Raising Cane: The Political Economy of Sugar in Western India 0813312876, 0813314283, 9780367284978

Citation preview

Raising Cane

Raising Cane The Political Economy of Sugar in Western India Donald W. Attwood

~ ~ ~~o~;!:n~~:up LONDON AND NEW YORK

First published 1992 by Westview Press, Inc. Published 2019 by Routledge 52 Vanderbilt Avenue, New York, NY 10017 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business Copyright© 1992 Taylor & Francis All rights reserved. No part of this book may be reprinted or reproduced or utilised 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. Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Attwood, Donald W. Raising cane : the political economy of sugar in western India / Donald W. Attwood. p. cm. Includes bibliographical references and index. ISBN 0-8133-1287-6. 0-8133-1428-3 (pbk.) 1. Sugar trade-India-Malegaon. I. Title. HD9116.I415M393 1992 338. l '7361'0954792-dc20 ISBN 13: 978-0-367-28497-8 (hbk)

91-17330 CIP

To family on both sides of the world

Contents List of Tables and Figures Acknowledgments

Xl XV

PART ONE INTRODUCTION l

The Problem

3

An Anomalous Industry, 4 Nationalism and Dependency Theory, 6 Dependent India? 10 Agrarian Regions in India, ll Political Economy, 14 Enterprising Peasant Families, 15 Looking Ahead, 18 Notes, 20

2

The Setting

22

Malegaon Village c. 1885, 23 Rural Life Before the British Raj, 29 Stratification, 31 Risk and Mobility, 33 Competition and Mobility, 36 Change in the 19th Century, 39 Conclusion, 45 Notes, 46 PART TWO SUGAR PRODUCTION BEFORE INDEPENDENCE

3

Irrigation and Imperialism

49

Origin of the Deccan Canals, 50 The Problem of Unwanted Water, 51 Fiscal Pressures and Wasted Water, 58 The Sugarcane Block System, 60 Cane Blocks and Famine Protection, 62

vii

viii

Contents

Impact of Peasants on Policy, 65 Notes, 66

4

Peasants Versus Capitalists

68

North India: Origins of the Industry, 69 The Bombay Deccan: Cane and Gut Production, 74 The High Cost of Sugarcane, 77 Management of an Expensive Crop, 79 Relations of Production, 81 Origins of the Bombay Sugar Industry, 85 Reorganization from Below, 88 Conclusion, 89 Notes, 90 PART THREE MALEGAON VILLAGE, 1900-1950

S

Old Elites and New Entrepreneurs

95

New Ideas and Movements, 95 History of the ]agir, 100 Shambhusingh Jadhavrao, 102 The Raja and the Sugar Economy, 104 Keshavrao Taware, llO The Nira Valley Sugar Company, ll3 Sugar Magnates, ll4 Idealists and Other Innovators, liB Notes, 121

6

Migration and Economic Mobility

122

Migrants with a Leg Up, 124 Drought and Migration, 126 Hazards in the Family Cycle, 129 Family Cycles and Upward Mobility, 130 Migrant Dhangars: A Success Story, 134 Less Fortunate Immigrants, 137 Local Land-Gainers, 139 Mobility for the Landless, 142 Local Land-Losers, 144 Overall Mobility, 146 Notes, 151

7

The Pattern of Inequality and Mobility Data and Methods, 153

153

Contents

ix

Economic Mobility and Inequality, 154 Caste and Mobility, 164 The Great Depression, 170 Political Economy So Far, 177 Notes, 181 PART FOUR COOPERATIVE SUGAR, 1950-1985

8

The Politics of Sugar

187

Sugar Interests at the National Level, 188 Sugar Interests at the State Level, 190 The First Cooperative Factories, 193 Factory Organizers Near Baramati, 196 Cooperative Organization, 199 Evolution of Factory Politics, 202 Cooperatives and State Politics, 208 Notes, 214

9

Performance and Impact of a Sugar Co-op

216

Economic Rationale, 217 Growth of the Malegaon Factory, 218 Malegaon's Other Investments, 225 Investments by Other Sugar Co-ops, 228 Impact of the Malegaon Factory: Employment, 230 Indirect Employment Effects, 234 Diversification and Mobility, 236 Changes in Relations of Production, 240 Changes in Land Distribution, 243 Nonagricultural Incomes, 250 Poverty, Inequality, and Economic Expansion, 255 Notes, 257 10

Why Do Some Cooperatives Work? Regional Differences in Sugar Production, 260 Technical and Economic Efficiency, 264 The Basis for a Class Alliance, 270 Comparisons with Other Types of Co-ops, 274 Competition and Efficiency, 276 Competition and Innovation, 280 Cooperative Factories in the North, 283 Competition and Accountability, 285 Notes, 287

260

Contents

X

PART FIVE CONCLUSION

11

Revolution from the Middle

291

Grass-Roots Development, 292 The Rise of Progressive Peasants, 297 Oligarchy and Bureaucracy, 301 West Bengal, 303 Uttar Pradesh, 307 "Living in a Revolution," 310 Notes, 318

Appendix on Methods Glossary Bibliography About the Book and Author Index

321 332 335 350 352

Tables and Figures TIJbles

2.1 Distribution of sample families by caste, Malegaon, c. 1900 2.2 Land distribution by caste and lineage, Malegaon, 1920 3.1 Millet production on dry and canal-irrigated plots 3.2 Subsistence crops under canal and well irrigation,

27

28

56

Ahmednagar district

57

Ahmednagar district

63

3.3 Foodgrain output in dry and canal-irrigated villages, 4.1 Land rental values along the Deccan canals 6.1 Distribution of sample families by caste and geographic origin

6.2 Caste and occupation of immigrants to Malegaon, c. 1900-1950 6.3 Economic mobility of landless immigrants to Malegaon 6.4 Caste and economic mobility among landless immigrants

75 125 127 131

to Malegaon

135

in Malegaon

139

families, Malegaon

140

6.5 Economic mobility of indigenous landless families 6.6 Caste and economic mobility among landless local 6.7 Origin and mobility of landless families, Malegaon, 1920-1950 6.8 Mobility between landed and landless strata, Malegaon, 1920-1950 6.9 Caste and mobility between landed and landless strata, 1920-1950 7.1 Correlation of initial with subsequent land distributions 7.2 Aggregate changes in land distribution, Malegaon, 1930-1950 7.3 Correlations between 1930 landholdings and transactions in the previous and following decades

xi

143 146 148 156 159 161

xu

7.4 7. 5

7.6

7. 7 7.8 7.9

Tables and Figures

Linear measures of association between 1930 landholdings and those in later decades Aggregate changes in landholdings by caste groups,

1930-1950

Cross-tabulation of caste and landownership categories, 1930 and 1950 Analysis of variance for landholdings, broken down by caste categories Analysis of variance for net changes in landholdings, broken down by caste categories Total land purchased and sold by Malegaon sample

9.1 Sources of initial and expansion capital, Malegaon factory 9.2 Shares owned by producer-members, Malegaon factory, 1969 9.3 Performance of Malegaon cooperative sugar factory, 1957-1958 to 1989-1990 9.4 Performance of local cluster of sugar factories, 1987-1988 9.5 Financial performance of local cluster of cooperatives, 1987-1988 9.6 Deductions from cane price paid by Malegaon factory, 1986-1987 9.7 Immigrants to Malegaon after 1950 9.8 Sample families with jobs in the cooperative sugar factory, 1980 9.9 Principal occupation of heads of households, Malegaon, 1970 9.10 Land-gainers and land-losers in Malegaon, 1950-1980 9.11 Correlation of landholdings in 1950 with landholdings in later decades

9.12 Changes in land distribution, Malegaon sample, 1950-1980 9.13 Cross-tabulation of caste and landownership categories, 1950 and 1980 9.14 Changes in land distribution by caste categories, 1950-1980 9.15 Analysis of variance for landholdings, broken down by caste categories

162 166 167 169 169 171 219 220 221 224 225 227 231 232 235 237 245 246 248 249 250

9.16 Distribution of business income by land-size categories, 1980 9.17 Distribution of business income by caste categories, 1980 9.18 Distribution of job income by land-size categories, 1980 9.19 Distribution of job income by caste categories, 1980

251 252 253 254

10.1 10.2 10.3 10.4 10.5 10.6

261 261 262 263 265 265

Number of sugar factories in operation, by state and sector White sugar production by sugar factories Utilization of sugarcane for different products, by state Production of sugarcane, by state and region Sugar recovery rates, by state and sector Total sugar losses by sugar factories

Tables and Figures

xm

10.7 Capacity utilization in the sugar industry, by state 10.8 Average cost of a bag of sugar, by state 10.9 Distribution of shares owned by members of 57 cooperative sugar factories in Maharashtra, 1977-1978

266 267 272

A.1 Data on caste, migration, and landholdings of Malegaon sample, 1900-1980

326

Figures 2.1 The Bombay presidency in the British period 2.2 Sketch map of Nira valley and vicinity

24 25

3.1

53

Crops under irrigation on the Nira left bank canal

4.1 Gut prices in the Bombay Deccan, 1897-1945 4.2 White sugar prices, 1895-1941

86 87

7.1 Scatterplot of Malegaon sample landholdings, 1950 vs. 1930 7.2 Hypothetical shift toward class polarization 7.3 Land distribution in Malegaon sample, 1930 and 1950

155 158 163

8.1

191

Maharashtra state in 1960

9.1 Scatterplot of Malegaon sample landholdings, 1980 vs. 1950 9.2 Land distribution in Malegaon sample, 1950 and 1980

244 245

Acknowledgments Like any book, this one is part of a dialogue. Over the years, I have asked thousands of questions, of myself and others, and tried to answer some. Out of all this discussion, a written pattern has grown. It is certainly not a definitive pattern. Among those whose words have been woven into it, there are many who might have fashioned it better. There are some who would have selected different colors and textures, or who might have preferred a totally different pattern. I am conscious of their voices and wish that I could adequately present them all. First and foremost are the voices of farmers and other villagers, whose experiences I have tried to understand and represent. A few of them will read this book and decide whether I learned anything from all their patient answers. If they were so inclined, they could tell more about the subject than I ever can. Equally important are the voices of the Indian intelligentsia: social scientists, journalists, writers, and technical experts of every description. Nothing puts an anthropologist on his toes like writing to a collection of experts about their own society. Anyone who approaches India is conscious of the depth of this challenge. Few societies on earth have such a large, highly trained, politically diverse, and openly critical intellectual community. It is both daunting and gratifying to work among them, and this book would have been impossible without the generous collaboration of many. Third, there are my colleagues and students in North America, who have challenged me to think and write more clearly on this subject. Their patient scepticism has been most rewarding. Luckily, I did not have to do this work alone. For two decades, I have had the good fortune to work with Professor B.S. Baviskar of the University of Delhi. I first met him when I was a Ph.D. student just starting my research, and his friendly guidance has helped me along ever since. During the last decade, we worked jointly on a number of projects: organizing conferences, building a cooperative research team, and publishing books and articles. Our ideas have evolved together, and much of what is contained here is also Baburao's work. We bring complementary experiences and viewpoints to our common efforts. For me, he has been an essential guide to reality, letting me know whether ideas make sense in terms of the life he has lived. Baburao brings a remarkable perspective to his research: he is not only one of India's leading sociologists but also the son of a small farmer in

xvi

Acknowledgments

Maharashtra. When I get caught up in my own struggles, trying to sort out what intellectuals think about peasants and what peasants think about themselves, it is reassuring to hear from one who has lived as a peasant and as an intellectual. Most of all, I have benefited from association with a man of great warmth, discernment, and character. Baburao and his family are included in the dedication to this book. And so are D.P. Apte and his family. D.P. Apte was for many years the Registrar of the Gokhale Institute of Politics and Economics in Pune. He also taught courses in economics, and he designed and carried out numerous economic surveys in rural areas. From the very first, when I began my research, he offered valuable suggestions concerning field methods, and I wish I'd had the sense to follow them more carefully. Equally important were the friendship and hospitality offered by his whole family. Our familial association began in 1970 and continued with visits back and forth between India and Canada in the 1980s. I received not only intellectual guidance but also emotional and practical support from my adopted family in Pune. Thanks are also due to the families of Nirmala Purandhare and KA. Patwardhan for their hospitality in Pune. While working in Maharashtra, I was affiliated with the Gokhale Institute, receiving helpful guidance not only from D.P. Apte but from two of the Institute's directors, V.M. Dandekar and N. Rath, who are among India's leading scientists in the study of agro-economic development. Those who helped on a daily basis out in the villages included two excellent research assistants, R.P. Nene and Sharad M. Gupte. Without their dedication and good sense, I would have gotten nowhere. S.V. Kulkarni also helped as an experienced enumerator. (The value of their work is more fully explained in the Appendix.) I first met Sharad Pawar, now Chief Minister of Maharashtra State, when he was a young MLA (Member of the Legislative Assembly) from eastern Pune district. He took me on a visit to a number of villages in his constituency and introduced me to village leaders. On the basis of these contacts, I was able to select two villages which seemed to offer the size, diversity, and historical interest that I was looking for. From this auspicious beginning, the research in Malegaon and Supe went smoothly, since my purpose and character had been vouched for, tacitly, by Sharad Pawar. His friendship and encouragement have been most helpful over the long course of this project. (Special thanks also to S.B. Dharmadhikari.) I wish it were possible to mention all the friendly and hospitable people I met in the villages. In Malegaon, Keshavrao Taware offered vital information concerning the early history of the cooperatives, and his son, S.K Taware, provided both information and practical help. V.C. Jadhavrao and his family gave valuable information on their distinguished ancestors and on the history of the sugar factory. The village leaders, including N.B. Taware and his brothers, along with KD. Jadhavrao, M.G. Taware, B.S. Gophane, N.G. Saste, S.P. Taware, and G.V. Taware, among others, were all very helpful and informative. I hope they will accept my heartfelt thanks on behalf of all the villagers.

Acknowledgments

xvii

In the Malegaon sugar factory, I received all possible help, including living quarters. Over the last 20 years, various chairmen and managing directors, including M.G. Taware, S.S. Hiremath, V.C. Jadhavrao, C.K Taware, and A.M. Pisal, have done everything to facilitate the research. I am also grateful to the many staff members who gave of their time to provide information. These include M.G. Dhole (Chief Accountant), B. G. Malegaonkar (Labour Officer), H.P. Wable (Storekeeper), S.K Taware (Agricultural Officer), and many others. Thanks are also due to others in the Baramati area, including D.G. Shembekar and S.G. Date. Although there has not yet been time to work out a complete analysis of my data from Supe village, the research done there greatly influenced my understanding of the changes taking place in Malegaon. (Supe is dry and poor, while Malegaon has become irrigated and prosperous.) I hope that V.S. Kelkar, R.B. Khaire, B.P. Kalkhaire, V.B. Kutwal, M.T. Bhondwe, and D.B. Chandgude will accept my thanks on behalf of all the friendly and helpful people of Supe. Many experts in Pune, and elsewhere in Maharashtra, have taught me about the sugar industry. Particular thanks are due to C.S. Mujgule, whose practical experience as a farmer and managing director of cooperative factories has been most enlightening. Likewise, I am grateful to M.S. Marathe, V.L. Tanpure, and V.S. Baviskar for all the information they have provided. Many other experts-factory chairmen, managing directors, and sugar technologists-have given generously of time and information. Thanks are due to the officers and staff of the following organizations: the Directorate of Sugar (Government of Maharashtra) in Pune; the Deccan Sugar Institute near Pune; the Maharashtra State Cooperative Sugar Factories Federation in Bombay; the Deccan Sugar Factories Association in Bombay; and the National Federation of Cooperative Sugar Factories in New Delhi. For their friendship and insights into other aspects of Indian life, I would like to thank S. Bandyopadhyay, S.M. Batra, Shanti George, A.R. Kulkarni, Dharma Kumar, Jayant Lele, A.S. Patel, Satish Saberwal, and V.M. Sirsikar. Likewise, thanks to my North American colleagues, Stewart Gordon, Gail Omvedt, Donald Rosenthal, Richard Tucker, and Eleanor Zelliot, who shared their perceptions of Maharashtra during the period when we were all doing field work there. That first, two-year stint of field work in Maharashtra was ably assisted by Rhoda Klein Attwood, who provided intellectual, emotional, and practical support. in circumstances ranging from pleasant to difficult. In the process, she taught me much about life. A small, remote and visionary institution, Deep Springs College, provided my first lesson in irrigation and in many other important matters. My colleagues here at McGill have taught me much about social theory, particularly theories of development, and they have given valuable comments and suggestions on earlier portions of this analysis. Special thanks, then, to Dan Aronson, Laurel Bossen, John Galaty, Philip Salzman, and especially

xviii

Acknowledgments

to Don Von Eschen, whose research on West Bengal has been a constant source of ideas and puzzles. I am grateful to the agencies that provided financial support for this research: the Canada Council, the Shastri Indo-Canadian Institute, the Social Sciences and Humanities Research Council, the International Development Research Centre (Ottawa), les Fonds pour Ia Formation de Chercheurs et l'Aide a Ia Recherche (Quebec), and the Faculty of Graduate Studies and Research at McGill University. (Special thanks to P.N. Malik of the Shastri Institute and David King of IDRC.) Aida Jain and Patricia Loveridge worked with me for years, helping with data coding and analysis and manuscript preparation. Valuable assistance was also provided by Carole Blackburn, Fern Brunger, Jerzy Depa, Teresa Gray, Paula Fleisher, Colleen McVeigh, Virginia Mills, Patrick O'Neill, Peter Orr, Henry Rono, Sherry Tepper, and Donna Winslow. Rose Marie Stano provided invaluable support. Particular thanks go to those who have read earlier portions of the analysis and commented on them: S. Bandyopadhyay, Harry Blair, E.W. Coward, Alan Gelb, A.O. Hirschman, Dharma Kumar, David Landes, Jayant Lele, Michael Lipton, David Ludden, Michelle McAlpin, Barbara D. Miller, Sidney Mintz, Mick Moore, Morris D. Morris, Lloyd and Susanne Rudolph, T.W. Schultz, Miriam Sharma, Judith Tendler, Norman Uphoff, and Robert Wade. I am especially indebted to those who read and commented on the entire manuscript: D.P. Apte, Donald Von Eschen, Stewart Gordon, Morgan D. Maclachlan, Baldev Raj Nayar, Deborah Sick, and, of course, B.S. Baviskar. It causes me deep sorrow that, due to his untimely death, my old advisor and mentor will not see this book published. Richard Salisbury provided advice and encouragement in the early years, when they were most needed. He also offered a rich framework of questions and ideas that I tried to adapt to this research. In later years, as a senior colleague, he offered more: an example of one whose humanity, competence, and good sense were equal to his formidable intelligence. His work and life embodied compassion and insight. This book would never have been written without the support of my parents, sisters, wife, and friends. May life bestow abundant blessings on them all. Donald W. Attwood

PART ONE

Introduction

l The Problem In western India, there is a sugar industry owned and operated by peasants-hundreds of thousands of peasants. The majority of shareholders are small growers, producing just an acre or two of sugarcane. Yet their industry, owned and managed cooperatively, is huge. In 1987-88, cooperative sugar factories in Maharashtra state produced 2.7 million (metric) tonnes of sugar. Measured in terms of value, their output in 1980-81 was almost as large as that of the state's iron and steel industry. 1 In recent years, development experts have become increasingly interested in "development from below," "grassroots development," and "peoplecentered development"-that is, in leadership, innovations, and experiments which try to solve development problems at the local level. There are many such experiments in the Third World (as elsewhere), and some have been quite successful. Some experts see experiments of this type as providing viable alternatives to the large-scale, centrally planned, bureaucratically managed, top-down development projects which have produced expensive and demoralizing failures all over the world. . The world has been learning, one way or another, that some development problems cannot be solved bureaucratically. The current dismantling of the centrally planned economies of Eastern Europe lends urgency to the study of alternative experiments in development. However, the failure of communism does not mean that capitalism is the simple cure for poverty and stagnation in the Third World. Capitalist bureaucracies (in international aid agencies, for example) have been nearly as prone as their communist counterparts to sponsor large-scale, centrally managed development projects-projects which have often been economically wasteful, socially and culturally inappropriate, politically oppressive, and environmentally destructive. It need hardly be mentioned that the great privately owned bureaucracies (national and multinational corporations) have often achieved similar results. It is the nature of all bureaucracies, public or private, to avoid accountability; and it is this quality which makes them dangerous to people who are relatively poor and powerless. The alternatives pursued by locally initiated organizations are one way of escaping the dead hand of private and public bureaucracy. In fact, local organizations may be the only effective way for Third World people to cope 3

4

The Problem

with two opposing threats: that is, with market forces dominated by great corporations, on the one hand, and with state forces controlled by giant bureaucracies, on the other (see Esman and Uphoff 1984). Contrary to those who would have the Eastern Europeans, the Chinese, and everybody else choose between one threat or the other, I suggest that all will need to steer a middle course by promoting local development initiatives. This is one reason why we need to understand how such initiatives have arisen and, on occasion, triumphed. The first cooperative sugar factory in Maharashtra was established in 1950, shortly after India became independent. By 1990, there were 92 co-ops operating in this state, producing 30 percent of India's white sugar. Since India and Brazil are the world's largest producers of cane sugar, it is evident that this peasant-owned industry has contributed to the nation's economic growth. The cooperative factories are remarkable in several ways. Most were founded and operated by leaders from ordinary village backgrounds. Many factories have fostered other kinds of cooperatives, plus educational institutions, irrigation projects, and other facilities to deal with the needs of their local areas. Many are efficient and innovative, which is surprising, since local politicians compete actively for seats on the factory boards, in order to build power in state politics. Competitive politics seems inconsistent with good business management. Nevertheless, many factory leaders, who often lack even a high-school education, have proven to be astute businessmen as well as tough politicians. As we shall see, these apparent incongruities actually help to explain the remarkable growth and efficiency of the sugar co-ops. The purpose of this book is to understand the historical and social background of the cooperative leaders, the causes and results of their innovations, and the paradoxes of this peasant-owned sugar industry. An Anomalous Industry Sugarcane has been known in some parts of the world as a crop which exploited slave labor to grow fortunes for colonial plantation owners-and which, more recently, has been taken over by huge industrial plantations cultivated by gangs of hired laborers. Sugarcane has been emblematic, in other words, of the great disparities in ownership and power which characterize many Third World agrarian regimes. Thus a peasant-owned sugar industry is an anomaly, strikingly different from the industries found, for example, in northeast Brazil, Central America, or the Caribbean. One purpose of this book is to explain why Western India is such an exception by drawing on the history of the region, its environment and cropping systems, the dynamics of village society, and the impact of national and international markets. Explanations are formulated and tested through comparative analysis, using data from other parts of India and from sugar industries elsewhere. Given its size and anomalous character, it is surprising how little is known of this industry in the world at large, even by development experts and

The Problem

5

sugar experts. There seem to be several reasons for this. First, India is seldom perceived as an important sugar producer, simply because it exports so little. (Indian sugar is sold on the huge domestic market.) Second, cooperative sugar factories are not dependent on foreign aid of any kind. This is in contrast, for example, to India's dairy co-ops, which have attracted attention in many international development agencies. The dairy co-ops have absorbed billions of rupees in foreign aid, and aid donors wish to be seen as promoting beneficial and important projects, so the donors gladly propagate claims made by Indian dairy organizations concerning their own achievements. Sugar cooperatives, on the other hand, are not dependent on foreign aid and do not require an international public-relations machine. A third reason that the sugar co-ops are not better known is rooted in the political history and culture of western India. Local journalists and academics-those who would be best equipped to study these cooperatives and assess their achievements-have a disdainful attitude toward them. There are two reasons for this. First, rural leaders in general, and chairmen of cooperative sugar factories in particular, have become increasingly powerful in state politics since independence (that is, since 1947), displacing the urban intellectuals who led the freedom movement and who naturally resent this displacement. Second, successful sugar co-ops are anomalous not only in the international context but also in the Indian context. Unlike most institutions created since independence, they are not simply a product of state planning and are not managed by state bureaucracies. As elsewhere in the world, it is the bureaucracy which gives power, prestige, and employment to the urban middle class. This class directly manages two-thirds of India's large-scale industrial production (in state-sector enterprises) and regulates the rest of the industrial economy through a vast array of state agencies (Rudolph and Rudolph 1987; Nayar 1989). Since the cooperative sugar factories are not under direct state management, the urban middle class regards them with suspicion. 2 Some of the disdain of local intellectuals for the sugar cooperatives has been picked up and amplified by foreign intellectuals. I have met many who seem appalled by the success of the co-ops. They assume that a huge and successful industry must be based on an especially sinister form of exploitation-probably exploitation of small farmers as well as factory workers. It is perhaps no exaggeration to say that intellectuals love miserable victims. On the other hand, they seem to have little tolerance for peasants who build institutions to serve their own needs and manage them without supervision from the middle class. India is known to foreigners, first, for poverty, and second, for bureaucracy and inefficiency. Many find it impossible to believe that some Indian farmers have found their own way out of poverty and have done so without the control of a state bureaucracy. There must be something sinister about this; it disturbs the natural order of things. The problem of disdain and disbelief goes even deeper: it is rooted in what has been for decades the dominant academic theory of development

The Problem

6

and underdevelopment in most of the social sciences. Known as "dependency theory" or "world system theory," this outlook rests on certain core assumptions. First of all, the world capitalist system exploits and impoverishes all ordinary people-poor peasants and laborers-who come into contact with it. The greater the contact, the greater the immiseration of the peasantry. Thus a flourishing sugar industry ought to be driving small farmers into debt, dispossessing them of their land, and exploiting the labor of an evergrowing landless proletariat. As we shall see, there is considerable evidence that none of this has happened. However, theorists of any persuasion generally prefer to dismiss anomalies rather than try to understand them. Against mountains of readily available evidence they believe, and will likely continue to believe, that the cooperative sugar factories serve only the interests of "rich peasants," or "kulaks." This book examines a wide range of historical and contemporary evidence bearing on such questions. At bottom, though, a simple test is best: Small peasants are not coerced into growing sugarcane, neither are they coerced into joining the sugar cooperatives; on the contrary, over the last few decades, hundreds of thousands have eagerly joined the older co-ops and supported leaders who propose to establish new ones. Unless they are suffering from a peculiar form of mass delusion, they must know whether other small peasants have already benefited from membership. Thus the anomalies multiply. This is a sugar industry which does not conform to patterns of organization found in other countries. It is a set of large-scale development projects which are not managed by a state bureaucracy. It is a cash-cropping system which has not impoverished the small peasantry. Small wonder the sugar cooperatives are generally ignored or condemned: they cannot possibly exist. Nationalism and Dependency Theory

History results from the impossible becoming possible-that is, from events which do not conform to familiar trends. My purpose is to understand how this could happen in rural Maharashtra. It is only a century since the first large-scale irrigation canals provided the base for a flourishing cash-crop economy. During that century, drought-ridden communities of subsistence farmers have transformed themselves into powerful centers of agro-industrial production. Capital and technology were brought in from outside: by cooperative banks and Bombay industrialists, for example. But the main driving force behind the growth and reorganization of the sugar economy was innovation by local villagers. In order to describe and analyze this case, I must draw attention to assumptions which cloud perceptions of sugar co-ops in particular and cash cropping in general. However, it is not my intention to engage in a general theoretical debate, nor to replace one grand theory with another. If dependency theory gives a good explanation of other cases, that is fine. 3 What is necessary, in that event, is to understand why some regions and agrarian systems are so different from others; and that is what this book is for.

