers, or, for that matter, adivasis, and other low-­status categories, say from rural Uttar Pradesh, or the children of agricultural workers. These qualitative interviews are consistent with the results of a field experiment in which the researchers sent several replies to job advertisements, that were substantively the same in terms of qualifications but differed only in the name. Names were chosen to be either typical of dalits, high-­caste Hindus or Muslims. The researchers then analysed the difference in initial callbacks for interview (i.e. passing the very first round of the application process.) They found that candidates with the same qualifications but a dalit or Muslim name had a statistically significant lower probability of callback than one with a high-­caste Hindu name – 67 per cent for the dalit, 33 per cent for the Muslim (Thorat and Attewell 2007). The main question in this kind of case concerns how a ‘discriminatory’ outcome can be a long-­term equilibrium, that is a critical element of an inequality trap. Why don’t employers discover ways of finding high-­productivity dalit workers, now screened out of the process? And if this is essentially a situation of asymmetric information, in which high-­productivity dalit individuals know their potential, why don’t they work out ways of distinguishing themselves from other low-­productivity dalits? The general answer is that a self-­enforcing equilibrium is often feasible, through a variety of mechanisms. We illustrate this with three examples. In a recent paper, Basu (2006) develops a model in which individuals with the same innate productivity but from different groups may experience discrimination in economic life. The central idea is that the ‘economic life’ concerned involves a series of contracts that are complementary to each other and may be affected by the identity of the individual. These contracts may be for sales, for provision of working capital, and so on. Following Basu’s formulation this is represented by the revenue line OY in Figure 10.2, that displays increasing

Culture and poverty   207 Y

Income, costs

C

O Number of contracts

n*

nmax

Figure 10.2 The effects of increasing returns in the number of contracts of a potential entrepreneur (source: Basu (2006)).

returns in the number of contracts for a relevant range. However, contracts also have costs. If costs are fixed per contract, they can be represented by the straight line OC. The figure is drawn for the case in which there is a critical value n* below which the individual makes losses, and above which it makes profits. nmax simply represents the maximum number of contracts the individual can handle in the period in question. The position and shape of the curve will be influenced by the innate productivity of the person, but differences in productivity are not required to generate an unequal equilibrium. All that is required is for potential providers of contracts to hold distinct beliefs over the probability of others providing contracts. This may reflect a belief about the underlying productivity, a belief about the extent of social networks of different groups, or about discriminatory behaviour of society at large. For whatever reason, suppose potential contract-­providers believe that a small firm owned by a dalit will get less than n* contracts, whereas one owned by an upper-­caste Hindu will get greater than n*. Then the dalit entrepreneur will get zero contracts and the upper caste Hindu one will get nmax contracts. And this will be a sustainable, unequal equilibrium. Furthermore, if the market is not saturated, this will have lower output than an alternative, full-­information equilibrium in which there is no discrimination against dalits. This suggest a prima facie case for policy to push the system to the other equilibrium even if the only concern is with efficiency. Now consider a different case, developed in the classic article on the economics of caste by Akerlof (1976). This paper is also concerned with how an unequal equilibrium can be sustained as a rational outcome of incentives –

208   M. Walton without resorting to a ‘taste’ for discrimination. The essence of the argument is as follows: different groups can work in different occupations, some of which are higher productivity than others. There exists a social custom that allocates individuals from castes to occupations, and there are sanctions that impose costs (in his model through being ‘out caste’) for both employers and workers that strike contracts outside this customary allocation. If these sanctions are strong enough, there will again exist an equilibrium in which there is no reallocation of jobs across castes, as a rational response to the social structure and associated customs. And this is in spite of the fact that there would be efficiency and output gains from reallocation. If a sufficient number of employers and workers can get together to start up a production system that ignores the social custom, then they can move to the less equal, more efficient equilibrium. Akerlof derives the required size of such a breaking coalition, in terms of the other parameters of the model (differential productivities between occupations, sanction costs, etc.) However, even if such a breaking coalition potentially exists, and is consistent with underlying preferences, it will only occur if the actors can solve their collective action problems. As a third variant, there may be induced effects on the behaviour of the group with an ‘inferior’ identity. Loury (2002) proposes several conditions under which this can occur for blacks in the United States. If there are lower returns to effort, owing to lower probability of getting a contract, this is likely to induce lower effort – potentially leading to a self-­fulfilling belief system. For example, if returns to education are lower, because of beliefs over future discrimination in work, incentives to send children to school – and to exert effort in studying – will also be reduced. This can be (and is) introduced into models such as Basu’s in which there are differential outcomes in markets. This has a more profound impact on the self-­enforcing nature of the equilibrium, since it implies that the potential productivity of a person from an ‘inferior’ group at a point in time is indeed lower than others. It is again intuitive that this is likely to be inefficient: if, say, dalits expected to have access to better-­paid, and higher productivity, occupations, they would have stronger incentives to keep their kids in school, leading to more skills formation and higher productivity. But this could take at least a generation, and potentially longer, given the significant intergenerational transmission of incomes and education even in relatively equitable developed European countries.14 From information to culture and power. An information-­based approach is of great value in using (or stretching) the tools of economics to aid in understanding how group-­based identities sustain an inequality trap. In some cases, such a trap is also likely to be more inefficient than an alternative, less-­unequal, equilibrium. However, this is incomplete as an account of cultural processes. It takes group-­based identities as given, rather than products of underlying social histories of difference. We suggested above that aspects of culture should be treated as reflecting and sustaining power differences. This takes us to approaches that treat the subordinate, socially stigmatized, position of a group as essential to the cultural pro­

Culture and poverty   209 cesses at work. Within this it is useful to distinguish two approaches, depending on whether the practices and beliefs of the subordinate group internalize the difference. In some cases, subordinate groups have considerable awareness of the injustice of the social hierarchy, yet the repertoire of behaviours does not directly challenge this owing to the awareness of the costs of so doing. Scott (1985), in his interpretative study of a Malaysian peasant community, argues that there is extensive awareness of the injustice of existing power structures, often drawing on an older normative framework of moral economy, but the avoidance of direct challenges is because of the perceived futility of doing so. There is rather a repertoire of practices of daily resistance, interpretations and irony in internal discourse in what he calls the ‘weapons of the weak’. He argues against a Gramscian view of a hegemonic culture imposed on the poor by the dominant group. We term this ‘rational submission’, since there is an underlying calculus around the costs of confronting the social order. Patterns of behaviour and notional ‘preferences’ of the poor may be rational within the system since alternative, more aspirational, preferences lead to lower welfare or violence – at least within the existing system. The existence of caste-­based violence and the need to pass the 1989 Scheduled Caste and Scheduled Tribe (Prevention of Atrocities) Act, is at least indicative of the costs of lower castes crossing social boundaries even in contemporary India. Yet there is also a range of work in anthropology and sociology supporting the view that social and cultural practices do get created by histories of social differences, and become part of the perpetuation of such differences in inter-­ group relations. This lies at the core of Bourdieu’s work, and his concepts of habitus and cultural capital referred to above (Bourdieu 1990). In Loury’s account of racial inequality in the United States, he argues that racial ‘stigma’ is central to the interpretation of the long-­term persistence of racial differences, and that this works both through the social identity of African-­Americans created by a society, and the associated creation of practices within significant (generally poorer) parts of the African-­American community. He sees the group as suffering from a ‘spoiled identity’, a product of a historical ‘racial dishonor’, or ‘an entrenched if inchoate presumption of inferiority, of moral inadequacy, of unfitness for intimacy, of intellectual incapacity, harbored by observing agents when they regard the race-­marked subjects’ (Loury 2002:  70). Urban ghettos exemplify the working of racial stigma. ‘These black ghetto dwellers are a people apart, ridiculed for their cultural styles, isolated socially, experiencing an internalized sense of despair, with limited access to communal networks of mutual assistance’ (ibid., p. 77). In an Indian context, these internalized consequences of social and cultural structures are important elements of Appadurai’s concepts of the ‘terms of recognition’ and an unequally distributed ‘capacity to aspire’ (Appadurai 2004). But we turn here to a different kind of evidence, from experimental game theory. In two sets of field experiments in Uttar Pradesh, a poor part of India where caste is clearly salient, Hoff and various associates have explored manifestations of social identities in behaviours.

210   M. Walton The first set of experiments involved a game undertaken with children of dalit and high-­caste background. It involved solving mazes, with payment for either for the number of mazes solved, in a piece rate treatment, or for the child who did the best in six mazes, in a tournament case (see Figure 10.3; Hoff and Pandey 2005). The children did not know each other. Under some treatments the game was run with no announcement of caste, while in others the caste and village of the children was publicly announced. In the first case – with no announcement of caste – there were no differences in performance between dalit and upper-­caste children. However, when caste was made salient, the scores of dalit kids dropped significantly below those of upper-­caste kids – when they were together (for the piece rate treatment) and when segregated (for the tournament treatment). We don’t know the underlying cause – whether this is due to a form of stereotype threat, or to a conscious or unconscious fear of punishment. But the consequences are that of underperformance for dalit children. While this is an experimental setting, this may indicate the consequences of caste being salient in real-­life contexts involving performances that affect economic outcomes.

Average number of mazes solved, by caste, in five experimental treatments 8

Piece rate

Tournament

6

4

2

High caste Low caste

0

Caste not announced

Caste announced

Caste Caste not announced announced

Caste announced and segregated

Figure 10.3 Results on children’s performance when caste is announced (source: Hoff and Pandey (2005)). Note A vertical line indicates statistically significant differences.

Culture and poverty   211 A more recent set of experiments explores the potential impact of caste on behaviours in a contractual setting (Hoff et al. 2008). This is based on the view that social norms are a necessary ingredient of economic cooperation both in informal and formal settings – in the latter they complement formal legal structures. An important example of a norm is the propensity to punish behaviour considered uncooperative in an economic transaction. The hypothesis is that a history of denial of basic rights to a social group could lead to lower payoffs to economic cooperation and consequently reduced gains from cultivating such cooperation-­supporting norms. This could lead to a weakened ability of the social group to enforce contracts and so lower economic performance. This is explored in a sequential exchange game involving three individuals. A first actor has an initial quantity of money, that he can keep or give to a second actor. If sent to the second actor it triples in value, simulating the benefits of economic exchange. The second actor can either keep the proceeds or share them with the first actor – sharing simulates cooperative behaviour in an economic transaction. The focus of the experiment is on the behaviour of a third party, who observes the outcome of the game, and can punish opportunistic behaviour, at a personal cost. The experiment was run with individuals from both dalit (Chamar and Pasi) and upper-­caste (Brahmin and Thakur) backgrounds, from different villages. Names are identified, effectively revealing caste status. The experiment is run with various combinations of the three actors with respect to caste. The key result is shown in Figure 10.4, that shows the probability of punishing opportunistic behaviour as the difference that the probability that a third party punishes a defectors and the probability he or she punishes a cooperator. Higher-­caste third party actors are substantially more likely to punish non-­cooperative behaviour than low caste ones – irrespective of whether the observed transaction is between two members of the same caste as the observer, or is between individuals from higher and lower caste groups.15 There is also greater common knowledge of the High caste punisher

Low caste punisher

6

Average pd-pc

5

4.8 4.0

4 3

2.6 2.1

2 1 0 HHH

HLH

LLL

LHL

Figure 10.4 Experimental results on the influence of caste on punishment norms for opportunistic behaviour (source: Hoff et al. (2008)).

212   M. Walton strength of the norm amongst upper-­caste groups, in terms of reported anticipation of punishment. This provides experimental evidence of differences in norms that, if also applied in real economic transactions, would relatively inhibit economic interactions of the deprived group. This is interpreted to be a product of social histories, with caste differences the contemporary manifestations of such histories. In this section we’ve looked at various mechanisms linked to culture that could lead to the persistence of inequality traps. Some of these can be interpreted as the consequences of behaviour with limited information, in which individuals make use of their beliefs on group characteristics (or stereotypes) to make decisions. Others more directly reflect current manifestations of unequal relations and stigma, or their consequences in internalized preferences and social behaviours. Both draw on histories of unequal power and social difference, that underlie contemporary structures of social stratification. In practice resultant behaviours of subordinate groups are likely to be a mixture of internalized beliefs, cultures of interaction and rational responses to diminished economic prospects – or retributions. The structure of interactions can potentially support an inequality trap – an equilibrium that is self-­enforcing, with cultural factors interacting with economic and political ones. Note that one consequence is a set of behaviours of deprived groups that diminish their economic advance. In ‘freeze frame’, at a moment in time, this looks like a culture of poverty in the way that Oscar Lewis originally explored the concept. But the behaviours and ‘preferences’ are products of histories of inequality. This makes an important difference for diagnostic, normative and policy purposes. Dynamics and transitions So far we have deliberately structured the questions in terms of long-­term equilibria – that is an essential feature of an inequality trap in a formal account. But we are keenly interested in change, and in particular the determinants of transitions between equilibria. When we come to policy, this is the issue, and for this it is important to have an analytical basis of how change occurs. We highlight two issues here: the potential drivers of change within a system; and the plasticity of identity itself, in terms of interactions with economic conditions and political structures. How a culturally shaped inequality trap may change. The essence of the concept of an inequality trap is that the system is in a self-­enforcing equilibrium, in which cultural influences interact with economic and political structures to create self-­fulfilling mechanisms. There is another feasible equilibrium – otherwise it cannot coherently be described as a trap. But something new has to occur for the system to be pushed out of its initial equilibrium. There are a number of possible drivers of change. First, there can be changes in information. This is particularly relevant to the mechanisms in which information and coordination was important. If there were better information on the potential for people from dalit families, then employers would not have to place

Culture and poverty   213 so much (explicit or implicit) reliance on stereotypes. Suppose there was a cost-­ effective psychometric test, that did a great job of predicting the trainability and future performance of individuals. This could lead to quite different screening mechanisms. Second, the system itself could evolve in ways that changed the structure or parameters that sustained the initial equilibrium. Take Akerlof ’s (1976) model. He derives a formula for the size of the ‘equilibrium-­breaking coalition’ that would be sufficiently large that the gains from hiring workers into higher productivity activities outweighed the costs of being ‘out caste’. With economic change the structure of interests and payoffs could change, providing the conditions for such a coalition to exist. Another example is when the extension of education increases the capacity of subordinate groups to overthrow elites, whether through greater consciousness, organizing capacity or bargaining power. This may lead to interesting choices for elites, who both gain from the higher productivity of a more educated workforce and face greater risks of loss of power (Bourguignon and Verdier 2000). The general point is that the parameters that shape decisions can change as an economy evolves. Third, there may be potential for agency within the system, irrespective of underlying changes in economic and social structure. This could flow from processes of social mobilization that lead to cultural shifts, solving collective action problems, and consequently changing the costs of actions faced by elite groups. Or there may be different elite factions, with some, out of long-­term interest or ideology, favouring actions that seek to break inequality traps. The constitutional and legislative policy of India after independence is a case in point – though this was also influenced by the increased electoral salience of lower-­caste groups with the rise of the Congress Party and democratization. The plasticity of identity. In introducing our concept of culture above, we emphasized that identity is dynamic and contested. There are potentially two-­ way interactions between cultural structures and both economic and political factors. We provide illustrations of both. Interactions between identity and economics. The discussion to date has mainly been concerned with assessing the influence of culture on economic behaviour. But there is also a tradition of work that emphasizes the influence of economic structure on culture. Interactions are likely to be two-­way. Both are illustrated in a study of the interactions between caste, economic opportunity and schooling choice in Mumbai (previously Bombay), India’s major economic city (Munshi and Rosenzweig 2006). The context for this is the historical role of specific caste categories (at the level of jatis) in channelling working-­class boys and men into specific occupational categories in this industrial city. This is interpreted as providing network externalities to job-­seekers. The exogenous ‘shock’ to the system is the substantial rise in the premium to English education that occurred in Mumbai after the growth acceleration and steady economic liberalization since the 1980s. The authors examine the pattern of allocations of boys and girls to education in the local language (Marathi) and English schooling. Their key result is the contrasting experience across genders. Boys’ schooling

214   M. Walton amongst children of parents in working-­class jatis displays considerable inertia in the allocation to Marathi education – a sign of the continued channelling of boys into traditional occupations that do not require English. This is likely to both sustain inequality of opportunity for these groups and be inefficient, given the shifting patterns of demand. By contrast, girls were not traditionally allocated to specific occupations, since labour force participation was low. They display a greater responsiveness to changing economic opportunities, with significant rises in attendance at English-­speaking schools. This study does not track the children into jobs, but is prima facie evidence of both of identity-­slowing adaptation for boys, and of adaptability in response to economic forces for girls. Interactions between identity and politics. There is also evidence of influences in both directions between culture and politics – of particular interest because of links to policy formation. The idea that caste is a pliable concept that interacts with political processes is an old one: for example, in the 1960s, Rudolph and Rudolph (1965) argued that caste associations were playing a distinctly modern role in organizing for democratic engagement in India. We first look at political influences on caste, and then on how caste can interact with political parties. Empirical evidence for the plasticity of caste identity, and of probable political influences, is provided by Rao and Ban (2007): they exploit a natural experiment that occurred with the 1956 reorganization of Indian states. The intention of this reorganization was to get greater social uniformity within states, using dominant linguistic patterns as a proxy for similarity. But linguistic boundaries were not sharp, or ‘mistakes’ may have been made, with the result that similar villages were allocated to different states. Rao and Ban undertake a comparison of villages along the boundaries between Andhra Pradesh, Karnataka, Kerala and Tamil Nadu, matching villages by linguistic similarity. Additional evidence on landlessness – that is closely associated with caste – corroborates the similarity in social structure. The central result is that caste structures appear to have become more dissimilar between villages that were allocated to different states. They interpret this as evidence of the political construction of caste, owing to the varying political histories and policies across the four states. These statistical results are consistent with a tradition of historical interpretive analysis of the influence of political processes on the structure and salience of caste. It is an important corrective to the view of caste as an especially durable institution of cultural structure (with a vintage in the thousands of years!). Two-­way interactions are also relevant to more directly political analyses. Mehta (2003) argues for the centrality of inequality in political and social life in India, and the rising salience of caste as a political category because of the weakness of other mechanisms for mobilizing groups and aggregating interests. In similar vein, Chandra (2004) argues that caste has greater political salience in a ‘patronage-­democracy’, in which political support is maintained primarily through delivery of private goods and services (jobs, subsidies, public works targeted at particular groups) rather than commitments to programmatic policies

Culture and poverty   215 and public goods. Under these circumstances, a potential ethnic or caste-­based grouping can serve as a vote-­bank, with a mutual interest between political elites in offering patronage and members of the ethnic group in offering votes. This can lead to increased salience of the group identity that is mobilized for political purposes, in relation to alternative identities. It is consistent with a finding of Rao and Ban who find rising identification of individuals with more aggregate caste categories in some villages in their study, that they interpret as an indicator of rising political mobilization. A particular example of the political salience of caste is the rise of the Bahujan Samaj Party (BSP), that emerged as an explicitly dalit-­based party, though it has over time developed electoral alliances with other social groups (Chandra 2004). It is currently in power in Uttar Pradesh, India’s largest state. By contrast adivasi groups have been less successful in terms of independent political organization: Banerjee and Somanathan (2007) find that political constituencies with adivasi concentrations continue to display a bias towards voting for the Congress Party – that in the post-­independence period sought to offer a place for all social groups. By contrast dalits have broken with such a bias, consistent with the development of alternative forms of independent political mobilization. This section has argued that historically shaped inequalities can show substantial persistence, and that cultural processes can be an important contributory factor, even after legal, or other formal, structures of differentiation have been removed under the banner of equality of opportunity. An inequality trap can usefully be analysed as a self-­enforcing equilibrium in which there are interactions between cultural, economic and political forces. Of equal interest is understanding shifts to different equilibria, where dynamic interactions between these factors can be at play. Culture – for example in the form of social identity – is intrinsically malleable, but just how change occurs, and whether it is initiated from economic structure, political forces, or transformations of a more directly cultural kind, such as through social mobilization, will be contingent on internal and external forces, and the potential for the exercise of agency within the system. Links to growth The core unit of analysis so far has been the group – or more precisely a set of groups interacting in a social, political context. The analytical frameworks we have used from both economics and anthropology function in this realm. Jumping to the level of a nation or country looks like a leap of faith. It is nevertheless of interest here for a couple of reasons. Culture of poverty advocates have gleefully applied their perspective to the question of the wealth of nations. And if anything useful can be said on links with aggregate growth, then this is of substantial importance, since in a narrow numerical sense, the effect on material poverty of inter-­country differences in average income levels and long-­term growth rates is immense.