The Problem

7

One problem with many theories of development and underdevelopment is that they are assumed, by their proponents, to be universal. This mindset inhibits the study of anomalies and variations-inhibits comparative analysis, in other words. Many people who claim to do comparative research do not understand what the term means. Comparisons are instructive only if we explore the differences as well as similarities between cases. However, fundamental differences among Third World societies are assumed not to exist by those who believe they have a universal theory of development. I began my research in Maharashtra in 1969. At that time, I was an enthusiastic admirer of the new school of dependency theory, which was emerging in Western social science (see, e.g., Baran 1957; Frank 1969). I had read the available literature-economic histories and village studieswhich purported to show the devastating effect of cash cropping and British colonial policy on Indian agriculture. I went to Maharashtra to study these effects at close range. Unfortunately, the results of my research did not conform to expectations. Not knowing what to do, I avoided the problem. My Ph.D. thesis (1974) concentrated on local politics, and I postponed the problem of deciding how to interpret the economic data. After years of uncertainty, I finally published a preliminary analysis (1979) and then returned to India for more data. Among anthropologists, sociologists, political scientists, and historians concerned with development, dependency theory was the dominant paradigm of the 1970s and 1980s. It just about demolished "modernization theory," a loose category of ideas whose common thread was the notion that developing nations would learn, through increasing contacts with the West, to mimic Western technology, politics, culture, agriculture, and industry. There were already some countries (China and India being examples) which were, in effect, saying "no" to modernization theory; and dependency theory emerged as an articulate ideology of disengagement from the capitalist nations of the west. Having fulfilled its historic mission by undermining the cre4ibility of modernization theory, dependency theory became so successful that it closed off important lines of inquiry for a decade or so. Recently, new questions and ideas have been emerging. (See, e.g., Paige 1975; Lipton 1977; Evans 1979; Hart 1982; Toye 1987; De Soto 1989.) Many development experts now recognize (again) that the impact of external forces (expanding markets, new technologies, foreign domination) is always mediated by the structure of internal forces (environmental, historical and social factors rooted in indigenous political economies). The resulting interaction between internal and external forces is not predictable by any general theory, at least not so far. My purpose, then, is not to attack dependency theory in general, but to use it as a source of questions and hypotheses. In science, the value of a theory is measured, in part, by the extent to which it is testable and thus falsifiable. The intellectual roots of dependency theory go back much farther than most people are aware. Nowadays, the theory is used primarily to discuss

8

The Problem

problems in Mrica and Latin America, but its roots are in Asia. Toward the end of the 19th century, Indian nationalists began assembling an economic indictment of British rule. This work culminated in an impressive twovolume study by Romesh C. Dutt (1901, 1903), The Economic History of India. Dutt's work contained a number of arguments which have become standard items in dependency theory today: • Through various mechanisms, British rule removed capital from India: capital which could have been invested in Indian industry and agriculture. • Through one-sided trade regulations, the British compelled India to export raw materials and import industrial products, helping to promote British and undermine Indian manufacturing. • Through heavy land taxes, through periodic increases in the tax rates, and through inflexible tax collection during bad crop years, the British forced farmers to grow more cash crops (such as cotton), to sell more of their produce, and to fall more heavily into debt. Moreover, an increased share of their produce was exported abroad. Thus the villages had a smaller buffer-stock of foodgrains for coping with occasional droughts. As a result, there was a series of devastating famines at the end of the 19th century. My study begins where Dutt's leaves off-that is, with famines in the late 19th century. Large-scale irrigation canals were built to protect Maharashtra from famine, but their unintended result was to stimulate a cashcrop economy centered on sugarcane. One purpose of this study is to consider whether this process conformed to the analysis of cash cropping put forth by Dutt and others. Dutt's work was brilliant and original in its day, and it became the intellectual stock-in-trade of the nationalist movement. As a socialist ideology also grew within the nationalist movement, the same basic ideas were elaborated using Marxist categories (see, e.g., R.P. Dutt 1940; Patel 1952; Mukherjee 1957, 1974; Chandra 1966). The result was an analysis based on Marx's (1887) and Lenin's (1956) theories concerning the impact of commercial agriculture on small farmers. The basic premise for this analysis was that rural society, before the British period, consisted of a relatively homogeneous mass of peasants (plus various artisans, priests, etc.). Village communities were self-sufficient and minimally engaged with the outside world. Peasants, too, were self-sufficient: they grew crops with their own labor and had more-or-less equal access to the community's land, which was not the exclusive, private property of a few. Landless agricultural laborers were scarce or non-existent, and so were the rich peasants or landlords who might have prospered by exploiting them. This idyllic portrayal of pre-British India provided the basis for a negative assessment of British rule. In addition to heavy taxes and pressures to grow crops for export, the British brought laws dividing the land into private

The Problem

9

plots which could be bought, mortgaged or sold. These changes forced many farmers into debt: their lands were mortgaged and then sold in order to repay these debts. Thus a new class of rich peasants and landlords emerged, a class which prospered at the expense of the majority, while the latter were forced into poverty and, in many cases, into the new class of landless laborers. The myth of a golden age before British rule, an age in which villages were self-sufficient, self-governing, and homogeneous, was understandably compelling to nationalist leaders. After independence, Indian sociologists and historians have gradually dismantled this myth (see, e.g., Srinivas and Shah 1960; Habib 1963; D. Kumar 1965; Raychaudhuri and Habib 1982), but it lingers on in various forms. (We shall confront it, as applied to Maharashtra, in the next chapter.) The object of this study is not to debate the general nature of pre-British society and the impact of British rule. The purpose, rather, is to focus on one region and one crop: on the impact of irrigation and sugarcane in Maharashtra. In order to understand the social impact of this crop, however, it is necessary to consider the assessments which are already current and to understand how they connect with the broad sweep of ideas on the impact of commercial agriculture. Thus I refer in later chapters to the "standard theory" of commercialization, which is the set of ideas just outlined-rooted in the work of Marx and Lenin and elaborated in the Indian context by scholars who were consolidating their nationalist and socialist ideals into a general theory of agrarian change. My purpose is to see whether this theory helps us to understand a specific case. In opposition to the standard theory I offer . . . the stories of some peasants. The core of this book rests on interviews with a sample of 130 village families, plus many local leaders and interested observers. I have relied particularly on descriptions of their personal and family histories. The risks, values and events which shaped their lives are spelled out, from their point of view, in these narratives. Inevitably, my own point of view intrudes in the process: in the selection of questions and interpretation of answers. Nevertheless, when you live a long time in a village, you inevitably learn some of what is on people's minds, despite your initial assumptions. In contrast, the standard theory (like other grand theories) rests mostly on secondary data: on statistical surveys of dubious validity, on documents assembled by official committees, on political tracts issued during times of unrest, on books and articles written by indigenous intellectuals. I have used these sources myself, attempting to construct an analysis which takes account of both points of view: that is, the views of officials and scholars, on the one hand, and those of villagers on the other. At bottom, though, an important test of the data and ideas gleaned from other sources is whether they would make sense to the villagers with whom I lived. Anything original in this book is based directly on what the villagers already knew, though they might not have discussed these matters until a nosey foreigner came to pester them. Since the voices of ordinary peasants

10

The Problem

are seldom heard amid debates on development theory, my bias is to lean toward what they accept as real. Dependent India? India overall poses many challenges to dependency theory, at least since independence. Many writers have described India as if it were a banana republic, mere clay in the hands of foreign adventurers. The reality is that_ India's rulers, whether foreign or domestic, have had to accommodate the complex roots of this huge and ancient civilization. And India's economy now belongs to Indians: it is not a convenient illustration of the perils of nco-imperialism. India began to industrialize during British rule. Presently, India ranks among the top 15 industrial nations in terms of total output and among the top ten in terms of ability to manufacture high-tech products. Its products run the gamut from handicrafts to satellites and atomic reactors. Its heavy industry is largely owned and managed by the public sector; and its consistent policy since independence has been aimed toward technological and industrial self-sufficiency (Nayar 1972, 1983, 1989). Its industrial and financial economy is dominated by state enterprises, and there is thus no politically-powerful class of private capitalists who might bend the economy toward foreign interests (Rudolph and Rudolph 1987). Moreover, India is no longer dependent on foreign food aid (and has only recently become heavily indebted to foreign banks). These achievements, which are impressive, are partly the result of an ideological consensus on the perils of dependency-a consensus grounded in the analysis of scholars such as R.C. Dutt. Because this consensus serves such an important function (protecting national economic policy from foreign influence}, anything which is blatantly inconsistent tends to be dismissed. Many Indian intellectuals think of the state as protector and builder of their industrial independence. Maharashtra's cooperative sugar factories, which are not under state management and which act, in some respects, like competitive, capitalist enterprises, are not pleasing to their eye. They expect the co-ops to fulfil their worst expectations concerning the impact of commercialization on rural society. The standard theory of commercialization, which evolved, as we have seen, as a critique of British colonial policies, took on new life in the 1960s and 1970s, as a critique of the government's new agricultural policies. There was a heated debate on the social impact of the "green revolution." The cause of this "revolution" was a package of technically-advanced inputs (hybrid seeds, chemical fertilizers and pesticides) which enabled farmers with irrigation to produce higher yields. Many social scientists (e.g., Frankel197l; Gough and Sharma 1973; Griffin 1974) expected that this technological change would create social disruption: that it would enrich big farmers, impoverish small ones, and create massive unemployment among agricultural laborers. All this would engender greater class polarization and propel the countryside toward a "red revolution." These predictions were based on the standard theory.

11

The Problem

The essence of the theory was that big landowners would find it easy to adopt the new inputs, because they had access to easy credit and had enough resources to buffer themselves from the risk of price fluctuations in the marketplace. Middle and small peasants, on the other hand, could not afford to buy the new inputs and would assume great risks in attempting to do so, by going into debt. Fluctuations in crop prices would make it difficult to repay these debts and drive the smaller peasants deeper into insolvency. Merchants, landlords, and rich peasants would take advantage of their increased credit needs and then foreclose on their mortgages. Thus the new technology would drive middle and small peasants into the landless proletariat. This theory was the basis for a vigorous attack on the government's decision, in the mid-1960s, to promote the new technology-a decision made in response to chronic food shortages, dependence on foreign food aid, and occasional famines. Ultimately, the critics were forced to shift ground, acknowledging that the green revolution had raised crop yields, but only in a few regions (principally those where wheat could be grown under irrigation). Empirical studies in these regions found, moreover, that small farmers were enthusiastically adopting the new inputs and were not being driven off the land. (See, e.g., Sen 1974; Bhalla 1974; Dasgupta 1980; Sarma 1982; Etienne 1982; Leaf 1984; Hayami 1984; Lipton 1989.) Confident predictions that the green revolution would turn into a red revolution gave way to more sober analyses of the roots of agrarian radicalism (e.g., Beteille 1974; Bouton 1985). The green revolution raised crop outputs in some regions but not others. In the more favored regions, small farmers actually became more viable because they could produce higher yields and larger incomes. By raising incomes in some regions, the green revolution increased disparities between these and other regions which remained technically backward and impoverished. This was a different kind of disparity, however, from the class polarization which was supposed to occur within any village which came into contact with new markets and new technologies. While the debate on the green revolution was raging, I wrote on the relevance of Maharashtra's sugar economy to these issues. 4 The green revolution began in the 1960s, but large-scale sugar production in Maharashtra started a century ago, thus providing an opportunity to study the effects of intensive commercialization over the long term. Sugarcane in this region is fantastically expensive by comparison with subsistence crops, and thus it should be extremely risky for small farmers. For this reason, and because the sugar economy has evolved over a long time, it provides a good test of the standard theory. Agrarian Regions in India In order to understand sugarcane as a crop, we must understand the environment, technology, and social order in which it is embedded. Agrarian systems are highly disparate: not only are different crops adapted to different

12

The Problem

environmental conditions, they also have technical requirements which influence the social organization of production in various ways. The standard theory of commercialization tends to assume the opposite. It assumes that the social dynamics of agrarian systems are essentially the same everywhere and vary in just one basic respect: the degree to which production is oriented toward external markets. Thus in the 1960s and 1970s, it was relatively easy for researchers to draw parallels between a wide range of agrarian movements and class systems with the help of a few deductive categories (see, e.g., Wolf 1969; Alavi 1973; Scott 1976). These were followed, however, by more sophisticated analyses recognizing the specificity and diversity of agrarian systems in different parts of the world, and thus the variable processes of commercialization and politicization which they foster (see, e.g., Paige 1975; Popkin 1979; Hart 1982; Bouton 1985). Dependency theory tends to assume that external factors (international markets, new technologies, colonial administrations) are overwhelmingly powerful and that regional and local factors are of negligible explanatory significance. Local relations of production and power are assumed to be passive and malleable, readily transformed by new laws of property or new connections with distant markets. 5 Unless they join a revolution, ordinary villagers are assumed to be merely the passive victims of history. This is a condescending attitude toward people who are, in fact, creative, innovative, and sometimes powerful in their responses to new external conditions. This attitude of condescension links dependency theorists with their avowed enemies, modernization theorists, since both tend to assume that the West (for good or ill) generates the impetus for all significant change. Thus dependency and modernization theory are like Siamese twins: divergent above but joined below the waist. This shared assumption of Western prepotency is rooted, of course, in 19th-century European ideology; but ironically it is now embraced by many Third World elites. It is a convenient assumption for development researchers and planners, for they can locate the causes of change in official documents shelved in capital cities, instead of having to inquire among the peasants. This book takes a different tack, arguing that regional and local relations of production and power are at least as important as external influences. The specific character of a region's political economy is shaped by its history, its natural environment, and the technologies used for production, distribution, and political control. Consequently, the impact of new capital, crops, and markets is determined not by universal laws of history but by the specific interactions of technology, environment, and social organization. If we have to give this explanatory approach a label, let us call it "interaction theory." The Third World has a great diversity of agrarian systems, some of which are prosperous and dynamic, while others are poor and stagnant. One of the most interesting challenges facing social science today is to explain this diversity. India will continue to generate good research on this question, since it has rich archives dating back to the precolonial period, combined with a strong corps of indigenous social scientists, plus a tradition of open

The Problem

13

inquiry and debate-a combination which has no parallel in other parts of the Third World. Moreover, India has a rich diversity of agrarian systems, which facilitates comparative analysis and hypothesis testing. One of the fundamental causes of diversity (but by no means the only one) is variation in the supply of rainfall and irrigation. (See, e.g., Epstein 1962, 1973; Beals 1974; Stein 1980; Stone 1984; Ludden 1985, 1988a). Speaking very broadly, one can say that India has coastal plains and inland river valleys where rainfall is plentiful, where yields per acre (with rice as the principal crop) are high, and population density is therefore also high. On the other hand, there are vast inland plains and plateaus where rainfall is scant and uncertain, where crop yields (mostly of sorghum and millet) are lower, and population densities are also lower. There are, of course, many transitional and mixed zones. Nevertheless, it is possible to generalize about some of the contrasts between wet and dry regions. Wet regions have long had secure subsistence bases with dense populations, large towns and cities, and strong urban influences in the countryside. Dense population meant that land values were high, while labor was cheap. Consequently, it became possible even for petty landowners to subsist on the labor of tenants, sharecroppers, and landless laborers. In terms of social stratification, then, these regions had large populations of low-caste, landless laborers, along with small, high-caste village elites of non-cultivating landlords, whose lifestyles were strongly influenced by urban cultural values. The dry regions, on the other hand, had lower population densities with insecure subsistence economies. Water was scarcer than land, yields were lower, and the supply of landless laborers was therefore smaller. Villages tended to have smaller populations of low-caste laborers and fewer noncultivating landlords. Most of the rural population consisted of "yeoman peasants" cultivating their own lands and belonging to the middle castes. Village headmen belonged to the same middle castes and shared the same cultural values as most of the smaller peasants. The dry regions comprised some of the most impoverished and economically insecure portions of the subcontinent; but these were also regions where social stratification tended to be more flexible, since chances for economic prosperity and social mobility depended partly on the uncertain rains. The most favored agricultural zones today are those portions of the dry regions which have benefited from large-scale irrigation systems during the last century or so. These areas include much of the northwest (Punjab, Haryana and western Uttar Pradesh, the heartland of the green revolution) and scattered portions of western and southern India. They have several advantages. Irrigation reduces the risk of subsistence crises and multiplies the returns which farmers can obtain from their land. Irrigated land can therefore support a denser population, and small farms become more viable. At the same time, these areas start with lower population densities, so there is less pressure to expand production of traditional subsistence crops. Instead, farmers have room to switch to high-value cash crops like cotton, wheat and

The Problem

14

sugarcane. In addition, the prevailing stratification systems are flexible and open to economic mobility. As a result, there are fewer barriers to innovation and enterprise. This is only a crude sketch of the variation in agrarian systems from one part of India to another. Its purpose is simply to put Maharashtra, which is a dry region, into context: to show why certain kinds of socio-economic processes may occur in such a region but not in others. To date, only preliminary efforts have been made to compare the diversity of interactions between external and internal forces in different parts of India. This book tries to explain why cooperative sugar factories arose successfully in one region but not in others. In order to arrive at such an explanation, it is necessary to understand what makes Maharashtra economically and politically distinct. Political Economy This endeavor fits neatly into a recent trend in research: empirical studies on the political economy of India and its component parts. The last few years have seen a number of significant publications in this field (e.g., Frankel 1978; Bardhan 1984; Desai, Rudolph and Rudra 1984; Bouton 1985; Rudolph and Rudolph 1987; Kohli 1987; Herring 1987; Nayar 1989). By the end of this book, I hope to have clarified not only an important part of Maharashtra's political economy but also some of the factors which make this region distinct from others. At this level of generalization, my conceptual framework is drawn from some of the works just cited. A brief sketch of this framework will prepare the reader for what lies ahead. The economic and social impact of British imperialism led to the emergence of two politically powerful new classes in India. The most obvious, before independence, was the urban, educated middle class, recruited largely from among the traditionally literate higher castes, such as Brahmans and Kayasthas. Members of this class were employed in the civil service and in the professions of law, journalism, teaching, medicine, and politics. In addition, they founded and led the independence movement. After independence, this class expanded and became ever more powerful through its control of public-sector industries and government bureaucracies in general. (See Bardhan 1984; Rudolph and Rudolph 1987; Nayar 1989.) Another class was also emerging toward economic power and political consciousness under the British, a class which I call the "commercial peasantry." This corresponds to what Bardhan (1984:46-47) labels as "the class of rich farmers" plus "the intermediate class of primarily family farmers." It also corresponds with those whom the Rudolphs (1987:49-55) call "bullock capitalists." The latter phrase captures the hybrid nature of family farmers who grow crops for their own subsistence and also for the market. My phrase is intended to do the same. I prefer the phrase "commercial peasantry" because I want to remind the reader that, even as these cultivators adopt new technologies and strategies for production, they remain imbedded in a

15

The Problem

social and cultural context which has deep historical roots. 6 This context has a profound impact on their actions as economic and political entrepreneurs. The commercial peasantry came into its own after independence, for two main reasons. First, starting in 1952, the country was ruled by governments elected under universal adult franchise. Since the overwhelming majority of voters live in the countryside, rural politicians came into power after 1952, particularly in state and local governments. Second, India (unlike many countries in Mrica and Latin America) committed itself in the 1960s to meet its own food needs from domestic production. This encouraged a growing trend (which was already strong in some regions) toward agricultural commercialization; and this in turn generated more prosperity in some rural areas, strengthening the commercial peasantry. By the end of the 1970s, the commercial peasantry became a self-conscious class, lobbying for rural interests in opposition to those of the urban middle class. These two classes, one urban and one rural, correspond to what Nayar calls the "intermediate strata"-that is, the classes which have been most powerful in India's "intermediate regime" since independence. Intermediate regimes are governments dominated not by big capitalists or landlords but by the urban middle class, plus (in this case) the "rich and middle peasantry" (Nayar 1989:112-18, 175-80). Bardhan (1984:40-45) includes the "industrial capitalist class" or "industrial bourgeoisie" along with the urban middle class and the rich and middle farmers in his triumvirate of"dominant proprietary classes." However, both Nayar (1989) and the Rudolphs (1987) take considerable pains to demonstrate that big industrial capitalists are politically marginal and economically hemmed in by the state sector. I agree with the latter assessment, though it is not essential to debate the issue here. Granted, industrial capitalists are economically powerful in Bombay city; but as we shall see, they have yielded power over the sugar economy in the hinterland. The internal cohesion and relative power of the intermediate classes vary from one region to another. In my opinion, these variations explain much about the divergent political economies of India's states. This point will emerge more clearly at the end of the book. Enterprising Peasant Families This book is about a class: the commercial peasantry. It is also about families, whom I choose to describe as "enterprising peasant families." The commercial peasantry emerges with the sugar economy, starting around the turn of the century. The enterprising peasantry, on the other hand, is much older. In later chapters, we shall see how peasants built a new political economy to suit their own interests. This story only makes sense if we notice that the same peasants were using sophisticated economic and political skills, risk-taking abilities and survival strategies, long before the growth of the sugar economy. They did not enter naked and defenceless into the commercial economy.

16

The Problem

The enterprising peasant is a different creature from the dependent peasant. The whole thrust of dependency theory is to argue that modern markets and technologies exploit and destroy the peasantry, turning peasants into proletarians. While not denying that there are dependent and victimized peasantries in some regions of the world, this book will argue that enterprising peasants are long overdue for recognition. The reasons are both practical and theoretical. While there are, as yet, few political economies in the Third World which give adequate scope to the creative abilities of peasant entrepreneurs, I suspect that the future of rural development hinges on their activities (see Hart 1982). That is the practical reason for giving them attention. The theoretical reason is that enterprising peasants have, in the past, played a vital role in the development of modern economies. The yeoman farmers of 16th-century England were a notable example (Wallerstein 1974:247-56). What exactly are enterprising peasant families? The full answer will emerge over the course of this book, but it may be helpful to provide a brief sketch in advance. Pemt~nt Ft~milies Are Enterprising

By this I mean that they consciously evolve skills and strategies to cope with risk, to broaden and diversify their economic base, to take advantage of new opportunities. This was as true in the 18th century as it is today. The only thing which has changed is the structure of risks and opportunities. According to some versions of dependency theory, pre-modern peasantries lived in almost idyllic circumstances, in which they were not forced to compete or take risks, in which they were protected by their communities and by the patronage of landlords and officials. On this reading, commercialization destroys communities and their web of secure relationships, turning peasants into proletarians. My image of the enterprising peasant, on the other hand, suggests far more continuity with the past. In the 17th, 18th and 19th centuries, as we shall see in the next chapter, villagers in Maharashtra were faced with high levels of risk. There were risks of crop failure, of infant mortality and other threats to family continuity, and of political competition and instability. At any given time, each family had to cope with a combination of risks and opportunities, which were not precisely the same as those of their neighbors. As a result, a flexible repertoire of strategies and survival skills had to be evolved to deal with a range of contingencies. Those who survived learned how to manage risks and respond to new opportunities. These same skills were later put to use in coping with the commercial economy of the 20th century. One of the recurrent failures in development theory has been the tendency to ignore the decision-making abilities and managerial skills by which peasants survive in a hostile world. Peasants are often seen as just another variety of proletarians, doing mindless, repetitive jobs like factory workers. But in the words of an eminent economist:

The Problem

17

Farmers the world over, in dealing with costs, returns, and risks, are calculating economic a.gents. Within their small, individual, allocative domain they are finetuning entrepreneurs, tuning so subtly that many experts fail to recognize how efficient they are. (Schultz 1980:644)

Pemsnt Enterprises Are Fsmiluu By this I mean that peasants are not individual "economic men" or women, nor are they communal conformists. The study of peasant economies has long been hampered by a polarization between theories about individual actors and theories about institutional structures. Classical and neo-classical economics provide the image of the economic man, the individual decisionmaker who allocates scarce means to alternative ends. Marxism and dependency theory provide the opposing image of a deterministic structure, in which individual strategies and innovations are of little significance. 7 These are the polar positions, which might be labelled "individualist" and "structuralist." (The well-known debate between Popkin [1979] and Scott [1976] illustrates these polarities.) As economic anthropologists have increasingly recognized, both positions are unproductive in terms of understanding much of what goes on among peasants in the real world. The family enterprise is coming to be seen as the crucial middle term, one which avoids the atomism of the first position and the determinism of the second. (See, e.g., Maclachlan 1987; Ellis 1988; Barlett 1989; Wilk 1989.) Chayanov (1966) was perhaps the first economist to recognise that family farmers are engaged in coordinating two enterprises: the family as a social unit and the farm as a productive unit. The family must be maintained and reproduced: it demands social continuity (which can only operate through the cultural values of a given community and region). The farm must also be managed so that soils, crops, rainfall, labor and markets are combined in ways which are technically feasible and economically coherent. Individualist theories tend to concentrate attention on the technical and economic requirements, while structuralist theories tend to concentrate on the maintenance or destruction of communities and classes. Economic anthropologists need to see in both directions. As a result, family enterprises are coming to the fore as a focus of analysis. What are family enterprises? Let me first describe them as they often appear today, in rural Maharashtra. The idesl family is the patrilineal, patrilocal joint family, consisting of a senior married couple, their sons, and their sons' wives and children. Owing to various contingencies of fertility and mortality (and economic opportunity), this ideal is often not attained in practice; and when it is, it appears only as a phase in a long cycle of family growth and dissolution. Nevertheless, there are good economic reasons why this family type serves as a conscious goal for many villagers. In brief, the family can be more productive and more secure if it combines the skills and experience of a senior couple with the vigorous toil of one or more junior couples. It is vital to combine the specialized efforts of men and women, and equally vital to combine those of the senior and junior generation of adults. 8 (The details are beautifully expounded in Maclachlan 1983.)