216   M. Walton Culture of poverty at the national level: argument and critique The extension of culture of poverty views to nations has the following thesis: some nations have cultures that are conducive to the drivers of economic growth, such as thrift, entrepreneurial dynamism, a high value of education, or innovation. They grow fast. Others don’t. They grow slowly. Max Weber in his The Protestant Ethic and the Spirit of Capitalism (1930) is often cited as an early proponent of this perspective, with the view that Calvinism was good for capitalism and therefore for the pursuit of long-­term growth.16 This view has proved surprisingly popular over time, re-­emerging in the context of debates over the East Asian Miracle, with some arguing that success was a consequence of ‘Asian Values’ of thrift, family support and so on. In similar spirit, Deepak Lal argued that rigid social structures generated a low ‘Hindu rate of growth’ of about 3 per cent per annum (Lal 1989). A more recent version of this view is evident in many of the papers in the Harrison and Huntington volume. This is vividly illustrated by an opening comparison between South Korea and Ghana by Huntington. He considers how these two countries could start with similar levels of income in the 1950s and then experience such extraordinarily divergent paths, and suggests that an important part of the explanation must lie in cultural differences: ‘South Koreans valued thrift, investment, hard work, education, organization, and discipline. Ghanaians had different values. In short, cultures count’ (Harrison and Huntington 2000: xiii). Unfortunately for the proponents, these arguments are refuted by even a broad brush look at history and growth patterns. As Sen (2004) argues, cultural determinists had to regularly update which were the privileged, pro-­growth cultures as new countries took off: from Protestant culture to Christian culture, when Catholic countries of Southern Europe took off; to Japanese culture when the Japanese miracle started; to the benefits of first Confucian, and then more broadly ‘Asian’, values as the East Asian Miracle spread to Buddhist Thailand and predominantly Islamic Malaysia and Indonesia. And with India, previously condemned to a Hindu equilibrium, the post-­1980s growth takeoff – associated in more recent years with a large rise in savings – can hardly be explained by an independent shift in culture! On culture, inequality traps and growth Does the apparent empirical vacuity of a culture of poverty diagnosis of slow growth imply abandoning any role for cultural influences on growth processes? While somewhat more speculative, we suggest it is useful to explore interactions with specific growth-­related processes through the prism of inequality traps. Lost growth of a subordinate group. Some of the theory and evidence presented above supports the view that subordinate groups, embedded within inequality traps, will have lower rates of accumulation of physical and human capital, slower reallocations to more productive activities, weaker pro-­social norms or lower induced efforts. This is in principle a lost source of aggregate

Culture and poverty   217 growth. There is microeconomic evidence of the importance of these processes, though we do not have evidence of how important this is in the aggregate.17 Inefficiencies of unequal institutional structures. A second strand of work brings together the recent emphasis on the role of historically shaped institutions on growth, with the influence of unequal structures on institutional design (Acemoglu et al. 2001; Engerman and Sokoloff 2002; World Bank 2005). In one story, predatory, rent-­seeking institutions may evolve that serve the interests of dominant groups, but do a poor job of providing protection of property rights for all, provisioning of public services in social and economic infrastructure that can underpin equality of opportunity, or developing broad financial institutions. There are intersections between this kind of story and group-­based differences to the extent that the latter become connected with or magnify inefficient economic institutions or processes. One example is Chandra’s analysis of the interaction between a patronage-­democracy and group identity as a vote-­bank (Chandra 2004). There is the potential for both increased salience of a group identity (as an example of the political construction of caste, say) and the workings of patronage-­based rent-­seeking, diverting policy action away from provision of the public goods that are necessary for growth. Inter-­group conflict and bad equilibria. A different set of mechanisms involves distributive conflicts, that frequently coalesce on group-­based lines. One interpretation of Bihar’s long-­term malaise is of a situation in which group-­ based patronage interacts with weak provision of public goods – from decent infrastructure to law and order – to support a low growth, high inequality, high corruption equilibrium. Bihar may now be experiencing a transition to a new political equilibrium in which a broader coalition is associated with an electoral mandate to deliver public goods, with the election of the government headed by Chief Minister Nitish Kumar in 2005. In more extreme cases, failures of inclusion and state action have provided the conditions in which the, ostensibly Maoist, Naxalite group has flourished. In some areas adivasi groups are caught in the middle of an inadequate state and such groups, with rising violence on both sides (Guha 2007). Transitions. Cultural processes may be relevant to transitions from inequality traps at a society-­wide level. If we treat Indian states as analogous to national economies, one interpretative strand of explanation behind the more rapid growth of southern than northern states is that the south experienced a much earlier transition from a deeply unequal caste and class-­based social structure. An element of this was action related to caste-­based social mobilization, that led to the confrontation of unequal cultural structures and a re-­definition of the role of group-­related identities in economic life. Kerala is an illustration. It historically suffered some of the more extreme restrictions on social and economic living in India. But it then exemplifies the interaction between the exercise of agency from below, through caste and class-­based processes of mobilization, and from above, through the enactment of pro-­equity laws and policies by elected governments (Heller 1999). Especially in early phases of the process, this involved action to support cultural transformations, including use of theatre and

218   M. Walton public performance as means of confronting and changing unequal socio-­cultural structures and attitudes related to caste and class. The presence of an organized, and relatively inclusive, communist party was an important force in providing a political party base for social movements, but of equal importance was the regular alternation in power between this party and the Congress Party. There is controversy over effects on long-­term growth. A common view is that the vigour of public action, the power of unions and an excessive regulatory burden had a dampening influence on private investment. However, in Heller’s diagnosis, the deeper changes involved the formation of institutional structures that, over the longer term, will be more effective mechanisms for the management of distributive conflicts within an essentially capitalist system. Normative analysis The discussion so far has been essentially positive – an analytical survey of how culture influences how the world works. But normative considerations have been frequently lurking around the corner, and are rarely far from the centre in discussions of culture and development. This section discusses them explicitly. As in the discussion above, the question of culture enters in two conceptually different ways: first, in terms of the normative treatment of pre-­specified groups with different circumstances; and second, in terms of what cultural processes imply for normative assessment. Normative approaches A natural place to start is with approaches in moral philosophy and welfare economics that are concerned with equality of opportunity.18 Consider a society that is organized into different, culturally defined groups – by caste category, adivasi, religion and so on. Most contemporary societies formally take the position that there should be equality of opportunity across such groups, and especially so when most people don’t choose which group they are in. Which group you are in is an aspect of circumstances that is out of an individual’s control, and is morally irrelevant to outcomes. This was certainly at the core of debates at Indian independence, and was incorporated into India’s legal structures. This approach is within traditions of thought that flow from Rawls, Dworkin and Sen in moral philosophy, and Roemer in welfare economics. These have in common a focus on the types of lives that individuals can choose to follow – that is, on possibilities. Sen, for example, argues for equality of ‘capabilities’, the potential set of ‘functionings’ people can have, where a ‘functioning’ is an activity that people value and do, ranging from quite basic ones such as living a healthy life, to more complex ones, such a participating in the community (see Sen 1992). Individuals may choose different effort levels or pursue distinct activities, but they should not face different capability sets – a girl born in a rural dalit family should have the same life-­chances as a boy born in an urban high-­ caste Hindu family.

Culture and poverty   219 A formal account of what an equal opportunity approach implies for development objectives (that broadly follows Roemer 1998) is given in the Appendix. Which group a person is in should have no effect on their possibilities, whereas their effort should. But effort may also be a product of their group’s position, especially when a group has been stigmatized, or more broadly when structures of inequalities have induced different preferences, for example over expectations on future economic possibilities. Thus equal opportunity is defined to occur when individuals from different groups, who exert the same relative effort within their group, enjoy the same advantage level. Thus high-­effort dalits or adivasis would enjoy the same advantage levels as high-­effort upper-­caste Hindus; low-­ effort individuals within each type would also have the same outcomes as each other; and so on. An alternative, much less stringent (and less interesting), condition is that average levels of advantage are equalized across groups. As seen in Table 10.4, this is clearly not satisfied in India. As already emphasized, societies are interested in levels as well as distributions, and for this we need to take the equal opportunity principle to an objective that can be maximized. One way of doing this, in the spirit of Rawls’ maximin criterion, is to define the objective in terms of policy choices that seek to maximize the condition of the most deprived group in a society. This is one approach to an ‘equitable development policy’ and a precise definition is given in the Appendix. How would this objective compare with alternatives? This can be illuminated by Figure 10.5. This portrays an ‘advantage possibility frontier’ that defines for two groups the range of potential advantages given the production technology, if resources are deployed efficiently. This assumes some complementarity at very low levels of advantage of each group, but a trade-­off at P

O2

B

X

R

A

E

C

O1

Figure 10.5  Different choices along an ‘Advantage Possibility Frontier’. Note Figure 5 is a reproduction of Box Figure 4.1 in World Bank (2005) and, as acknowledged there, is drawn from Buchanan (1976), through Atkinson and Stiglitz (1980).

220   M. Walton higher levels of advantage. The equal opportunity outcome is given by E, where both groups achieve the same level of advantage. What was just termed the equitable development policy is achieved at R, where the relatively disadvantaged Group 1 maximizes their advantage. The highest total advantage is reached at B – this is analogous to an output-­maximizing position. Note that a move from R to B is not Pareto-­improving, since Group 1 loses. But we can think of B as a position of Kaldor–Hicks efficiency, since the Kaldor–Hicks criterion differs from the Pareto criterion exactly in that it requires only that it be possible for a social planner to compensate any losers in the move from A to B, rather than that there be no losers. Consider finally a position such as X, that is within the advantage possibility frontier and also displays an unequal distribution of advantage in favour of Group 2. From here there is the possibility of using policy to shift to positions that are both more equitable and efficient (though not necessarily Pareto-­ improving, in the case of a move to, say, R, since Group 2, a beneficiary from structures prevailing at X, loses out.) This is why an inequality trap in a position such as X has particularly severe costs for a society. In the survey of positive analyses, a distinction was drawn between approaches that introduce culture through taking culturally defined groups as given, and those that seek to integrate a richer account of dynamic cultural pro­ cesses. The same contrast is appropriate here. The normative account presented so far here is based on existing groups that are salient to social and political concerns within a society. This is again useful. But it fails to take account of cultural process. This requires taking additional steps. A good place to start is the endogeneity of effort. In a Roemer-­based framework this is taken account of via the equal opportunity principle described above. This implies the same outcomes for individuals in the same percentile of the effort distribution of their respective type, even if average effort levels are different across groups. Yet there is something unsatisfactory about this. It fails to deal with the origins of differences in average effort levels in terms of the intertwined effect of power and cultural processes. We saw above that these can reflect a variety of processes, flowing from historical inequalities, stigma, adaptive responses and privately rational behaviour in an environment of limited information. Failing to account for the cultural origins of effort differences (in this example), can lead to policy proposals that indeed compensate for differences in circumstances, but with potentially large policy actions – in spending, regulation and so on – that have little to do with the origins of the underlying differences. ‘Effort’ is an abstraction from this simplified account, which can be thought of as covering the complex array of practices, beliefs, aspirations and behaviours shaped by unequal cultural structures. For these reasons in Rao and Walton (2004b), we suggested a distinct normative concept of equality of agency. By this we mean that actors from different groups suffer from unequal, culturally shaped capacities to conceive of and pursue alternative designs for their futures. This mirrors the earlier positive analysis, and specifically Appadurai’s two concepts, in that it would imply:

Culture and poverty   221 1 a societal context that did not stigmatize or diminish particular social identities ­– that is in which there existed equal ‘terms of recognition’ across groups; and 2 an equally distributed ‘capacity to aspire’ across groups. We can think of this as an extension of an equal opportunity approach, which integrates the cultural capacities into the capability set of different groups. Groups, again, are not given, but are themselves products of social interactions, or even purposive ‘group-­making projects’. So it is more accurate to think of (in) equality of agency as being distributed across individuals, but this being shaped by the culturally shaped manifestations of differences in power and status, that manifest themselves in shifting patterns of inter-­group relations.19 Implications for evaluation What does this imply for the evaluation of culturally shaped inequality traps? The general approach illustrated by Figure 10.5 suggests how we can assess alternative situations, or, in more formal language, rank alternative equilibria. One way to do this would be through strict application of what we termed above ‘the equitable development’ criterion, that would lead to a ranking of alternatives in terms of the discounted value of the mean level of advantage of the (cultural) group with the lowest level of advantage. In an Indian context, this would imply choice of a development path that led to the greatest present value of average welfare of, say, dalit or adivasi groups. In terms of practical choices this could well involve departures from pure equality of opportunity across groups, if this allowed greater levels of potential advantage for the poorest group. This is precisely the difference between point E and point R in Figure 10.5. Alternatively there may be a political consensus for departing from the equitable development policy, trading off advances for the poorest group for the sake of greater aggregate development, moving to a point such as B. This approach would in principle allow an assessment of how far an inequality trap is inferior, through the effects on advantage of poorest groups. A richer evaluation would then require at least three additional considerations. First, as noted above, there will be a distribution of effort levels within the poorest group. We may not only be interested in the average levels of advantage, but in the distribution of advantage within the group. Given the concern with poverty, an obvious preoccupation would be with potential outcomes for the lower parts of the within-­group distribution. But it is quite plausible that there would also be concern that the very best of the group could excel to the same extent as any group in the society.20 Second, this concern becomes more important once we recall that ‘effort’ is itself at least partially endogenous, a cultural product of historically shaped structures of expectation, opportunity and advantage. It may well be easier to design policies that lift a relatively small group of dalits or adivasis into the social and economic realms of the advantaged groups – in the sense that for

222   M. Walton these groups the disadvantages flowing from information, social and cultural capital were effectively removed. But different approaches – that could be more or less costly – could be needed to maximize the potential outcomes for, say, the bottom half or quarter of those within dalit or adivasi groups. Third, the question of who should take action has been left within the abstraction of the policy ‘choice’. Here there is a sharp contrast between a culture of poverty and an inequality trap view. If a group of the poor are poor because of ‘their’ culture, in some sense the fault lies with ‘them’. Now, a government may choose to act for a number of motives: paternalism; a desire for social and national integration; or concern with negative externalities of such group-­based differences. But this is quite distinct from the moral responsibility for action when the reasons for contemporary deprivation lies in a historical process based on unequal socio-­cultural and political structures. There remains the political economy question of whether policies will be adopted, but the moral basis for action and policy design differ. Furthermore, public support for concerted action can be strengthened by recognition of the injustice of social arrangements.21 Both the last two points are linked to the previous subsection, that argued that cultural processes matter for normative purposes, in addition to potential outcomes for different culturally defined groups. Achieving equality of opportunity is likely to require equality of agency, and this is likely to involve forms of engagement, mobilization, or rights for all dalits or adivasis. Policy implications So far the chapter has argued that culturally shaped inequality traps are relevant in many (most?) socially differentiated societies and that this can provide a normative basis for shifting from a higher to lower inequality position. This case is especially strong where there are also efficiency costs of a high inequality trap. It also suggested that transitions between equilibria may be feasible, but that it is important to understand the potential drivers of change, whether these come from exogenous sources, endogenous changes in underlying economic structures, or from the agency of groups within the social system. This takes us to questions of policy. There is not the space in this chapter to undertake an evaluation of specific policies. We rather conclude with a survey of categories of policies that flow from the preceding analysis. While we focus on culturally-­shaped inequality traps, these are understood as interacting with both political and economic structures. And as in both the positive and normative discussion, it is useful to distinguish between policies, of all kinds, that seek to improve the position of pre-­existing, culturally salient groups, and policies that intrinsically involve cultural processes. Interventions can be organized along two dimensions. First, they can be categorized by their proximate objectives: whether they seek to directly redress a group-­based difference in opportunities or outcomes; or whether they seek to enhance the agency of a deprived, subordinate group (that is, to directly tackle inequality of agency). Second, they can be organized in terms of the approach

Culture and poverty   223 Table 10.5  A categorization of policies to tackle culturally shaped inequality traps Policies intended to redress Policies intended to group-based opportunities increase the agency of and outcomes deprived groups Preferential policies conditional on group membership

Reservations in jobs and schooling ‘Tribal’ policies

Political reservations for specific groups Targeted political mobilizations from above

‘Universal targeted’ policies, aimed at specific conditions of groups but with no group membership condition

Access to land, education, credit, communal resource management etc. in areas and activities important to dalits and adivasis

Universal suffrage. Increased information and accountability

Actions influencing cultural Education with attention to processes identity, language of deprived groups Land policy adapted to adivasi practices

Social mobilization from below, deliberation and political culture

Source: author.

used, for which we suggest three categories: preferential policies that are conditional on being a group member; ‘targeted universal’ policies, that are in principle universal, but oriented to the particular circumstances and needs of the subordinate group at point in time; and policies that directly engage with cultural process. Table 10.5 summarizes, with examples of the kinds of policies implied. Proactive policy to tackle group-­based difference has been a notable feature of post-­independence Indian policy. While the post-­independence political elite was dominated by individuals from the socio-­cultural elite, there was a strong impulse to take redistributive action in support of equality of opportunity. This flowed from the ideology and conviction of the leadership, the nature of the independence struggle, and political calculations around political support. Intellectual debate and action by individual actors also mattered: both Mahatma Gandhi and B.R. Ambedkar played major roles in building the case for action to tackle the deprivations of dalits, albeit with very different policy positions. Ambedkar, himself a dalit, led the drafting of India’s constitution. In terms of the discussion of transitions above, all this implied a degree of agency from above. In some parts of the country, notably Kerala, this was complemented by effective pressure from below. But this occurred in the context of embedded and unequal socio-­cultural and economic structures. In terms of action it is striking that Indian governments chose to work on several fronts – indeed across virtually all the categories in Table 10.4. First, there was explicit action in the political realm. In addition to the important universalist work of the election commissions, political reservations were introduced to assure representation of scheduled castes and tribes: through reserving

224   M. Walton some constituencies for political representatives from these groups in line with their population shares. Second, some of the broad-­based policies in social provisioning, in basic education for example, aimed to bring all groups into the system, including deprived categories, notably dalits and adivasis. Redistributive land policies could potentially have helped the position of poorer groups (though in practice they tended to benefit middle farmers). Third, to tackle potential discrimination in higher education and jobs, reservations, linked to population shares, were introduced in both tertiary education and government (though not private) jobs for both scheduled castes and tribes.24 Results have been mixed. The broad brush outcomes summarized in Tables 10.1–10.4 indicate some absolute progress but continuation of relatively worse outcomes for dalits and adivasis some 60 years after independence, with limited signs of convergence for dalit groups but not for adivasis. In terms of policy impacts of political reservations, Pande (2003) found a significant influence of scheduled tribe reservation on spending directly targeted to tribal policies, and of scheduled caste reservation on job quotas. But there was no significant influence on the general quality of service delivery. Banerjee and Somanathan (2007) also found both absolute and relative gains in access to a set of public goods for dalits, but not for adivasi groups in the 1980s. They hypothesize that this was linked to greater political mobilization of dalit groups. However, this occurred in the context of slow progress in public service delivery by international standards. In education, for example, the PROBE report (1998) found dismal quality of public education in the mid-­1990s, and continued salience of caste and gender as sources of educational deprivation. Since then there has been important progress in enrolments, but low quality continues to be a major concern – and there is some evidence that access to quality schooling is worse for dalit and adivasi groups. Policies on job and educational reservations remain controversial. In theory, effects are ambiguous. If they help solve information problems (as in Basu’s model) they can help push a system to a better equilibrium on both efficiency and equity grounds. But if there are adverse effects on job-­matching or effort, efficiency can be hurt. There appears to have been a significant impact on the shares of individuals from deprived groups in these organizations. A recent study from Bertrand et al. (2008) for engineering colleges in an Indian state, found that beneficiaries of reservations amongst scheduled castes and other backward castes indeed came from less financially advantaged households than those displaced. However, they came from households with, on average, about three times the average incomes of their respective groups in the overall population, and much higher levels of parental education. Despite much lower entrance exam scores, the marginal lower-­caste entrant does benefit: the study finds a strong, positive economic return to the engineering education. Lower-­caste graduates may still experience relative difficulties in finding jobs, according to a study (of a different sample) by Deshpande and Newman (2007), but there are real economic gains. This work supports the view that this type of reservation brought gains to the individuals concerned – and potentially broader symbolic benefits. But given

Culture and poverty   225 the relatively small numbers in tertiary education and public sector jobs, they only reach the top segment of the distributions of dalits and adivasi individuals. One of the major issues facing India is whether the explicit link of group-­ membership to benefits has interacted with the continuation of a patronage democracy to foster the rent-­seeking for state-­based privileges rather than genuine equality of opportunity. This is of particular concern given the contrast between reservation-­based mechanisms that favour better-­off individuals within traditionally deprived groups with continued dismal conditions, and slow progress, on generalized service delivery, including what in Table 10.4 was referred to as ‘targeted universal’ policies for deprived groups. Moreover, policies linked to group-­based membership can change the salience and significance of groups – and lead to incentives for particular categorizations. A topical example is the campaign of members of the Gujjiar group to be classified as a ‘scheduled tribe’ in order to receive the benefits of this classification.25 As a final comment, we would emphasize the importance of complementarity between action in the economic and political spheres, and action that works with cultural processes, that seeks to achieve more equal terms of recognition, and fosters the aspirational, organizational and symbolic capacities of subordinate groups. There is case study evidence of the importance of these processes, for example from self-­help groups and the work of the Self Employed Women’s Association. These will need to be systematically explored in future work.

Conclusion This chapter has surveyed the relationship between culture and poverty. It has argued that the concept of a ‘culture of poverty’ is empirically and normatively problematic. It is empirically flawed since it is based on a view of culture as a self-­standing attribute of poor groups. Following central perspectives in anthropology, it is argued that cultural processes are fundamentally relational, dynamic and contested. Cultural processes are indeed part of the lives of poor people, but these are products of histories of unequal, group-­based interactions. The experience of deprived groups in India vividly illustrates this. Equally problematic is the view that differences in national cultures explain differences in growth rates across countries. This is inconsistent with the historical experience, in which growth takeoffs clearly occur across countries with distinct cultures, and not in ways explicable by independent cultural change. As an alternative to a culture of poverty view, the chapter proposes the concept of ‘inequality traps’ within which cultural processes play a role, interacting with both economic and political processes. As an analytical abstraction it is useful to explore these as self-­enforcing equilibria, sustained by a range of mechanisms, from the persistence of stereotypes under conditions of limited information, to the internalization of practices, aspirations and belief systems created by histories of social difference and unequal economic opportunity. As a complement to this positive analysis, the chapter proposed a normative approach that draws both on ethical traditions based on equality of opportunity

226   M. Walton (or equality of capabilities), and recognizes cultural influences on the capacity of different groups to shape their lives. The concept of equality of agency expands the traditional approach to equality of opportunity to include culturally shaped capacities to conceive of, aspire to, organize for, and undertake projects for economic and social advancement in a society. Policy analysis needs to integrate both the positive and normative approaches. India is again an example of particular interest. India has pursued a range of policies that seek to move towards both greater equality of agency across groups and greater equality of social and economic opportunity. But these polices interact with unequal economic, cultural and political structures. Progress in reducing group-­based differences has been modest at best, at least with respect to average group outcomes. Affirmative action through reservations of jobs and tertiary education has brought clear benefits for some – especially those from better off backgrounds from within lower-­caste groups – but may have interacted with India’s patronage democracy to induce greater rent-­ seeking, rather than the creation of the conditions for genuine equality of opportunity and agency.

Appendix: formal treatment of inequality traps and equality of opportunity In this appendix we give a formal account of both inequality traps and the implications of an equality of opportunity principle, based on Bourguignon et al. (2007). As noted in the text, this applies to cases where groups are given, and not to endogenous group formation. Inequality traps We develop a simple definition in terms of the dynamics of a general notion of ‘advantage’. A society is partitioned into various ‘types’, that are socially relevant groups with different circumstances outside the control of individuals, such as place of birth, caste, ethnicity, etc.26 However, within any one type, there can be a range of efforts exerted by different individuals. Outcomes are also influenced by ‘policy’. An individual i of type j has advantage level utij = u(ctj, etij, φt) where c j represents the circumstances of group j, e ij is the effort of individual i in group j, φ represents policy, and t denotes a particular time period. For there to exist an inequality trap, two conditions must hold: first, that there has to be persistence in relative positions in a distribution across time periods; second, this persistence is (at least partly) a product of relations between groups, i.e. in this abstract account, is a function of the whole distribution. This might come about because the circumstances enjoyed by group j today (e.g. dalits in the year 2000), depends in part on the advantage levels enjoyed by the preceding generaj tion (e.g. dalits in the 1960s): ctj(u t–1 , ξt), where ξt denotes an innovation at time t; and an individual i’s effort levels today may also depend both on past advantage (their own or of a previous generation), and on the distribution of past

Culture and poverty   227 ij advantages: etij(u t–1 , Fφ t0–1,t–1(u), ζt), where Fφt–1, t–1(u) denotes the entire distribution of advantages at time t – 1, and ζt denotes another innovation at time t.27 What of policy? It is plausible that the policy in place at time t has also been determined in part by the prevailing distribution of advantages (reflecting unequal influence) at time t – 1: φ(Fφ t–1,t–1,θt), where θt denotes a third innovation at time t.28 For now, we treat policies as entirely endogenous, returning to the question of policy as a control variable in the last section of this chapter. In this formulation, there is both intertemporal transmission of circumstances, and dependence of both efforts and policies on the previous distribution of advantages. All three form elements of the transmission processes of the distribution of advantage from between periods. Thus a general reduced form of the dynamic process of advantage can be written as:

(

)

utij = Γ utij−1 , Ft −1 ( u ) , γ t

(1)

where γt is a vector of innovations. Equation (1) simply defines, at a general level, a first-­order stochastic dynamic process, in which individual (or lineage) i’s current advantage depends on her own past advantage, as well as on the entire distribution of advantages in the preceding period (or generation). Now assume that this process is characterized by multiple equilibria, in each of which the distribution of advantages converges to a well-­defined long-­run distribution F∞k(u), where k denotes a particular long-­run – or limiting – distribution of advantages, each of which corresponds to a particular equilibrium of the dynamic process. As with any dynamic process with multiple equilibria, the idea is that differences in initial conditions (in the initial distribution of advantages and circumstances, for instance) may imply convergence to different long-­run equilibrium distributions. If K is the set of equilibria for this process, a particular long-­run equilibrium k*(k* ∈ K) is an inequality trap if for any two types, j and l, the distribution of advantage for type j, F ∞kj (u), is dominated by that of type of l l, F ∞k (u), provided there exists some alternative equilibrium m(m ∈ K), in which j l no dominance relationships exist between F ∞m (u) and F ∞m (u), for any j, l.29 In plain English, this classification defines an inequality trap as a long-­run distribution of advantages in which a particular social group does persistently worse than some other social group, even though an alternative equilibrium exists where no two social groups can be similarly ranked. Equal opportunity and the equitable development policy Following Roemer (1998), a standard element in ethical approaches is to treat the distinction between circumstances – that are outside an individual’s control – and the efforts they choose to exert, as having normative significance. Efforts, but not circumstances, can lead to morally justified differences in outcomes. As with the positive discussion of inequality traps, a person’s advantage level, u, is produced by the interaction between circumstances, c, effort levels, e, and policy φ. If we then assume that policy is a control variable, contrary to the earlier

228   M. Walton positive discussion, then we can think of a society choosing policies that seeks to further the pursuit of equality of opportunity. This is consistent with the discourse and legal designs in Indian post-­independence thinking. Now take a population divided into groups that differ only according to their circumstances – such as the socio-­cultural groups that are the concern of this chapter. The possibilities an individual i in group j faces is a product of their initial circumstances and policy, with actual outcomes in advantage levels, uij depending on his or her effort – uij = u(c j, e ij, φ). For each group there will be a distribution of advantages associated with the within-­group effort distribution. Let each type be characterized by a cumulative distribution of advantage given by F φj(u), where the subscript φ indicates that this distribution may be affected by the set of policies chosen by the government. Since circumstances are morally irrelevant, a strong criterion for a situation of equal opportunity is a situation in which: F j (u) = F k (u),∀j,k

(2)

This criterion requires that the within-­type distribution of advantages be identical across all types, implying that their means and Lorenz curves also be identical. Two people in identical percentiles of the distribution of effort, but with different circumstances, would have exactly the same advantage level. Intuitively, high-­effort dalits or adivasis would enjoy the same advantage levels as high-­effort upper-­caste Hindus; low-­effort individuals within each type would also have the same outcomes as each other; and so on. A weaker criterion, which is implied by Equation (4) but does not imply it, is that the mean advantage levels across all types be equalized: µ j (u) = µk (u),∀j,k

(3)

But equality of opportunity is typically not the exclusive aim of societies. There is typically a keen interest in expanding the level of advantage, and this is often a particular concern where there are intense levels of deprivation. A natural way of formulating this – in the tradition of Rawls and Roemer – is to maximize the level of advantage of the most disadvantaged group. In Bourguignon et al. (2007) we proposed one definition of an equitable development policy as the policy φt that solves, at time t, the following: ∞

∫

max min eδ (t − s ) µ sj(ϕ ) ds ϕt ∈Φ

j

(4)

t

subject to utij ≥ ut ∀i, ∀j , ∀t

(5)

where δ is a discount rate, assumed constant for simplicity. Policies not only affect within-­period economic processes and outcomes, but also saving and accumulation incentives, which in turn affect the growth process of μ j. The equitable development policy has the following features:

Culture and poverty   229 1 2

3

it accounts explicitly for the dynamic nature of the development process, since the optimal policy is the one which maximizes the presented discounted value of advantages for the least-­advantaged type;30 provided that it is consistent with the absence of severe deprivation (defined by having no individual advantage below some critical level ūt, which is allowed to vary over time, as a society’s definition of deprivation changes31); and provided that the policy belongs to some permissible set Φ. We interpret this permissible policy set as a subset of the set of technically feasible policies, reflecting social choices about the legitimacy of various policies. For example, forced labour, forced fertility control, or expropriation of property might all be feasible policies, but might be deliberately excluded from the permissible set Φ by social choice.

This is not the only interpretation of what an equitable development policy may be – as discussed in the text, some may argue that full equality of opportunity is preferred to a maximin criterion.

Notes   1 Centre for Policy Research, Delhi and Kennedy School of Government, Harvard University, [email protected]. Arguments in this paper draw on earlier work with Vijayendra Rao, see Rao and Walton (2004b). This paper was prepared for the Fifth Annual AFD/EUDN conference in Paris on 5 December 2007. I am grateful for comments from, or discussions with, Ashwini Deshpande, Karla Hoff, Stuti Khemani, Stéphanie Mahieu, Rinku Murgai, Vijayendra Rao and Rohini Somanathan. Both views expressed and errors are mine.   2 Note that economic characterizations of poverty traps generally apply only to absolute, not relative poverty. See below.   3 See Baland and Platteau (1996) for an extensive review.   4 See Rao and Walton (2004b: 3–4) for further discussion on the general approach.   5 The classification system in the classical Hindu texts was into five hierarchical groups, including four varnas – brahmins (scholars), kshatriyas (warriors), vaishyas (merchants), and shudras (artisans and labourers) – plus the untouchables that formally fall outside the system. In the modern period there exists a much larger number of specific, endogamous groups or jatis typically defined in terms of more specific occupational, geographic and linguistic categories, with a rough allocation to the five broader categories. There are some 2,000–3,000 jatis in India For a brief account of caste see Deshpande (2001 and 2007).   6 See Somanathan (2010) for a review of the formation of policies for both scheduled castes and tribes. Reservations for scheduled castes were given under colonial rulein the 1930s.   7 Though note the metric for making comparisons of changes at different parts of the distribution is unclear – is a ten percentage point improvement in literacy larger or smaller in value from an initial value of 80 and 20 per cent? Thus these changes are only indicative.   8 See Rao and Ban (2007) for a brief discussion of the literature.   9 Such poverty studies typically combine quantitative data on the dimensions, incidence and depth of poverty, with an account of the characteristics of typical groups in poverty.

230   M. Walton 10 Evidence for this particular channel is weak (see Deaton 1997 for discussion), but the point here is how economists have generally approached theorizing of poverty traps. 11 See World Bank (2000) for a seminal account. 12 See various essays in Volume III of Rudolph and Rudolph (2008). The issues were considerably more complex than the anti-­Brahmin focus, of course. One other dimension of concern and action, also around the upper parts of the distribution, concerned the desire to promote the development of a local business class and avoid dominance by predominantly Gujarati and Marwari business groups from northern India. 13 In his treatment of race in the United States, Loury (2002) uses the distinction between racial stereotypes and racial stigma to contrast the first approach and the others. 14 Recent work by Jalan and Murgai (2007) also finds significant, but declining, intergenerational transmission of education across generations, with greater transmission (that is more persistence) for low-­caste groups, and, especially, for groups with lower wealth. 15 This result holds after controlling for various socio-­economic characteristics. For upper-­caste observers there is greater likelihood of punishing when the victim of defection is the same subgroup as the punishers; but not in the case of lower-­caste punishers. 16 In fact Weber himself did not argue for a straight causal relationship but for a more complex ‘elective affinity’. 17 See World Bank (2005: ch. 4) for a summary of microeconomic evidence. Banerjee and Duflo (2005) argue that microeconomic distortions, reflected in large observed differences in the return to capital within the same economy, can be of large aggregate importance. But it is harder to form a view of losses associated with the particular processes under consideration here. 18 The first part of this section draws closely on Bourguignon et al. (2007). 19 A concept of equality of agency can go beyond cultural factors. It could naturally include equal access to political process as an intrinsically desirable end, beyond the instrumental influence on equality in social and economic opportunities. 20 See Somanathan (2010) who cites B.R. Ambedkar as explicitly advocating that that the best (amongst dalits) should be able to rise to the highest positions in society. 21 This is an important theme of Loury’s discussion of the case for societal action for the contemporary manifestation of historical injustice against black Americans (Loury 2002). 24 These were later extended to ‘other backward castes’, traditionally above dalits but below upper castes in the social hierarchy. 25 Somanathan (2010) introduces her exploration of the demand for disadvantaged status with an account of the Gujjiar’s claim to scheduled tribe status. 26 The distinction between circumstances and effort is important to normative concerns, to which we return below. 27 Piketty (1995), for example, writes a model in which beliefs about the pay-­off to effort in terms of mobility depend on one’s family history, and on that family’s outcomes relative to those of others. Compare the discussion of future aspects of culture and the capacity to aspire above. 28 A large class of political economy models yield results in which policy variables depend on the distribution of wealth or incomes at an initial period. 29 Dominance can be by a demanding first-­order stochastic dominance criterion; or by the weaker, and probably more reasonable, second-­order stochastic dominance criterion. 30 The formulation of the problem implies selecting the type with the lowest present value of advantage over time, which is not necessarily the same as picking today’s least advantaged type. Obviously, under uncertainty one would replace variables by their expectations.

Culture and poverty   231 31 The critical deprivation level ut is not necessarily the same as a society’s chosen poverty line: if an individual has a level of income below the poverty line she is categorized as deprived within the society, whereas if her level of advantage falls below the critical deprivation level the society takes action to lift her back above this level. In very poor societies this may be a very low level, such as preventing death from lack of food in famine conditions.