18

The Problem

To some extent, the family is an end in itself. The goal of economic strategies is to ensure continuity and prosperity for the family. However, it is also useful to think of the family as a device for achieving economic efficiency and security. Nowadays, a common strategy is diversification. Farming families with two or more sons will often try to educate at least one, preparing him for entry into factory or white-collar employment. Diversification offers more legs to stand on in case the farm goes broke, and it also offers a source of savings and investment for increasing farm output. In the 18th century, diversification meant encouraging a son to seek military service with a local Raja, or supporting a father's efforts to build a political following and claim the headmanship of a village. The risks and opportunities were different, but the underlying goals and strategies were similar. The basic goals were the security and continuity of the family; and the principal managerial task was the allocation of skills and resources to meet a variety of ever-changing circumstances. For more than a century, Marxists and other experts have been predicting the demise of the peasantry as a result of commercialization. Meanwhile, enterprising peasants have flourished, at least in some parts of the globe. (In the final chapter, I will discuss why they do not flourish universally.) Peasants in Maharashtra have been lucky, since they have not faced warfare or armed revolution in this century. Moreover, they have benefited from new markets, public investments in infrastructure, and new technologies. However, dependency theorists do not consider these as benefits; they seem to regard them as more destructive than wars and famines. The plausibility of this outlook will be tested in the chapters ahead. Looking Ahead Chapter 2 introduces the geography and history of rural Maharashtra. In order to understand the impact of sugar production, we must know something about the society which it affected. Sugarcane became a significant crop only after 1874, when the first irrigation canal was constructed. Economic histories of 18th- and 19th-century Maharashtra, plus government reports and early village studies, tell us about village life before canal irrigation. In addition, oral family histories tell us much that is not found in print. These sources help to understand the character of village life before the rise of the sugar economy. Was this a static, homogeneous society, in which economic risks were minimized by sharing and reciprocity? The answer will tell us whether cash cropping caused a crisis in the social order or merely accentuated processes which were already in place. The second part of this book, consisting of Chapters 3 and 4, gives an overview of the institutions, policies, trends, and problems which shaped the sugar economy of Maharashtra from about 1900 to 1950. Chapter 3 asks who took up growing sugarcane and why? Was the crop monopolized by big landlords and capitalists? And why did the irrigation department

The Problem

19

promote the growing of sugarcane? Did this decrease the irrigation of subsistence crops and threaten the security of ordinary villagers? Chapter 4 compares Maharashtra with northern India, where sugarcane was grown on a much larger scale. Why were sugar factories more common in the north, and why were they organized differently from those in other countries? Why were factories in Maharashtra not organized on the north Indian pattern? In addition to these questions on industrial organization, there are similar questions concerning crop production. Why were cane growers in Maharashtra eager to adopt new technologies and new relations of production, while those in north India were more conservative? What were the relations of production which evolved in Maharashtra? The third part of this book consists of a micro-level study of the sugar economy, c. 1900 to 1950, in one locality-a village called Malegaon and its surrounding area. While the whole book is shaped by interview data from this locality, Part Three is devoted almost exclusively to the oral family histories collected in and around Malegaon. Chapter 5 is concerned with old elites and new entrepreneurs. What was the character of the old landed elite, and how did it affect the growth of the commercial economy? What role did it play in the emergence of new institutions, such as credit and marketing cooperatives? Who were the new sugar entrepreneurs and how did they interact with older elites? Chapter 6 is concerned with ordinary farmers and laborers in Malegaon. Who were the laborers and how were they recruited? Why did farmers migrate to Malegaon from other villages, leaving the subsistence economy behind? Who was profiting from sugarcane? What were the causes of economic mobility? How did some landless laborers become landowners? Chapter 7 is concerned with the overall pattern of stratification in Malegaon and how it was changing from c. 1920 to 1950. What was the pattern of land distribution, and how did it change? What was the pattern of economic mobility, and how did it connect with the caste hierarchy? What was the impact of the great depression (which caused a crisis in the sugar economy)? The fourth part of the book is concerned with the political economy of sugar since independence, starting with a review of changes in official policy at the national level. Chapter 8 discusses the politics of sugar. Why were cooperative sugar factories started in the 1950s, and by whom? How did political competition emerge in these factories, and how was it organized? What connections evolved between factory politics and state politics? How did the dynamics of patronage politics impinge on factory management? Chapter 9 is concerned with the economic performance and impact of the Malegaon sugar factory. What were its achievements in terms of production, investment, expansion, and diversification? How efficient was its technical performance, and how did it compare with other factories nearby? What were its problems, and how did it attempt to solve them? This chapter is also concerned with the impact of the sugar factory on Malegaon and other nearby villages. What were its effects on employment, migration, land distribution, and economic mobility? And how did all this relate to the caste hierarchy?

20

The Problem

Chapter 10 considers the economic performance of all cooperative sugar factories in Maharashtra and compares them with private sugar factories in northern India. Are the cooperatives really efficient? Are they heavily subsidized? What are their effects on the growth of other institutions serving the villages? What makes them cohesive and innovative? How can they survive the fissiparous effects of political competition among their leaders? Chapter ll steps back from the specific features of co-ops and sugar production to ask why Maharashtra seems to be different from other regions, both inside and outside India. What are the characteristics of this society which seem to encourage innovation from below? What encourages industrial pioneers to emerge from village backgrounds? The answer to these questions, which are necessarily somewhat speculative, involves comparing the political economies of other regions. In particular, interactions between the urban middle class and the commercial peasantry are explored on a comparative basis.

Notes l. Maharashtra, which includes the city of Bombay, is India's most industrialized state. (See Muthiah et al. 1987:180,196). 2. This suspicion is not altogether a bad thing, since it provides a check on the political machinations of the "sugar barons"-the politically powerful chairmen of some cooperative factories. However, suspicion also perpetuates aversion and ignorance. It is surely no coincidence that the only major field study by an Indian of a cooperative sugar factory was by a scholar whose father was a small farmer, not a member of the urban middle class (Baviskar 1980). 3. Judging by the literature on Central America (e.g., Bossen 1984; BulmerThomas 1987; Brockett 1990), dependency theory fits this region fairly well. I am tempted to suggest that there is a Latin American style of commercialization, in which peasants are driven off the land to make way for (and supply cheap labor to) ever-larger plantations and haciendas. This process, which is well documented for Central America, is rooted in a set of political economies which are radically different from most of their Asian counterparts. 4. This was later published as Attwood l988b. 5. Exception to this one-sided theoretical approach has been taken even by some Marxists and dependency theorists (see, e.g., Laclau 1971; Cardoso 1977; Evans 1979; Warren 1980; Wolf 1982). Dependency theory comes in many colors (see Chilcote 1984), and the critique outlined here applies mainly to the more extreme versions. Some dependency theorists have emphasized the need for a dialectical approach to the interaction between external and internal forces. 6. It would be tedious to use a phrase such as "commercial peasants" throughout the book, so I often write simply "farmers" or "peasants." I use the term "peasant" when referring to the 19th century or earlier (before the rise of the sugar economy) and also in contexts involving traditional structures and values. Thus I refer to the cultivating middle castes as the "middle peasant castes." I use the term "farmer" in the context of the sugar economy, when I am thinking of cane growers as commercial peasants. 7. Before the rise of neo-Marxism and dependency theory, the "substantivist" school of economic anthropology tried to argue that pre-commercial peasants were

The Problem

21

guided by what amounted to rules of social and cultural conformity in their economic behavior. This was a non-Marxist version of structural determinism, denying that peasants make decisions of any significance. 8. Women's specialized activities (working in the fields, tending cattle, cooking, tending children) are vital to the family enterprise. So are their contributions to decisions concerning family management and organization-decisions regarding the education and marriage of children, co-ordination of labor within the family, and so forth. Unfortunately, my data on this side of the enterprise are incomplete, for two reasons. First, when the research was planned in the late 1960s and early 1970s, I was still ignorant of many issues concerning women in development. Second, it was difficult for a foreign male to arrange interviews with village women in a setting where they would feel comfortable and free of supervision by male family members. I have systematic data concerning women's work in the fields, but little on their activities and decisions inside the home.

2 The Setting For those who have travelled the backroads of Maharashtra, it is not difficult to imagine what the region may have looked like in the middle of the 19th century. The land would have looked much the same: wide river valleys and rolling plateaux, separated in places by mountain ridges and steep scarps. Except for a few trees and fields of grain, growing in the folds of hills, at the foot of scarps, and along the rivers, the land was dry and rocky. This sun-bleached and dusty land was the home of sturdy peasants who were also, on occasion, redoubtable warriors. Villages and hamlets were clustered near streams and rivers, which were dry for much of the year. Houses were built of mud, brick and stone, and temples of carved stone were found in every large village, as were mansions belonging to notable families. Within a day's walk of most villages, there was a market town, which also served as the administrative center for a group of perhaps 50 or 100 villages. The climate and economy of this region were shaped by the coastal mountains which guarded its western marches. Clouds sweeping inland from the Arabian sea collided with the mountains and released their rain, which drenched the steep and humid coast. Further east, the Deccan plateau lay in the rain shadow of the mountains. A wide belt, running parallel to the coastline just east of Pune city, was known as the famine zone. In this zone, the average annual rainfall varied between 12 and 20 inches and was so uncertain that agriculture was a "gamble in the rains." The rains, if they came at all, fell almost entirely from June to November. Some crops, like bajri (spiked millet, Pennisetum typhoidemn, here referred to as "millet") were sown in June, while others, like jowar (Indian millet, Sorghum vulgare, here referred to as "sorghum") were sown around October to ripen in the winter. Different localities tended to rely more on kharif (summer monsoon) or rabi (winter) crops, depending on whether they tended to get rain early or late in the year. The staple crops, sorghum and millet, were sturdy and drought-resistant, as they had to be. Here and there, they were protected against the uncertain weather by small-scale irrigation works. Farmers built small weirs and diversion channels on nearby streams. They also dug deep wells, which if they were lucky, could irrigate two or three acres. Water was drawn out of these wells 22

23

The Setting

in big leather buckets by teams of bullocks. In some cases, the wells watered vegetable gardens, fruit trees, and even small plots of sugarcane. However, most fields were unprotected by irrigation. They produced at most one crop a year and sometimes less. The inland plateau region of western Maharashtra, including the famine zone, lay in an administrative territory known under British rule as the Bombay Deccan, a part of the Bombay Presidency. (See Figure 2.1.) The Deccan districts were subject to direct administration by the British, with orders emanating from Bombay via Pune (or Poona), the capital city of the Deccan. Interspersed among these districts were various princely states, where the British ruled indirectly. British rule began with conquest of the Deccan in 1818 and ended in 1947 after a long campaign for independence. Prior to British conquest, Pune was capital of the Maratha empire, ruled in the 18th century by hereditary prime ministers, known as Peshwas. In 1820, there were hardly any roads and thus very few bullock carts on the Deccan plateau. "Pack bullocks remained virtually the only means of land transport until about 1830" (Guha 1972:21). Pack bullocks were gradually replaced by carts, which could haul heavier goods over longer distances. Carts and roads were both introduced, in mid-century, by deliberate government efforts. Prior to this period, the region was accessible to longdistance trade in high-value goods; but owing to the scarcity of roads and carts, there was little trade in goods which would have been familiar to most peasants. Malegaon Village c. 1885

~uch of this book is concerned with one village, called Malegaon, located in eastern Pune district, about 100 kilometers from Pune city. It is situated next to Baramati, a market town and administrative headquarters for Baramati taluka. (A taluka is an administrative subdivision of a district; it consisted in 1981 of llO villages with a total population of 220,390 and a market town with a population of 37,121.) Baramati is connected with Pune city by paved roads and a railway spur, which were crucial for the growth of the sugar economr.. )See Figure 2.2.) Malegaon is large both in area and population, especially since the Malegaon Cooperative Sugar Factory was established there in 1957. When I first visited the village in 1969, there were two features which attracted my attention. ~ne was the sugar factory, which was causing rapid change in the village and its surrounding area. The other was the mansion of an old aristocratic family, the ]agirdars, or Rajas, of Malegamillhese two structures represented the past and the future; and since I was pr6posing to study change, I thought it might be interesting to discover how they were connected. I noticed that the next-to-last jagirdar of Malegaon, Shambhusingh Jadhavrao, was one of two men memorialized with statues in the guest house of the sugar factory. At this point, I will attempt to describe the social organization of Malegaon toward the end of the 19th century (just before the irrigation canal was

24

t N

()

/

)

I

rv, "'l

c:::::::::======---• km

0

"",

'

100

\,'

')GUJA RAT

\

~..J

_ _ '_J

c..-,

'

200

)

/-J"'\ l_ I\

'

( "'

r

,__ _\

,.,--v 1

r

,..

I...J

.,.,

\

\,

I') DECCAN

I

J Ahmednogor

\

• Poono

l..

......

\

~')

I

0

z ( )::>

l

r'- /

z, /

\

......,) ....... ./ \

\

~......,\/'

"""'-< ~ijopur

• Kolhopur

) SOUTHERN MARATHA ~ ""'\ COUNTRY ro.J

'\.._)

"""

,..( FIGURE 2.1 The Bombay Presidency in the British Period

J \

'--Sholapur e

So!ora •

\

~

?

• Dhulio

'-}

l

~

~' '-)

)

I

FIGURE 2.2

To Bombay

100

Sketch Map of Nira Valley and Vicinity

0

Km 200

To Ahmednogor

�

26

The Setting

built), relying primarily on oral family histories collected in the village. Then I will review the historical literature on 18th-and 19th-century Maharashtra and consider some interpretive issues which bear on this study. The old village elite was composed of several groups. At the top in terms of prestige and power was the Jagirdar, along with a handful of his distant relations. The Jagirdar collected land taxes from the farmers and had a number of other official and quasi-official functions, some of which are discussed in Chapter 5. Around 1900, he owned a sizeable portion of the village lands: roughly 250 irrigated and 200 dry acres, which were mostly rented out to farmers and herders. Some of his former lands had been given to loyal retainers. Another powerful group was the Taware lineage, which, like the Jagirdar, belonged to the Maratha caste-a loosely-defined caste of farmers, warriors, officials, and princes, dominant in most villages throughout the region. The Tawares had settled in Malegaon long before the Jagirdar's family and were the largest lineage, owning the largest share of village lands. Table 2.1 gives a list of castes and lineages in the village c. 1900, ranked according to ritual status. In this table, and throughout the book, the Tawares, along with the Jagirdar and a few of his distant relatives, comprise the "major Maratha lineages" in Malegaon. Leading Taware families held the right to share the headmanship or patilki watan of the village. Patils supervised the collection of taxes and were also peace officers. Only a few Taware families were, in fact, big landowners and endowed with individual power and influence. The majority were small farmers, economically poor and insecure, though compared to most other villagers they shared in the higher status and authority of their lineage. The Tawares, as the largest and perhaps oldest kinship group in the vilhige, had long constituted the social and economic center of village life. Next to them, other lineages and castes were quite,small and, except for the Jagirdar, owned only small portions of village land. The Maratha caste also included minor (non-Taware) lineages. Some had originally been affines of the Tawares who immigrated during famines or political unrest. One minor lineage, the Sastes, were fairly well off, probably due to the length of time they had been settled in Malegaon; but some were no better off than the other peasant castes (Malis and Dhangars), who also lived in the village. Table 2.2 gives a clear idea of the differences in landed wealth between major Marathas, minor Marathas, and other peasant castes, as of about 1920.1 The data, based on oral family histories, are approximate; but the differences among these categories are clear. The major Maratha lineages held the largest share of the village land, the Jagirdar alone claiming about 265 standard acres. 2 The average holding of major Marathas was much greater than that of minor Marathas or other middle castes. In addition to the Jagirdar and leading Taware families, there were other, numerically small, elites with specialized occupations. One such group consisted of families belonging to the Brahman caste. This traditionally

The Setting

27

TABLE 2.1

Distribution of Sample Families by Caste, Malegaon, c. 1900

Caste Category

Number of Sample Families•

Specific Caste

Common Ancestral Occupation

Brahman Gurav Gujar Lingayat

Priest, accountant Shaivite priest Merchant Merchant

Maratha

Major Minor

Farmer Farmer

Q.thu

Mali Dhangar

Farmer, gardener Farmer, herder

3 7

Mid-lowb

Sutar Lohar Kumbhar Bhoi · Nhavi Par it Ramoshi Wadar

Carpenter Blacksmith Potter Palanquin bearer Barber Washerman Watchman Construction

1

Muslim

Various services

1

Chambhar Mahar Mang

Cobbler Village servant Ropemaker, musician

3 6 3

Peasants

3 1 1 2

23 14

7

1

76 • Indigenous families captured in the sample survey of 1970-71. Though not all were captured in the sample, many small, specialized castes were living in the village c. 1900. Source: Sample survey in Malegaon.

b

literate caste of priests and officials served both the state and the Jagirdar. One Brahman family provided hereditary accountants or kulkarnis for the village: they kept the all-important land-ownership and taxation records. Another family provided priests and managers for the Jagirdar's court and estate. In both cases, these families had acquired large landholdings before the turn of the century. One other group was part of the village elite: the merchant castes, or at least their wealthier families. There were two such castes: Lingayat Wanis, originally from Karnataka, south of Maharashtra, and Gujar Jains, originally from Gujarat. (It is interesting that the important merchant castes in rural Maharashtra came from other regions.) Merchants tended, of course, to concentrate in market towns like Baramati. However, since Malegaon was a

28

The Setting

TABLE 2.2

Land Distribution by Caste and Lineage, Malegaon, 1920

1920 Landholdings Caste & Lineage Categories•

swab

Low (Untouchable) Mid-low (Artisan & Service) Other Peasant Castes Maratha: Minor Lineages Major Lineages• High (Brahman & Merchant)

17.48 34.75 33.34 198.57 923.82 305.83

1.09 1. 74 2.08 8.63 41.99 43.69

16 20 16 23 22

1513.79

14.56

104

N

7

• See Table 2.1 and text for explanation of caste categories . Areas are in standard irrigated acres. c The Jagirdar is included among the major Marathas. ~: Sample survey in Malegaon.

.b

big village with a Jagirdar in residence, it supported a few merchant families. The wealthier ones must have had economic influence, particularly as creditors, but it is unlikely they had open political influence. In Table 2.2, Brahmans and merchant castes are lumped together as "high castes." (This lumping is explained in Chapter 7.) These few Brahman and merchant-caste families had average landholdings larger than any other category in the village, even slightly larger than the major Marathas. (If the Jagirdar had been excluded, as an exceptional case, the major Marathas would have had an average landholding of 31.34 standard acres, considerably less than the high-caste average of 43.69 acres.) The landed wealth of the high castes was due to their literacy and their ability to deal with external systems of power and exchange. Of the four groups who together made up the old village elite-the Jagirdar's family, the Brahmans, the wealthier merchants and the bigger Taware landowners-only the latter were connected with a sizeable kinship group inside the village, a point which was to be of great importance in the evolution of local politics after independence. The other three depended for their status, wealth and power on connections with the political economy of the region: the Jagirdar and Brahmans as administrators and deputies of the state, the merchants as links with regional trade and markets. During the British period, these externally connected elites probably grew more powerful, due to the expansion of transport and communication, which caused Malegaon to be more intensely connected with the world at large. As we examine the final half century of British rule, 1900-1947, we shall find the Jagirdar and Brahmans at the height of their local influence. We

29

The Setting

shall also see some of the processes which led the Tawares, along with Maratha farmers throughout the region, to become politically active and eventually (after independence) major players in local politics. Aside from these elite groups, the rest of the village consisted of small and middle farmers, plus artisans and laborers. About 30 percent of the small farmers were Tawares, poor cousins of the elite Tawares. Another 40 percent or so belonged to minor Maratha lineages and other peasant castes (Malis and Dhangars). Malis tended to be gardeners, growing fruits, vegetables and flowers for the market; and Dhangars were often shepherds. Some Dhangars were hired by the Jagirdar to tend his sheep and cattle. Most families of the artisan (mid-low) castes followed specialized, hereditary occupations: Sutars were carpenters, Lohars blacksmiths, Nhavis barbers, and Parits washermen, for example. (In Marathi, these terms name both the caste and its occupation.) Malegaon had a large variety of specialists who could serve the needs of the Jagirdar as well as the farmers. 3 For example, the numerous Ramoshis provided watchmen and hunting guides for the Jagirdar, and the Bhois provided palanquin bearers. Some Muslim families also provided servants for the Jagirdar. (Strictly speaking, the Muslim minority was outside the Hindu caste system, but it may be lumped with the midlow castes in terms of general social status.) At the bottom of the status hierarchy came the castes which were considered untouchable. Mahars were among the larger non-Maratha castes; but as landless laborers and village servants of low status, they had no economic or political power. Mahars provided general messenger and watchman services for the headman and helped to clean the village, while Mangs were musicians (drummers and buglers) in the Jagirdar's service. Another untouchable caste was Chambhar (cobbler), which provided whips, shoes, and leather buckets to the farmers, plus harnesses for the Jagirdar's horses. These castes had to live in a separate area outside the village proper and were not allowed to enter the village Shiva temple or to draw water from wells used by middle and higher castes. Part of the cultural rationale for their low status was that Mahars and Chambhars handled dead cattle, which were polluting, and that all three castes ate beef or pork. Their weak economic position is shown in Table 2.2. The occupations and relative wealth and power of these castes were typical for the surrounding countryside-except that Malegaon was a large village with a Jagirdar in residence, which increased the variety of castes and occupations. Most villages, including Supe, were dominated by large, landowning Maratha patrilineages, like the Tawares; but there were a few villages near Malegaon dominated by Dhangar lineages. (See Orenstein [1965] for a description of a Dhangar village near Baramati; also Carter [1974] and Dandekar [1986] for Maratha villages in a nearby district.) Rural Life Before the British Raj With the example of Malegaon in mind, we can look back at the historical record before 1900, concerning rural life in western Maharashtra. 4 The

30

The Setting

economic and social history of the precolonial period has often been sketched from 19th-century records written in English by colonial officials. There are obvious problems with these sources; but fortunately, recent historians have delved into the masses of 17th and 18th-century Marathi documents which are housed in the Peshwa Daftar in Pune. Descriptions of precolonial rural society, based on primary sources in Marathi, have been essayed by Kulkarni (1967, 1969), Fukazawa (1972, 1982), Perlin (1978) and Gordon (1979, n.d.). The core of any village was its farmers, usually of Maratha caste. Above the farmers, as we have seen, there were a variety of local elites connecting them with the regional economy and the state. Peasant-caste headmen and Brahman accountants, at the least, were found in nearly every village. Village occupations and status were governed to some extent (but not entirely) by hereditary principles. The majority of farmers usually belonged to one or two Maratha lineages. Non-farmers tended to pursue occupations to which their castes had hereditary claims. Most villages in Maharashtra had a collection of village servants and artisans known as the bare balutedar, or "twelve holders of baluta rights." (Their number was not always twelve.) These ranged in status from the Brahman priest, astrologer and accountant at the top to the untouchable rope-maker and servant at the bottom. Among the middle-status balutedar castes who were essential to the village economy, there were the Sutar (carpenter), Lohar (blacksmith), Kumbhar (potter), Nhavi (barber) and Parit (washerman). Each caste was endogamous; and some, such as Sutar and Lohar, were considered higher in status than those, such as Nhavi and Parit, who were concerned with removing bodily wastes for other castes. Each caste often had a monopoly on its occupation within the village and was expected to provide services on demand or on a regular schedule. In the absence of a hereditary blacksmith, however, the carpenter might do the blacksmithing, or a Lohar without hereditary rights might be brought in from another village (Fukazawa 1972). Each balutedar was given an annual or semi-annual harvest payment by the farmers, the amount being more or less fixed by custom and depending only roughly on the amount of work that had been provided to a given farmer in a given season. They also received various minor payments and sometimes plots of inam (revenue-free) land. The baluta system was similar but not identical to the jajmani system of northern India (Fukazawa 1972, 1982:252). Baluta services were owed to, and regulated by, the village as a whole. For example, Mahars were expected to act as messengers and watchmen for the headman, to clean the village (particularly of dead cattle or other animals), and generally to assist its administration as guides, porters, etc. In return, they were often assigned small plots of land and given harvest payments. However, there was probably not enough community work to provide subsistence for all low-caste families. Mahars and Mangs must have been employed as agricultural laborers by individual farmers and paid with harvest shares.