References Acemoglu, D., Johnson, S. and Robinson, J. (2001) ‘The Colonial Origins of Comparative Development: An Empirical Investigation’, American Economic Review, 91 (5), pp. 1369–401. Akerlof, G. (1976) ‘The Economics of Caste and of the Rat Race and Other Woeful Tales’, The Quarterly Journal of Economics, 90, pp.  599–617. Also available in Akerlof, G. (1984) An Economic Theorist’s Book of Tales, New York: Cambridge University Press. Appadurai, A. (2004) ‘The Capacity to Aspire: Culture and the Terms of Recognition’, in Rao and Walton (2004a). Arizpe, L. (2004) ‘The Intellectual History of Culture and Development Institutions’, in Rao and Walton (2004a). Atkinson, B. and Stiglitz, J.E. (1980) Lectures on Public Economics, London: McGraw-­ Hill International Editions. Baland, J.M. and Platteau, J.P. (1996) Halting Degradation of Natural Resources. Is there a Role for Rural Communities?, New York and Oxford: Oxford University Press. Banerjee, A. and Duflo, E. (2005) ‘Growth Theory through the Lens of Development Economics,’ in Durlauf, S. and Aghion, P. (eds) Handbook of Economic Growth, Volume 1A, Amsterdam: Elsevier Science Ltd.; North Holland. Banerjee, A. and Somanathan, R. (2007) ‘The Political Economy of Public Goods: Some Evidence from India’, Journal of Development Economics, 82, pp. 287–314. Basu, K. (2006) ‘Participatory Equity, Identity and Productivity: Policy Implications for Promoting Development’, CAE Working Paper #06–06, Cornell University. Becker, G. (1971) The Economics of Discrimination, Chicago: University of Chicago Press. Bertrand, M., Hanna, R. and Mullainathan, S. (2008) ‘Affirmative Action in Education: Evidence From Engineering College Admissions in India’, NBER Working Paper No. 13926, Cambridge, MA: National Bureau of Economic Research. Bourdieu, P. (1990) The Logic of Practice, Stanford: Stanford University Press. Bourguignon, F. and Verdier, T. (2000) ‘Oligarchy, Democracy, Inequality and Growth’, Journal of Development Economics, 62 (2), pp. 285–313. Bourguignon, F., Ferreira, F. and Walton, M. (2007) ‘Equity, Efficiency and Inequality Traps: A Research Agenda’, Journal of Economic Inequality, 5, pp. 235–56. Brubaker, R. (2002) ‘Ethnicity Without Groups’, Archives of European Sociology, 43 (2), 163–89. Buchanan, J.M. (1976) ‘A Hobbesian Interpretation of the Rawlsian Difference Principle’, Kyklos, 29 (1), pp. 5–25. Chandra, K. (2004) Why Ethnic Parties Succeed: Patronage and Ethnic Head Counts in India, Cambridge, New York and New Delhi: Cambridge University Press. Deaton, A. (1997) The Analysis of Household Surveys, Baltimore and Washington, DC: Johns Hopkins University Press and the World Bank. Deshpande, A. (2001) ‘Caste at Birth? Redefining Disparity in India’, Review of Development Economics, 5 (1), pp. 130–44.

232   M. Walton —— (2007) ‘Caste’, in Basu, K. (ed.) The Oxford Companion to Economics in India, Delhi: Oxford University Press. Deshpande, A. and Newman, K. (2007) ‘Where the Path Leads: The Role of Caste in Post-­University Employment Expectations’, Economic and Political Weekly, 42 (41), 13–19 October. Engerman, S. and Sokoloff, K. (2002) ‘Factor Endowments, Inequality and Paths of Development among New World Economies’, Economía, 3 (1), pp. 41–110. Guha, R. (2007) ‘Adivasis, Naxalites and Indian Democracy’, Economic and Political Weekly, 42 (32), 11–17 August. Harrison, L. and Huntington, S. (eds) (2000) Culture Matters: How Values Shape Human Progress, New York: Basic Books. Heller, P. (1999) The Labor of Development: Workers and the Transformation of Capitalism in Kerala, India, Ithaca: Cornell University Press. Hoff, K. and Pandey, P. (2005) ‘Opportunity is Not Everything. How Belief Systems and Mistrust Shape Responses to Economic Incentives’, Economics of Transition, 13 (3), pp. 445–72. Hoff, K., Kshetramade, M. and Fehr, E. (2008) ‘Does Social Exclusion Produce Cognitive Shifts? Experimental Evidence on Group Differences on Social Restraints on Opportunistic Behavior’, mimeo, Washington, DC: World Bank. Jalan, J. and Murgai, R. (2007) ‘Intergenerational Mobility in Education in India’, mimeo, Delhi: World Bank. Jodhika, S.S. and Newman, K. (2007) ‘In the Name of Globalisation: Meritocracy, Productivity and the Hidden Language of Caste’, Economic and Political Weekly, XLII, 41, pp. 4125–32. Lal, D. (1989) The Hindu Equilibrium: Volume 1: Cultural Stability and Economic Stagnation, Oxford: Clarendon Press, pp. 1500–980. Lewis, O. (1959) Five Families: Mexican Case Studies in the Culture of Poverty, New York: Basic Books. Loury, G. (2002) The Anatomy of Racial Inequality, Cambridge, MA: Harvard University Press. Mehta, P.B. (2003) The Burden of Democracy, New Delhi: Penguin. Munshi, K. and Rosenzweig, M. (2006) ‘Traditional Institutions Meet the Modern World: Caste, Gender, and Schooling Choice in a Globalizing Economy’, American Economic Review, 96 (4), pp. 1225–52. Pande, R. (2003) ‘Can Mandated Political Representation Increase Policy Influence for Disadvantaged Minorities? Theory and Evidence from India’, The American Economic Review, 93 (4), pp. 1132–51. Piketty, T. (1995) ‘Social Mobility and Redistributive Politics’, Quarterly Journal of Economics, CX (3), pp. 551–84. PROBE team (1998) The Public Report on Basic Education, India, New Delhi. Rao, V. (2006) ‘On “Inequality Traps” and Development Policy’, Development Outreach, 8 (1). Rao, V. and Ban, R. (2007) ‘The Political Construction of Caste in South India’, mimeo, Washingon, DC: the World Bank. Rao, V. and Walton, M. (2004a) Culture and Public Action, Stanford: Stanford University Press. —— (2004b) ‘Culture and Public Action: Relationality, Equality of Agency, and Development’, in Rao and Walton (2004a). Roemer, J.E. (1998) Equality of Opportunity, Cambridge, MA: Harvard University Press.

Culture and poverty   233 Rudolph, L. and Rudolph, S.H. (1965) ‘The Modernity of Tradition: The Democratic Incarnation of Caste in India’, American Political Science Review, 59 (4), pp. 975–89, reprinted in Rudolph and Rudolph (2008). —— (2008) Explaining Indian Democracy: A Fifty-­Year Perspective, 1956–2006, New Delhi: Oxford University Press. Scott, J. (1985) Weapons of the Weak: Everyday Forms of Peasant Resistance, New Haven: Yale University Press. Sen, A.K. (1992) Inequality Reexamined, Cambridge, MA: Harvard University Press. —— (1998) Reason before Identity, Oxford: Oxford University Press. —— (2004) ‘How Does Culture Matter?’, in Rao and Walton (2004a). Somanathan, R. (2010) ‘The Demand for Disadvantage’, paper presented at the Fifth Annual AFD/EUDN conference, Paris, October 2007 and Chapter 6, this volume. Swartz, D. (2000) Culture and Power: The Sociology of Pierre Bourdieu, Chicago: University of Chicago Press. Thorat, S. (2007) ‘Human Poverty and Socially Disadvantaged Groups in India’, Discussion paper #18, Delhi: United Nations Development Programme. Thorat, S. and Attewell, P. (2007) ‘The Legacy of Social Exclusion: a Correspondence Study of Job Discrimination in India’, Economic and Political Weekly, XLII: 41, pp. 4141–4. Weber, M. (1930) The Protestant Ethic and the Spirit of Capitalism, trans. T. Parsons, London: Allen and Unwin. World Bank (2000) World Development Report 2000/01: Attacking Poverty, New York: World Bank and Oxford University Press. —— (2005) World Development Report 2006: Equity and Development, New York: World Bank and Oxford University Press.

Comments on Michael Walton’s chapter Stuti Khemani

Summary of chapter This chapter reviews a broad literature across disciplines on the relationship between culture and poverty, and makes a specific argument for why and how culture matters for poverty. In my reading, the argument is that cultural processes convey unequal ‘advantages’ to individuals that are persistent over time, and have large effects on economic opportunity available to individuals. Some of the theoretical and empirical examples offered in support of this argument are as follows. One, individuals can hold beliefs that some groups have better networks than others, which leads them to discriminate against members of marginalized groups when entering into productive contracts. Two, social customs can relegate individuals from specific groups to low productivity jobs, with costs for both employers and workers if they attempt to assign these individuals to other occupations not customized for their group. Third, internalizing the discriminatory features of the markets for economic opportunity, individuals belonging to the marginalized groups can under-­invest in their own capacity because they expect low returns. The chapter also offers thoughts on how the distribution of advantages might become more equal over time, or how the advantage of the most marginalized group might increase sufficiently so that they can access the economic opportunities needed to move out of poverty. One form of change is endogenous – cultural identities can themselves shift, on their own through social mobilization or in response to changing economic and political conditions. In terms of the analytical tools of the examples above, it may be that the elite or the marginalized find it in their interest to undertake the collective action needed to move a group to higher productivity jobs. In another example, the marginalized may respond to new economic opportunities available and invest in their own capacity. Public policy, of course, is an important determinant of both absolute and relative levels of advantage, and the chapter devotes some time to discussing how policy design may be informed by cultural features of poverty. The range of policies that are argued to be consistent with the positive and normative cultural approach to poverty are provided in Table 10.4, with the matrix of rows and columns providing the broad characterization of types of policies, and the cells providing specific examples of each.

Comments   235

Comments on the chapter I found the arguments provided in the chapter generally convincing and useful, and greatly enjoyed reading it. Some comments follow below. The chapter’s argument, and all the examples it harnesses, seems to imply that culture matters for poverty largely in situations where poverty is concentrated in specific social groups. That is, when we observe poverty concentrated in specific groups (where the poor share a particular social identity that is different from the rich) we can find cultural reasons for it, as provided in the chapter. Before applying the approach of this chapter and its policy implications, isn’t a straight empirical exercise warranted on the distribution of poverty across social groups? If poverty is spread across multiple social groups, no one of which contains a majority of the rich, then it’s not clear that the idea of ‘inequality traps’ provided in the chapter applies to its understanding. If it does not, then the group-­targeted preferential policies derived in the chapter could be less than desirable if such policies inhibit collective action or promote coordination failures among the poor when they are ethnically fragmented. The chapter seems to argue that, in the Indian case it focuses on, poverty is indeed concentrated in a few social groups. However, simply going by the Indian numbers on growth and poverty reduction (even at the lower end of the range in the great Indian poverty debate), it would seem that there is quite a bit of ‘churning’ in the distribution, and it is far from ‘stable’. Could this churning have led to growing inequality for instance within dalit groups, and could it be greater than inequality between poor dalits and (say) poor thakurs (upper-­caste groups)? Distributions are also complex in that they are not bipolar, with the rich on one end and the poor on the other, because the vast majority of people likely fall somewhere in the middle, and perhaps skewed towards the poor end, with real risks of falling into poverty with even small income or health shocks. That is, it seems plausible that in the multi-­ethnic, multi-­religious, multi-­linguistic societies of the Indian sub-­continent and Africa, if we overlay some sort of a distribution of social groups onto a distribution of poverty we would find a substantial set of people sharing characteristics of poverty but fragmented along the lines of cultural identity. There is a growing literature correlating ethnic fragmentation across and within countries with lower investments in public goods, and with poorer quality fiscal policies that contribute to macroeconomic instability, each of which has substantial implications for poverty.1 The central idea shared between models in this literature is quite different than that of ‘inequality traps’ – it is more akin to costly bargaining between groups with equal or near-­equal ‘power’. Some aspects of the policy implications of a costly bargaining perspective might be directly opposed to those of ‘inequality traps’. Specifically, group-­identity targeted preferential policies, which seem to come out of the normative analysis in this ‘inequality traps’ chapter, might exacerbate inefficiencies and inequalities in resource allocation if we are in a world of costly bargaining between groups. The argument cautioning against group-­based targeting, or at least showing a

236   S. Khemani trade-­off between group-­based targeting and universal provision of public goods, is explicitly political. The chapter cites work by political scientists that convincingly establish the view that the main currency of political competition in India is the provision of direct transfers and benefits to individual households (subsidies, welfare payments, jobs), often at the expense of broad public services (public/preventive health, quality education) that simultaneously benefit many. This argument has been made in broader contexts in the form of ‘clientelist’ or ‘patronage’ politics. Political science research takes the existence of patronage politics as given, or a left-­hand-side variable, and then goes on to explore its impact on political mobilization on the basis of ethnic identity, the right-­handside variable. Keefer and Khemani (2004, 2005) reverse the order – we seek to explain clientelism or patronage politics, using social fragmentation and polarization as one of several explanatory variables. In doing so we develop a new perspective on the nature of the inefficiency in resource allocation, which is different from the traditional view of clientelism as the capture of resources by the ‘elite’ in exchange for delivering the votes of the non-­elite who can be persuaded or coerced in various ways. We argue that social fragmentation combines with other forces (information asymmetries among voters on actions of political agents; credibility constraints of political candidates) to create incentives for poor or non-­elite voters to mobilize to demand policies that deliver targeted and verifiable benefits to their individual households, at the expense of broad public goods that would provide greater benefits to their and other groups. To exemplify the hypothesis in the context of the Indian case – dalits may organize to demand greater provision of in-­house latrines to dalit households (a classic transfer targeted at dalits), at the expense of organizing to demand public sanitation improvements in concert with other poor voters from other social groups. Political mobilization around social identity might have the perverse effect of increasing within-­group inequality by allowing some members to escape from poverty by availing of group-­targeted transfers, with not much effect on increasing general advantage for the group because of fiscal constraints in reaching all eligible beneficiaries. Yet, the hypothesis continues, all group members would vote for the group-­targeted transfer because they think they have a better probability of accessing benefits from public resources through their group membership than if resources were allocated to universally access­ ible public goods. Devarajan et al. (2009) argue that some fiscal policies, of decentralizing bits and pieces of budgets to village-­level governments (panchayats) in India can exacerbate the problem of non-­elite mobilization for private transfers to members of their group, at the expense of mobilizing for broad public goods. Decentralization to panchayats is skewed in a particular way – they have considerable responsibility and discretion in beneficiary selection for redistributive schemes that are group-­targeted, and this responsibility is well-­publicized, in contrast to their residual and under-­emphasized roles in maintaining local public goods that are constructed under national schemes. We argue that when non-­elite citizens

Comments   237 face such a local government they have incentives to mobilize to demand a greater number of beneficiaries be selected from their group, but at the expense of demanding better maintenance of local public goods, like schools or sanitation from which more members of the groups could receive greater advantage. Local elections would then turn on performance in providing targeted benefits at the expense of performance in maintaining public goods. Walton discusses political reservations for dalits and adivasis as a policy response to cultural inequality traps, and cites a growing empirical literature which concludes that reservations increase the likelihood of targeting government programme benefits to members of these groups. The extent to which reservations improve overall welfare of the targeted group (e.g. whether increased targeted services come at the expense of public good provision) is still unclear. Keefer and Khemani (2009) shed some light on this by examining a unique public spending programme in India – the Member of Parliament Local Area Development Scheme (MPLADS) – which entitles every member of the national parliament, elected from single-­member constituencies, to substantial resources to spend on local public infrastructure in their districts. We find that in districts that are reserved for scheduled tribes and scheduled castes, politicians spend less effort in using their MPLADS entitlement to bring public goods to the district, especially when they are politically dominant. Scheduled tribe and scheduled caste politicians who have been elected for consecutive terms from a district, spend 14 percentage points less of their entitlement than other politicians. That is, dominant incumbents from reserved districts are not dominant because they exert great effort in providing public infrastructure to their constituencies. Their dominance likely comes from other kinds of identity-­based services, which may not be the optimal way to improve distribution of advantages across and within groups. Future research could valuably focus on empirically distinguishing between competing hypotheses of the impact of group-­targeted preferential policies and its possible trade-­off with universal public goods policies.

Note 1 Habyarimana et al. (2007) is a recent contribution and reviews the public goods literature; Persson et al. (2007) and Khemani and Wane (2008) contribute new theory and evidence on the quality of fiscal policy when there are multiple groups.

References Devarajan, S., Khemani, S. and Shah, S. (2009) ‘The Politics of Partial Decentralization’, in Ehtisham, A. and Giorgio, B. (eds) Effectiveness of Decentralized Strategies and Outcomes, Cheltenham: Edward Elgar. Habyarimana, J., Humphreys, M., Posner, D. and Weinstein, J. (2007) ‘Why does Ethnic Diversity Undermine Public Good Provision?’, American Political Science Review, 101 (4), pp. 709–25.

238   S. Khemani Keefer, P. and Khemani, S. (2004) ‘Why do the Poor Receive Poor Services?’, Economic and Political Weekly, February 2004, 39 (9), pp. 935–43. —— (2005) ‘Democracy, Public Expenditures, and the Poor’, World Bank Research Observer, Spring 2005, pp. 201–27. —— (2009) ‘When do Legislators Pass on “Pork”? The Determinants of Legislator Utilization of a Constituency Development Fund in India’, Policy Research Working Paper No. 4929, May 2009, Development Research Group, Washington, DC: World Bank. Khemani, S. and Wane, W. (2008) ‘Populist Fiscal Policy’, Policy Research Working Paper No. 4762, October 2008, Development Research Group, Washington, DC: World Bank. Persson, T., Roland, G. and Tabellini, G. (2007) ‘Electoral Rules and Government Spending in Parliamentary Democracies’, Quarterly Journal of Political Science, 2 (2), pp. 155–88.

Comments on Michael Walton’s chapter Stéphanie Mahieu

In his chapter, Michael Walton convincingly argues that, although culture does matter to understand poverty mechanisms, there is no such thing as a ‘culture of poverty’. He develops interesting tools to understand the changing patterns of inequality traps throughout history and different political contexts, based on Indian examples. However, the question is not totally solved: why is there persistency of some group-­based differences? Is it because of internal or external mechanisms? As a social anthropologist, I will comment here on two of the topics related to the discussions about culture and its implications for development and poverty reduction, which have recently been subjected to vivid discussions: first, temporality and the issue of cultural distance between societies, and secondly, meaning, that is, can culture explain collective behaviours, including poverty mechanisms?

Temporality and cultural distance Walton pertinently refers to Arjun Appadurai’s critical view on the definition of culture usually given by anthropology (Appadurai 2004). For the American anthropologist, culture has too often been seen as primarily concerned with the past, that is with traditions. Appadurai sees such an approach as problematic, especially for development, because it gives the false view that culture, seen as the staunch guardian of ancestral traditions, would be the enemy of development. Appadurai suggests the concept of the ‘capacity to aspire’, that is to ‘conceive of, plan, organize and take action with respect to future paths’ (Appadurai 2004). For him, this capacity is unequally distributed across different groups in a society. The issue raised here, the ability of a given society (or of certain groups within a society) to launch itself into the future, is crucial not only for the dialogue between economists and anthropologists, the latter being often accused of not succeeding in making predictions (and therefore policy recommendations), but also for development policies. This applies not only in poverty reduction, but, more generally, in development interventions. Very often, development actors tend to consider that many ‘traditional societies’ have no historical memory (in the Western sense) and in particular that there is no memory of the

240   S. Mahieu previous development projects, that every new project takes places in a milieu with no traces of previous interventions (Olivier de Sardan 2005: 139). This – alleged – lack of cultural memory would prevent efficient ways to anticipate the future, and passivity towards external interventions would then become the norm. This vision is of course not accurate, as the local beneficiaries often have a deeper memory of past projects than development agencies themselves and precise anticipations of what they can expect from a given intervention. But this tendency to distinguish between societies with and without history, which refers to the classical distinction between stationary and cumulative history made by Claude Lévi-Strauss (Lévi-Strauss 1973) appears in many anthropological ana­ lyses and development interventions. These evolutionist premises, even though they have been radically criticized, notably by postcolonial studies, can yet be spotted in many contemporary anthropological analyses (Thomas 1989: 178), but are even more present within development. The ambiguity stems from the very project of development, as it was defined in the 1940s, which had (and still has) a strong positivistic and normative dimension. It is doubtless that much effort has been given in the last decades to widen a strictly growth-­based definition of development, to encourage more participation of the beneficiaries and to take into consideration elements such as indigenous knowledge (Briggs 2005). However, one has to recognize that normative macroeconomic-­type theories still prevail over other approaches, in what Jean-­Pierre Olivier de Sardan calls the ‘developmentalist configuration’ (Olivier de Sardan 2005: 25). Such an approach, where growth is a key element, is underpinned by a specific vision of how human societies have developed so far and of how they should develop in the future. What sometimes underlies this vision of several phases of human development is an ideology which French anthropologist Françoise Paul-­Lévy has called the primitivist ideology. Its origins are to be found in the late nineteenth century, when sociology was created as a science of ‘modern’ societies and anthropology as a science of ‘primitive’ societies, with both positivist and evolutionary visions on what is the model of reference and the goal to reach. This has often been made with laudable (at the time!) civilizing intentions. For the primitivist ideology, the obvious cultural and geographical distance between Western and non-­ Western societies has been analysed as a temporal distance, according to a hierarchical grading of human societies, from the least developed (‘primitive’) to the most developed (Western) (Paul-­Lévy 1986: 314). According to this vision, the existence of ‘primitive’ cultures in the contemporary world is to be explained by their lack of historical memory. Cultural and geographical distance becomes a temporal distance, with the view, sometimes used to qualify certain tribes in Papua New Guinea, that ‘they still live in primitive stone age’ (Lemonnier 1999). This vision has recently (as of May 2008) been reactivated with the ‘discovery’ of a ‘lost tribe’ in Brazilian Amazonia. Yet, such a vision tends to leave traditional societies ‘out of time’ (Thomas 1989) and doesn’t help understanding how poverty and exclusion mechanisms perpetuate in changing political contexts. This is why Walton’s reference to Appadurai’s focus on the future is very

Comments   241 pertinent and his historically contextualized analysis of the Indian case convincing. I will now turn to the second part of my comment, based on Walton’s attention to the ‘relational and dynamic’ dimension of culture.

Meaning and charity. Can culture explain collective behaviour? Social anthropology is often seen as a ‘sciences of cultures’. This is one of the reasons why social anthropologists have become significant actors of development after it became clear that a strictly technique-­based approach to development had failed; hypotheses were made on the cultural reasons impeding development, etc. (Escobar 1991). However, defining culture is a hard task, as Walton rightly stresses. In his chapter, he considers culture in its identity dimension, that is ‘how groups define themselves in relation to other groups’, and provides a fluid, relational definition of culture, which since Fredrick Barth’s famous essay Ethnic Groups and Boundaries (1969) appears consensual in social sciences. The concept of identity, however, is not less problematic than culture: besides its polysemy, identity, as Brubaker suggests, is often used in two very distinct perspectives, like in the example of the interaction of identity and politics commented on by Walton (pp. 214–15), in this case the interaction between caste and political parties in India. Brubaker and Cooper (2000) suggest that categories of ethnopolitical practice (driven by ethnic entrepreneurs) are distinct from categories of social analysis (identity, ethnicity for social scientists). This distinction was first theorized by Bourdieu and his distinction between categories of practice and categories of analysis (Bourdieu 1977). Yet, the two conceptions of the word identity are confused by many social scientists. In addition, even with a relational, multiple view on culture or on identity, there is often the implicit assumption that people behave the way they behave because they belong to a ‘culture’ (even a fluid, dynamic, contested, relational one). Their belonging to a certain culture(s) provides them with collective meanings about the world, which lead to (and explain) collective actions. As social anthropologists, we often witness actions which appear profoundly foreign to our mode of understanding since anthropology was created as a science of cultural difference (Bazin 2003: 422). This applies in particular to poor, marginalized populations, the traditional focus of interest of the discipline. According to anthropologist Jean Bazin, when these actions appear as weird, irrational, counter-­productive, we assume that the driving force behind them is a set of meanings, a shared view on the world, a ‘culture’ (Bazin 2003: 430). The implicit assumption about the rationale of their actions is: ‘They behave like this because their culture makes them believe this is right, correct, desirable (and even though it may lead to poverty and exclusion).’ This assumption is often made in the name of a ‘principle of charity’. It is often here that positive analysis becomes normative, as rightfully suggested by Walton. Trying to understand is indeed a basic principle of charity, to show that other cultures have their coherence: people are not stupid, they just see things

242   S. Mahieu differently. But often anthropologists have defended cultural peculiarities, in the name of this principle of charity (and even when these peculiarities are ethically questionable), when economists had much more normative positions. Yet, it is hard to recognize that all (apparently) irrational practices have their own logic, without falling into cultural relativism. This has been formulated by Bazin about the questions of meaning and belief, in the field of religious anthropology: Saying, while watching pilgrims, ‘all those nice people are wrong’, is judging them and not studying their behaviour. Explaining why they are wrong, what misleads them is to justify their judgment, not understand their practice. But saying that they are not wrong (or not very wrong, or not more wrong than the others) is not to gain a deeper understanding of what they are doing (it is at best empathize with them). (Bazin 1991: 507) This applies as well to the mechanisms of poverty presented by Walton in his chapter, because they appear particularly foreign to our modes of understanding: why do some groups perpetuate practices which obviously maintain them in a precarious situation? There should be an explanation, because it seems inconceivable than certain groups would consciously keep on choosing poverty over other options. And this explanation, after having examined other hypotheses, is often culture, because at least it gives a meaning to otherwise incomprehensible practices. But the problem is that such an explanation is often tautological, and culture appears as a ‘black box’ which can explain any – apparently – irrational behaviour. One could argue, as Walton rightfully does, that the situation of groups trapped in poverty mechanisms is mainly the result of external exclusion mechanisms from other groups of the society and more generally from external institutions. This appears particularly true in the case of the Dalits in India; one could also cite the Roma in Europe. But external exclusion mechanisms cannot explain all differences between groups and the issue of culture comes back immediately. What can we do? Do we have to resign ourselves to the powerfulness of cultural explanations? According to Bazin, however, anthropology is not condemned to remain knowledge of cultures. What we can do as anthropologists (but I would add that this applies as well to development practitioners) is to provide a description of human actions that helps us understand collective actions, that is to describe a set of rules that are sensible, accepted, plausible, expected. If we, as social anthropologists, economists (in the perspective of Douglass North) or development practitioners, establish the rules of the game, prediction becomes easier. Establishing the rules of the game also helps questioning the ‘altruist paradigm’ (Olivier de Sardan 2005: 71), so present within development, but, as we have seen about the charity/denunciation dilemma, often too normative.

Comments   243

References Appadurai, A. (2004) ‘The Capacity to Aspire: Culture and the Terms of Recognition’, in Rao, V. and Walton, M. Culture and Public Action, Stanford: Stanford University Press, pp. 59–84. Barth, F. (1969) Ethnic Groups and Boundaries. The Social Organization of Culture, London: Allen and Unwin. Bazin, J. (1991) ‘Les fantômes de Mme du Deffand: exercices sur la croyance’ Critique, 47, pp. 492–511. —— (2003) ‘Questions of Meaning’, Anthropological Theory, 3 (4), pp. 416–34 (translated from French, first published 1991). Bourdieu, P. (1977) Outline of a Theory of Practice, Cambridge: Cambridge University Press (translated from French, first published 1972). Briggs, J. (2005) ‘The Use of Indigenous Knowledge in Development: Problems and Challenges’, Progress in Development Studies, 5 (2), pp. 99–114. Brubaker, R. and Cooper, F. (2000) ‘Beyond Identity’, Theory and Society, 29, pp. 1–47. Escobar, A. (1991) ‘Anthropology and the Development Encounter: The Making and Marketing of Development Anthropology’, American Ethnologist, 18 (4), pp. 658–82. Lemonnier, P. (1999) ‘La chasse à l’authentique. Histoires d’un âge de pierre hors contexte’, Terrain, 33, pp. 93–110. Lévi-Strauss, C. (1973) Antropologie structurale deux, Paris: Plon (Structural Anthropology, Vol. II, trans. Monique Layton, London: Viking, 1976). Olivier de Sardan, J.P. (2005) Anthropology and Development. Understanding Contemporary Social Change, London: Zed (translated from French, first published, 1995) Paul-­Lévy, F. (1986) ‘A la fondation de la Sociologie: l’idéologie primitiviste’, L’Homme: Anthropologie Etat des Lieux, 97–98, pp. 299–320. Thomas, N. (1989) Out of Time. History and Evolution in Anthropological Discourse, Cambridge: Cambridge University Press.

Part V

Conclusion

11 Revisiting the role of culture Jean-­Philippe Platteau and Robert Peccoud

In this concluding chapter, we want to point out a number of important lessons that can be drawn from the dialogue between economists and other social scientists that has unfolded in the preceding contributions.

Culture exists yet it evolves The first lesson is that the relationship between culture and development, where the latter is understood as a process of change leading to improvements in the well-­being of the people, involves a two-­way interaction. Running counter to the ‘essentialist’ view of culture propounded by Samuel Huntington (1993, 1998), Francis Fukuyama (1995), David Landes (1998), and others, an approach advocated most forcefully by Jean-­François Bayart in this volume (Chapter 3) emphasizes the endogenous character of culture. Culture is a construct that is easily modified and shaped by political, social and economic forces or agents. As a consequence, it cannot form by itself a genuine impediment to development. Opposition to the ‘essentialist’ view is all the stronger as, in the words of Héla Yousfi (Chapter 2), culture is suspected of bringing back tradition. In reality, a middle ground is quite possible between these two opposite views of the role of culture, and it rests on the idea that it is useless to look for one-­way causality. At the same time, it is important to stress that cultural factors or agencies are not instantaneously modifiable or malleable at will. To some extent and for some time, they ‘stick’. As remarked in Chapter 1, they are more subject to inertia than other variables such as prices and quantities on a market. Understood as social norms, they actually belong to the realm of slow-­moving institutions in contrast to political institutions and, to a lesser extent, legal systems which are fast-­moving institutions (Roland 2004). When grafted into social universes characterized by different ‘local conditions’, which arise from each country’s slow-­ moving institutions, imported fast-­moving institutions will fail to take root. This key point is well-­recognized by Michael Walton (Chapter 10) when he says that social norms can outlive their usefulness, such as when severe stigmas outlive the abolition of formal barriers to equality of opportunity, or even proactive measures to redress them. This illustration is particularly apt, indeed, since there are innumerable examples where deliberate efforts, say through the channel of

248   J.-P. Platteau and R. Peccoud legislation, have failed to change the behavioural patterns and the associated belief systems associated with them within a reasonable span of time (for more examples, see Aldashev et al. 2009). The contribution of Timur Kuran in this volume (Chapter 4) also provides a vivid illustration of how changes in comparatively fast-­moving institutions have been blocked in the Middle East by attitudes and behaviour patterns rooted in ancient practices and principles. On the whole, however, it is correct to say that economists tend to ignore or underestimate the role of culture. It is only recently, following the glaring failures of external interventions in Russia and other transition economies, then in Iraq and Afghanistan, that they have become suddenly aware of the ‘stickiness’ of culture. Some scholars had, however, warned us for a long time against the pitfalls of ‘institutional imperialism’ and the naïve belief that institutions from the West can be easily and successfully transplanted in developing countries. Especially noticeable in this regard is the critique of Hungarian liberalism by the historian Szekfu as reported by Andrew Janos (1982). For Szekfu, the profound dysfunctioning of Hungarian society in the first part of the twentieth century was to be ascribed not to liberalism per se, but to ‘the transplantation of culturally alien liberal ideas into the ostensibly unreceptive soil of Hungary’ (p. 267). More precisely: The results of this transplantation were not only grotesque, they were outright harmful, because the gap between form and substance could only be closed by corrupting the electoral process in politics and by perverting the market mechanism in economics. In turn, the prevailing methods of corruption – the purchase of votes by the government, and the purchase of favours by the entrepreneur – had a deeply corrosive effect on national morale. They undermined the self-­respect of the elites and the legitimacy of the nation state, and in this manner they paved the way for the revolutions of 1918–1919, and for the ‘national tragedy’ of the peace treaty. (Janos 1982: 267) Notice that the same conclusion applies, mutatis mutandis, to Afghanistan where the first election of President Hamid Karzai, in spite of the superficial appearances of a free democratic election such as imposed by dominant Western countries, followed traditional procedures rooted in the country’s past. These procedures include hard bargaining between candidates and the elders leading the main tribes, the use of money and other material privileges to influence their behaviour, and strict collective discipline of tribal people once their leader has stated which presidential candidate to support. To explain variations in performance between countries, it is necessary to go behind the veil of formal institutions and understand how historical experience has shaped the set of expectations, beliefs, and social norms that form the cultural fabric of societies. For example, it is striking that Tatarstan, a large Muslim republic in the heart of Russia, is a stable political entity that has followed consistent economic policies since the time of the dissolution of the

Revisiting the role of culture   249 Soviet empire. It is ruled by an autocrat (Mr Shaimiev) who has managed to keep Tatarstan within Russia’s borders while maintaining a considerable measure of independence vis-­à-vis the Kremlin. Transfer of political power from Mr Shaimiev to his annointed successor (Mr Minnikhanov) is going to take place orderly without shaking the strong regime inherited from the Soviet era. By contrast, the republics of the north Caucasus – Ingushetia, Dagestan, and Chechnya – have been subject to great instability, tension, and violence that make them largely lawless places on the verge of civil war. Insofar as stability has been brought back to Chechnya, it is only because the Kremlin has eventually entrusted the government of the place to a local ruthless warlord (Kadyrov) who agrees to maintain Chechnya within Russia’ borders. One plausible explanation for the differences between Tatarstan, on the one hand, and the republics of the north Caucasus, on the other hand, is historical: while the former was conquered by Ivan the Terrible as early as in the 16th century, and then successfully integrated into the Russian empire, the latter were colonised only in the second half of the 19th century, and were never fully assimilated. Russia’s reliance on force and repression to pacify the region has contributed to undermine the legitimacy of its rule, which explains why after the break-­up of the Soviet Union in 1991 Chechnya demanded full independence (The Economist 2010: 36–7). An interesting parallel can be drawn here with the situation prevailing in the heart of Western Europe in the sixteenth century, more specifically in the so-­ called Low Countries. As a matter of fact, the separation of this region into two separate entities which were to become Belgium and Holland is usually attributed to the outcome of the force of arms: the duke of Alva, commander of the Spanish troops, succeeded in conquering the Flanders and Brabant while the north-­east was captured by Maurice and William Louis of Nassau and thus brought over to the rebel side. The real picture is actually more complex, and geographical differences in the state-­building process seem to have played a major role. More precisely, the central government acting on behalf of Philip II, king of Spain, made its presence felt rather strongly in the dynamic, urbanized west which depended on commerce and industry (in Brabant, Flanders and Holland-­Zeeland) while the Walloon provinces (Hainault, Artois, Namur) constituted the heartland of the numerous nobility, who provided one of the pillars of the Burgundian state and the ties binding these provinces to the centre. By contrast, the north-­eastern provinces from Utrecht to Groningen had only recently been added by Charles V to the hereditary lands belonging to the Spanish Habsburgs, which explains why they demanded a separate status and far-­reaching autonomy within the commonwealth. Since such a request was deemed unacceptable by the Spanish crown, war was inevitable (van Nierop 2001: 34–5). Conflict caused by the Protestant Reformation had the effect of compounding this problem. There is thus no escape from the conclusion that inasmuch as countries have followed different historical paths resulting in specific sets of norms, expectations and beliefs, they must find within their existing slow-­moving institutions

250   J.-P. Platteau and R. Peccoud (their cultural patrimony) the roots for changes in their fast-­moving institutions (Roland 2004; Platteau 1994). Such an approach does not lead to a deadlock since, although they change slowly, social norms tend to undergo continuous rather than abrupt transformations. There is actually ample evidence to show that in developing countries institutions that involve deep-­rooted social norms and beliefs have succeeded in changing with stunning regularity over time. Two examples readily come to mind here. The first one concerns reproductive behaviour. As a matter of fact, we are presently witnessing a rapid move to replacement-­level fertility at world level, a move that the Economist weekly has recently called ‘one of the most dramatic social changes in history’, since it means that demographic growth slows down to such an extent that world population will stabilize in the forthcoming future (2009: 29). By about 2020, according to predictions based on present trends, the global fertility rate is going to dip below the global replacement rate (2.33) for the first time, reflecting fundamental drops in fertility in poor countries that used to have large families. Between 1950 and 2000 the average fertility rate in developing countries fell by half from six to three, meaning that each family had three fewer children within the short span of just 50 years. Such a fall is close to what happened in Europe during nineteenth and early twentieth-­century industrialization process. But what took place in Britain over 130 years (1800–1930) took place in just about 20 years in South Korea (1965–85) and in Bangladesh (1980– 2000). And things are moving even faster today since the same decline took place in Mauritius in just ten years (1963–73), and the fall of fertility has been even more sensational in Iran where the number of children per family fell from seven in 1984 to below replacement level in 2006 (the average fertility rate is only 1.5 in Tehran). The second example is closely related to the first one as it deals with the transformation of the family structure. This is a question that has been addressed by both economists and other social scientists, and can be cited as a convincing illustration of the complementarity of their respective contributions. Thus, the celebrated thesis of Ester Boserup (1965), which had a large resonance among development economists and was further refined by them (Binswanger and Rosenzweig 1986; Pingali and Binswanger 1986; Binswanger and McIntire 1987; Pingali et al. 1987; Binswanger et al. 1989; Hayami and Otsuka 1993), is grounded in incentive considerations the incidence of which is assumed to change with ecological conditions. In a nutshell, the rise of peasant farms is explained by growing land scarcity because, as land pressure increases, farmers are induced to shift to more intensive forms of land use, which implies that they adopt increasingly land-­saving and labour-­using techniques. An important characteristic of these techniques is that labour quality, which is costly to monitor, assumes growing importance. Given the incentive problems associated with care-­intensive activities (sometimes labeled ‘management diseconomies of scale’), the small family or peasant farm in which a few co-­workers (spouses and their children) are residual claimants, appears to be the most efficient farm structure.