31

The Setting

Communal regulation of the baluta system may have weakened during the 19th century (Fukazawa 1972:40), though two village studies in the early 20th century report that balutedars were still servants of the village (Mann 1917; Mann and Kanitkar 1921). However, Mahar men in large numbers were seeking jobs in the cities, leaving their women behind to retain their land and baluta claims (Mann 1967:73-81). After the sugar economy took firm root in Malegaon (by 1910 or 1920 at the latest) agricultural laborers, who were often migrants, were paid mainly in cash. Traditional baluta arrangements continued for some specialists, but these declined after the 1950s, when the cooperative sugar factory gave a new boost to the commercial economy. New specialists were in demand (to repair bicycles and irrigation pumps, for example), and artisans with new skills wanted cash payments. Farmers came to depend less on traditional crafts and services and were able to pay cash to meet their requirements for equipment repairs and household utensils. Laborers and artisans began to seek wage-paying jobs in the formal sector, and many left for the cities. Stratification Having sketched the social order in Malegaon in the late 19th century, and having noted some parallels with descriptions of village life in the 18th century, we now turn to interpretive issues which bear on the subject of this book. As noted in Chapter 1, there has been a tendency to idealize village life in the pre-British era. The ideal "Village Community" has been described as a self-enclosed, self-sufficient, and self-regulating society, which functioned quite independently of the outer world, although compelled to pay tribute to it. Passages from the early records of the East India Company have been quoted time and again to support this image (see, e.g., Mukherjee 1974: 156-57). The image of the Village Community has reappeared in recent scholarly work on 18th-century Maharashtra. The village has been described as a "miniature world, self-sufficient in itself" (R. Kumar 1968a:16). While admitting that there were status differences inside the village, serious inequality in wealth or power is denied: "The absence of sharp differences in incomes in the villages of Maharashtra created a climate devoid of conflict and strife." Even the untouchable Mahars "possessed an important voice in the community despite their low social ranking" (R. Kumar 1965a:205, 209). This appealing image of the precolonial Village Community was based on 19th-century documents written in English by colonial officials. The image has been seriously challenged by historians using primary sources in Indian languages from earlier periods. For example, the myth of a community of equals has been demolished by research on south India (D. Kumar 1965) and on the north: There were large cultivators, using hired labour, and raising crops for the market; and there were small peasants who could barely produce foodgrains

32

The Setting

for their own subsistence. Beyond this differentiation among the peasantry, there was the still sharper division between the caste peasantry and the "menial" population, a primitive landless proletariat, which served as the reserve for supporting peasant agriculture. (Habib 1982:247) The situation was similar in 18th-century Maharashtra, where there was "remarkable economic differentiation among the peasants" (Fukazawa 1982:259). For example, the village offices of headman (Ratil) and accountant (kulkarni) "provided the opportunity for significant economic and social differentiation, for each of the important posts offered its own source of power" (Charlesworth 1985:23). The literate Kulkarni, usually a Brahman, controlled the land and tax records. The headman, usually a Maratha, enjoyed a variety of perquisites, including a share of revenue collections, various gifts and services, sometimes tax-free land, and the right of precedence during village festivals. These advantages gave leverage to their holders, which enabled some to accumulate land, wealth and power. Even great officials and nobles claimed the headmanship of certain villages and made a point of maintaining these claims. There were also hereditary district officers, known as deshmukhs, who lived in the countryside and were active in its administration. In addition, there were deshpandes, jagirdars, inamdars, saranjamdars, and rajas, who also lived in the countryside and wielded various kinds of power. The Jagirdar of Malegaon was just one example. Many notables held large landholdings, often as inams: that is, as revenuefree estates. From the point of view of the Peshwa's government in Pune, such lands were "alienated" to their holders for a variety of purposes. Some alienations were to support military commanders or officials for their services to the state, others to support religious institutions. There can be no doubt that during the late eighteenth and early nineteenth centuries there was a massive extent of alienated land and revenue throughout the Bombay Presidency. Besides the whole villages and groups of villages held within jagirs and inams, a significant proportion of the land of the typical village was alienated. (Charlesworth 1985:28) Most of the alienated lands were held by those who stood above the ordinary ruck of peasants. It used to be assumed that the structure of state power and the structure of village society were two distinct entities. However, recent research has shown that state officials were actively engaged in "penetrating" village life (Perlin 1978). Families with one official title or another would use their leverage to extend control over other resources. Leverage operated in various ways: by interceding with other officials, settling disputes, extending credit, and employing armed troops. These devices enabled some families to buy more lands, acquire official appointments, and even buy offices. For example, the patilki watan (village headmanship) could be divided, bought and sold.

The Setting

33

Mann (1967:133-34) records a case where half a headmanship was bought in the 18th century for Rs.3,000-which seems too large a sum for a village title and 36 acres of land. Mann reasons that this was an investment in prestige and in leverage: that is, in "the opportunity for other and less legitimate forms of gain." A variety of office-holders were engaged in credit and patronage transactions with all kinds of villagers. Their networks extended into many villages, even where they had no official titles or lands (Perlin 1978). Like the Jagirdar of Malegaon, they encouraged artisans, merchants and servants to settle wherever they resided. The Jagirdar rewarded many specialists and servants with gifts of inam land. Dhangars received land in return for tending his flocks, Muslims for tending his orchards, Ramoshis for acting as watchmen and hunting guides, Mangs for providing music on ceremonial occasions. This process rewarded and protected specialists who were not in great demand by ordinary farmers. At the same time, it also contributed to social and economic differentiation, since the Jagirdar gave much larger pieces of land to his Brahman priest, and to a Brahman doctor, than to his lower-caste servants. Thus it makes little sense to think of 18th- or early 19th-century villages as homogeneous communities of middle peasants, nor of the state as a superstructure whose operations lay outside the community.

Risk and Mobility The social order of precolonial Maharashtra was also not as static as the Village Community image suggests. For example, Mann (1917:43-44) published a series of figures from the revenue records of one village, dating from the late 18th and early 19th centuries. As Fukazawa (1982:254) has noted, the data show considerable fluctuation in land distribution. Moreover, the total number of landholders varied widely, nearly doubling in 20 years. Similar instability was evident in another 18th-century village, where there were also large fluctuations in the annual land revenues (Mann and Kanitkar 1921:40-42; Mann 1967:128). A variety of factors probably contributed to this unstable pattern. At times, the villages were devastated by plague, famine, banditry, or warfare. On such occasions, there was considerable migration and resettlement. Western Maharashtra is a semi-arid region with frequent, severe droughts. The present government has an efficient network of roads and famine-relief plans to prevent deaths from major crop failures. However, the Peshwa's government of the 18th century lacked these resources, and migration was the only option for many villagers when their crops failed. This was true even in the late 19th century. In Ahmednagar district, after the famine of 1877, "Over 18 per cent of the district's population had migrated to the neighbouring district of Khandesh in search of food and employment" (Baviskar 1980:21). One of the worst famines on record, in 1630-31, was described by several observers.

34

The Setting

According to the contemporary poet-saint Ramdas, most of the people were reduced to beggary, which rendered begging fruitless. Villages were depopulated as some people died of starvation and others were forced to leave their homes in search of livelihood elsewhere. (Kulkarni 1973:560)

Some migration was temporary and reversible; but prolonged droughts undoubtedly led to permanent resettlement by those who survived. In an unpredictable way, then, famines weakened or dispersed the demographic strength of families, lineages, and villages, and thus undermined their economic and political strength. Subsequently, the survivors were forced to rebuild a social order with scattered fragments of the old groupings. Moreover, famines were not the only general crises which could have these effects. Gordon (1979) has shown that prolonged political instability could be more devastating than occasional famines, leading to the depopulation of entire districts. Early British revenue officers found areas where the best lands were only recently settled; and there were some large and fertile districts, such as Khandesh, which remained partly depopulated. If we look at the kinship units, the families and lineages, which make up villages in Maharashtra today, we do not see the residue of a static social order. Most villages have one or two dominant lineages, which are the largest in population, own most of the land, and formerly held rights to the headmanship, the patilki watan. These are ramified, extended patrilineages of great genealogical depth: they have obviously dwelt in the same place for a long while, giving them time to grow and divide into many branches. On the other hand, villages also consist of an accretion of minor or "broken" lineages (see Carter 1974; Dandekar 1986). Often, lineage members recall that their ancestors migrated from another place, the mulgaon or root village, which they occasionally revisit for religious observances. These minor lineages are found in all types of villages, including those which have been relatively untouched by commercial agriculture. (I did intensive fieldwork during 1969-71 in one such village-Supe, not far from Malegaon.) Fukazawa (1982:251) and other historians mention the "strangers" or uparis, who were a recognized category of cultivators and tax-payers in revenue documents from the 18th century. Some uparis were surely ancestors of the minor lineages found in many villages today. State officials sometimes encouraged uparis to migrate from other regions by offering them concessional tax rates, in efforts to compensate for temporary declines in population and revenue brought on by famines and periods of political instability. Migration had a lot to do with shaping the society of precolonial Maharashtra; and Ludden (l988b) has concluded that the whole of peninsular India was reshaped by large-scale migrations after about 1300 A.D. Migration was not only a response to crises but also to better opportunities for employment, patronage, trade, irrigation, or military campaigning. In Malegaon, the Tawares say they migrated from a root village in Satara district, as did other Tawares living in nearby Sangvi village. They entered the new area as uparis, but they eventually displaced whoever had dominated

The Setting

35

Malegaon before their arrival. The earlier inhabitants, after whom the village may have been named, either died out or were driven out. If environmental and political instability were threats to the continuity of the village social order, they were even more threatening to the continuity of individual families. The eldest villagers living in 1970 could remember a time when famines and plagues were common, when infant mortality rates were high, when life expectancy was low, and when, as a result, family continuity was more problematic. In these respects, conditions at the end of the 19th century probably resembled those a century earlier. For a family to farm successfully, it had to have a continuous supply of young men and women to take over the heavy tasks. Likewise, for a family to become or remain powerful, it had to have a continuous supply of young men, guided by elder men of experience. Even in recent generations, a father who died young bequeathed to his heirs a much lower chance for economic success. (See Chapter 6 for examples.) Elderly parents with no living sons would also be in bad straits. Sons were often lacking due to high rates of infant mortality. There were several tactics for dealing with this problem. In the first place, women were expected to marry at or before puberty, and their highest social and cultural goal was to bear sons. If a wife bore no sons, or if they did not survive, a man could legitimately take a second wife. Another solution was to adopt a son, perhaps a brother's son (if he had several). Otherwise, if the family had a marriageable daughter, it could select a son-in-law from a family with many sons and less land, and invite him to live and work with his wife's family as a ghar jawai ("house son-in-law"). Marriage patterns also reflected the need to cope with threats to family continuity. Following what appears to be a long-established pattern, families from dominant lineages generally marry similar families in villages at some distance: a separation of 10 to 50 miles is common (Carter 1974:94-95). The two families tend to maintain long-term social ties and may arrange further marriages with each other, since matrilateral cross-cousin marriage (a man marrying his mother's brother's daughter) is considered a good idea. If your affines provided a good wife .to you, they will probably do equally well for your son. In other words, the families see a value not just in arranging one marriage (which is the limit of their relationship in northern India) but in maintaining a connection over time. In-laws are considered potential allies in case of trouble: they are often used as a source of credit, for example. Moreover, in-laws have a foothold in distant locales, where famines or political disturbances might be less severe. This helps explain, I believe, why connections with the same families were maintained over generations (as in south India), while affines were often sought in rather distant villages (as in North India). This combination of tactics was suited for maintaining lifelines in other districts, where refuge might occasionally be required. Oral family histories from Malegaon tell repeatedly that, in times of localized crises, people fled first to their in-laws. These oral histories do not

36

The Setting

extend back to the 18th century, but there is little reason to doubt that the same pattern prevailed then. Migrants would have stayed temporarily as uparis or "strangers" with their in-laws, and if conditions in the new village were favorable, they might have settled permanently. For example, in Malegaon there is a hamlet called "Pawnewadi." This means, literally, hamlet of the in-laws, pahune. This place is inhabited by a minor Maratha lineage, the Sastes, who are settled in such numbers that they must have been there for at least a century. Their ancestors were affinal immigrants, and might have been registered in the tax roles, initially, as uparis. In short, it seems unreasonable to assume that the precolonial village had a stable, even rigid, structure, as the Village Community image suggests. Rural society in Maharashtra had to be flexible in order to cope with environmental and political crises. Some localities were more affected by instability than others: for example, the rains were more reliable near the coastal mountains than in the famine belt farther east; and wells were more reliable in some areas than others, due to variations in topography and drainage patterns. Consequently, there were some areas where instability reached its peak, while in more favored areas there was probably more social continuity. Nevertheless, the whole social order was affected to one degree or another by the necessity to adapt to an unstable environment. Competition and Mobility

The previous section, on risk and mobility, is consistent with the widely held assumption that peasants are essentially "risk avoiders." But there was another side to their response to uncertainty and instability. Many villagers were actively engaged in seeking opportunities through political and military enterprise. This can been seen from the "job description" of those political elites who were in close contact with the peasants. Above the village headman was usually a district officer, a deshmukh. According to Gordon (n.d.), the deshmukh occupied a pivotal position in the countryside, interacting on a face-to-face basis with both the state ruler and the villagers. A deshmukh was a hereditary grant holder who was expected to mobilize troops at the request of his ruler. His family often claimed a variety of "nested" rights, such as the headmanship of specific villages, the command of a fort, and the ownership of revenue-free lands. In times of peace, he served as a revenue and judicial officer for the state. In less settled times, which were frequent, he acted as a military entrepreneur, taking sides in factional disputes and wars of succession. His powers and privileges were subject to constant renegotiation and often depended on the fortunes of war. Villagers also counted on the fortunes of war. The Maratha armies, which fought almost continuously in the 17th and 18th centuries, and which sooner or later campaigned in every region of the subcontinent, were peasant armies. Particularly in the 17th century, most of the troops returned home during the monsoon season to plant their crops (Gordon n.d.). Military action was

The Setting

37

thus a potentially profitable sideline, which might lead to opportunities to command troops in the field and thus win land and revenue grants from the ruler. Gordon portrays the Maratha polity as a field of intense competition, from top to bottom. But what were people competing for? Not land, which was abundant and relatively unproductive, nor control of trade. At least in the 17th century, the Deccan region had a low level of trade and monetization, and no large-scale trading cities. The main objects of competition were various offices yielding shares in the produce of the land. Ordinary villagers were striving to become headmen, headmen to become deshmukhs, and so forth. The chief tools at their disposal were political and marital alliances, plus military skill and leadership ability. A number of Maratha (and Dhangar) princely families-the Gaikwads of Baroda, the Sindhias of Gwalior, the Holkars of Indore, and the Bhonsles of Nagpur-started out in the early 18th century as ordinary field officers. For their successes in battle and their loyalty to the Peshwa, they were granted large estates outside the core of the Maratha polity. Originally granted as saranjams (non-hereditary estates for the support of troops), these were later consolidated into princely states. Shivaji, the 17th-century founder of the Maratha kingdom, also started out as a minor field commander and grant holder (Gordon n.d.). Political competition worked with migration to alter the composition of the villages. For example, one Taware sub-lineage recalls that their ancestors were forced to migrate temporarily to Nagpur, due to disputes with the Jagirdar. Shifts in local political power were both the cause and result of migration and of fluctuations in the demographic strength of competing families and lineages. Some shifts were no doubt rapid and dramatic, others slow and insidious. Every village seems to harbor stories of old, protracted struggles between families, sub-lineages, lineages, and even castes over ritual precedence and political power. The holding of power and precedence was manifested, for example, in the acknowledged right to lead a procession of bullocks round the village in the annual bail pola festival. This was not simply a right granted by inheritance nor by action of the state authorities, though both these factors played a role in the succession to the headmanship. Power was also the outcome of a competitive process. Power meant the opportunity to buy a share of a headmanship, to collect taxes, to hold revenue-free land, to build a following by advancing credit, to mobilize troops. Aside from the gratification of lording it over their fellow villagers, competitors were seeking to protect their families against the political and economic hazards of the regional system. One of the principal techniques of competition was building patronage alliances up and down the ladder, just as is done today. Competition and mobility had their effect on the caste system. The latter is generally assumed to be supremely rigid, with perfect continuity in membership and status from one generation to the next. Yet some 40 to

38

The Setting

50 percent of the rural population now belong to one caste, the Marathas, which seems to have flexible boundaries. A century ago, most Marathas were called "kunbis," which simply meant "peasant." At that time, the "pure Marathas" or "Kshatriya Marathas" were considered to be members of highstatus lineages with hereditary claims to high office. However, in the early 20th century, these two types of Marathas joined forces in the non-Brahman movement. (See Chapter 5.) Previous status distinctions, which were ambiguous in any case, were submerged; and the term "kunbi" fell out of use. That such a process could occur in response to a political movement suggests that the categories of Maratha and kunbi were not very distinct in earlier times. Gordon (n.d.) observes that the term "Maratha" emerged centuries ago as a designation for military leaders and grant holders; and there was much mobility in these categories. It has been suggested that there was formerly a hierarchy of Maratha and kunbi lineages, ranging from royal to common, each intermarrying hypergamously with lineages adjacent in the hierarchy (Karve 1968). This may have been true, but it suggests a picture which is far too simple. In reality, every lineage had its cadet branches, poor cousins, adopted heirs, illegitimate sons, and broken segments, which overlapped in status, power, wealth, and connections with others (Omvedt 1976:131). Local lineages were internally stratified, with some families holding the village headmanship, for example, while others were just scraping by. Hypergamy, far from establishing a clear hierarchy among lineages, probably served as a device for competitive mobility and thus as a source of complexity and fluidity. A village family newly risen to wealth and power would attempt to marry its daughters into an established patil or even deshmukh family. At first, only an impoverished deshmukh would agree to such an offer, in exchange for a commodious dowry. In the course of a generation or two, however, the successful pursuit of such a strategy could lead to connections with deshmukh families in better circumstances. Finally, after another generation or so, there might come a strategic triumph: the reverse offer of a deshmukh daughter to the newly-risen family, at which point its status would be unassailable (Baviskar n.d.) Ordinary gossip about some deshmukh families suggests that such processes used to be common. Consequently, I doubt if the terms "Maratha" and "kunbi" ever had very distinct referents, and I take this as another indication of a flexible social order. Even today, for example, there is a small, local caste of farmers known as Karekars in Ahmednagar district, who are not normally considered true Marathas; yet some of the more successful Karekar families have intermarried with Marathas (Baviskar 1980; n.d.). I suspect this process has occurred continuously over the last few centuries and take it to mean that the web of social relationships was always ready to accommodate enterprising and upwardly mobile families, even if the vocabulary of status, kinship, and caste was inherently conservative. We might recall that the founder of the Maratha kingdom, Shivaji (16301680), had a grandfather who was a cultivating village headman. Because

39

The Setting

he was not from an aristocratic, "twice-born" family, Shivaji faced some delicate problems when he decided to be crowned as a monarch, a Kshatriya king. He hired a distinguished Brahman from Benares, who "found" a genealogy showing that Shivaji was descended from Raj put rulers in Rajasthan. Once this problem was settled, the Brahman more or less invented a coronation ceremony-since, under the Mughal empire, no independent Hindu king had been crowned for centuries (Gordon n.d.). Shivaji's career illustrates not only the potential rewards of military enterprise but also the opportunities for manipulating symbols of "ascriptive" prestige. Maharashtra, during the 17th and 18th centuries, was a competitive, mobile and highly stratified society, just like today-though it operated on different principles from the more bureaucratized and commercialized society of the 20th century. Uncertain rainfall was one cause of flexibility in the social order, while the tradition of political competition and military enterprise was another. Change in the 19th Century There is an increasingly sophisticated historical literature, covering a wide range of issues, on changes in 19th-century rural Maharashtra. McAlpin (1983), Guha (1985) and Charlesworth (1985) have pored through the voluminous Survey and Settlement Reports, published in the late 19th and early 20th centuries by the Government of Bombay. These reports, prepared separately for every taluka and updated every thirty years or so, contain masses of information on local prices, rents, land revenues, weather patterns, crop yields, investments, land sales, and so forth. Thus these historians have based their conclusions on local-level economic data, often supported with quantitative analyses of trends and variations. Far more than previous studies, their work is sensitive to the mixed effects of macro-level changes in, for example, government policies and transportation. They are also more sensitive than previous historians to the economic impact of ecological variation. I shall not attempt to summarize all the issues and debates concerning change in the 19th century. Instead, I shall touch on a few points which lead toward the origins of the sugar economy at the end of the century. In order to assess the impact of the sugar economy, we need to know whether 19th century Maharashtra was radically transformed by the British Raj. Historians are not yet agreed on whether crop yields and average incomes were rising or falling in the 19th century (see McAlpin 1983; Charlesworth 1985; Guha 1985). Given the limitations in official statistics concerning crop yields, this issue may never be settled. However, there may be a growing consensus on land revenues, which were once thought to be a major cause of impoverishment. It seems clear that revenue rates were set at high levels before the middle of the century, causing a general contraction in the cultivated area, but that the rates declined thereafter (Charlesworth 1985). One reason was that foodgrain prices rose, meaning that the same revenue could be paid in cash by selling a smaller portion of the crop (Guha

40

The Setting

1985:144). (Revenue rates were fixed for thirty years at a time.) Moreover, the cultivated area expanded. Though total revenues increased, they did not grow as fast as the cultivated area, so the average revenue per acre declined (McAlpin 1983:249-60). The expansion of trade, due to construction of roads and railroads, has been well documented; but opinions are still divided on how this affected ordinary villagers. Railways enabled farmers in remote districts to produce and sell cash crops, such as cotton, which were then shipped to markets overseas. One question is whether this also depleted villages of the foodgrain reserves needed to tide them over during severe droughts. The last quarter of the century, when more railways were being built and more crops exported, also included severe famines, with great loss of life. Did the railways increase the threat of famine? It used to be assumed that, when farmers were tempted or coerced into growing cash crops, such as cotton, they grew proportionately less of their food crops. However, Harnetty (1971) has shown that the acreage of both cotton and foodgrains rose during the 1860s, when there was a boom in the cotton export market, thanks to the American civil war. Cash cropping stimulated foodgrain production, since there was more money available to invest in bullocks, carts, implements, and wells. McAlpin (1983:145-47) finds that the proportion of land devoted to foodgrains remained stable from 1860 to 1900. Thus the export market did not tempt farmers to abandon their subsistence crops. The expansion of the railway network stimulated exports, but McAlpin (1983:151-53, 175-76) has shown that the railways also served a protective function, moving foodgrains from surplus to deficit areas when crop failures occurred. Indeed, she finds that the volume of foodgrains shipped internally, from one part of the country to another, grew at a faster rate than the volume exported. Other studies have found that the railways caused prices to converge in the districts which they connected (Hurd 1975, 1983). Previously, prices could be extremely high in one district, due to crop failure, but nearly normal in other districts not far away. Karl Marx summarized the problem in one of his letters on India: Nowhere, more than in India, do we meet with social destitution in the midst of natural plenty, for want of the means of exchange. It was proved before a Committee of the British House of Commons, which sat in 1848, that "when grain was selling from 6s. to 8s. a quarter at Khandesh, it was sold at 64s. to 70s. at Poona, where the people were dying in the streets of famine, without the possibility of gaining supplies from Khandesh, because the clay roads were impracticable." (Marx 1853) The distance from Khandesh to Poona was only about 200 miles. Thus the convergence of prices in distant areas, after the introduction of the railway, showed that foodgrains were being moved to districts where they were most needed in times of famine (see Keatinge 1912:150).