Revisiting the role of culture   251 It is puzzling to observe, however, that even in conditions of technological stagnation the individualization of the farm-­cum-family structure may occur. For example, in the San-­Koutiala-Sikasso region in Mali, although there is no clear evidence of technological change, collective farms coexist with mixed farms – in which individual plots are cultivated by members alongside collective fields – and small farms born of the break-­up of large family farms. It is also evident that the latter two forms have become more widespread as time elapses. Attempting to solve the puzzle, Catherine Guirkinger and Jean-­ Philippe Platteau (2009) have written a model of patriarchy in which the family head decides whether he prefers to keep the family whole or split it into several units, and in the former eventuality whether he prefers to keep the farm completely integrated or to grant individual plots to members who are also required to continue to work on the collective field. They show that individualization of the farm-­cum-family structure in the form of a mitigated form (the mixed farm) or an advanced form (the branch households formed as a result of the splitting of the stem household) occurs when land scarcity increases beyond a certain level or when exit opportunities for members improve above a certain point. A striking parallel is thus established between the twin processes of individualization of farm-­cum-family structures and individualization of land tenure rules (i.e. the move from corporate ownership of land, possibly including the granting of long-­term use rights to individual households, to individualized forms of tenure ranging from less to more complete private property rights) since the same forces, increasing land pressure and growing market integration, are predicted to trigger them. In the case of tenure rules, the effect operates through better internalization of externalities and stronger incentives to conserve and improve land (Demsetz 1967; Hayami and Kikuchi 1981; Feder and Feeny 1991; Feder and Noronha 1987; Baland and Platteau 1996; Platteau 2000: chs. 3–4). It is remarkable that individualization of land tenure rules, and the concomitant individualization of family-­cum-farm structures, have proceeded at breakneck pace in sub-­Saharan Africa compared to leading industrialized countries of the Western world. What it took England several centuries to achieve (from as early as the fourteenth century to end of the eighteenth/early nineteenth century) thus required less than one century in African countries. The above example, but also the detailed analysis of family and kinship systems proposed by Eliana La Ferrara in this volume (Chapter 5), may incidentally serve to illustrate a point that is a source of considerable misunderstanding between economists and other social scientists. For the purpose of analysing a given institution, economists often use static models, and this creates the impression that they are only preoccupied with immutable aspects of institutional realities. In fact, the comparative static model allows them to obtain powerful insights into what factors are susceptible of causing changes in the institution considered. By following this method, they are able to understand why institutions may have to change – in the above example, why African farms and families are bound to

252   J.-P. Platteau and R. Peccoud evolve towards the peasant farm model typical of Eurasia and, in La Ferrara’s chapter, why the family gradually loses importance when markets and state intervention develop – yet, admittedly, they are unable to predict how such changes will precisely take place. To detect paths of institutional change, dynamic models are needed and these are much more difficult to write in a way that remains analytically tractable.

The need to understand the others’ universe of meaning The second lesson that emerges from this book is that, since culture is ‘a producer of meaning’, a reference system that people use to make sense of their own actions and of the world in which they live, we may not assume that inhabitants of other parts of the world see things in the same way as we do. In the words of Héla Yousfi (Chapter 2), ‘all cultures establish criteria by which individuals denote, classify, identify, evaluate, connect and order’. Therefore: the question is not whether or not developing countries should change their culture to promote progress but rather what is the framework of meaning in each society that initiates an effective social transformation that does not compromise the locally shared conception of a ‘good’ relationship between individual autonomy and collective order. If it is added that the framework of meaning provided by a culture is susceptible of adapting continuously, albeit comparatively slowly, to new economic, technological and social realities, Yousfi’s statement closely resembles the above-­ reached conclusion based on the distinction between fast- and slow-­moving institutions. At this juncture, it is useful to give the reader two striking illustrations so as to better understand what is at stake here. In Kibera, the biggest slum area of Nairobi (Kenya), many so-­called self-­help groups are in existence, many of whom are women’s groups. When one of the present authors looked into the internal mode of functioning of these informal groups, he was struck by the fact that two radically different organisational features almost uniformly characterised them. In the first place, almost all groups have a written constitution in which precise rules are laid down regarding the composition of the executive committee, the responsibilities of the committee’s members, the way to elect them, and accountability procedures including the holding of regular meetings of all group members and decision-­making processes. In the second place, also written in the constitution of a large number of self-­help groups are what members call ‘rules of politeness’, which stress the proper manner in which to address the chairperson and other committee members in general assembly meetings and the gradual sanctions to be applied to any person violating the politeness code. The sanctions consist of fines to be paid in cash to the whole group and exclusion from the meeting and, eventually from the group in the event of repeated unattended warnings (Abraham et al. 1998).

Revisiting the role of culture   253 The existence of the first set of provisions was really surprising since they are at odds with customary practice in Africa. Upon inquiry, it actually appeared that groups had borrowed them from the guidelines set by an international organization (a United Nations agency) which group leaders felt they had to comply with in order to be eligible for receiving financial support. It is therefore no surprise that they do not take these externally imposed rules too seriously, and apply them rather perfunctorily. Quite different is the situation regarding the rules of politeness, which are seriously considered and rigorously enforced. This is the consequence of the fact that those rules were decided and designed by the group members themselves. They evolved endogenously from problems with which African people are very familiar but that Western experts did neither anticipate nor really understand. In the African setup, indeed, relationships are highly personalized and there is no neat separation between the sphere of private life and the public or professional domain. Because of this interpenetration of various levels of a person’s actions, there is always a risk that frustrations or resentments accumulated in one domain spill over into another domain. In particular, a group member who has a grudge against the chairperson, say, because his own family has an unresolved land conflict with the chairperson’s family, will always be tempted to vent out aggressive feelings in a meeting in which completely different problems concerning the whole group are being discussed. To counter that risk and prevent personal animosities from hampering the effective working of the group, a code of politeness is laid down that members take very seriously because they understand too well the stake behind these rules. In the same connection, it is interesting to observe that in SubSaharan Africa as well as in other countries where personalised relationships are pervasive, any leader of a grassroot organisation is expected to spend a lot of time dealing with members’ personal problems, trying to arbitrate not only interpersonal conflicts that may have arisen between them but also intra-­family problems with which particular members are struggling. In Senegal, for example, the director of a Non-­Governmental Organisation connected to rural populations spent more time and devoted more energy to settle disputes between some members and their co-­ wives than to fulfill the mission of the organisation as such. In point of fact, the leaders of grassroot organisations find themselves in the position of traditional leaders whose one important function is to settle conflicts so that they do not disrupt the social order of the community and undermine its collective activities (Abraham and Platteau 2004: 213). Let us now turn to our second example which we borrow from a study of relief aid in Southern Sudan by Simon Harragin (2004). In this study, the author ­persuasively argues that there is a fundamental conflict between the allocation criteria applied by Western donors for food relief operations in famine-­stricken areas, on the one hand, and the principles of fair treatment prevailing among local populations, on the other hand. In a world characterized by acute tensions among tribes (especially the Dinkas and the Nuer) who severely compete for grazing and water, the existing political equilibrium is necessarily precarious,

254   J.-P. Platteau and R. Peccoud liable to be broken by any occurring event that may cause a new imbalance. In the words of Harragin: One of the major priorities of local leaders, since the start of large-­scale food relief deliveries in southern Sudan in 1989, has been to ensure that the distribution of such food does not exacerbate existing tensions. The targeting of aid by relief agencies to certain zones in preference to others has sometimes caused tension. . . . The local perception is that large, powerful family groups contain a majority of the population, and so deserve the largest share. Having a large family is the ideal to which Dinka aspire. . . . Penalising those who have attained the kind of success to which people aspire and rewarding the unsuccessful has a questionable logic in a war where aspiration and hope of better times sometimes seems to be all that remains. The feeling is that all kin groups are separate economic units and should have equal rather than preferential rights to access resources – whether in the form of grazing or relief – be they rich or poor . . . [therefore], it is not seen as politically fair or wise to sustain the welfare system of some lineages and not others purely on the basis of differentiated need . . . Favouring the vulnerable could legitimise a system that favours a particular kin group and encourage the kind of preferential treatment that could be used to justify corruption. (Harragin 2004: 310–11) Clearly, as this last example attests, ‘an individual cannot be seen in isolation from a complex web of kinship and territorial obligations’, and individuals, whichever their level of needs, cannot make claims to resources separate from those that come from membership of a specific kinship group (Harragin 2004: 311). This is just one example among many others that illustrates the conflicting views of the autonomy of the individual in modern market societies and in tribal societies where traditional modes of dependence on the environment are still largely prevailing. The above discussion doubtless leads to the same conclusion as the aforementioned collective work edited by Vijayendra Rao and Michael Walton (2004), namely that it is dangerous to apply ‘best practice’ guidelines that are not well informed by the social and historical context, and that ‘good interventions are very difficult to design ex ante’. To sum up, ‘development is not easy. It is, at its core, a social and cultural process that requires a slow process of learning from the ground up in order to be effective and sustainable’ (2004: 363–6).

Culture and business There are two themes addressed in the book that deserve to be singled out in this last part of our concluding comments. The first of these themes is the difference of business culture between countries in general, and between advanced and developing countries in particular. It is evoked in the chapters written by Héla

Revisiting the role of culture   255 Yousfi (Chapter 2) and Philippe d’Iribarne (Chapter 8), with special reference to sub-­Saharan Africa. What is stressed by these authors is the existence of social norms born in the universe of lineage societies that tend to continue to govern relations between employers and employees and among the latter themselves within the setup of modern enterprises. Insofar as these norms exert pressure on managers and workers so that they redistribute their incomes towards kith and kin, they may discourage effort and lower workers’ morale. Spontaneous redistribution out of a sense of family solidarity does not have adverse consequences on behaviour, but forced redistribution backed by hostile attitudes of envious and jealous relatives does obviously create an unhealthy environment for the application of effort and the accumulation of capital (Platteau 2000: ch. 5; 2009). In his contribution to this book (Chapter 7), Marcel Fafchamps has chosen to emphasize the lack of familiarity of African entrepreneurs with technical and organizational innovations such as they have developed outside the African continent. Hence the continuing presence of weak market institutions and practices in much of this region. Yet, at the same time, he also points out that African business communities are often segmented along ethnic, social and religious lines, which may have the negative effect of fuelling political tension and frightening foreign investors whose presence is badly needed. Interestingly, his own research also shows that relationships between ethnicity and business networks tends to be loose even when the latter are ethnically concentrated, in the sense of being located in areas dominated by one ethnic group (Fafchamps 2004: 307; see also Platteau 2009). Moreover, he has evidence that not only trade with relatives and friends is extremely rare in Africa but also that, whenever it happens, it harms firm performance. African entrepreneurs often complain that it is indeed difficult to keep business with relatives within the confines of an economic transaction. This finding echoes several of our above-­stated conclusions regarding the limited sphere of individual autonomy and the intermingling of all domains of life, whether private or public. Migration and religious conversion have been argued to provide convenient escapes from family pressures for entrepreneurial individuals eager to seize economic opportunities and improve their standard of living (for references and examples, see Platteau 2000: ch. 5; 2009). Héla Yousfi mentions another possible strategy on the basis of the experience of a large business firm in Cameroon: the enactment of strict rules that constrain individual behaviour and enable them to invoke some externally imposed discipline to withstand family and ethnic pressures. However, although one can well see how this solution helps promote efficiency within the firm, it is hard to understand how it can protect dynamic individuals outside the firm when they have to decide how to allocate and use their incomes. Culture and affirmative action A natural and radical solution to the above problem of individual emancipation seems to lie in affirmative action policies aimed at moving whole groups or

256   J.-P. Platteau and R. Peccoud c­ ommunities forward by granting them privileged access to public sector employment, political representation and educational opportunities. In India, a country in which state-­led affirmative action has to be actively and consistently pursued, the recognition of group-­based identities to link individual advancement to group mobility has thus been based almost entirely on caste classifications. Such policy can also be advocated on the ground that it helps marginalized communities to break out of the poverty trap in which they have got stuck due to the mutually reinforcing effects of social stigma, low expectations and limited opportunities (see Walton, Chapter 10 in this volume). Things are not so simple, however. As argued by both Rohini Somanathan (Chapter 6) and Stuti Khemani (comment to Chapter 10), affirmative action has led to disappointing results in India, partly because relatively poor communities have spent much time and energy in jockeying on the political stage to get recognition as backward communities eligible for the benefits of reservation quotas (‘caste was after all a social construction’, writes Rohini Somanathan), partly because many poor people are fragmented along the lines of cultural identity, and partly because more attention has been paid to advantages obtainable under the policy of reservations than to the learning processes involved. To sum up, the quota mechanism enforced by the Indian government has produced important perverse effects in addition to benefiting only a small new elite lifted up from their marginal background and status. A political system in which cultural identities of politicians and parties matter more than their performances, and a growing polarization of society that strengthens incentives for shirking and rent-­ seeking are among the worst unintended effects of the particular mingling of culture and politics that has happened under affirmative action policies. In a statement presented in comments concluding the conference that gave rise to this book, the anthropologist Olivier de Sardan made the point that accounts invoking cultural factors are ‘lazy explanations’. This sounds strangely familiar to the ears of economists who are used to being told that cultural explanations are ‘ad hoc’ and are therefore better avoided. There is no denying that for too many analyses where cultural factors figure prominently such a harsh criticism is not unfounded. We nonetheless hope to have succeeded in persuading the reader that there is a great need for methodologically sound research which does not hesitate to venture into the cultural domain. This implies that culture is granted an explicit role, yet only within an analytical framework in which reciprocal causality and feedback effects between cultural and other variables are allowed for. Culture evolves and is historically constructed, but its ability to constrain behaviour and to determine outcomes alongside other factors in the short and medium run must be unambiguously recognized lest policy-­makers and donors should continue to make costly and avoidable mistakes. This is clearly the message conveyed by Hilary Synnott (2009), for example, when she writes that aid workers and diplomats offering assistance to Pakistan ‘need to operate as close to the ground as possible’, and that ‘their skills should include knowledge of local languages and an awareness of local history, custom and social norms.

Revisiting the role of culture   257 This will help them to avoid advocating policies and processes that conflict with local culture’ (p. 171). One important implication is that policy-­making must be backed by strong research that is not shy of placing cultural determinants at the core of the analysis. And it must be borne in mind that it is not because a researcher assumes culture as given while addressing a particular problem that he or she believes that culture is stagnant or immutable. In the same manner as it is not because a researcher analyses the evolution or the adaptation of culture that he or she believes that culture may never constrain behaviour.

References Abraham, A. and Platteau, J.P. (2004) ‘Participatory Development: Where Culture Creeps In’, in Rao, V. and Walton, M. (eds) Culture and Public Action, Stanford: Stanford University Press, pp. 210–33. Abraham, A., Baland, J.M. and Platteau, J.P. (1998) ‘Groupes informels de solidarité dans un bidonville du Tiers-­Monde: le cas de Kibera, Nairobi’, Non Marchand, Bruxelles: Editions De Boeck Université, 2, pp. 29–52. Aldashev, G., Chaara, I., Platteau, J.P. and Wahhaj, Z. (2009) ‘Modern Law as a Magnet to Reform Unfair Customs’, Working Paper, Centre for Research in the Economics of Development (CRED), Belgium: University of Namur. Baland, J.M. and Platteau, J.P. (1996) Halting Degradation of Natural Resources – Is There a Role for Rural Communities?, Oxford: Clarendon Press. Binswanger, H., and McIntire, J. (1987) ‘Behavioural and Material Determinants of Production Relations in Land-­Abundant Tropical Agriculture’, Economic Development and Cultural Change, 36 (1), pp. 73–99. Binswanger, H. and Rosenzweig, M. (1986) ‘Behavioural and Material Determinants of Production Relations in Agriculture’, Journal of Development Studies, 22 (3), pp. 503–39. Binswanger, H.P., McIntire, J. and Udry, C. (1989) ‘Production Relations in Semi-­Arid African Agriculture’, in Bardhan, P. (ed.) The Economic Theory of Agrarian Institutions, Oxford: Clarendon Press, pp. 122–44. Boserup, E. (1965) The Conditions of Agricultural Growth: The Economics of Agrarian Change under Population Pressure, London: Allen and Unwin. Demsetz, H. (1967) ‘Toward a Theory of Property Rights’, American Economic Review, 57 (2), pp. 347–59. Economist, The (2009) ‘Go Forth and Multiply a Lot Less’, 31 October, pp. 29–32. —— (2010) ‘Beyond the Kremlin’s Reach’, 30 January, pp. 36–7. Fafchamps, M. (2004) Market Institutions in Sub-­Saharan Africa –Theory and Evidence, Cambridge, MA and London: The MIT Press. Feder, G. and Feeny, D. (1991) ‘Land Tenure and Property Rights: Theory and Implications for Development Policy’, The World Bank Economic Review, 5 (1), pp. 135–53. Feder, G. and Noronha, R. (1987) ‘Land Rights Systems and Agricultural Development in Sub-­Saharan Africa’, Research Observer, 2 (2), pp. 143–69. Fukuyama, F. (1995) Trust – The Social Virtues and the Creation of Prosperity, London: Hamish Hamilton. Guirkinger, C. and Platteau, J.P. (2009) ‘Transformation of the Family under Rising Land Pressure: A Theoretical Essay’, Working Paper, Centre for Research in the Economics of Development (CRED), Belgium: University of Namur.