The Setting

41

That being the case, it is still not entirely clear why famine mortality reached an abrupt peak at the end of the 19th century and declined thereafter. Unusual weather patterns may have been responsible for the great famines of the 1890s (McAlpin 1983). Moreover, the government evolved a muchimproved famine-relief administration after the turn of the century. McAlpin also believes there was general improvement in economic conditions, which enabled farmers to save, invest, and ride out crop failures after 1900. Charlesworth (1985), though more cautious, also seems to believe that there was overall improvement up to about 1920. However, others emphatically disagree, taking the great famines as indicators of long-term economic decline under the British raj (Guha 1985). One must ask, then, why famine mortality declined so markedly during the final half-century of British rule. This book offers no general answers to these questions, but it does offer some points worth pondering. The sugar economy began to flourish after 1900, during the period when famine mortality declined sharply. Concerning regional security from famine, sugarcane was part of the solution, not part of the problem (as explained in Chapter 3). While the causes of demographic trends are still in dispute, the trends themselves have been clearly documented by decennial censuses since 1871. Starting with the first census, the population of western India alternately rose and fell, due to famines and epidemics in the 1870s, 1890s and 1910s. After 1921, the population curve took a steady upward course, due to declining infant and general mortality rates (Visaria and Visaria 1983). The causes of this remarkable transition, due to some combination of economic improvement, better nutrition, better famine relief, and better public health measures, do not seem to be fully understood. If there was a general improvement in economic conditions, as some believe, this probably did not last into the depression of the 1930s, yet mortality continued to decline. The last great killer was the influenza epidemic of 1918, remembered by some of the older villagers in 1970. Thus the connections between economic and demographic trends are not yet fully agreed on. Charlesworth (1985) may have pointed the way toward better understanding by disaggregating the data, both spatially and temporally. He shows that there were medium-term cycles of improving and deteriorating economic conditions and also that trends varied considerably in different ecozones. What, then, was happening to land tenures and general levels of living in those parts of the Bombay Deccan affected by irrigation and commercial agriculture? And what was the fate of the drier areas? The answers are not certain, but recent work suggests that the trends were less alarming than once supposed. Some historians believe that small peasants became heavily indebted and sold off much of their land in the late 19th century. In 1875 the government became intensely concerned with this perceived problem, due to the outbreak of the "Deccan Riots." These riots erupted in eastern Pune and Ahmednagar districts, spreading rapidly among 33 villages in the famine zone. (Supe,

42

The Setting

where I did fieldwork in 1969-71, was one of the first villages involved in the riots.) The pattern followed by the rioters was consistent from village to village: generally they broke into the shops and homes of village moneylenders, sometimes looting the contents; but they always concentrated on seizing and burning the mortgage deeds and other records which the moneylenders had in their possession. There was remarkably little violence against persons. The government immediately launched an inquiry into the causes of these disturbances. (See Deccan Riots Commission 1876). Far from settling the question, the Commission's report became a basis for controversy among historians concerned with agrarian change. (See Kumar 1965b, 1968a; Catanach 1966, 1970; Charlesworth 1972, 1985; Guha 1985). Long before the British period, villagers had depended on creditors to tide them over the years of recurrent drought. However, the government feared that, by 1875, moneylenders (sa.vka.rs) were foreclosing on mortgages at an increasing rate, forcing cultivators to sell off ancestral lands and become their tenants. Later historians have used the riots, and the government's anxiety over their cause, to argue that there was a social revolution occurring in the countryside, brought about by British laws of private property and the British court system, through its enforcement of mortgage contracts (see Kumar 1965b, 1968a; Banaji 1977; Natarajan 1979). There is little statistical evidence to support this argument. It is clear, at least, that moneylenders were resorting increasingly to the new legal system to enforce their contracts. In Supe in 1867, there were only seven mortgage deeds registered, while in 1874 there were 33 (Deccan Riots Commission 1876: 317). In the district courts, the number of lawsuits relating to agricultural debts doubled in about the same period. Moneylenders were nearly always the plaintiffs and over 90 percent of cases were decided in their favor (Charlesworth 1972:406-07). The great majority of cultivators were illiterate and at a disadvantage in dealing with the alien procedures of the courts. Nevertheless, there is little evidence that lands were becoming increasingly concentrated in the hands of moneylenders. By 1875, moneylenders belonging to merchant castes owned only a small fraction of the cultivated land in the region, nor was their share greater in the riot areas than in nearby locations. Moreover, their ownership was extremely patchy, concentrated near a few market centers and otherwise sparse (Charlesworth 1985:104-06). The riots thus did not signal the arrival of an agrarian transition which was dispossessing the peasants of their lands. Even fifty years later, there was no evidence of such a trend: "despite all the dire prophecies of concerned officials, all the land in the Deccan did not pass into the hands of sowkars [sa.vka.rs ], to whom the peasants were indebted" (Guha 1985:145). Even up to 1936, there seems to have been no net addition to the share of landholdings owned by moneylenders from the merchant castes. Landownership, in fact, was not a particularly good investment for them (Guha 1985:143-47). As we shall see in Chapter 4, merchants probably did not want to tie up their capital in land, when it

The Setting

43

would earn more from trade and moneylending. Foreclosure was probably the last resort in recovering a debt, and the land was usually resold to another cultivator. There was an old saying in the villages that, "the Kunbi is the Marwari's [moneylender's] cow ... and is too valuable an animal to be allowed to perish" (Charlesworth 1985:103-04). If the moneylenders were not grabbing a large share of the land, what was the cause of the riots? Economic conditions in the riot areas had gone up and down. The political peace and stable revenue system which the British had established by the middle of the century brought increasing prosperity to many villages. Improved transport and marketing facilities strengthened this trend, particularly during the U.S. civil war, when cotton prices boomed. For example, in Petha Supa [a revenue unit like a small taluka, centered on Supe village]-the area where the riots first broke out in 1875-the cultivated area increased from 114,000 acres at the time of the first survey settlement [1843-44] to at least 142,000 acres in 1873. In the same period, it was calculated, population increased by 43 per cent, working bullocks by 10 per cent, carts by 220 per cent, ploughs by 31 per cent, and wells in use by 71 per cent. (Catanach 1970:16)

To some observers, the riots indicated that financiers more than producers were taking advantage of the expanded market system. The Deccan Riots Commission looked for, and discovered, evidence of increasing agricultural indebtedness, which they viewed with alarm. However, this indebtedness was perhaps the result of the contraction of trade after the bonanza conditions of the American Civil War period, a contraction which coincided with conditions of scarcity, if not of famine, in 1870-71 and 1871-72; these conditions, therefore, did not necessarily represent a long-term trend. (Catanach 1970:19)

Moreover, in 1867, the government had started introducing a new revenue assessment in nearby areas, demanding higher taxes in a period which was turning from a boom into a recession. (The new rates of assessment were based on crop prices from the 1860s, and were thus over-optimistic about the future prosperity of the area.) There was little correlation between areas hit by the new assessments and villages where the riots occurred, but fear of the new rates may have triggered the disturbances. What characterized the riot villages more than anything else was their poor soil and precarious climate, since these villages were in the heart of the famine zone. One official wrote in 1843 of an area which would prove highly turbulent in 1875: 'there is a tradition amongst the people, that the country between Jejuri and Baramati was formerly cursed, the ill-effects of which remain to this day in the failure of the periodical rains.' (Charlesworth 1985:110-11)

44

The Setting

This was precisely the area around Supe and Malegaon. These villages had shared to some extent in the price boom of the 1860s, but conditions in the 1870s were worse. Their high level of debt was an inevitable result of precarious rainfall, combined with a slump in crop prices. The government attempted to promote a more even balance in credit relations by passing, in 1879, the Deccan Agriculturalists' Relief Act, which inhibited the transfer of land from peasant debtors to their "nonagriculturalist" creditors. This had the effect of drying up some of the credit available from merchants and of encouraging "moneylenders of peasant origin" (Guha 1985:147). Thus we come to the question of the "rise of the rich peasants." It has been argued that a class of rich peasants arose in Maharashtra at the end of the 19th century, taking advantage of new opportunities for cash cropping and investment, and ultimately depriving other peasants of their share of the land (see Kumar 1968b; Banaji 1977; Charlesworth 1978). One way the rich peasants could increase their landholdings was by extending credit to others, then foreclosing. Here again, there is little evidence of a trend toward land concentration. Small peasants do not seem to have been more encumbered by debt than large ones; and those who were heavily in debt in 1875 had mostly recovered by 1889. Small peasants benefited, as well, from the expansion of off-farm employment, and from the fact that "market wages were considerably higher than the earnings of self-employed farmers" (Guha 1985:151-54). Guha notes that smallholders would go to great lengths to retain ownership of their lands. On the other hand, "when a holding grew beyond a certain size it would be handicapped by the need to employ increasing amounts of expensive wage labour-something that the small holder could to a large extent avoid." Moreover, large holdings would be divided on inheritance, since sons were entitled to equal shares. Consequently, the statistics show a decline in the total share of land concentrated in large holdings (Guha 1985: 155-58). In western India, cash cropping led to what Charlesworth (1985:14951) calls "differential commercialisation," which was not the class polarization suggested by classic Marxian analysis. Due to variations in soil and climate, some areas were able to take up cash-cropping, while others were not. Parts ofGujarat and the Southern Maratha Country (as well as the Central Provinces) specialized in cotton production. Tobacco was mostly grown in central Gujarat, groundnut in Satara district, and sugarcane along the new Deccan Canals. In the Deccan districts outside the canal tracts, the dry villages stagnated. Levels of prosperity changed more between villages than within them. Charlesworth seems to have revised his earlier position on the rise of the rich peasantry. He still concludes that, "Initially, commercialization seemed to favour, as traditional orthodoxy has it, rich peasant elites" (1985:293). However, these rich peasants did not emerge from a homogeneous society; they replaced an hereditary elite of district and village officers who had been

45

The Setting

on the decline (due to the centralization and bureaucratization of the British raj). Charlesworth's conclusion regarding the later effects of cash-cropping is less orthodox: After f900 commercialisation, in the more advanced regions and localities, had eaten into the heart of the village economy and its operation, but, coincidentally, the distribution of rewards from commercial agriculture and economic opportunity in general seems to have become more broadly based .... In the Bombay village a range of groups were increasingly able to establish some stake in commercialisation and poor peasants maintained their position in the village rather than running proletarianised from it. (1985:293, 299)

In fact, Charlesworth concludes that there was actually too little social transformation of the countryside, resulting in the dissipation of the possible advantages of commercial agriculture. Instead of creating concentrations of capital resources, the income from cash-crop cultivation was used, in a type of conservative egalitarianism, predominantly to preserve the village social system. (1985:299)

This may have been true for the drought-ridden Deccan as a whole, but surely not for the canal villages, which produced strong concentrations of capital and entrepreneurial talent, as we shall see. Conclusion Archival records will eventually provide a rich portrayal of the strategies employed by enterprising peasant families in the precolonial period. These families sought to avert climatic and political risks; to expand their families and alliance networks; and to compete for wealth, prestige and power. Our assessment of the 19th century obviously depends on what we think carne before it. If we cling to the Village Community image, it is easy to imagine the 19th century as a period of great social disruption. I believe it was more fundamentally a period of continuity. Granted, Maharashtra was politically more peaceful and bureaucratic, economically more commercialized. This did not mean, however, that the villagers were exposed for the first time to risk, competition and inequality. Their way of life was already imbued with strategies for avoiding the negative consequences of risk and competition, while grappling for the positive ones. As we shall see, the framework of risks and opportunities changed dramatically with the rise of the sugar economy. Yet many of the social, economic and political skills which villagers had put to use in the 18th century could be adapted to the new regime. More than that, these skills could reshape the political economy to suit the rising commercial peasantry.

The Setting

46

Notes l. There are two reasons why the number of families in each caste category is slightly different in tables 2.1 and 2.2. Table 2.2 includes some families who immigrated to Malegaon between 1900 and 1920. On the other hand, some families are missing from this table, due to incomplete data on their landholdings in 1920. 2. In tables throughout the book, landholding data are compiled in "standard irrigated acres." The nature of this procedure, and the reasons for it, are explained in Chapter 7 and the Appendix. 3. Table 2.1 enumerates only families which happened to be included in my sample survey of 1970-71. Thus it under-represents the small, specialized castes which were serving the Jagirdar at the end of the 19th century. 4. Malegaon is not presented as a typical or average village, though I have pointed out its similarities to other cases. It simply gives us a concrete picture of one village, c. 1900, which can be used as a basis for questions and comparisons.

PART TWO

Sugar Production Before Independence

3 Irrigation and Imperialism What are the conditions which encourage or compel subsistence farmers to undertake cash cropping? And what are the consequences for their subsistence security? There are two main schools of thought on these questions: one maintains that new markets and technologies have, on balance, brought more benefits to village cultivators; and the other maintains that such changes have done far more harm than good. Western market economists form the core of adherents to the former school, while dependency and world-system theorists are among those committed to the latter. Market economists tend to assume that villagers are independent decision-makers, seizing on new entrepreneurial activities as opportunities arise, while dependency and worldsystem theorists tend to assume that villagers are coerced into new productive activities by powerful landlords, merchants, industrialists, or state officials (see, e.g., Wallerstein 1974, 1980). The historical role of coercion in shaping some agrarian systems, such as the sugar plantations of the Caribbean or the cotton plantations of the US south, is undeniable. However, it is the argument of this chapter, and the next, that there are other cases in which coercion plays a minor part in the transformation of subsistence agriculture. The rise of sugarcane cultivation in the Bombay Deccan is such a case. As we have seen, agricultural production was precarious in the Deccan districts and relied mainly on drought-resistant varieties of jowRr (sorghum) and bt~jri (millet). Starting in the late 19th century, parts of the Bombay Deccan were transformed by the construction of large-scale irrigation canals, which were intended to provide famine protection through the extensive irrigation of subsistence crops. However, the canals failed to provide water effectively for this purpose. In 1901, a new set of irrigation policies was tried out, leading to a more intensive use of canal water for an expensive and thirsty cash crop, sugarcane. Ever since, sugarcane has been the basis of increasing prosperity in the canal villages, while the majority of villages remain dry and poor. The first objective of this chapter is to understand why there was a policy shift toward intensive irrigation. A second and related objective is to determine whether this shift was a threat to the subsistence security of the region. As noted in Chapter 1, Indian nationalists have argued that the British government 49

so

Irrigation and Imperialism

encouraged cash crop production at the expense of subsistence crops, thus rendering the country more vulnerable to famines. The same theme has been taken up and elaborated by dependency theorists and others in recent years. For example, Whitcombe argues in her (1972) book on northern India that peasants were coerced into growing cash crops under canal irrigation; and as a result, the region became more vulnerable to famines. In response, this chapter will attempt to answer the following questions: First, whose interests were at stake and what were the causes leading to the shift toward a more intensive irrigation policy in the Bombay Deccan? Second, were the cultivators coerced into accepting this policy or did they prefer it for their own reasons? Third, what were the consequences of this policy for subsistence crop production and famine protection? Origin of the Deccan Canals

As the British developed an interest in improving Deccan agriculture, their attention was naturally drawn to irrigation. Twenty years after conquest of the Deccan, the Bombay government published a compilation of reports from district officers concerning indigenous irrigation works-wells, tanks and weirs (Bombay Government 1838). After the British Crown took over from the East India Company in 1858, attention was devoted to surveying and constructing new irrigation works. In Pune district, for example, engineering surveys were made in 1863-64 of more than twenty proposed sites for tanks, weirs and canals, and one old tank was restored as a famine relief work (Beale 1901:236). 1 The impetus to build government irrigation works came partly from the high cost of well irrigation. Because they required a lot of bullock labor to lift the water, wells were expensive to operate (Taylor 1856). Since they often had to be sunk 40 feet or more and lined with stone, they were also expensive to build. By 1900, a well cost about Rs.1,000 to 1,500 to construct (Indian Irrigation Commission [IIC] 1902:221). Hence it was concluded that the government should attempt to supply cheaper water by building tanks and weirs. Nevertheless, since wells could be more widely distributed than other sources of irrigation, since they could be built and operated without government supervision, and since about one-third of their operating expenses went for hired labor (thus providing a major source of employment}, the building and repairing of wells was encouraged by the government, especially during famines. The instrument for encouraging well construction was the takavi advance, a loan from the government to the cultivator with repayments collected through the land revenue machinery. In 1877-78, following a bad famine, some Rs.300,000 were advanced as takavi loans, principally for digging new wells and repairing old ones, but also for buying seed-grain, cattle, and fodder. In a series of famines from 1899 to 1902, takavi advances rose to over Rs.20 million, including some 3.75 million for building and repairing wells (Mann 1925:7; McAlpin 1983:179-84).

Irrigation and Imperialism

51

This pattern of assistance extended well into the 20th century, in a period when the government was also investing in several major canal systems. Between 1896 and 1912, the number of irrigation wells in the Bombay Presidency increased by 30 percent, from 200,000 to 260,000 (Keatinge 1921:50). In the famine of 1918-19, about Rs.l5 million were distributed as takavi advances, of which about Rs.4 million were for wells. These figures indicate that the Bombay Government never developed a one-sided commitment to canal irrigation. Wells and canals were the two largest sources of irrigation; and of the two, wells continued to irrigate the largest area by far (Irrigation Inquiry Committee, Bombay [IICB] 1938:6,82). By the 1860s, officials were well aware that small, scattered works, such as wells, tanks and weirs, all suffered from one dismal flaw: in the worst droughts they dried up. Even in normal years, minor rivers ceased to flow in the hot season (March through May). Wells and tanks were recharged annually by monsoon rains, and if no rains fell, many went dry. Where could an assured supply of water be found? Obviously, in the coastal mountains, where the headwaters of the rivers flowing eastward across the Deccan are replenished by two to four meters of rainfall each year. Most of this rain falls from June through October. Thus, in order to provide irrigation for the important rabi or winter-season crops, it would be necessary to store the runoff behind dams in the coastal mountains and then transport it 100 miles and more into the famine belt after the rainy season. Irrigation systems of this magnitude were not attempted in India until the middle of the 19th century. Work on the Mutha canal system, which was the first in the Bombay Deccan to employ a storage dam, began in the late 1860s. The system came into operation in 1874, costing some Rs.l.8 million and capable of irrigating nearly 17,000 acres. Construction of a much larger system, the Nira Left Bank Canal, was started in 1876 as a famine relief work. This canal went into operation in 1885 with an estimated capacity of 113,000 acres. By 1900-01, with the addition of interest and other charges, the total capital cost of the Nira canal was almost Rs.5.7 million (IIC 1902:223). The Problem of Unwanted Water

An assured supply of water should have been irresistible to the cultivators. As a former Director of Agriculture wrote: "It might be thought that when a canal is opened in these arid regions the owners of the land under command would lose no time in making use of the water provided. But this is not the case" (Keatinge 1921:80). After investing so much in them, the government discovered that the canals were not wanted on a regular basis. The Deccan canals were conceived and built as "protective" works, which in the vocabulary of the Indian administration meant two things: first, the canals were intended to protect food crop production over a wide area against droughts; and second, they were not required to be self-financing. In other words, the authorities did not expect to recover the full interest on capital

52

Irrigation and Imperialism

costs from irrigation charges. In these two respects, protective works like the Deccan canals contrasted with canals built in northwest India, which were classified as "productive" works because they could eventually pay for themselves (Stone 1984:23-26). The fundamental paradox of the Deccan canals was that, although they were built in response to a devastating series of famines, they failed initially as protective works; and it was this failure which caused them to be treated more as productive works, with markedly greater success. The unsatisfactory demand for water along the Deccan canals can be indicated by comparison with other regions. In all other provinces except Bengal, the area actually irrigated by major government works was said to be more than 80 percent of the area which could be irrigated. In the Bombay Presidency, on the other hand, this proportion was only 33 percent. 2 During the period from 1891-92 through 1901-02 (which was a decade with unusually frequent and severe famines), the area actually irrigated by the Nira canal ranged from 16 percent to 46 percent of the estimated irrigable area (IIC 1902:223, 236). These results brought criticism from the central government: 'Viewing again the large expenditure incurred in the construction and upkeep of these works, the Government of India cannot regard as satisfactory the results obtained, and will be glad to receive from the Government of Bombay an explanation of the reasons why greater progress has not been achieved in extending the irrigated areas .. .' (quoted in IIC 1902:225-26)

The Nira canal was by no means useless in a famine, as shown by Figure 3.1. In the famine of 1891-92, for example, the irrigated area under rabi jowa.r (winter sorghum, the main subsistence crop) expanded rapidly from about 4,500 to 14,000 acres. This shows that the canal was fulfilling one important function of a protective work: providing emergency irrigation for food crops in the event of drought. During the severe famines which followed later in the 1890s, the area of irrigated sorghum soared to around 30,000 acres. However, this area continued to rise and fall quite irregularly, varying inversely with the amount of rainfall. When the rains were adequate, the cultivators cut back drastically on their irrigation of 'sorghum and other subsistence crops. During the period 1890-91 to 1931-32, for example, there was a negative correlation between the area of irrigated sorghum and the rainfall from September to December, this being the rainfall which most affected the winter crops. 3 Because many or perhaps most of the cultivators did not take water on a regular basis, it was not easy to provide them with emergency irrigation in a drought. Fields not regularly irrigated were, for the most part, not adequately levelled, diked and ditched, with the result that the water would drown some corners, not reach others, and generally be wasted. Moreover, the cultivators usually waited until the last minute to request canal water, in the hope that the rains might come after all; "and as no canal is constructed

Irrigation and Imperialism

20

Ill

& 15

l

10 5

t\ I \ 1\

I\ I\ 1\f\~

~

~

1\

~

1\

25

I

..

I ..

I \ I \ I I l I

I

A

30

I \I I \I

\

I

I', "",,\ ,' \\

I

\.I :. \ 1

1\ r'.. I :". •·. .: ···. " .. ,, \ I : ..

\I

~

I\ I \SORGHUM 1 '

I\ I \

~

35

~

"

~

40

I

53

\1

\ \

\

,_

I Jl

\ \..,

:.

\j/ .

····v'. . . ··....

~

=

... ..·

MILLET

0~. . . .'---~--~---,----,----r----r----r--~~

1885-86

FIGURE 3.1

1890-91

1900-01

1910-11

1920-21

1930-31

Crops Under Irrigation on the Nira Left Bank Canal (Source: Inglis and Gokhale 1934:34-35, 44-45.)

to satisfy sudden demands on a large scale, many of the crops will wither before they can be irrigated" (Beale 1901:30). The Executive Engineer for Irrigation, Pune, elaborated on the problems of meeting sudden surges in demand: The capacity of the canal staff is taxed to the utmost in order to ensure a fair distribution and mature as large an area as possible. . . . Tens of thousands of acres of rabi crops require water at the same time and all in 12 or 15 days. On such occasions the demand is most intense and an equitable distribution of the areas becomes well-nigh impossible. (Visvesvaraya 1903:4)

Thus for a number of reasons, the canal could not supply water equitably and efficiently to a much larger area in famine years than was served in normal years; consequently, there were a great number of cultivators who either could not get water when they most needed it or else could not make effective use of what they got. Because they were not using much of the canal's capacity on a regular basis, neither they nor the irrigation officials were able to cope effectively with severe droughts. The critical question, then, was this: why were the cultivators not irrigating subsistence crops in normal years? Perhaps the irrigation rates were too high? The canal rate for winter sorgham, the main subsistence crop, was Rs.2.25 per acre in 1901 (Visvesvaraya 1903:12). This was just twice the maximum land revenue charged in many canal villages, a rate which every cultivator could pay in a normal year

54

Irrigation and Imperialism

(Bombay Government 1912:136). Moreover, McAlpin (1983:198-202) has calculated that the real value of revenue payments was declining as prices rose in the late 19th and early 20th centuries. If the revenue rate was becoming more affordable, then the irrigation rate should not have been prohibitive. At the turn of the century, the market value for one acre of sorghum grain under canal irrigation was about Rs.50. 4 The irrigation charge was thus less than 5 percent of the gross value of the crop, which seems a reasonable rate for what was, if nothing else, insurance against drought. This is especially true if we also include the value of sorghum fodder, an excellent cattle feed, which added 20 to 40 percent to the value of the grain (Patil 1928a:64-70). Since an acre of dry sorghum, on the other hand, was worth only about Rs.6 (Mann and Kanitkar 1921:80-81), the potential returns from irrigation were high. Perhaps the irrigation rates were biased in favor of cash crops? The rate charged for sugarcane (in 1903) was Rs.25 per acre, 11 times the rate for sorghum. As with sorghum, the irrigation charge was less than 5 percent of the gross value of the cane crop, which sold for about Rs.550 per acre in 1908-09 (Visvesvaraya 1903:12; Inglis and Gokhale 1928:15). By the 1920s, the charge for sugarcane was increased to more than Rs.50 per acre. The rates for seasonal grain crops were also increased, but the cane rate stayed nine times higher (Inglis and Gokhale 1928:15; 1934:21). The much higher cane rates reflected different water requirements, for cane needed an estimated eight to ten times the amount of water required for a seasonal grain crop (IIC 1902:251). Thus the charges per volume of water were held roughly equivalent. The rates charged for foodgrains did not even pay the average cost per acre of canal administration, let alone produce a repayment on capital costs (IIC 1902:230). So why were farmers reluctant to irrigate their subsistence crops? One genuine cause of hesitation was the cost of preparing a field for irrigation: at least Rs.l7 per acre for levelling, ditching and diking (IIC 1902:220). This was not a great sum for those going into cane production, for they would expect to spend annually about Rs.500 per acre, for a profit of Rs.lOO or more (Inglis and Gokhale 1928:15). On the other hand, Rs.l7 was a much larger share of the value of an acre of irrigated sorghum. Another basic problem was the chronic shortage of plough cattle (Charlesworth 1985:140,212), which made any shift to permanent irrigation more expensive, since the heavy black soil, when irrigated, required several pairs of bullocks for plowing. In addition, most of the sorghum and millet were intended for home consumption, and farmers were reluctant to incur cash costs which they could not recover through profitable sales. The low elasticity of demand for coarse foodgrains no doubt led them to expect that an aggregate increase in production would cause crop prices to fall. Aside from these constraints, the principal cause of irregular demand for canal water was the nature of the soil, the climate, and the crops which

Irrigation and Imperialism

55

were adapted to them. In the river valleys along the canals, much of the soil was Deccan black soil, also called black cotton soil, which was very moisture-retentive. If the rains did not fail altogether, the cultivators could usually manage to bring in a crop. Conversely, it appears that adding irrigation in seasons of normal rainfall did not increase the yields of these crops in proportion to the additional cost of irrigation, labor, bullocks and equipment (IIC 1902:227). The staple foodgrains had been adapted over centuries to the semi-arid climate. Above a certain minimum, extra doses of water contributed mostly to growth of leaves and stem, not the grain head; and longer stems increased the danger of lodging. To substantiate this point, we need data on the results of irrigating subsistence crops in years of normal rainfall. Patil and his associates (1928a, 1932) observed cultivators growing sorghum and millet (as well as other crops) in both canal-irrigated and dry villages over three crop years. All three were years of normal rainfall, so the dry plots gave adequate yields while the canal plots showed the value of irrigation under non-famine conditions. In the dry villages selected for study, sorghum was not the most important foodgrain; the figures on millet are more informative, since more plots were studied. The results are listed in Table 3.1, and below the raw data are the results of my calculations. 5 First, we may note that, while canal irrigation increased the mean yield by 66 percent, costs rose by 87 percent. In other words, cost efficiency was lower on the canal-irrigated plots: these produced about 12.6 pounds of grain per rupee as compared with about 14.2 pounds per rupee on the dry plots. Cost efficiency is a somewhat artificial concept with regard to peasants growing crops for home consumption, but it clearly mattered to those who were considering cash payments for irrigation-especially since canal irrigation would boost production beyond their own subsistence requirements, and the extra cost would have to be recovered through market sales. 6 Another striking feature in Table 3.1 is the increased variability of net profits under canal irrigation. The coefficient of variation (the standard deviation divided by the mean) rises from 6.84 for the dry plots to 12.20 for the wet ones. Moreover, the variability of yields does not decline under irrigation. If profits are more variable for canal-irrigated millet, this means that the greater costs do not consistently lead to proportionately greater yields. This hypothesis is supported by the correlations between yields and costs, which are strikingly different in the two sub-samples. The dry plots show a positive correlation of0.57, suggesting that when more labor, fertilizer, etc., were added to a plot, the yield rose accordingly; but the canal-irrigated plots show a negative correlation of -0.20, suggesting that irrigation of dryland subsistence crops was financially risky (except, of course, in a drought). 7 Likewise, the correlation between profits and costs is negative for the wet plots, but slightly positive for the dry ones. In a year of normal rainfall, canal irrigation raised millet yields to some extent, but apparently not enough to compensate for the additional costs and risks. The likely cause of this problem was that yields were not very

56

TABLE 3.1 Millet Production on Dry and Canal-Irrigated Plots

Dry Plots

Irrigated Plots Net Profit (Rs./ acre)

Plot No.

Yield (lbs./ acre)

Cost (Rs./ acre)

1 2 3 4 3A 4A SA 6A 138 148 158

538 285 88 611 124 588 433 260 615 205 300

37.91 24.95 17.11 31.96 33.43 30.38 20.78 21.00 28.96 15.74 23.71

+ 5.14 - 0.11 -10.79 +15.02 -25.06 +15.06 +16.66 + 2.33 + 8.54 - 1.98 - 4.66

25.99 7.09 0.27

+ 1.83 12.52 6.84

Mean: 367.91 Std. Dev.: 197.30 Coef.of Var.: 0.54

Plot No.