258   J.-P. Platteau and R. Peccoud Harragin, S. (2004) ‘Relief and an Understanding of Local Knowledge: The Case of Southern Sudan’, in Rao, V. and Walton, M. (eds) Culture and Public Action, Stanford: Stanford University Press, pp. 307–27. Hayami, Y. and Kikuchi, M. (1981) Asian Village Economy at the Crossroads, Tokyo: University of Tokyo Press, and Baltimore: Johns Hopkins University Press. Hayami, K. and Otsuka, K. (1993) The Economics of Contract Choice – An Agrarian Perspective, Oxford: Clarendon Press. Huntington, S.P. (1993) ‘The Clash of Civilizations?’, Foreign Affairs, 72 (3), pp. 22–50. —— (1998) The Clash of Civilizations and the Remaking of World Order, New York: Simon and Schuster. Janos, A. (1982) The Politics of Backwardness in Hungary, 1825–1945, Princeton: Princeton University Press. Landes, D. (1998) The Wealth and Poverty of Nations – Why Some Are So Rich and Some So Poor?, London: Little, Brown and Cy. Pingali, P.L. and Binswanger, H.P. (1986) ‘Population Density, Market Access and Farmer-­Generated Technical Change in Sub-­Saharan Africa’, Research Unit of the Agricultural and Rural Development Dept, Washington, DC: The World Bank. Pingali, P., Bigot, Y. and Binswanger, H.P. (1987) Agricultural Mechanization and the Evolution of Farming Systems in Sub-­Saharan Africa, Baltimore and London: Johns Hopkins University Press. Platteau, J.P. (1994) ‘Behind the Market Stage Where Real Societies Exist – Part I: The Role of Public and Private Order Institutions’, Journal of Development Studies, 30 (3), pp. 533–77, ‘Part II: The Role of Moral Norms’, Journal of Development Studies, 30 (4), pp. 386–422. —— (2000) Institutions, Social Norms, and Economic Development, London: Routledge. —— (2009) ‘Institutional Obstacles to African Economic Development: State, Ethnicity, and Custom’, Journal of Economic Behavior and Organization, 71 (3), pp. 669–89. Rao, V. and Walton, M. (2004) ‘Conclusion: Implications of a Cultural Lens for Public Policy and Development Thought’, in Rao, V. and Walton, M. (eds) Culture and Public Action, Stanford: Stanford University Press, pp. 359–72. Roland, G. (2004) ‘Institutional Change: Fast-­Moving and Slow-­Moving Institutions’, Studies in Comparative International Development, 38 (4), pp. 109–31. Synnott, H. (2009) Transforming Pakistan – Ways out of Instability. A Publication of the International Institute for Strategic Studies (IISS), New York: Routledge. Van Nierop, H. (2001) ‘Alva’s Throne – Making Sense of the Revolt of the Netherlands’, in Darby, G. (ed.) The Origins and Development of the Dutch Revolt, London and New York: Routledge, pp. 29–47.

Index

Abidjan 180–1 Abraham, A. 253 Acemoglu, D. 143–4, 217 Adelkhah, F. 67, 79 adivasis 196–8, 201–2, 204–6, 215, 217, 219, 224–5, 228, 237 Adler, N.J. 42 advantage possibility frontier 219 affirmative action 125, 225 Afghanistan 248 Africa(n): 68, 73–4, 107, 143–5, 150–7, 175–6, 181–7, 251, 253, 255; central 120, 154; East 120; West 154; Americans 209 African-American deprivation 200, 209 Afro-Brazilian(s) 183 Afro-Latino groups 199 Agence Française de Développement (AFD) 58n23, 59n31 Agra 38 Ahidjo (Président) 178 Akans 115 Akerlof, G. 9, 207–8, 213 Akresh, R. 117 Al Azhar university 105 Aldashev, G. 248 Aleppo 93 Alesina, A. 109–11, 121n1 Alexander, P. 5 Algeria 4, 45, 67 alienation 68 Allah 66 altruism, altruistic 108, 118–19 altruist paradigm 242 Alva (Duke of) 249 Ambedkar 138n5, 223 Ambrus, A. 113 American: ideal 161; model 59n29; NGO, 177; society 161 Americans 90, 230n21

Amin, S. 24 amoral familism 119 Amselle, J.L. 71 Anatolia 67 ancien régime (France) 162 Anderson, S. 113 Andhra Pradesh 214 Anglo-Saxon 161 Anherer 180 Ankara 67 ankyeni 189n5 Anomie 68 Anthropological Survey of India 130 Aoki, M. 8, 10 Appadurai, A. 4, 29, 201, 209, 220, 239 Arab: nation 98; scientists 89; victories 90; world 174 Arabic 59n29, 68, 90 Arabs 90 Argentina 165 Arizpe, L. 199 Armenians 156 Artois (province of) 249 Aryanism 72 asabiyya 69 Asia 107; Asian businessmen 153; Central and East 184 Atatürk 67 Athos, A. 42 Attewel, P. 206 Audrain, X. 78 Austin, J.L. 44 authoritarian: rule 105; style 184 autonomy 41, 48, 53, 55 Backward Castes 38 Backward Classes Commission 130, 136; Movement 138n11 Badini, J. 5, 6 Baghdad 175

260   Index Bahia (Brazil) 6 Bahujan Samaj Party 133 Baland, J.M. 114, 229n3, 251 Baldwin, R. 3 Balkans 90 Ballantyne, T. 72 Baluchi 70 Bamenda 189n6 Bamiléké 73 Ban, R. 214, 229n8 Banerjee, A. 131, 215, 224, 230n17 Banfield, E.C. 23, 119 Bangladesh 113, 116, 120, 250 bankruptcy laws 32 Bardhan, P. 7, 10, 11, 12 Barr, A. 153 Bastide (Roger) 183 Basu, K. 9, 11, 57n3, 206, 208 Bates, R.H. 71, 120 Bauer, P. 3, 4 Bauman, Z. 21 Bayart, J.F. 13, 14, 16, 50, 59n30, 59n32, 65–7, 72, 74–6, 79, 172, 187–8, 247 Bayly, C.A. 70 Bazin, L. 172, 180, 182, 241–2 Beaujard, P.H. 174 Becattini, G. 37 Beck, T. 144 Becker, G. 201 behavioural rule 8 Belgium 249 beliefs : 39; belief systems 248 Benin 78, 176 Ben-Porath, Y. 121n6 Bentolila, A. 117 bequests strategic 121n4 Berckowitz, D. 36 Berman, E. 119 Bernheim, D. 121n4 Bernstein, L. 147 Bertrand, M. 120, 224 Bertrand, R. 72, 75 Besley, T. 118 Bessis, S. 26 best practice 41, 44, 49 Béteille, A. 127, 139n14 Bhabha, H. 25 Bhuyias 133 Bihar 131, 135, 139n20, 217; Legislative Assembly 132; politics 133 bilateral inheritance system 115 Bill of change 146 Bing, W. 66

Binswanger, H. 250 birth control 69 Bisin, A. 108–9 Biya (President) 178, 189n6 black box 242 Blair, H.W. 133 Blanc, A. 57n1 Bliss, C.J. 28 Bloch, F. 113 Bloch, M. 45, 179 blood relationships 107 blood: line 114; ties 116 Bollinger, D. 171 Boltanski, L. 79 Bond, M. 43 Botticini, M. 121n2 Bourdieu, P. 4, 29, 202, 209, 241 Bourguignon, F. 213, 224, 230n18 Bowles, S. 9, 10 Boyacigiller, N. 42 Brabant (province) 249 Brahmins 133, 204, 211, 229n5; antiBrahmin 230n12 Brass, P. 72 Braudel, F. 76, 186 Brazil 36, 88; Brazilian Amazonia 240 Breman, J. 4, 5 Breton 181 bridal family 113 bride 111; burning 113 Briggs, J. 240 British 53, 71–4, 125–6, 132, 199; monarchy 173 brotherhood 78–9, 166 Brown, P. 68 Brubaker, R. 197, 241 Bruton, H. 3 Buckley, P.J. 39 Buddhist Thailand 216 Bureaucratic: impersonality 47; system 58n25 Burgundian state 249 Burkina Faso 117, 175 Burma 66 Burundi 72 business culture 147 Calvinist ethic 57n6 Cameroon 47, 51, 69, 73, 76, 163, 166, 176–8, 189n6 Canada 166 capability sets 218 capacity to aspire 29, 202, 221, 239 capitalism 26, 33

Index   261 caravan routes 97 care-intensive activities 250 Carré, O. 69 Caselli, F. 119 Casson, M. 37, 39 Caste system 129; structure 130; low 130; status 211 Caste-based contests 125; caste classifications 256 Castel, R. 167 Castoriadis, C. 66 Catholics 108 Caucasus 70, 249 Central Africa 186 Central Asia 70, 76, 184 Chad 74 Chamars 133, 211 Chandra, K. 214–15, 217 Chang, H.J. 35 charitable mission 97 charity: principle of 241–2 Charles V 249 Chauvet case 173 Chechnya 249 Cheddadi, A. 69 Chehabi, H.E. 69 Chiapello, E. 79 child fostering 117 China 36, 43, 53, 66, 69, 70, 74, 156, 174, 184 Chinese 68 Chrétien, J.P. 71 Christian culture 216 Christianity 68, 72, 89 Christians European 90 Church 69 civil society 97–8, 103 clan 119, 166 Clegg, S.R. 43 Clifford, J. 65 Clignet, R. 185 coalition equilibrium-breaking 213 Coate, S. 117–18 cocoa tree planting 121n5 co-evolution (of culture and economic development) 4 Coleman, W.J. 119 collateral (loan) 144 collective action 5, 35, 69 collective opinion 6 collective programming 38, 42 collectivism 53 colonialism 24–5, 71–5, 77, 137n3, 182, 184

Colson, E. 71 Comaroff, J. 73 commitment 37 common property 5 communalism 72 Communist 66, 218 communitarian societies 41 compartmentalization principle 180 confessionalism 69 Confucianism 43, 57n5, 66, 216 Congo-Brazzaville 68; Zaire 185 Congolese culture 68 Congo-Zaire 68 Congress Party 215, 217 consensus-seeking logic 167 Constituent Assembly 127 context-specificity 56 contract law 104; breach of 146–7; compliance 146; enforcement 147 conventions 8 Cooper, F. 241 cooperation, cooperative practices 5, 30, 38, 46, 51, 117–18, 165, 211 coordination 8, 9, 37 Copenhagen 67 Copernician revolution 65 Cordell, J. 6 core-periphery 27 corporate governance 32 corporations 97–8 corruption 166, 178, 185, 254 cost–benefit calculations 8 Côte d’Ivoire 74, 78, 180 court procedures 104 court proceedings 144 Cox, D. 121n1 craft guilds 97 craftsman 48 credit rating agencies 146 Cromer (Lord) 89 Cromer, E.B. 100n6 Crooke, W. 133, 137n2 cross-cultural comparisons 21 cross-cultural management 171 Crozier, M. 47, 58n25 cultural: attributes 8; capital 4; change 24; complexity 43; determinism 216; distance 239; environment 49; inertia 23; interactions 27; norms 8; relativism 242; representations 8; shifts 213; traits 21; transformation 217; transmission 111 custom 10, 32 customary norms 115

262   Index d’Iribarne, P. 13, 15, 35, 43–7, 49, 53, 55, 57n1, 59n31, 189n8, 255 Dagestan 249 dalits 196–8, 204, 206–12, 219, 221–5, 228, 230n24, 235–7 Damascus 100n19 Darwin 10 Das, T.K. 39 Das, V. 138n12 Dasgupta, P. 39 Davis, D. 69 De Boeck, F. 68 de Bony, J. 46 de Certeau, M. 76 De Rozario, P. 33, 46 de Sardan, O. 13, 240, 242 de Tocqueville, A. 44–5, 161 de Vries, J. 77 Deaton, A. 230n10 Debray, R. 59n30 decentralized despotism 73–4 Dei Ottati, G. 37 Deininger, K. 144 Deliège R. 137, 139n17, 189n3 democratic liberties 96 Demsetz, H. 251 dependency theory 24, 80 Depressed Classes 129 Deshpande, A. 224, 229n5 Devarajan, S. 236 development path 221 developmentalist configuration 240 Dhobis 133 Dimitrievic, D. 178 Dinkas 254 discrimination 7, 29, 125–37, 208; taste for 208 disempowerment 29, 201 distributional shifts 204 Dobbin, F. 167 dolat 79 Doms 133 Douala 176 Douglas, M. 4, 28, 30, 35, 56, 160, 161 dowry payments 111, 121n2; prohibition act 113 Dozon, J.P. 71 Drèze, J. 130 Drucker, P.F. 42 Du Bois, W.E.B. 137 Duflo, C. 230n17 Dumont, L. 4, 173, 189n3 Durkheim, E. 57n4, 171, 189n3 Dusadhs 133

Dushkin, L. 126 Dworkin 218 dynastic links 118–19 Dzakpasu, R. 58n21 East Asia 184; East Asian miracle 216 eBay feedback 146 Eccles, R. 37 economic rationality 8 Edlund, L. 113, 115 educational deprivation 223 Egypt 67 Eisenstadt, S.N. 23 elective affinity 230n16 Ellis, S. 172 Elyachar, J. 74 embeddedness 79 emotions 10, 41 empathy 108 empowerment policies 160 Engerman, S. 217 England 251 English 133, 213 Enke, S. 3 enlightenment 21, 24, 59n30 enrolment rates 131 Ensminger, J. 150 entrepreneurialism 38 entrepreneurship 36 equal division 114 equal opportunity outcome 220; approach 221 equal protection clause 128 equality of agency 30–1, 196, 220, 225 equality of opportunity 222–3, 225 equilibrium: configurations 10; full information 207; inferior 147; multiple 8, 147; political 217; unequal 207–8 equitable: development 221; policy 229 Escobar, A. 4, 25–6, 241 Ethiopia 145 ethnic: groups 27, 70–1, 107–8, 116, 119, 181, 255; culture 183–4; fragmentation 235 ethnicity 38, 69, 71, 153, 255 ethnoculturalism 74 Etzioni, A. 36 Eurasia 252 Europe 96, 108, 113, 166–7, 173, 208 European Union 67 Europeans 183 evolutionary: games 10; trajectories 10 exchange: personal, impersonal 88, 92–5, 99

Index   263 expectations 8, 9 extended families 116 Exterior Castes 129 external economies 10 external exclusion mechanisms 242 externalities 24, 251; network 156; negative 222 Eyadema 74 face-to-face relationships 37, 94 Fafchamps, M. 13, 15, 40, 117, 120, 121n1, 143, 145, 147–8, 150, 153, 255 family, family structures 48, 107–21; networks 117; see also nuclear, extended farm-cum-family structures 251 Farmer, R.N. 41 fatalism 57, 87–8, 95 Fauré, Y.A. 172 Fauvelle-Aymar, F.X. 176 favouritism 47 Feder, G. 251 feedback mechanism 9 Feeny, D. 251 female infanticide 113 female schooling 113 Ferguson, J. 25, 73 Fernandez, R. 108 Field, E. 113 financial development 93 Fischer, D.H. 100n10 fishery, fishing grounds 6 Fisman, R. 120 Flanders (province of) 249 Florence 93 focal equilibrium 8 Foner, E. 45, 161, 167 forward-looking: agents 202; parents 113 Foster, A. 117 fostering see child Foucault 70 Foucault, M. 25 France 45, 51, 53, 59n29, 59n30, 69, 111, 165, 167–8, 177, 181 franchising 146 Franke, R.H. 43 Freeman, R.B. 33 friendships ties 185 Fukuyama, F. 23, 37, 41, 144, 247 functionalist 70 Galanter, M. 127, 134, 136, 138n11 Galeotti, A. 155 Gambetta, D. 38, 119

game: theory 7; sequential exchange 211 Gandhi (Mahatma) 127 Geertz, C. 4, 45 Gemeinschaft 22, 160 Gemeinsinn 69 general equilibrium theory 7 generalized reciprocity 118 genos 69 Germany 111, 168 Geschiere, P. 68–9, 71, 74, 76 Gesellschaft 22, 160 Ghana 115, 118, 121n5, 216 Giddens, A. 28, 161 Giuliano, P. 108–11 global economy 105 globalization 20, 26–30, 43, 50, 68, 76, 175, 179, 188 Goitein, S.D. 100n16 Gold Coast 78 Goregaon 138n5 Government of India Act 127 Grameen Bank 120 Gramsci 72 Gramscian view 209 Granovetter, M. 37, 79, 120, 147 grassroot organisation 253 Greece 69, 111 Greif A. 8, 100n4, 100n12, 116, 144, 148 Griaule, M. 72 Grimard, F. 117 Groningen (province of) 249 groom 111 group-based: identities 256; targeting 235–6 group-making projects 221 group-targeted: preferential policies 237; redistribution 236; transfers 236 Gubert, F. 117 Guetto (black): 209; urban 209 Guha, R. 127, 217 guilds 97 Guiso, L. 23, 100n8, 109 Gujarati business groups 230n12 Gujars, Gujar community 125, 137n3 Gujjiar group 225, 230n25 Gulliver, P.H. 69 Guyer, J. 182, 187 Haastrup, K. 28 habits 4 habitus 201, 209 Habyarimana, J. 237n1 Haggis, J. 22, 57n3 Hainault (province of) 249

264   Index Hall, S. 25–6 Hamam 94 Hampden-Turner, C. 43 Harbison, F. 41 Harragin, S. 253–4 Harrell, S. 69 Harries, P. 71 Harrison, L. 4, 12, 23, 30, 195, 216 Havard, J.F. 78 Hayami, K. 250–1 headscarf 67 Hefner, R.W. 27 Heller, P. 217 Henry, A. 45, 47, 55, 59n31, 163, 169n2, 182, 185 Hibou, B. 74 hierarchies, hierarchical relationships 29, 45, 162, 173, 185, 202, 230n24 Higgins, B. 3 Himbara, D. 150 Hindi Speaking Belt 131 Hindu equilibrium 216 Hindu rate of growth 216 Hindu(s) 28, 70, 126, 132, 206–7, 218 Hirschman, A. 3, 4 history, stationary and cumulative 240 Hobsbawm, E. 73, 173,, 178, 187 Hoff, K. 209, 210–11 Hofstede, G. 38, 42–3, 49, 58n20, 171 Holden, N. 43 Holland 249 honour 37, 48, 53, 167 household formation 120 Hsu, M.Y. 66 Humphrey, J. 40, 57n1 Hungarian society 248 Hunt, L. 65, 73 Huntington, S.P. 4, 12, 21, 23, 30, 195, 216, 247 Hutton, J.H. 133, 137n2, 138n6, 139n17–18 Hutu 70 Ichino, A. 117 ideal society 162 identity 8, 28, 57n9, 65, 67, 71, 205–6, 215, 241; caste 214; cultural 235, 256; group 215, 217, 256; inferior 208; spoiled 209 ideologies 33, 58n19 Ife 176 ijtihad 91, 105 Iliffe, J. 72 Imperialism 80, 248

increasing returns 10 India 70, 74, 113, 125–37, 156, 195–231, 235–7, 241; Victorian 173 Indian: constitution 126–7, 198; government 256 Indirect Rule 73 individualism 43, 53 Indonesia 216 industrial clusters 40 Industrial Revolution 150, 152, 172 inequality 4, 25–6, 29, 56 inequality of opportunity 214 inequality trap 195–231, 235; culturally shaped 212 inertia: generalized 95; 248 infanticide 113 inferior identity 208 information-based approach 208 Inglehart, R. 100n1–2 Ingushetia 249 inheritance: Islamic 96, 98; patterns 115; practices 104; rules 114 see also bilateral; matrilineal; patrilineal institutional reform 104 institutional: design 217; imperialism 248 institutions: fast-moving, slow-moving 247, 252 insurance: informal 120; mutual 117 inter-vivos transfers 115 intermarriage patterns 108 intermediaries impersonal 96 internalization of externalities 251 International Monetary Fund 32 Iran 67, 69, 70, 79, 104, 250 Iraq 69, 248 Irschick, E.F. 72 Isar 180 Islam 23, 57n7, 72, 87–99, 174 Islamic: partnership 95–6; law 96, 185; inheritance 96, 98; thought 105; civilization 104; economics 103 Islamism (post) 104 ISO certification 153 Istanbul 67 Italy 107, 111, 119, 181 Ivan the Terrible 249 Ivoirization 182 Ivorian 181–2, 184 Jackson, M. 155 Jacobinism 68 Jaeger, A.M. 44 Jaffrelot, C. 70, 72, 134 Jains 174–5