Cost (Rs./ acre)

5 6 7 8 9 10 7A SA 9A lOA

540 360 928 493 760 140 720 300 600 1280

59.64 36.80 44.28 51.69 33.82 37.02 32.88 85.65 56.41 48.96

-18.64 - 9.60 +19.94 -14.83 +26.45 -23.52 +22.63 -59.65 -11.66 +43.88

612.10 330.45 0.54

48.72 16.06 0.33

- 2.50 30.51 12.20

66.37

87.46

-236.61

Increase with Irrigation (percent): Yield/Cost Ratio: 14.16 (lbs./rupee) Yield-Cost Corr.: 0.57 Profit/Cost Ratio: Profit-Cost Correlation: So!.!J::!OS!S:

Net Profit (Rs./ acre)

Yield (lbs./ acre)

7.04' 0.09

Patil 1928a:l0, 29-48; 1932:69-101.

12.56 (lbs.jrupee) -0.20 -5.13\ -0.68

57

Irrigation and Imperialism

TABLE 3. 2 Subsistence Crops Under Canal and Yell Irrigation, Ahmednagar District

Canal Villages (Canal irrigation mostly)

Dry Villages (Yell irrigation only)

Year

Irrigated Millet

Millet All Irri& Sor- Irrig. gated Sorghum ghwa Crops

Irrigated Millet

Millet Irri& Sorgated Sorghum ghum

All Irrig. Crops

1938-39 Acres Percent

17.30 11.96

8.40 5.81

25.70 17.77

145 100

118.80 8.57

116.60 8.41

235.40 16.98

1,387 100

1939-40 Acres Percent

14.50 4.10

37.50 10.60

52.00 14.70

354 100

157.40 6.52

237.10 9.82

394.50 16.34

2,414 100

Mean Percent (1938-40)

8.03

8.21

16.24

100

7.55

9.12

16.66

100

~=

Gadgil 1948:26-27, 30-33.

responsive to extra water. With garden crops, on the other hand, this problem did not arise. In a study of 30 sugarcane plots along the Nira canal in 1920-21, costs and yields showed a positive correlation of 0.32. 8 To sum up, Patil's data show that the use of canal irrigation for growing millet in years of normal rainfall meant: (1) a decrease in cost efficiency, (2) an increase in the variability of net profits, and (3) a negative correlation between costs and yields. Although the number of plots is small, the data suggest why cultivators did not find it in their interest to use canal irrigation for subsistence crops on a regular basis. It might still be supposed that the fault for this situation lay with some feature of the design or administration of the canals. It is possible to test this hypothesis by comparing the cropping pattern under canal irrigation with the one chosen for fields under well irrigation. Wells were constructed, owned and operated by the cultivators, who were free to select crops according to their own needs and resources. If there were a bias against certain crops on the canals, this should manifest itself in a quite different cropping pattern under well irrigation. Table 3.2 presents the appropriate data for this comparison, taken from D.R. Gadgil's classic study on the Economic Effects of Irrigation (1948). Gadgil and his associates surveyed nearly 400 farms, located in matched pairs of dry and canal-irrigated villages, the villages in each pair being close neighbors and the whole set being spread over two canal systems in Ahmednagar District, just north ofPune District. The table summarizes cropping patterns for millet and sorghum grown under well and canal irrigation. 9 The patterns are remarkably similar: in both cases, the cultivators put about

58

Irrigation and Imperialism

16 percent of the irrigated area under these two crops. This shows clearly that the canals did not distort established cropping patterns chosen by the cultivators. Water was expe[lsive in either case, and it did not pay to use a large share of it on subsistence crops. This might sound as though the cultivators clung blindly to past experience. In some respects, canal irrigation provided new opportunities: it was far more reliable in a drought, and it could deliver abundant water to hundreds of acres in a compact block, which wells could never do. Thus it makes sense to ask whether the cultivators could have learned to make better use of canal water for their subsistence crops. If they could have, they must have been remarkably obtuse not to have done so. Returning to Figure 3.1, it is evident that, for a period of several decades, many cultivators rushed to increase their irrigation of sorghum and millet when the rains failed and then, when the weather returned to normal, promptly stopped irrigating these crops. As the experience was repeated, moreover, the fields were undoubtedly better prepared for irrigation; and yet the same cultivators discontinued irrigating these crops in normal years. This could only mean that the yields were never sufficiently high to offset the additional monetary and labor costs or the additional uncertainty in the relationship between costs and yields. 10 The situation was summed up by M. Visvesvaraya, the Executive Engineer for Irrigation, Pune District: Black soil has a peculiar property of resisting evaporation. . . . In good seasons the black soil of the Deccan yields a full harvest and in ordinary years a fair harvest .... It is only in a year of severe drought that irrigation of dry crops is really useful, and that there is any large demand for water for them. (IIC 1902:227) .

Thus it was unprofitable to apply canal water to subsistence crops in years of ordinary rainfall. This situation was brought about not by biased irrigation policies but by the nature of the soils, climate and crops. Consequently, so long as the canal administration aimed at irrigation of subsistence crops, it was doomed to waste a large share of the available water. In a semi-arid climate, this waste was unacceptable. Fiscal Pressures and Wasted Water If the cultivators would not pay for water on a regular basis, the government could not recover even a small fraction of its investment in the canals. In the year 1899-1900, for example, the irrigation works of Bombay generated a net revenue of 1.4 percent on capital outlay, as compared with an average of 6.4 percent for all of India and nearly 10 percent for Punjab (IIC 1902:225). Referring to these melancholy facts in testimony before the Indian Irrigation Commission, M. Visvesvaraya, the Executive Engineer, commented defensively:

Irrigation and Imperialism

59

These results which have been more or less the same in all recent years, have discredited the Bombay works in the estimation of the Government of India. The annual grants for new works have in consequence been curtailed and the strictest economy is enforced in the maintenance of existing works. {IIC 1902:225)

This financial problem had two causes: one was the lack of regular demand for irrigation; and the second was the much greater expenditure required for canals in the Deccan as compared with other regions. This additional cost was due to the lack of perennial rivers, which compelled the government to build large storage dams. As Visvesvaraya noted: Water is very expensive in Bombay. . . . I have shown that an expenditure of Rs.100 has provided facilities for irrigation of about five acres in Punjab, four in Madras and three in the North-Western Provinces. In Bombay the corresponding area is less than half an acre. It may be roughly stated that, on account of the great cost of storage, water-supply is three to six times more expensive here than on the other irrigation systems. {IIC 1902:225)

As custodians of the most expensive water in the country, Visvesvaraya and his colleagues were understandably keen to find a use for it. Before 1901, the protective policy meant that as much water as possible was kept in storage after the monsoon for irrigating winter foodgrains (sorghum and wheat). However, much of this precious store was wasted in normal years. Consequently, Visvesvaraya proposed a more intensive policy of cash-crop irrigation. There was certainly an element of fiscal self-interest behind this proposal. The rate charged for irrigating foodgrains was less than the cost of delivering the water, in part because extensive irrigation entailed high distribution losses through evaporation and percolation.n Consequently, the larger the irrigated area under subsistence crops, the more financial problems the canals had. "Increase in the irrigation of ordinary crops will never pay in Bombay. Water is too costly to be profitably applied to them," as Visvesvaraya pointed out. "If the Bombay works are ever to prove remunerative, perennial and other high class irrigation should be largely encouraged in ordinary years" (IIC 1902:227). Among the crops in this region, sugarcane grew well under canal irrigation and had one special advantage: it was flexible in the timing of planting, watering, and harvesting. Thus cane was suited to become the premier cash crop along the Deccan canals. As two irrigation officials observed later on: "the more flexible the crop requirements the better for all concerned. . . . Sugarcane which meets irrigation limitations and natural conditions best has been a very profitable crop to the cultivators and Government, and has brought in more than half the canal revenue" (Inglis and Gokhale 1934:21) These authors reviewed the results of irrigating a variety of other cash crops and found that none was as well suited to the soils, climate and canals of the famine belt. 12

60

Irrigation and Imperialism

As we have seen, sugarcane was charged a higher rate than other crops, since it needed water all year round, particularly during the hot season (March-May), when most other crops were not grown. As a result, it was evident that sugarcane would have a strong positive influence on the irrigation budget. On the Nira canal, sugarcane was generating 57.5 percent of the revenue by 1901-02, even though it covered only 11.6 percent of the irrigated area (Visvesvaraya 1903:8). Beset as they were with recurring famines, budget cuts, and criticism from the Government of India, the Bombay irrigation officials seized on sugarcane as a crop which would utilize their water and balance their budgets. That the government was ultimately more concerned about wasted water than fiscal losses was, however, demonstrated by the outcome of hearings held by the Indian Irrigation Commission in 1901. One result was to approve Visvesvaraya's plan to experiment with a more intensive policy favoring sugarcane. This decision might have been motivated simply by frugality. However, another recommendation by the same Commission, one with even more profound effects on the region's future, was in favor of constructing a whole new series of Deccan Canals (IIC 1903:72-79). The government decided to invest vigorously in a seven-fold expansion of the canal systems, even though there was no foreseeable date when they would pay back their costs. Even so, the fiscal interests of the government surely encouraged it to favor irrigation of sugarcane. Were there other interests in the British Raj pushing in the same direction? For the most part, such interests did not exist. Sugar was not an export crop, it was produced entirely for the home market. Consequently, British shipping and trading companies did not expect to earn profits, nor did the government expect to earn export taxes, by promoting sugar production. There was only one British-owned sugar factory ever established in the Bombay Deccan (at Belapur). All the rest of the privately-owned factories (there were 13 by 1941) were owned by Indians. In sum, there were almost no British firms with a stake in the sugar economy of the Deccan, nor did the government have a stake in export earnings from this crop. After sugarcane began to flourish along the canals, the government did develop an increasing attachment to the crop and to prospects for industrial sugar production, as we shall see. However, this attachment remained firmly rooted in concern over the utilization of expensive water. The decisive step toward solving that problem was the sugarcane block system, pioneered by Visvesvaraya. The Sugarcane Block System In his memorandum to the Irrigation Commission, Visvesvaraya explained the problems of irrigation management in the Deccan and followed with a proposal for a more effective irrigation policy: the sugarcane block system. This system, which Visvesvaraya tried out first on the Nira canal, included the following basic features (see IIC 1902; Visvesvaraya 1901, 1903):

Irrigation and Imperialism

61

l. By a signed contract, a cultivator would agree to take canal water on a certain area, which constituted the block, for six years; and the government would guarantee to supply this water. This would eliminate delays and uncertainties caused by the prevailing system of seasonal applications for water. The cultivators would be more willing to invest in the preparation of their fields if they were certain to receive water for several years. It was Visvesvaraya's belief, borne out by subsequent experience, that the temptation of sugarcane profits was the best inducement for getting ordinary peasants to demand irrigation on a regular basis. If irrigation were not guaranteed for several years, only rich peasants would dare to invest in this expensive crop. 2. Block irrigation would be charged at a fixed rate per acre, regardless of the crops grown. This would reduce the complications connected with charging different rates for different crops. 3. Perennial or twelve-month irrigation, suitable for cane cultivation, would be provided on one-third of each block. The other two-thirds would receive irrigation for eight months, suitable for double cropping in the monsoon (kharif) and winter (ra.bi) seasons. No irrigation would be supplied to this portion of the block during the hot season, March through May. This division of the block area into thirds fitted an established system of crop rotation. Blocks would be allotted in multiples of 1.5 acres, with half an acre for cane and one acre for seasonal crops. 4. In the 30 villages under the block system, the total block area would be about 21 percent of the cultivable area under command of the canal, and the area of perennial irrigation would thus be only 7 percent (one-third of 21 percent) in these villages (Visvesvaraya 1903:9). The total block area would be 18,000 acres, with 6,000 acres of cane planted each year. Distributing the blocks among all 30 villages would encourage local cultivators to take up sugarcane and correct the imbalance caused by immigrant entrepreneurs, clustered near the market towns, who were using a large share of the water. 5. The distribution of perennial irrigation on this basis could not be guaranteed below the first 65 miles of the canal. On the last 40 miles, serving 33 villages, preference would be given for irrigating seasonal crops. The stored water available for rabi (winter) crops would irrigate from 23,000 to 33,000 acres, depending on the amount of rainfall collected in the previous monsoon season. Under the extensive policy, by comparison, the maximum irrigated area of rabi crops had been 37,000 acres in the famine of 1900-01 (Visvesvaraya 1903:9).

Instead of storing the maximum volume of monsoon rain, in case there might be a drought in the rabi season, the block system would guarantee water for 6,000 acres of sugarcane, leaving a less certain supply for emergencies. However, as we shall see, the system could be adjusted in a crisis so that even larger areas of rabi crops could be irrigated. When Visvesvaraya appeared before the Irrigation Commission, he had already obtained the endorsement of many cultivators for the block system:

62

Irrigation and Imperialism

"During 1900, the cultivators were consulted as to their willingness to bind themselves to take water for selected areas for fixed periods of 5 or 6 years. There was a very ready response. Applications were received for about 18,000 acres in the upper half of the canal. They were willing to pay Rs.10 per acre" (IIC 1902:232). The rate was soon changed to Rs.12 per acre (Visvesvaraya 1903:12). A three-acre block was thus charged Rs.36, which was two rupees more than the previous rate for an acre of cane plus two acres of eight-month irrigation. At the time Visvesvaraya canvassed the cultivators, there were about 3,500 acres of cane on the Nira canal. Immediately after the block system was introduced, the area rose to between 7,000 and 8,000 acres, staying at about this level until the first world war (see Figure 3.1). Not only did the cane area double, but the total irrigated crop area increased by more than the additional cane acreage. The average for all crops rose from 42,700 acres in the five years preceding 1901-02 to 54,700 acres in the five years preceding 1915-16 and 76,500 acres in the five years preceding 1925-26 (Visvesvaraya 1903:24-25; IICB 1938:84-85). Thus overall capacity utilization rose dramatically: before 1900 the largest proportion of the estimated irrigable area ever watered by the Nira canal was 42 percent. By 1922-23, the proportion was 66 percent, and this was typical for the decade. Likewise, the rate of return on capital outlay rose from 1.5 percent in 1899-1900 to 8.28 percent in 1922-23 (IIC 1902:242-43; IICB 1938:8,84-85). In the history of agricultural development, few administrative reforms can have been so timely or effective, both in stimulating increased production and repaying public investment. Cane Blocks and Famine Protection The cane blocks made canal irrigation more intensive and less wasteful. This did not mean, however, that the Nira canal lost its protective function; indeed, as it turned out, the canal became more protective in various ways as a result of the block system. Food production and agricultural employment expanded after the block system was introduced. There were several reasons for this: First, between one-third and two-thirds of each block (depending on the amount of overlapping cane) was devoted to seasonal crops, generating a larger and more stable output of foodgrains (Inglis and Gokhale 1934:21). Second, seasonal crops on the block system gave higher yields because they were irrigated and also because they were rotated with sugarcane, benefiting from residual manure in the soil and from the careful levelling, ditching and diking required for cane (Patil1932:56,101). From the farmers' point of view, sugarcane generated their cash income, so this crop bore the entire cost of irrigating and improving their fields, with results which improved their subsistence crops as well. Third, expanding cane cultivation increased the demand for bullocks (for plowing, crushing, transport, etc.). Gadgil's data (1948:32-35) show that

63

Irrigation and Imperialism TABLE 3.3

Foodgrain Output in Dry and Canal-Irrigated Villages, Ahmednagar District

Villages and Years 12t:ll: V1llii!!ll 1938-39 1939-40 Total HSI' Villl&!i!l 1938-39 1939-40 Total

Net Cultivated Area (NCA] (acres)

Foodgrain Area (acres)

5,304 6,522

4,377 5,311

11,826

9,688

4,490 6,604

3,213 4,655

11,094

7,868

Foodgrains as ' of NCA

Foodgrain Output Valued in Rupees

Foodgrain Output per NCA (Rs./acre)

41,107 60,320 81.9

101,427

8.58

49,673 74,884 70.9

124,557

11.23

(Foodgrains- millet, sorghum and wheat.) Gadgil 1948:30-33, 42-45.

~:

investment in livestock was 55 percent higher (relative to net cultivated area) in canal-irrigated than in dry villages. More bullocks needed more fodder, and since sorghum was an excellent fodder, its production rose. These were all reasons why "the irrigation of seasonal crops follows sugarcane and does not precede it" (Inglis and Gokhale 1934:21, italics in original). This is shown in Figure 3.1, where the area under irrigated sorghum remains at higher levels following the introduction of the block system. These points are confirmed by Gadgil's (1948) study on the impact of canal irrigation in Ahmednagar district, north of the Nira canal area. As mentioned, Gadgil compared agricultural inputs and outputs in two sets of villages, canal-irrigated and dry. His data on the production of foodgrains (millet, sorghum and wheat) are summarized in Table 3.3. This table shows that foodgrains continued to occupy most of the cultivated area, even in the canal villages. It is true that a lower proportion of the cropped area was devoted to foodgrains in the canal villages. However, it is important to remember that higher yields were produced under irrigation. Consequently, total foodgrain output (valued in rupees per net cultivated acre) was 31 percent higher in the canal villages than in the dry ones, under normal rainfall. In a drought, of course, the difference would have been much greater. Gadgil's data also show that investments in buildings and equipment, as well as livestock, were higher in the canal villages (1948:35). Thus irrigated cash cropping relieved the basic shortage of capital which was, according to Charlesworth (1985:76-78,212), a fundamental constraint on Deccan agriculture.

64

Irrigation and Imperialism

The demand for labor was also much higher in the canal villages. Double cropping and perennial cropping, made possible by irrigation, required more labor at more regular intervals. In the irrigated villages, expenditures on hired labor (per net cultivated acre) rose 365 percent by comparison with the dry villages, while inputs of family labor remained about the same (Gadgil 1948:49-51,63-67). Extending Gadgil's data to the whole Bombay Deccan, it has been estimated that sugarcane cultivation generated paid employment for perhaps 100,000 laborers for varying periods each year (Guha 1985:139-40). The canal tracts absorbed many laborers from the dry villages, particularly when the latter suffered from crop failures. Thus the Deccan canals substantially increased employment opportunities in the region. This effect came about in two stages. First, a large supply of construction labor was needed to build the canals. Later, many laborers who migrated to construction sites found employment as agricultural laborers, and they were followed by still more migrants from the dry villages. The rising demand for labor in the canal villages pushed upward on real wages, to the benefit of the laborers (Shirras 1924; Patil 1932:3). Even so, large, short-run fluctuations in wage rates were caused by variable rainfall (Guha 1985:140-42). This was because much of the labor force consisted of seasonal migrants who were self-employed farmers in the dry villages. In good seasons, the demand for labor was high in these villages, forcing up wages in the canal tracts. When the rains failed, the supply of migrant labor was more abundant. The rise in labor demand, along with increased foodgrain production, helped protect the region against famines. Trends in famine mortality rates have been analyzed by McAlpin (1983:168-71), who argues, for example, that although the crop failures of 1899-1900 and 1918-19 were about equally severe and widespread, the latter one caused much lower mortality. The improved situation in 1918-19 was brought about partly by more efficient famine relief administration; but it was also helped, according to McAlpin, by better economic conditions in general. The Deccan canals contributed to this improvement by creating a larger and more stable demand for labor. As a result, the demand for employment on famine relief works declined. For twelve months in 1899-1900, the average daily number of persons on famine relief was over one million, peaking at 1.5 million. In 1918-19, on the other hand, the corresponding numbers were only one-tenth as large, and the crisis lasted for only about six months (Keatinge 1921:56-57). Many villagers from the dry areas, who would have sought employment on relief works under earlier conditions, migrated to the canal areas instead. Consequently, expenditure on famine relief actually declined along with mortality rates (IICB 1938:99). Moreover, the famine of 1918-19 showed that the block system could be modified in the direction of more extensive irrigation during a crisis. In that year, the government cancelled irrigation for overlapping cane. (Cane was often in the ground for 15 to 18 months; thus the old crop would overlap with newly planted cane.) As a result of this cancellation, the area

Irrigation and Imperialism

65

under sorghum and millet rose sharply, as seen in Figure 3.1. This brief modification of the block system maintained the canal's protective function and contributed to a record rise in foodgrain irrigation the following year. Impact of Peasants on Policy The standard theory assumes that cash cropping cuts against the interests of ordinary peasants by undermining their subsistence security. Consequently, peasants must be duped or coerced into growing cash crops. Moreover, since it is the world system which determines the fate of the peasantry, agrarian systems are shaped everywhere by external forces: by military conquests, colonial governments, unequal treaties, international markets, new technologies, and so forth. In this view, whatever differences now exist among agrarian systems have been primarily due to variations in the timing, intensity and interaction of such external forces, not the result of indigenous conditions or local initiatives. The results of this chapter point to the following conclusions: First, the cultivators had good technical reasons for not wanting regular canal irrigation for subsistence crops. There is no evidence that they were compelled to switch to cash crops on the canals, since the same proportion of irrigated land was devoted to sorghum and millet under wells. Irrigation was too expensive, and the financial returns too uncertain, for subsistence crops to receive much canal water. The second conclusion is that the Bombay government switched to a more intensive irrigation policy because the cultivators were simply not using the canal water. Except for fiscal pressures, the government had no particular stake in the growth of a sugar industry. Decisions about irrigation policy were pragmatic responses to decisions made by the cultivators themselves. The officials who decided to experiment with the cane block system were swayed by the example of a few entrepreneurs making profitable use of canal water while nobody else was using it regularly. This point has serious implications for those who prefer to see Third World peasants as the passive victims of history. The third conclusion is that a more intensive irrigation policy decreased rather than increased the region's vulnerability to famines, since the cane blocks encouraged regular irrigation of seasonal foodgrains as well as sugarcane. In addition, irrigated cash cropping multiplied the demand for labor, reducing the public employment which had to be provided on famine relief works. Even though the Deccan Canals were no longer regarded as purely protective works, they helped famine-proof the region to an extent which had been impossible under the older, extensive irrigation policy.I3 What would have happened to the Bombay Deccan if there had been no shift to a more intensive policy? In 1901, when Visvesvaraya persuaded the Indian Irrigation Commission to sanction the cane blocks, there were only two canal systems with storage dams in the Bombay Deccan. Had these canals continued to waste water and lose money at the prevailing rate, it is

66

Irrigation and Imperialism

likely that investments in future canal systems would have slowed down or ground to a halt. The prospect of reduced wastage enabled the Irrigation Commission to sanction investments for a whole series of new canals, so that by 1936-37 there would be six major systems in operation, irrigating more than seven times the area watered in 1899-1900 (IIC 1902:242; IICB 1938:84-85). These new canals vastly increased the region's resistance to famines; and it is likely that they never would have been built, or their construction would have been slowed by decades, if there had been no shift to irrigation on the block system. When the canals failed to serve their original purpose, and when they later became more efficient, these outcomes resulted from production decisions made by enterprising peasants. They calculated the costs and risks of irrigating foodgrains, as well as the possible returns from marketing the surplus. Their choice of which crop to irrigate was based on technical factors and economic returns, not on coercion. From their point of view, it made sense to irrigate a cash crop or not to irrigate at all in normal years. The irrigation engineers took about 15 years to grasp the logic of the situation and modify their policies accordingly. The new block system amounted to official recognition of what the peasants were willing and able to do with the water. In that sense, the policy was as much a product of peasant enterprise as the crop itself. Notes l. Famine relief works were construction projects (such as roads or canals) set in motion during a drought in order to provide employment to villagers. Most government irrigation works constructed in the 19th century were famine relief projects. 2. The data given on the other provinces seem too good to be true, but there is little reason to doubt that the Deccan canals were doing worse than those in other regions, particularly those in northwestern India. 3. Calculated from data in Inglis and Gokhale (1934:19,45, Fig.l4), this correlation (-0.32) is significant at the 5 percent level. It might be stronger if we had the rainfall series for the Nira valley alone, instead of an average for three weather stations across the Deccan. The same relationship holds between the area of irrigated millet and monsoon-season rainfall (Inglis and Gokhale 1934:18-19,44, Fig.l3). 4. The estimated average yield for irrigated sorghum was 1550 pounds per acre, and prices in Baramati averaged about Rs.3.18 per hundred pounds in the period 1893-1903 (Beale 1901:30; Bombay Government 1912:53). 5. Patil's sample villages were near Pune city, where summer millet was grown more than winter sorghum, and where rainfall was more secure than in the famine belt further east. Consequently, his findings may be slightly more favorable to the dry plots than they would have been in the famine belt. The three crop years, 192526 to 1927-28, are indicated by letters attached to the plot numbers in Table 3.1. Two "dry" plots have been removed from the raw data because bajri was preceded by canal-irrigated sugarcane, and the residual moisture and fertilizer raised the yields far above normal (Patil 1932:56,101). 6. Patil's cost data include imputed wages for family labor and imputed rents for bullocks and land, valued at prevailing market rates, plus all the normal cash expenditures. One can, of course, question the appropriateness of using market rates

Irrigation and Imperialism

67

to impute wages for family labor-indeed, Patil's boss did so at the time (Patil 1932:iii-iv). However, Patil's method does not seem to bias the comparison between dry and wet plots, nor the v~riations and correlations within each group of plots. Obviously, net profits would have been higher in both groups if family wages had been imputed at lower, more realistic, rates. 7. The difference between the yield-cost correlations is significant only at about the 10 percent level, since the number of plots was small. 8. Calculated from data in Inglis and Gokhale (1928:18, Plate IliA), this correlation is significant only at the 10 percent level, due to the low number of plots. 9. In the dry villages, wells irrigated about 4 percent of the net cultivated area, while in the canal villages, 28 percent was irrigated. Nearly all the latter area (82 percent) was watered by canals. (Eleven percent was watered by wells and 7 percent by canals and wells combined.) The figures in Table 3.2 include crops under all forms of irrigation in the canal villages. 10. This analysis is supported by more recent experience in a different region of the Deccan plateau-on a canal system built in 1956 in the state of Andhra Pradesh. Although the state made this system as extensive as possible, charged low water rates, and encouraged irrigation through a system known as "localization," the initial demand for water was low, since the villagers were accustomed to growing sorghum as their main crop. It was only after the canal area was linked to outside markets via a railway line, and immigrants from the coast had demonstrated the profitability of growing rice as a commercial crop, that the demand for canal water shot up (Ramamurthy 1989:21). ll. If distribution losses had been included while comparing the water consumption of foodgrains vs. sugarcane, the difference in their requirements would have been narrowed. Consequently, the rates charged for foodgrains would have appeared even more concessional. 12. It is interesting that cotton, which was an important crop in areas to the north, east and south of the Bombay Deccan, was scarcely grown at all along the Nira canal before 1925. Cotton required irrigation during the hot season (May in particular), so it competed directly with cane when canal water was scarcest (Inglis and Gokhale 1934:16-18). The same source also explains why wheat, fruit trees, groundnuts and vegetables were less popular than sugarcane. 13. See Stone (1984) for a similar rebuttal of Whitcombe's (1972) argument that canal irrigation in north India increased vulnerability to famines.