Index   265 Jalan, J. 230n14 Jameson, F. 65 Janelli, R.L. 172, 183–4 Janos, A. 248 Japan 41, 53, 66, 174 Japanese colonization 184; culture 216 jatis 197, 213–14, 229n5 Jensen, R. 113 Jewish merchants 116 Jews 108; ultra-orthodox 119 Jharkand 133, 139n20 job quotas 131 Jodhika, S.S. 206 Kabylia 4 Kadyrov 249 Kalberg, S. 75 Kaldor–Hicks efficiency 220 Kanungo, R.N. 44 Karnataka 214 Karzai (Hamid) 248 Kayasthas 133 Keefer, P. 143, 236–7 Kenya 74, 145, 252 Kerala 214, 217, 223 Kernen, A. 66 Kerr, C. 58n19 Keshavarzian, A. 69, 78–9 Khaldun (Ibn) 69 Khemani, S. 13, 16, 236–7, 237n1, 256 Kiggundu, M.N. 44 Kikuchi, M. 251 Kikuyu 73, 174 Kimball, M. 117 Kin groups 107 Kinshis 66, 68–9, 73, 76, 114, 116–20, 184–5, 254–5 Klein, M. 79 Kleinfeld, R. 58n15 Kluckhohn, C. 58n20 Knack, S. 143 Knorringa, P. 38 knowledge gap 156 Konrad, K. 120 Kopytoff, I. 67, 69, 182 Korea 66 Korean: case 184; conglomerates 172, 183; culture 183 Koutiala (district of) 251 Kroeber, A.L. 58n20 Kshatriyas 229n5 Kumar (Nitish) 21 Kumar, D. 138n12 Kuper, A. 57n13

Kuran, T. 9, 13–14, 30, 57n7–8 Kurds 70 Kurmis 131 Kuznets, S. 3 La Ferrara, E. 13–14, 251 Labour unions 98 Lagerlof, N. 113 Lake Chad 185–6 Lal, D. 216 Landes, D. 4, 247 land-saving and labour-using techniques 250 Larrain, J. 24 Latin America 107, 199 Law of corporations 98 law: pro-equity 217 League of Nations 70 Lebanese businessmen 153 Lebanon 69 legal order 36 legal treatises 100n14 legal: Islamic 88, 98; systems 58n15, 100n5 Lemonnier, P. 240 Lenartowicz, T. 43 Leonard D.K. 44, 58n21 Leonardi, R. 38 Lev, Y. 100n20 level playing field 167 Lévi-Strauss, C. 55, 240 Lewis, B. 90–2, 100n9 Lewis, O. 199 Lewis (Oscar) 195, 212 Lewis, W.A. 4, 22, 57n8 Ligon, E. 117 Lithuania 111 Litvak, M. 69 Liu, X. 66 Locke, J. 161 Lokur Committee 136 London 93, 176 Lorenz curves 131 Loury, G.C. 139n16, 200, 208–9, 230n13, 230n21 Low Countries 249 Luhmann, N. 40–1, 58n18 Lybia 67 M’Bokoki, E. 71 MacClelland, D.C. 22–3, 100n8, 160 MacGaffey, J. 185–6 MacIntire, J. 250 MacKean, M. 6

266   Index macLuhan 189n1 MacMillan, J. 148 Madras Presidency 126, 139n18 mafia 119 Maghribi traders 116 Mahieu, S. 13, 16 Maka 69 Makram-Ebeid, M. 13–14 Malaysia 67, 74, 216 Mali 251 Malkki, L.H. 71 Mallah caste 133 malleable culture 215, 247 Malthusian 76 Mamdani, M. 73 Manacorda, M. 111 management: models 41, 44, 48, 51, 58n21; practices 161–8 Mandal commission 128 Mankon 176–7 Mankon Cultural Development Association 176; museum 189n4 Maoist 217 Marabout 78 Marathi 213–14 market 7, 10, 33; conditions 148; embedded 147, 150; exchange 144, 150, 153; institutions 143–55; interactions 155; secondary 153 marriage payments 113 Marwari business groups 230n12 Marx, Marxism, Marxists 24, 65, 68, 171, 187 material culture 66 matrilineal: inheritance 114; institutions 121n5; norms 115; societies 115 Mauritius 250 Mauss, M. 66 Maximin criterion 219 meaning systems 28, 49, 56 Mecca 89 Medieval: Europe 111–14, 156; legal treatises 100n14; Middle East 95 Medina 98 Mediterranean 92, 116 Mehta, P.B. 214 Meier, G. 3 Meiji (post-meiji) 66 Meisel, N. 33 Mellat 79 Mendelsohn, O. 130 Mengin, F. 66 Menkhoff, T. 40 mental images 162

mental models 33 merchant community 96 Merdonite 173 Messick, R.E. 144 Metcalf, T.R. 71, 74 Mexico 165, 199 micro-lending 37 Middle Ages 93, 95, 99 Middle East 87–99 Minneapolis 176 Minnikhanov 249 Minten, B. 148 Moaddel, M. 100n2 Mobutu 74 modernity 20–2, 27, 31, 55 modernization 97 modernization theory 20–6 Moe, T. 34 Mollering, G. 39 Montesquieu 171–2 Montesquieun, C. 45 Montparnasse 181 Moretti, E. 111 Morocco 51, 145, 165, 167 mosque 89, 97 Muamalat 93 Mueggler, E. 66 multiculturalism 76 Mumbai 175, 213 Mundas 133 Mundimbe, V.Y. 26 Munshi, K. 120, 213 Murgai, R. 117, 230n14 Murid brotherhood 78 Muslims 38, 57n8, 70, 74, 126, 206 mutual aid 47 mutual insurance (help) 117, 166 Myers, C.A. 41 Myerson, R. 8 Mysore 126 Nairobi 174, 252 Namur (province of) 249 Nath, R. 41 nation-state 43 natural resources 5 natural selection 10 Navajo 175 Naxalite 217 Ndi (John Fru) 189n6 Needham, R. 68 Negro race 137 Netherlands 111 networks: externalities 156, 213; insurance

Index   267 120; production 38; social 32, 36–7, 41, 79 new institutional economics 35 New York 176, 181 Newman, K. 206, 224 NGO 74, 253 Nohria, N. 37 Noland, M. 100n8 norms (social) 4, 8, 9, 33, 39, 107, 120, 147, 155, 196, 247, 257; cooperative 109; customary 115; of reciprocity 37 Noronha, R. 251 North Africa 116 North Atlantic 172 North, D. 10, 32–4, 100n12, 144, 155, 160–1, 242 Northern Europe 166 Notué, J.P. 189n4 nuclear families 107, 115–16 Nuer 254 Obbo 185 occupational structure 130 Oduro, A. 153 OECD countries 111 Olson, M. 34 Ong, A. 66, 69 open-access social orders 34 opportunism 7, 39, 144, 165 opprobrium 5 Oraons 133 organizational theory 43, 46, 49 Orientalist 73, 77 Orissa 132, 135–9n20 Ortner, S.B. 65 Other Backward Classes 128, 139n21 Otsuka, K. 115, 250 Ottoman: Turkey 89; empire 156 Ouchi, W.G. 37–8, 42, 162 Pakistan 257 panchayats 236 Pande, R. 131, 224 Pandey, K. 210 Papua New Guinea 240 Parsons, T. 21, 58n20 participatory: budgeting 30; democracy 172 particularistic cultures 165 partnership: commercial 96; Islamic 95–6 Pascale, R.T. 42 Pasi 211 path dependence 10, 33 patrilineal inheritance 114

patronage 5; democracy 214, 217, 217, 225 Paul-Lévy, F. 240 Peel, J.D.Y. 72 Perez-Gonzalez, F. 120 Perrot, C.H. 176 Persian influence 89 Persians 90 personalism 93 personalistic attitudes 95 personalized relationships 46; see also exchange Persson, T. 237n1 Pezzin, L. 120 philia 69 Philip II 249 Phratry 69 Piketty, T. 230n27 Pingali, P. 250 Piore, M. 38 Piot, C. 67, 69, 71 planner (social) 220 plasticity of identity 213–15 Platteau, J.P. 7, 13, 38, 46, 100n12, 114, 164, 229n3, 250–1, 253, 255 Plissart, M.F. 68 Polignac, F. 69 politeness code 252–3 political mobilization 236 political policies 127, 223; group-targeted 237 political reservations 223 politics of identity 65 Portugal 111 Posner, R.A. 119 post-modernism 6, 28, 75 Pouillon, J. 178 poverty alleviation 31 poverty trap 196–231, 230n10, 256 power: inequalities 29; interaction with culture 205; relations 173; relationships 50, 66; structures 201–2; unequal 212 Powers, D.S. 100n20 preference (culturally shaped) 200 preference for children 108 preferences 10 preferential rights 254 primitivist ideology 240 primogeniture 114 principal-agent problems 34 production networks 38 property rights 32, 51 Prophet 94

268   Index Protestant: culture 216; Reformation 249 Proudhon 187 Prunier, G. 71 psychoanalysis 65 psychological games 10 Putnam, R.D. 36, 58n16, 73, 119, 147 Quisumbing, A. 115, 120, 121n5 quotas 138n11; job 131 Qur’an 89, 105 Rabin, M. 10 racial: dishonour 209; inequality 139n16; stigma 209, 230n13 Radcliffe-Brown 189n7 Radhakrishnan, P. 138n11 Rahman, M. 115 Raj (British) 71–4 Rajasthan 125, 130 Ranger, T. 73, 173, 187 Rao, V. 4, 11–12, 29, 57n12, 113, 203, 214, 220, 229n1, 229n4, 229n8, 254 rational economic activity 33; behaviour 220; bureaucratic principles 36; submission 105 rationality: economic 8, 35; limited 9 rationalizations 10 Ravallion, M. 117 Rawls 218–19 Ray, I. 7, 11–12, 57n3 RDPC 178 reciprocity 117 Redaelli, S. 115 Redding, S.G. 43 religions 23, 38, 67, 76; clericalized 70 religiosity 100n3 religious: authorities 69; conversion 255; institutions 68; minority 126; seminaries 119 Renan, E. 89, 100n7 rent-seeking (patronage-based) 217 repertoires (of behaviours, practices) 209 repertories (cultural) 59n33, 72 reputation (s): 6; mechanisms 146, 148 reservation of seats 138n8 residual claimants 251 Respeito 6 Richman, B.M. 41 risk sharing 117 Rist, G. 27 rituals 39, 120, 176 Robertson, R. 27 Robinson, D. 79 Rocca, J.L. 66

Rodinson, M. 57n7–8, 100n8 Rodrigues, V. 138n5 Rodrik, D. 29, 33, 35 Roemer, J.E. 219, 227 Roitman, J. 185–6 Roland, G. 247, 250 Roman Law tradition 144 Rosaldo, R. 67 Rosenzweig, M. 117, 213, 250 Rostow, W. 3, 22 Roth, K. 43 Roy, O. 69, 71, 76 Rudolph, L. 214, 230n12 Rudolph, S.H. 214, 230n12 rule of law 35 Russia(n) 70, 248 Rwanda 72–3 Sabel, C.F. 38 Sahara 74 Sahvesan 70 Said, E.W. 25, 57n7 Sako, M. 37 Salzman, P.C. 70 San (district of) 251 sanctions: self-reinforcing 9 Sandholm, W.H. 100n3 Santhals 133 Sarakhsi (shamseddin al-) 94, 99 Sarakhsi, S. 94–5, 99 Sardars 206 Saudi Arabia 67, 104, 145 Scandinavian countries 107, 111 Schatzberg, M.G. 76 Schech, S. 22, 57n3 Scheduled Caste and Scheduled Tribe Act 209 Scheduled: Tribes 125, 127–39; Castes 127–39, 196 Schildkrout, E. 69 Schmitz, H. 40 Schoar, A. 120 Schone, B. 120 Scotland 173 Scott, J. 4, 5, 209 Searing, J.F. 78–9 Sebsi 182 segmentation of market exchange 150; social 155 self-enforcing equilibrium 215, 225 self-esteem 7 self-fulfilling belief systems 208 self-help groups 225, 252 self-interest 5, 39

Index   269 Sen, A.K. 29, 30–1, 56, 130, 197, 216, 218 Senegalese 73, 76, 79, 181 Senegambia 79 Serer 78 Seurat, M. 69 Sforzi, F. 37 Shah, P.G. 138n10 Shaimiev 249 Shiism 69 Shiites 175 shudras 132, 229n5 Sicilian mafia 119 Sikasso (district of) 251 Sikh 206 Singh, K.S. 130 Siow, A. 121n2 slavery 59n27, 69 social: alliances 68; capital 58n16, 119; connectedness 121; exchange game 5; identities 130; inequalities 136; interactions 126; networks 32, 26–9, 50, 147, 152; organization 29; prestige 7; safety nets 116; segmentation, fragmentation 155, 236; shame 5, 7; uniformity 214 socialization choices 108–9 Soderberg, A.M. 43 Sokoloff, K. 217 solidarity: mechanical 57n4; networks 166, 183; organic 57n4 Somanathan, R. 13–14, 131, 215, 224, 229n6, 230n20, 230n25, 256 South Asia 111 South Korea 180, 183–4, 216, 250 Southern Europe 216 Soviet: Empire 76; threat 90; era 249; Union 249 Spain 90, 107, 111, 249 Spanish: crown 249; Habsburgs 249; troops 249 spillovers 10 splitting (of household) 251 Srinivas, M.N. 130, 137 Stache, F. 39 Stalinist 71 state-building process 249 stem household 251 stereotyped behaviour 50 stereotypes 205–13, 225, 230n13 Stiglitz, J.E. 38 stigma racial 129 stochastic dominance (first – and second – order) 230n29 Stock markets 96

Storper, M. 35 strategic syncretism 72 structural adjustment 71, 77 structuralism 24 structure 57n10 subordinate groups 209, 212, 216, 222, 225 Sub-Saharan Africa 77, 117, 143, 251, 253, 255 Sudan 253–4 Sundaram, K. 136 Sunnis 175 Superintendent of Census Operations (India) 139n18 Supreme Court of India 128 Swartz, D.L. 202 symbolic production 65; benefits 224; meaning 197 Synnott, H. 257 Syro-Lebanese 181 Szekfu 248 Tabellini, G. 109 Tadelis, S. 146 Taiwan 36, 66 Talibé 78 Tamil Nadu 138n11, 214 Tanzania 145 Tapper, R. 70 taste for discrimination 208 Tatarstan 248–9 Tayeb, M. 43 Teasong (conglomerate) 183 Tehran 250 Teng, B.S. 39 Tenure rules 251 terms of recognition 221 Terroirs 78 Texas 176 Thailand 66, 70, 88 Thapar, R. 70 Thomas, N. 189n7, 240 Thompson, E.P. 65 Thorat, S. 198, 206 Thornton, K.R. 113 Togo 74 Tönnies, F. 22, 160, 189n3 Tozy, M. 74 tradition 51–4; ethnic 182; loss of 188; monarchical 177 transaction costs 32 Triaca, B. 189n4 tribalist 74 Tripoli 69

270   Index Troeltsch, E. 75 Trompenaars, F. 43 trust 37–40, 58n17–18, 97; high-trust, lowtrust societies 41 trustworthy relationships 40, 109 Tunis 93 Tunisia 47–8, 67 Turkey 66, 70, 88 Turkish conquests 90 Turks 90 Turner, B.S. 100n8 Tutsi 71 Udovitch, A.L. 100n11, 100n13–15 Udovitech 92–3 Udry, C. 118 Uganda 145 UNC 178 unequal cultural structures 220 unequal economic opportunity 225 UNESCO 180 United Kingdom 111 United Nations 253 United States 41, 45, 48, 53, 67, 108, 111, 128–9, 162, 165, 175, 200, 208, 230, 13 universal provision of public goods 236 universal suffrage 223 untouchability 127, 138n5, 198 untouchables 126, 129 Upper caste Hindu 207; Brahmin and Thakur 211, 235; groups 212; kids 210 US Supreme Court 139n15 USSR 70, 76 utilitarian 41, 70 utilities 7 Utrecht (province of) 249 Uttar Pradesh 137, 139n19, 209 Vail, L. 71 Vaishyas 229n5 values 8, 31, 33, 36, 38, 45, 55 van Bruinessen, M. 70 van Nierop, H. 249 Varnas 229n5 Vega-Redondo, F. 155 Verdier, T. 108–9, 213 Veyne, P. 66 Vicziany, M. 130 Vietnam 145

Vijayan, P.P. 138n9 Vovelle, M. 71 Waaf 78, 97, 101n21 Wales 173 Walker, C.E. 73, 77 Wallerstein, I. 24 Walloon provinces 249 Walton, M. 4, 11, 12, 13, 16, 29, 57n12, 220, 229n1, 229n4, 247, 254, 256 Wane, W. 237n1 Warnier, J.P. 13, 15, 16, 66, 69, 73, 186 water-access 6 Weber, M. 22–3, 57n6, 72, 74, 100n8, 107–8, 160, 216 Wedeman, A.H. 66 Weick, K.E. 43 Weisskopf, T.E. 139n14 West Africa 181, 186 Western Sumatra 115 Western: countries 247–9, 251; Europe 48, 92–3 Whitley, R.D. 43 wife beating 113 Williamson, O.E. 39 Winick, C. 107 witchcraft 69 within-group: distribution 221; inequality 236 Wolof, Wolofization 78 Womack, P.J. 44, 49 Woodruff, C. 148 World Bank 37 World Values Survey 100n2, 109 Yadavs 131, 137n2 Yamey, B.S. 3, 4 Yeshiva 120 Young, P. 8, 9, 10, 155 Yousfi, H. 12, 46–7, 52, 165, 247, 252, 255 Zaire 74, 185 Zaria 176 Zeelan 249 zero-sum game 76 Zootechnical Research Institute 177 Zoufri 48 Zuker, L.G. 38, 58n17