4 Peasants Versus Capitalists When industrial crops were grown in the tropics under colonialism, what determined how production was organized, how resources were distributed, and how these patterns changed over time? Studies written from the perspective of dependency or world-system theory emphasize external causes: that is, the influence of world trade, foreign investment, colonial administration, western technology, and so on. From this perspective, external forces are considered highly potent, while indigenous social structures are passive and malleable. There are cases where this approach may be appropriate: some societies in the New World, for example, were shaped and reshaped by the demands of export crops such as sugar. However, proponents of the dependency approach seldom make comparative studies of regions where industrial agriculture could not be imposed freely on the peasantry-at least not in forms which would be most advantageous to capitalists. In many areas of the world, sugarcane is grown on large plantations whose ownership is tied to the factories that process the cane (Padhye 1924:38-39). This is so for a simple technical reason. When cane is harvested, the sucrose content begins to decline after a few hours. This means that cane cannot be stockpiled at the factory; it must be brought fresh from the field to the factory gate. Moreover, the cane should be harvested at peak maturity, when sucrose is at a maximum. For these reasons, efficient utilization of a factory's capacity requires careful scheduling of the harvest and delivery operations. Coordination of factory and field is easiest when the agricultural operations are under the control of the factory itself; hence the rise of the factory-owned plantation (Shlomovitz 1984). This plantation system predominates in many areas of the world, not so much due to the advantages of large-scale, mechanized farming but mainly due to the technical necessity for coordination between factory and field. 1 India is now the second largest producer of centrifugal (industrial) cane sugar, yet nearly all its cane is grown on small peasant holdings, not large plantations. This is true even in north India, where sugar is processed in factories owned by private capitalists. Moreover, the industry in western India is now dominated by peasants, as owners of cooperative sugar factories. Why does the plantation system not predominate in India? And why is a major segment of the industry now controlled by peasants? The answers lie 68

Peasants 'l&rsus Capitalists

69

not in external forces but in the ability of rural interests in the north to resist change, even when intimately associated with industrial production, and in the ability of cane growers in the west to take control of change through their own innovations. North India: Origins of the Industry Until after independence, the main center of the Indian sugar industry was in northern India, in what are now the states of Bihar, Uttar Pradesh (U.P.), Haryana and Punjab. In 1920, U.P. and Bihar produced 60 percent of the sugarcane in India and had 13 of the 18 sugar factories (Indian Sugar Committee [ISC] 1921:35-36, 61-62, 291). Thus we begin with an outline of how the industry got started in that region. 2 What is remarkable about this industry is the extent to which it grew without causing any fundamental change in the agricultural system that supplied its raw material. Indian villagers have long produced sugar for the Indian marketparticularly a form of crude, brown sugar known as gur in Hindi or gul in Marathi. Gur was (and still is) made by the cane growers themselves, using bullock-driven cane crushers and other small-scale equipment. In the 18th century, India was an important world supplier of sugar (primarily khandsari, an artisan sugar which is more refined thangur). "But the development of the sugar beet in Europe as an alternative raw material, under the patronage of Napoleon, virtually wiped out the Indian export industry" (Hirsch 1961:72-73). The sugar industry began a slow recovery after 1902, when the first modern mill was established. In that period, India was importing about 700,000 tons of white sugar per year, primarily from Java; and it was not easy for Indian sugar to compete. By 1914, if not earlier, sugar experts realized that the main problem with the industry was cane supply or, in other words, the relationship between factory and field (Sahasrabudhe 1914:22-23). In 1920, the Indian Sugar Committee made detailed inquiries into prospects for the industry in all parts of the country and concluded that: Indian factories speaking generally are both small and inefficient. If there is one cause more than another to which both the smallness and the inefficiency can be attributed, it is the difficulty they experience in obtaining an adequate supply of cane. (ISC 1921:292)

Why was cane supply such a problem in India? And why could the industry not follow the example of other countries in resolving this problem? In north-central India (eastern U.P. and Bihar), where the industry had its first growth, the factories obtained most of their sugarcane from small and middle peasants who grew other crops (chiefly rice) for their own subsistence (Amin 1984:134). Even by Indian standards, this region had a dense population and small operational landholdings. (There were large estates owned by zamindars, but the estates were cultivated by myriads of tenants and subtenants.) Thus cane cultivation was generally a small-scale,

70

Peasants

~rsus

Capitalists

low-cost, and low-yield enterprise. Average costs around 1920 were about Rs.80-85 per acre, for a yield of about 10 tons of cane. (Yields in the Bombay Deccan were about 26 tons per acre [ISC 1921:12,36, 61,160].) Costs per acre were about 70 percent lower in the north than in the Deccan, and even costs per ton of cane were lower. Around 1910, for example, cane cost about Rs.6.67 per ton in the north, as compared with Rs.8.75 in Bombay (Sahasrabudhe 1914:5). In 1930, the cost was about Rs.7.50 per ton in parts of the north, as against Rs.17.75 per ton in Bombay (Indian Tariff Board [ITB] 1931:55-59). (This difference was offset, to some extent, by the higher sucrose content of the Deccan cane: in other words, there was less disparity in the cost of growing a ton of sugar than a ton of cane.) The large supply of cheap cane stimulated the industry to grow first in the north. Instead of obtaining this cane from a compact, centrally managed plantation, the sugar factories bought it from a host of small, widely-scattered cultivators. 3 This was a cumbersome cane-supply system for a number of reasons. First, there was no overall coordination for the planting and harvesting of the crop in a given factory area, so there might be too much ripe cane available in one portion of the harvest season and not enough in others (Amin 1984:190-95). Moreover, the operations of harvesting and delivering the cane were plagued with uncertainties and cross-purposes. Cultivators wanted to harvest their cane at its peak and then quickly plant other crops, but they could not be sure that a nearby factory would take delivery or pay a reasonable price, since the market was likely to be glutted. One way around this problem was for the cultivators to ignore the factories and process the cane intogur, which they often did (at least in western U.P.). On the other hand, factory managers had no assurance of being able to obtain a steady cane supply throughout the harvest season, partly because of the coordination problem and partly because of the diversion of cane to gur production. Moreover, in attempting to buy cane at the lowest possible price, and in delaying payment, or otherwise cheating or manipulating the cultivators, the factories often discouraged the latter from increasing the production or sale of their crop. In short, the relationship between the agricultural and manufacturing branches of the industry was unstable and antagonistic, which made it difficult for the factories to obtain a steady and reliable cane supply and thus to lower costs and invest in improvements. Moreover, this relationship discouraged the cultivators from investing in new equipment and methods of production. Why was the industry saddled with this problem? Why was sugarcane not grown on factory plantations? Indian industrialists would have liked nothing better than to copy the example of factory plantations elsewhere. However, they were unable to do so because India, unlike most other sugarproducing countries, was not a land of new settlement during the European colonial expansion. (Most plantation systems were established in the New World or other regions where land was abundant and cheap.) In India, most

Peasants versus Capitalists

71

of the good land was already densely occupied by 1900. The ISC summed up the problems of buying land on the open market: in no part of India does there appear any likelihood of factories being able to purchase any large areas of land by direct negotiations with the existing holders. The extreme fragmentation of holdings generally throughout India, the multiplicity of interests in them, the frequent encumbrances on them and the consequent difficulty in obtaining a valid title, but above all the reluctance of holders to part with their land all present an almost insuperable obstacle to such purchase. (ISC 1921:293, emphasis added)

A large compact area of good land could not be obtained on voluntary purchase, simply because the land was densely settled. By offering very high prices, the factories might have bought enough land, but this was impossible, given their weak competitive position against the Javanese industry. (The Javanese factories, it should be noted, obtained their lands through a leasing system imposed by the government on the villagers [Geertz 1963: 86-87]demonstrating, once again, the diverse nature of cash-cropping systems, and the prime importance of political power in shaping them.) If it was impossible to rent or purchase enough plantation land on the open market, then the government might consider compulsory leasing or acquisition in favor of the sugar factories. The ISC discussed these possibilities at some length. Although there were some members who argued strongly in favor of this approach, the majority of the ISC concluded that it was not feasible: In spite of the weighty arguments in favor of compulsory acquisition of land for sugar factories, the majority of us are opposed to such acquisition .... In short, the logical conclusion of those arguments is that the agricultural production of India would be immeasurably increased to the benefit of the country and of the world at large, if all land were handed over to be exploited by capitalist enterprise. The truth of this proposition is as undeniable as the impossibility of putting it into practical application .... we cannot contemplate with equanimity the establishment of factories in the midst of an aggrieved and sullen peasantry, which we are convinced would be the inevitable outcome, if land were acquired by Government on any large scale in order to promote the development of the sugar industry. (ISC 1921:295-98, emphasis added)

This passage explains, with admirable clarity, that the goals of technical efficiency and industrial growth were pointing in one direction while the will of the peasantry was pointing in the opposite direction. The opinion of the policy-makers was that they had no choice but to follow the latter direction. Contrary to what many have imagined about the power of the colonial administration, the government recognized that it did not have the political strength to displace large numbers of peasants from. their lands in order to create a plantation system. At the time that the ISC was deliberating on these matters, the nationalist movement was just beginning to appeal effectively to various sources of discontent in the countryside. For the sake

72

Peasants

~rsus

Capitalists

of political stability, the ISC followed a long-standing imperial tradition of seeking to minimize direct government intervention in the countryside. This policy position stands in contrast to the history of sugar production in Latin America and the Caribbean, where the land was often sparsely settled and capitalists or latifundistas were given official encouragement to seize large estates by various means, including outright military force (Wolf 1982:ch.5). The Indian government was, of course, able to encourage the sugar industry by other means, without having to resolve the underlying canesupply problem. The most obvious method was a protective tariff on imported sugar. However, the influence of free-trade ideology was still strong in official circles in 1920; and the consensus seems to have been that subsidizing an inefficient industry out of the pockets of consumers would not lead to a healthy economy (ISC 1921:346)-though this consensus had begun to weaken after World War I (Dewey 1978). When the depression of the 1930s hit the countryside, all doubts were set aside and a protective tariff was imposed. The tariff, enacted in 1932, was a great success. The number of sugar factories in the country rose from 29 in 1930-31 to 140 in 1936-37, and the annual production of white (industrial) sugar rose even more rapidly, from 120,000 tons to 1,129,000 tons. As a result, imports fell from 809,000 to 22,000 tons per year. Most of this boom was concentrated in the north: U.P. and Bihar had 109 out of the 140 factories, producing 83 percent of the white sugar in India by 1936-37 (ITB 1938:18-22). Impressive as this growth was, it entailed no resolution of the basic problem of coordination between factory and field. Factories multiplied in number, but they remained relatively small and inefficient by world standards. This growth also brought no real change to the organization of cane production in the north. Instead of growing cane for a depressed gur market, the villagers were able to sell increasing quantities to the new mills. However, the organization of production remained essentially the same: cultivation of mostly small holdings by family labor, supplemented to varying degrees by hired labor. There was not much change in equipment or methods. Little investment was made in fertilizer, irrigation, or improved equipment, particularly in eastern U.P. and Bihar, where the standard of cane cultivation continues to be low, even at present (Ramana Rao 1979:120-21, 145). Amin (1984:125) points out, for example, that this area was slow to adopt iron cane crushers for gur production. Many of the cane growers still do not use improved methods or equipment, one of the main reasons being lack of investment capital (Pandey n.d.:61). North Indian sugar factories were concentrated in two agroclimatic zones: the north-central zone (consisting of Bihar and eastern U.P.) and the northwestern zone (consisting of western U.P. and Punjab). In the former, the humid climate made irrigation inexpensive and sometimes even unnecessary. The mild winters and abundant labor supply made it possible to grow cane without great effort or expense; but on the other hand, there

Peasants

~rsus

Capitalists

73

were problems of waterlogging, flooding, and plant disease that held cane yields to low levels. The very cheapness of the crop and of the labor supply probably discouraged efforts to invest in better equipment and improved methods of cultivation. In addition, many growers were small peasants, dominated by landlords and moneylenders, who kept them in poverty despite the expansion of cash cropping (Amin 1984:139-57,269). Sugar factories were also established in the northwestern zone, where the climate was drier and there was, in compensation, abundant canal irrigation. Here, cane yields and methods of cultivation were generally of a higher standard, and the growers were less dominated by landlords and moneylenders (Amin 1984:60-61). However, cool subtropical winters limited yields, and the crop had to compete for irrigation and labor with other cash crops, particularly wheat and cotton. Investments in improved cane cultivation were (and are) limited by the demands of these other crops, particularly by the demand for labor during seasonal bottlenecks (Batra 1988, 1991). In this zone, many cane growers remained independentgur producers, putting them in a stronger bargaining position with the sugar factories. As a result, the factories had to pay higher prices for their cane and purchase it from longer distances, adding significantly to their costs. Amin attributes the conservatism of the cane growers in eastern U.P. to an agrarian structure dominated by landlords and moneylenders (1984:13957). In order to guarantee a steady supply of cane, the factories first relied on contracts with middlemen. The latter were mostly za.minda.rs or moneylenders, who already had substantial claims on the crops grown by the villagers and who could use these claims to compel the growers to harvest and deliver their cane to a certain factory at a certain time. In this way, the factories were relieved of some of the problems of scheduling and coordinating deliveries from thousands of scattered growers and were also saved from competing over cane supplies and thus having to raise their prices. Cane prices were held down, even in a period of rapid expansion, and a large share of the payments went into the pockets of the middlemen. This system of cane supply, therefore, did not encourage growers to invest in new equipment or to reorganize production along more efficient lines. In the late 1930s, the government of U.P. tried to improve the situation by organizing the growers into cooperative cane supply unions and requiring factories to buy much of their cane through the unions. These unions helped the factories to regularize their cane supplies, but they left much to be desired for the growers. For one thing, they were organized and run entirely by government bureaucrats. Although there were officers elected from among the growers, these did not have an active role in managing the unions, and most growers felt that they had little or no control over union activities (Hirsch 1961:111-25). Moreover, the unions were dominated by the same landlords and moneylenders who already acted as middlemen for the factories and who subverted the controls which unions were supposed to exercise over cane deliveries and payments (Amin 1984:266, 272-77). Thus the cane supply unions brought little improvement for the majority of cane growers.

74

Peasants *rsus Capitalists

The Bombay Deccan: Cane and Gul Production Until the 1960s, the Indian sugar industry was centered mainly in the north. There was a small industry in the Bombay Deccan, but this was a sideshow. Nevertheless, it was an interesting sideshow, first, because it developed along markedly different lines and, second, because it eventually became more dynamic and efficient than the industry in the north. These differences were the result of the cane supply systems that evolved in western India. The initial slow growth of the Bombay sugar industry was due to the semi-arid environment. In past centuries, sugarcane (like other garden crops) was grown only on tiny plots under well irrigation. When the Deccan Canals came into operation in the 1880s, enterprising cultivators began to discover the profits in sugarcane, which they processed into gut. The prime innovators along the new canals were cultivators from the Mali caste, who migrated into the canal areas from their native town of Saswad, near Pune city. The Saswad Malis had been market gardeners for generations, and they had two special skills suited to cane production. First, they specialized in intensive cultivation of garden crops on irrigated plots; and second,. they were experienced in selling their produce on the market. Their inherited experience contrasted with that of most village farmers, since the latter grew mainly subsistence crops, usually without irrigation, and had less experience with market transactions. The Malis migrated from Saswad into the Nira valley after the canal was opened. They found along this valley a society unprepared for high-cost farming. As noted in Chapter 3, the canal water was heavily under-utilized. Local cultivators did not apply canal water regularly to their subsistence crops and were slow to take up sugarcane, since they lacked the capital and expertise needed for such an expensive crop. Consequently, the Malis found an abundant supply of irrigation and land waiting to be used. This does not mean that there were large tracts of good land unused, but rather that the land was under-utilized relative to its new potential. In this semi-arid region, population was not dense by Indian standards (only about 160 persons per square mile in 1901), because the productivity of the land was low. A sparse crop of sorghum or millet was often followed by a whole year of fallowing. Under canal irrigation, however, the land could yield two or three seasonal crops each year, or a perennial crop like sugarcane; and these would be crops of greater yield and higher market value. In terms of its productive value, the same piece of land under intensive irrigation would be worth many times its value than when dry. Mann and Kanitkar (1921:76-81) found that an acre of dry sorghum or millet would bring a gross income of about Rs.12. The average gross income from an acre of cane, on the other hand, was found to be Rs.618-.fifty times the value of the staple dry crops (Inglis and Gokhale 1928:15). The same point is illustrated in Table 4.1, where we note that the average rental value of a cane block was nearly twelve times the value of unirrigated

Peasants

~rsus

TABLE 4.1

Capitalists

75

Land Rental Values Along the Deccan Canals

Type of Land & Irrigation

Rs./Acre

Sugarcane block (12-month irrigation on onethird & 8-month irrigation on two-thirds) Two-season irrigation (8 months) One-season irrigation (4 months) Unirrigated land in canal villages Unirrigated land in dry villages

30.28 11.46 4.65 4.96 2.57

(These are averages over several canal tracts during the period from 1922-23 to 1926-27.) Source: Inglis and Gokhale 1934:8. land in a dry village. On this basis, we may safely argue that canal irrigation for sugarcane enhanced the value of land by at least ten times. Thus the addition of canal irrigation to a large portion of the Nira valley was equivalent to multiplication of the land area, since one acre under perennial irrigation was equivalent in value to many acres of dry land. This created what I term an "irrigation frontier." It was like an ordinary frontier, in that land was readily available to innovators at cheap rates. A Mali immigrant planning to grow sugarcane could offer a rent that easily surpassed the best return the owner had ever obtained from dry-cropping the plot himself. (The rental rate for cane land was about Rs.25 per acre in 1914.) Paradoxically, however, from the point of view of the Mali tenant, rent was only a small fraction (on the order of 5 percent) of his total costs, since sugarcane in the Deccan was extraordinarily expensive. Total costs ranged from Rs.500 to 600 per acre, which was five to six times more costly than an acre of cane in northern India, although the yields were also at least 2.5 times higher (Knight 1914:33-34; Inglis and Gokhale 1928:15). The reasons why Deccan cane was so expensive are explained below: the high costs of irrigation, labor, bullocks and fertilizer were responsible. Land rent was one of the less important expenses. Thus land was cheap along the irrigation frontier. This situation began to change in the Nira valley after World War I, when the Malis were copied by local cultivators and the demand for land was rising. Fortunately, a series of new canals were opened in nearby districts; and by the late 1920s, nearly all the Saswad Malis had left the Nira canal, following the frontier. The scale of this frontier should be explained, since from some points of view it was large and from others small. By 1936, there were six major canal systems in the Bombay Deccan, irrigating 242,000 acres, including 40,000 acres of sugarcane (IICB 1938:88-89). For a dry province, this amounted to a hefty increase in irrigated land, though it was less than 2 percent of the net sown area of the Bombay Deccan. It was also a great

76

Peasants

~rsus

Capitalists

deal smaller than the irrigation frontier of northwestern India, which was opened up by large-scale canal building in the middle of the nineteenth century. By 1935, the snow-fed rivers of the Punjab were watering 11 million acres under government irrigation works, equivalent to about 3 7 percent of the total sown area (IICB 1938:81). Only a relatively small area of the Bombay Deccan, then, was directly affected by the irrigation frontier. Nevertheless, this frontier was large enough to stimulate cash cropping, innovation, migration, and industrialization on an unprecedented scale. The Malis and other entrepreneurs had plenty of room to move around in, at least until the 1940s. At first the Malis cultivated garden-sized plots with family labor, but they used their gul profits to rent more land, expand operations, and hire labor. In 1890, only 300 acres of sugarcane were grown along the Nira Canal, but thirteen years later there were 8,500 acres. Though the Malis started small, quite a few of them expanded to at least 20 acres of cane (sufficient to require the use of a power-driven cane crusher), and some families eventually cultivated hundreds of acres. The irrigation frontier lasted only for a brief period along any given canal. When a new canal was opened, Malis and other innovators moved in. Within a few years, local cultivators (primarily Marathas) began to imitate the Malis, with varying degrees of success. Within two decades, then, the available irrigation would be in use, and the Malis and their imitators would be looking for new horizons. Large-scale cane growers operating on rented lands emerged among local Marathas, Dhangars, and Brahmans. The frontier made room, as well, for thousands of small cane growers, as we shall see in Chapters 6 and 7. Farms which had been too small for viability under dry crops could become prosperous under canal irrigation. Medium-scale farmers with limited capital could rent out part of their land and cultivate the remainder intensively, yielding higher incomes than before. The frontier made it possible for small farms to coexist with those that were larger and more heavily capitalized. In contrast with northern India, the cane growers of the Bombay Deccan were not under the domination of landlords or moneylenders. One reason they were not dominated by big landlords was due to the irrigation frontier. Malis and other cane growers sometimes rented land from Rajas, Jagirdars, and other big landlords, but on the same terms as they obtained from ordinary villagers. Because land was widely available at reasonable rates, big landlords could only collect a competitive rent, and they had no control whatever over the production process. (This will become clear as we go along.) The irrigation frontier not only made land available for innovation and expansion but it also freed the land from past tenure and labor relations. It became possible to rent land in a competitive market outside the control of peasant lineages or princely patrons. Entrepreneurs were free to mold new production relations to suit the needs of their crop. Not only was land freed from prevailing networks of social control, labor also became available on a competitive market. Much of the hired labor

77

Peasants l&rsus Capitalists

employed by the Malis was supplied by seasonal and permanent migrants from the dry villages. The availability of these laborers helped to dissolve old relations of production within the canal tracts. Thus the irrigation frontier functioned as frontiers are supposed to, providing not only access to new resources but freedom for social innovation; and it yielded its first fruits to Mali immigrants who were not bound by the social networks of indigenous villagers. Malis and other entrepreneurs were likewise not dominated by the moneylenders (savkars) who financed their crop and acted as brokers in selling the gul. This may seem like an implausible assertion. It is a widely held assumption that village cultivators, especially those growing cash crops, are under the economic domination of their creditors (see, e.g., Whitcombe 1972:171). Amin (1981) has argued that this was the case for cane growers in eastern U.P.; and the extreme inequality and stagnation in that region make this a plausible argument. In the Bombay Deccan, however, the situation was different. There was a great deal of innovation, investment, and economic mobility by the cane growers, making it less likely that they were dominated and exploited by their creditors. Evidence bearing on this question is discussed below and in later chapters. If merchants did not control all the profits from sugarcane, this was partly due to the irrigation frontier. Where land is available at low cost, it is difficult for creditors to coerce their clients by putting liens on their lands. A farmer in financial trouble could sell out, pay his debts, and, following the example of the Malis, start over as a tenant. This might be troublesome, but perhaps less so than dealing with a coercive creditor. Moreover, in an expanding commercial economy, the merchants earned more profits from doing a large volume of business, asgul brokers, than they might have earned from unbearable interest rates. Merchants wanted to attract more clients and encourage them to expand production, rather than bleed them dry. The High Cost of Sugarcane Still another explanation is that the financing of this crop was so expensive, it gave the debtor an advantage. Big debtors always get better credit terms. In considering the impact of sugarcane on the organization of production, and on rural society in general, we must bear in mind the technical requirements of the crop, plus the high cost of meeting those requirements. Irrigation was expensive, since sugarcane was a thirsty crop, demanding regular watering every ten days or so throughout the year. Around 1900, the rate for year-round canal irrigation was Rs.25 per acre (Visvesvaraya 1903:12). This was three-and-a-half times the total cost of growing foodgrains on dry land two decades later (Mann and Kanitkar 1921:76,80). However, the cost of irrigation was by no means the heaviest charge on the cultivator's working capital. The largest share of the costs (some 45 to 50 percent) went for farm-yard manure, oil-cake fertilizer, chemical fertilizer, and seed cane-costing in all about Rs.230 per acre (Knight 1914:33; Inglis and Gokhale 1928:15).

78

Peasants versus Capitalists

Additional heavy charges went for hired labor and the hiring or upkeep of bullocks, since sugarcane stayed in the ground from 12 to 18 months, requiring labor over several seasons. Before planting, the crop needed three plowings, interspersed with harrowing, then manure spreading, ridging and furrowing. Planting was followed by five weedings, by another manuring, and by earthing up (mounding earth around the young plants to prevent lodging later on). In addition, labor was needed every ten days for irrigation. And finally, a large team effort was required for cutting the cane, carting it to the crushing mill, crushing it, boiling the juice, and making it into gul, which was then carted off to market. (Gul was made with bullockdriven cane crushers and big, shallow iron pans for boiling the syrup.) The labor cost of all these operations, including hired bullocks as well as men, came to a total of about Rs.160 to 220 per acre around 1910, or 35 to 40 percent of total recurrent costs (Knight 1914; Inglis and Gokhale 1928). The small grower could do many operations with his own family labor and bullocks, thus reducing cash expenditures. However, there were several operations, such as ridging and furrowing, earthing up, and cutting, carting, crushing and boiling, which were better done by contract teams, costing a total of at least Rs.135 per acre (Knight 1914:33). Labor costs were high, not only because the crop demanded lots of work, but also because labor was hard to get in this region of relatively low population density, especially during the harvest season. In addition to all these running costs, cane growers needed improved implements, stronger bullocks, more powerful cane crushers and better boiling furnaces. These they could either rent or purchase. Purchase costs are discussed in the next section; for the moment, we may note that the rental cost of a bullock-driven cane crusher was only about Rs.8 per acre in 1905 (Knight 1905:14). Thus the cost of machinery was not very high compared to other running costs, such as those for labor and fertilizers. This provided a strong incentive for fixed capital investments which would reduce working capital expenditures. To summarize, during the period before World War I, recurrent costs ranged from Rs.300 to 500 per acre, which was something like 50 to 70 times the cost of growing foodgrains on dry land. At that time, sugarcane in the Bombay Deccan was probably the most expensive crop grown in India. There were two main reasons why costs were so high. First, there was low rainfall: more irrigation was required, and irrigation was more expensive than in the north. Second, the local variety of cane required lots of fertilizer, especially when grown on canal irrigation. The canal-irrigated valleys had heavy black soils which became sticky and waterlogged under copious irrigation; consequently, large doses of organic manure were needed not only for nitrogen, but also to maintain the physical quality of the soil. Careful experimentation by the Department of Agriculture established that at least Rs.138 worth of fertilizers per acre were required for a good yield. Other costs (for labor, bullocks, and equipment) were in proportion to this amount.

Peasants

~rsus

Capitalists

79

Even with the thriftiest methods, the cost of cultivation was still about Rs.320 per acre (Knight 1914:12-13, 34). Management of an Expensive Crop A capital-intensive crop demands increasingly careful management in order to reduce costs and risks. An indication of careful management is the willingness to adopt new inputs, methods and implements, in order to economize on running costs. 4 Some improvements were tested and promoted by the Bombay government's Department of Agriculture. However, many improvements were developed by the cane growers themselves. For example, there was an entrepreneur in the Baramati area, G.G. Shembekar, who helped to introduce a whole new crop cycle to the region. Shembekar, who is profiled in the next chapter, made a number of systematic experiments with new farming techniques, consulting researchers in Pune and Coimbatore. Prior to World War I, cane was grown on a twelve-month cycle, with planting usually in December. Shembekar found that cane would grow better if planted at the start of the rainy season, in June, though it would still have to be harvested in the winter. He began experimenting with an 18-month cycle. The first trials were disappointing, but in the second year he earned a sizeable profit (Mahajani 1986). Growers flocked to his farm to learn the secrets of what they called adsati (year-and-a-half) sugarcane. The practice spread across the region and is one feature that made sugarcane yields higher in the Bombay Deccan than in the north, where the twelve-month cycle is still standard. The cane growers also found other avenues for improvement. From about 1890, heavy iron plows began to replace the indigenous wooden plow in the canal areas. The iron plow, which cost about Rs.45 in 1905, turned the soil deeper, improving its texture under heavy irrigation. This plow required three teams of bullocks, but it did the work two or three times faster than the wooden plow, saving on the cost of labor. Iron harrows, which cost about Rs.IOO to 200 by 1914, were better than indigenous implements for mixing manure into the heavy soil (Knight 1905:4; 1914:6-11). Crushing the sugar cane, boiling the juice and makinggut were operations where careful management and improved equipment paid good dividends. Gut from the Bombay Deccan was known for its high quality, which brought premium prices. The quality of the gut depended on the timing of the harvest and the care with which it was made (Knight 1914:30-31). As the Saswad Malis migrated into new canal areas, they brought better equipment and more efficient ways of organizing contract teams to do the harvesting and processing operations (Patil 1928b:l7). The adoption of new cane crushers was rapid. In the late 19th century, most of the indigenous crushers were wooden. Bullock-driven iron crushers became common along the canals by 1906, the Department of Agriculture promoting their use and helping to improve their design (Patil 1927:4-7). Good-quality sugarcane has a tough stem, like bamboo, requiring powerful

80

Peasants

~rsus

Capitalists

machinery to extract all the juice. The iron crushers extracted about 14 percent more juice, and they also worked faster and saved labor (Knight 1914:25). By 1927, an iron crusher cost about Rs.200 to 300 (Patil1927:78). In his study of the Godavari and Pravara canals in the late 1930s, Gadgil (1948:121-25) found that a bullock-driven iron crusher, plus the related gut-making equipment, cost from Rs.300 to 650, the average being about Rs.500. There were about 375 of these crushing units along the two canals. Owned by large cane growers, they represented a total investment of about Rs.200,000 and crushed about 3,000 acres of cane. Smaller growers rented this equipment, supplying their own bullocks. Cane crushers powered by oil engines were also introduced by the Agriculture Department. In 1914 there were only three or four in the whole region, but by 1927 the number was more than 45. These power crushers saved human and bullock labor and extracted about 10 percent more cane juice than the bullock-driven units. Power crushers were needed by growers with more than 20 acres of cane. A small power crusher, along with multiple furnace and other equipment suitable for processing 20 to 45 acres of cane, cost about Rs.6100 in 1927. Larger units ranged up to Rs.14,200, with capacity for 100 acres (Knight 1914:26; Patil 1927:10-11,18). Along the Pravara and Godavari canals, in the late 1930s, Gadgil found more than a hundred power-driven units, crushing about 6,300 acres of sugarcane. These larger gut-making units were usually owned by persons holding a considerable acreage under sugarcane who use their equipment primarily for the conversion of the sugarcane from their farms into gut; but they also manufacture gut for other farmers in the area. . . . The entire capital and labour equipment in this case is that of the owner of the gut factory. The farmer brings the sugarcane to the factory and takes away the gut. (Gadgil 1948:117-18)

The investment in each of these power-drivengut "factories" was Rs.6,000 to 7,000, giving a total investment of about Rs.700,000 for 100 units. Compared with the working capital required just to grow the crop, this processing equipment was only moderately expensive. From the above figures, for example, the purchase cost of power-crushing and gut-making equipment comes to just Rs.111 per acre of cane. Consequently, growers were eager to adopt new technologies, making capital investments (or obtaining equipment on hire) in order to reduce their working-capital expenditures. There was a tendency, of course, for different factor proportions to be used on different-sized holdings, the larger ones becoming more capitalintensive and the smaller ones remaining labor-intensive. New equipment enabled large growers to save on the cost of human and bullock labor while the demand for labor was rising. The manufacture of gut employed about 5,500 workers for four months per year along the Pravara and Godavari canals. The total employment generated by irrigated farming along these

Peasants *rsus Capitalists

81

canals was estimated to be more than 22,000 workers, with a wage bill costing Rs.4 million (Gadgil 1948:124). The larger growers were under pressure to substitute capital equipment for labor because hired labor was not only costly, it was sometimes hard to get. Most of the harvest and gul-making labor was supplied by migratory contract teams, consisting of cultivators with bullock carts from the dry villages. However, when the rains were good, these farmers preferred to stay at home tending their own crops, even when gul prices and wages were high (Patil1927:10). Moreover, wages sometimes rose whengul prices fell, catching the growers in a painful squeeze (Patil1932:22). Consequently, the growers used family labor, or substituted capital equipment for labor, in whatever quantities they could manage. 5 What can we say, then, about the level of managerial skills employed on this crop? Clearly, these skills varied from one grower to the next. Inglis and Gokhale (1928:3-5,13) stressed that the skill of the grower was a more important determinant of variations in yields and profits than the type or condition of the soil. There was no question that the Saswad Malis and some of their imitators were skilled entrepreneurs. On the other hand, there were local cultivators who tried to imitate the Malis without knowing quite what they were doing. By the early 1930s, the prolonged drop ingul prices had forced most to become stricter in their management of expenses. Those who were slow to adjust were saddled with heavy debts, and some were forced to sell off their lands. However, it is interesting to note that cane production fell off more sharply along the newer canals, where the Mali immigrants were concentrated (Inglis and Gokhale 1934:34-35). Perhaps it was easier for them, as tenants, to cut back production and leave land idle. Along the older Nira canal, where local cultivators had taken over after the Malis moved on, the total area under sugarcane remained remarkably stable. This suggests that local cane growers had learned their business and were able to economize well enough to keep going. Considering that they were caught in a very tight squeeze, this is tribute enough to their managerial abilities. An expensive crop stimulates careful management, innovative techniques, and new investments in order to reduce costs and risks. Cane growers in the Bombay Deccan achieved the highest yields in the country, and many adopted new methods and technologies to increase output, economize on working capital, and raise the productivity of labor.

Rdations of Production The unusual capital and managerial requirements of this crop also exerted a strong influence on the organization of production-or, in other words, on class relations. In addition to the adoption of new methods and equipment, the cane growers introduced a new pattern of crop production, based largely on market transactions. Of course, long before the opening of the canals, there were instances of lands being rented for cash and labor hired for wages.

82

Peasants

~rsus

Capitalists

Nevertheless, the cane growers established a new pattern and intensity of such transactions, combining land, labor, and capital in specific ways to meet the needs of a new cash crop. They introduced this pattern into a regioti that had been devoted largely to subsistence cropping, where land and labor had been allocated mainly through familial and patron-client relationships. What specific relationships emerged among the suppliers of land, labor, capital, and human capital? Firs; consider the suppliers of lsnd. In the late 19th century, land was largely in the hands of village cultivators, who did not recognize its commercial potential when canal irrigation became available and were glad to rent it to Mali immigrants or other entrepreneurs. At a rental rate of about Rs.25 per acre (Knight 1914:330), land was simply not the most expensive or scarcest factor of production, at least not while the irrigation frontier lasted. Consequently, mere ownership of land did not provide effective control over the production of sugarcane. The most enterprising cane growers were usually tenants, renting land from villagers and from a few scattered Rajas and other big landlords. As we shall see in the next chapter, the Rajas encouraged tenants who would improve their land, offering long leases at fixed rents. The irrigation frontier meant that they lacked the power to coerce their tenants, since there was plenty of cheap land available. Moreover, the high cost of cane cultivation meant the tenants had managerial autonomy, since landlords could not afford to put working capital worth Rs.SOO per acre into the hands of tenants who might decamp. The tenants had to provide the working capital and take the risk of losing it. As a result, while other crops were grown on a sharecropping system, sugarcane never was. If the Malis and other enterprising tenants were in control of cane production, why did they not demonstrate this by buying land? The answers are simple. Some expected to exhaust the land and move on, so long as new canal tracts were being opened up. Moreover, buying land would have meant tying up their working capital. Working capital on the order of Rs.SOO per acre was harder to obtain than rental land at the rate of Rs.25 per acre, so why sink capital into land? The pattern of renting instead of buying prevailed until the 1930s, as long as new canal tracts were being opened up. Renting also had the advantage that tenants could easily reduce their operations when prices fell, without leaving large fixed investments idle. Second, consider the suppliers of cspitsl. Most cane growers had to borrow from gut brokers and share profits with them (Patil 1932:21-22). Though they were often accused of cheating the growers, it would be a mistake to assume that the brokers were able to bleed. the farmers and rake off all profits through interest charges, brokerage commissions, and creative accounting. Living and working in town, as they had to, the brokers could not supervise the day-to-day production of the crop. On the other hand, without such supervision, lending large amounts of working capital to a partner who did not expect to share in the profits would have been foolhardy. Consequently, the management of the crop (along with its attendant risks and potential profits) had to be left in the hands of the actual cultivator. Would any

Peasants Versus Capitalists

83

cultivator have assumed the burden of such management, if all surplus were extracted by the broker? (The convenient alternative would be to grow subsistence crops, thus vastly reducing his credit needs.) Good management was essential in growing this crop, and managerial experience was still scarce. Could it then have gone unrewarded? This argument leads to a seeming paradox: the larger the credit needs of the crop, the more control the cultivator

will have over its production.

What evidence is there to support this reasoning? Direct evidence on the sharing of profits and losses between moneylenders and cultivators is difficult to obtain, since both sides are generally reluctant to discuss credit transactions with anyone else. However, the indirect evidence seems conclusive. Many small cane growers were able to expand their scale of operations, in a few cases to hundreds of acres. (Examples are described in Part Three.) Moreover, many were also able to invest in cattle, implements, cane crushers, irrigation wells, houses, business enterprises, education, and so forth. In these cases, at least, it is evident that the brokers did not extract all the surplus from cane production. We might also note that most brokers did not attempt, or were not able, to parlay their control of credit into control of land: that is, they did not become big landlords. The reasons have already been suggested: ownership of the land alone did not mean control over cane production; and merchants, like enterprising tenants, did not want to tie up their working capital in land-not even at the artificially low prices which might be obtained by foreclosing on mortgages (cf. Musgrave 1978). It was not the direct cost which inhibited them, but rather the opportunity cost. This leads to an interesting conclusion. If the brokers, who had access to capital and who could have bought plenty of land (especially during the depression), did not become the dominant class in this production system, this must have been because human capital was more important than monetary capital. In other words, skill was the strongest determinant of variation in yields and profits. This rather abstract formulation had good social visibility, due to the outstanding abilities of a particular caste, the Saswad Malis. The crucial importance of expertise in managing large amounts of capital meant that farmers, not moneylenders or landlords, emerged as the class in control of cane production. Third, consider the suppliers oflabor. For the Malis and other entrepreneurs, the supply of labor came initially from their own families, supplemented increasingly by hired labor. Cane growers worked out an elaborate system of labor arrangements to suit the requirements of their crop: • Gang laborers, often women, were hired on a daily basis to do weeding and other comparatively light tasks. A daily laborer was called a rojandar, from roj, meaning "daily." • Contract teams of men were hired to do heavy, specialized operations, such as ditching, planting, and ~arthing-up, on a per-acre basis. Such a team was called a toli or, in the Baramati area, a phail. These teams

84

Peasants versus Capitalists

consisted mostly of young men, with strength and skill, who could complete their contracts in minimal time. Each was organized by a mukluulam, or foreman, who did the same work as the other team members and negotiated contracts with the cane growers. • Individual men were hired on annual contracts for more continuous operations, such as irrigation and cattle tending. An annual laborer was call a saldar, from sal, meaning "year." While the other types of hired labor were used by small farmers on an occasional basis, annual laborers were hired only by middle and large farmers. Those with experience were sometimes hired as mukkadams (foremen) to supervise other laborers. • Migrant teams of harvest workers brought their bullock carts and offered an array of coordinated services for harvesting cane and making gul. Such a team was known as a ravi. These specialized contract teams were employed near Pune as early as 1905, and they usually came from the dry villages (Knight 1905:9; Patil 1928b:1). The number of hired laborers varied with the size of the farm and the supply of labor within a farmer's family, but a farmer with five acres of cane (that is, 15 acres under irrigation) required all four types of hired labor. He might also buy his own cane-crushing and gut-making equipment (which was otherwise rented from neighbors). At the top end of the scale, a very big cane grower might even hire a "master" (this was the local term) to supervise laborers and manage farm operations in consultation with the grower. As the cultivation of sugarcane expanded, so did the demand for labor, stimulating seasonal and permanent migration from the surrounding dry villages. Most of the seasonal migrants were landowning peasants, who brought their bullock carts along for harvest operations. As a consequence, they were not dependent laborers, subject to ready coercion. In a good season they often stayed home tending their own crops, regardless of the wages offered by the cane growers. On the other hand, droughts brought them flocking to the canals in large numbers, forcing wages down. That the labor market was less than coercive is shown by the trend in wages. Cash wages for agricultural labor in the Bombay Deccan nearly tripled between 1900 and 1921, while real wages (that is, cash wages deflated by the cost of living) rose by 45 percent (Shirras 1924:23; cf. Keatinge 1912:6873). Calculations over a longer period show no overall trend (Guha 1985:14142). However, the same source shows that real wages reached their peak in 1931-33, which was precisely the period of greatest difficulty for the cane growers. High wages at that time were due to good seasons in the dry villages; and the cane growers' inability to cut back on wages shows how weak was their control over the labor supply. The low levels of coercion in the labor control system are also illustrated by the careers of some permanent migrants, who came to the canal areas as construction workers and daily laborers. After a while, some were hired on

Peasants vhsus Capitalists

85

annual contracts and later promoted to foremen. Later still, they became tenants, and a few eventually bought out their landlords and became cane growers in their own right. (See Chapters 5 and 6 for details.) These striking cases of upward mobility not only bear out the lack of coercion in the labor system, they also reinforce the point that human capital was the most crucial factor of production. While working on annual contracts, and then as foremen, these laborers were acquiring skills in the management

of cane production, skills which were the basis of their subsequent upward mobility.

Thus the Malis and other entrepreneurs introduced a more complex and specialized organization of production into the canal villages, a system based on market transactions among independent suppliers of land, labor, capital, and expertise. These cane growers were different in several respects from those of north-central India: they were not under the domination of landlords or moneylenders; they were free to innovate and invest in new methods and equipment; and they were able to reorganize production relations to suit the needs of their crop. Many of these differences were the result of the irrigation frontier, which made land, labor, and water available in new combinations. The high cost of cane cultivation was also important, since it stimulated technical and organizational innovation. Origins of the Bombay Sugar Industry Before 1932, innovations in the sugar economy of the Deccan centered mostly on the production of gul. Industrialists also saw the potential for a modern white-sugar industry. However, they were preoccupied with solving the cane supply problem, which presented even greater difficulties than in northern India. Three sugar factories were started in the Bombay Deccan during the sugar boom of the 1920s. Two followed the north Indian pattern of cane supply and quickly failed, primarily for three reasons: First, the cost of cane cultivation was higher than in the north, so the factories had to pay high prices to the growers but still sell their product in competition with Java sugar. Second, because the growers could not be controlled through middlemen as they were in the north, it was difficult for the factories to obtain a steady supply of sugarcane when gut prices were high. (See Figure 4.1.) Third, sugar prices declined after 1924 and fell further with the onset of the depression (see Figure 4.2). Only after the sugar tariff was enacted in 1932 did the Bombay industrialists find a way around these problems. The production of cane and gul had stagnated during the depression, so there was less demand for canal water. Meanwhile, the irrigation frontier had expanded along several new canal systems. This allowed the industrialists to try something that was usually impossible in India: a system of factory-run plantations. These were established only along the new canals and in a new area of perennial irrigation on the Nira Left Bank Canal (whose capacity had been expanded by a new storage dam). In these areas, land and irrigation were under-utilized and were available at rates that the industrialists could afford.

Peasants "!&rsus Capitalists

86 70

60 Ul

J:l

..... 0 '01' N

50

(

~

30

4

Column Ttl. 16 \ltd. Total (41)

4

1 4 4

2 4

48 2 9 15 11 7 15 107

(129) (5) (15) (26) (17)

(7)

(18)

(216) 8

B. 1950 Landless 0.01-1.0 1.01-2.5 2.51-5.0 5.01-15 15.01-30 >30

7 2 6 1

Column Ttl. 16 \ltd. Total (41)

3

5 2 3 3 5 4 22 (28)

1

35 14

13

2 2 2 7 (7)

10 12 11 12 107

(111) (22) (30) (16) (15) (11) (12) (216) 8

(For 1930, Spearman's Rho - 0.51 [unweighted], significant at 0.01.) (For 1950, Spearman's Rho- 0.43 [unweighted], significant at 0.01.) • Weighted Ns are rounded to the nearest integer; hence the total N is slightly less than the sum of the partial Ns. ~: Sample survey in Malegaon.

gories. In this table, we see how the middle peasant castes (Marathas and other peasants) were distinct from those above and below them. The low and mid-low castes had a restricted range of economic variation in 1930, being mostly landless laborers or small farmers. High caste families, conversely, were almost entirely large landowners. The middle peasant castes were the only ones to vary across the entire range from landless to big farmers. Thus inherited caste status partly determined economic wealth, but primarily at the extreme ends of the hierarchy. For the majority in the middle (60 percent of the sample, unweighted, and 55 percent weighted), inherited caste status had less influence on the size of a family's landholding. This helps to explain why the relationship between caste and class in Indian rural society has been discussed at length but inconclusively (see, e.g., Beteille 1971; Sharma 1984; Jayaraman 1981; Omvedt 1982). Some

168

The Pattern of Inequality and Mobility

analysts (e.g., Mencher 1974) have argued that caste is equivalent to economic class. In Table 7.6, the restricted range of variation at the extreme ends of the caste hierarchy lends support to this view. On the other hand, the almost unrestricted variation within the middle peasant castes supports a contrary argument, that caste status has no necessary connection with economic wealth for the majority of villagers. The problem with both arguments, as they are usually presented, is that they are not based on empirical comparisons of variation within and between castes. My own argument, outlined in earlier chapters, is that caste helped to generate and maintain fairly extreme economic inequalities, particularly at the high and low ends, long before the rise of the sugar economy. Moreover, the data under consideration here suggest that the commercial economy has loosened some of these prior inequalities. The lower part of Table 7.6 .shows some evidence for this conclusion. By 1950, the low and mid-low castes had moved upward in a few cases. Likewise, some of the other peasants had worked their way higher up the ladder. On the other hand, a couple of high-caste families had moved downward, though remaining on the whole distinctly well-off. Between 1930 and 1950, there is a decline in the measure of association (Spearman's Rho), indicating a looser relationship between caste and land by 1950. Another way to test the changing relationship between caste and land is by F ratios, derived from the analysis of variance. 14 The object here is to compare economic variation between strata against variation within them. If caste categories were internally homogeneous, then any apparent differences between them (as in mean landholdings, for example) would be statistically significant, and castes would be more or less equivalent to classes, as Mencher argues. Conversely, if caste categories were internally heterogeneous, they would overlap in terms of land distribution, and their apparent economic differences would be less significant. The F ratio is a measure of the variance between categories divided by the variance within them. When the F ratio is high, this means that economic differences between castes are stronger than differences within them. Conversely, as the F ratio declines, there is greater variation within castes than between them. In Table 7.7, the F ratio for 1920 equals 14.20, while for 1950 it is 2.98. Both values are significant at five percent: in other words, there is an association between caste status and size of landholding throughout the period. However, the strength of this association declines dramatically, indicating that the relationship started strong and became weaker over time. (Weighted F ratios show the same pattern.) 15 High-caste families tended to be big landowners and low-caste families small ones, but with less certainty as time went on. In 1920, knowing a family's caste and lineage would have given a fairly good prediction of its landholding, but this was much less true by 1950. 16 (After 1960, the F ratios rise again, though nowhere near their original level: this is a point £Or discussion in Chapter 9.) Table 7.8 is comparable to 7.3, which tested the association between initial land size and subsequent changes, only Table 7.8 takes caste category

169

The Pattern of Inequality and Mobility

TABLE 7.7

Analysis of Variance for Landholdings, Broken Down by Caste Categories

Year

F Ratio

1920 1930 1940 1950 1960 1970 1980

14.20• 7.04• 3.07 2.98 2.96 4.58• 5.12 8

Degrees of Freedomb 96 101 101 101 101 101 101

Correlations R

Eta

Weighted F Ratio•

0.51 0.39 0.31 0.31 0.32 0.35 0.33

0.65 0.51 0.36 0.36 0.36 0.43 0.45

(31.00) (17.93) (6.84) (6.60) (6.56) (9.91) (10.39)

(N- 102 for 1920, 107 for later decades.} • Significant at 0.01; other F ratios in this column significant at 0.05. b Degrees of freedom within caste categories (unweighted sample). • All F ratios in this column are significant at 0.01. Source: Sample survey in Malegaon. TABLE 7.8

Analysis of Variance for Net Changes in Landholdings, Broken Down by Caste Categories

Decade

F Ratio•

Degrees of Freedomb

1920s 1930s 1940s 1950s 1960s 1970s

1. 91 2.27 0.25 0.19 1.14 0.92

96 101 101 101 101 101

Correlations R

Eta

Weighted F Ratio

-0.18 -0.15 -0.01 -o.03 0.04 -0.11

0.30 0.32 0.11 0.10 0.23 0.21

(2.45) 0 (5.37) 0 (0.46} (0.25) (1.89) (1. 28)

(N- 102 for the 1920s, 107 for later decades.) • No F ratios in this column are significant at 0.05. b Degrees of freedom within caste categories (unweighted sample). • Significant at 0.05. Source: Sample survey in Malegaon.

as the initial condition. In all decades, the F ratios are low: none of the unweighted ratios, and only two of the weighted, are significant at five percent. In other words, there was not much relationship between caste and the net amount of land gained or lost in these decades. The largest F ratio is for the 1930s. However, we note that the R correlation is negative, which suggests an inverse relationship. Moreover, Eta is much higher than R,

170

The Pattern of Inequality and Mobility

suggesting that the assoctatton was nonlinear. In other words, there may have been a relationship between caste and mobility in the 1930s, but not in the direction predicted by the standard theory. Higher-caste families tended to lose land and middle or lower ones to gain in that period. Defenders of the standard theory might argue that, even if the association between caste and landholding weakened after 1920, it was already high because of changes caused by the first phase of commercialization, from 1885 to 1920. I think the data on caste and land actually point in a different direction. They suggest that caste and wealth were already closely linked before 1885, particularly at the extreme ends of the status hierarchy. As we have seen, the strongest connection between caste and land was in the highest and lowest castes-that is, among castes which had for centuries specialized in non-agricultural occupations and had acquired their lands through patronage and the baluta system. If we removed these castes from the tables, there would be almost no association between hereditary status and land ownership. Let us bear in mind that the middle peasant castes constitute the majority (55 percent weighted or 60 percent unweighted). In 1930, they owned nearly 70 percent of the land; and in 1950, they owned nearly 80 percent (weighted or unweighted). The fate of the middle peasant castes is thus crucial to any test of the "differentiation of the peasantry." If cash cropping had caused greater differentiation between 1885 and 1920, we would certainly see the effects within the middle peasant castes. However, Table 7.6 shows that middle and small farmers were not disappearing within these castes. Moreover, the majority of landless families in these castes, as of 1920, were recent immigrants to Malegaon. Local landless families were just 11 percent (unweighted) or 39 percent (weighted) within the middle peasant castes. We are faced, then, with a pattern of striking inequality between the higher and lower castes, while the middle castes show that ordinary peasants were not being driven off the land. The extreme differences between the higher and lower castes owed their origin to the Jagirdar's patronage and other non-commercial processes of the 18th and 19th centuries. High and low-caste families were relatively numerous in Malegaon because they provided specialized services to the Jagirdar. Hence the initial level of caste inequality, around 1885, was unusually high in this village, thanks to the number of families at the extremes. In other canal villages, where middle-peasant families were relatively more numerous, the pattern of change under commercialization probably resembled that found among the middle castes in Malegaon. The Great Depression

The great depression was a cause of rapid upward and downward mobility. The standard assumption is that these processes worked primarily to the disadvantage of middle and small farmers, and agricultural laborers. We have just seen that this was not true for Malegaon over the whole period from 1920 to 1950. It remains to take a closer look at the 1930s and explore

The Pattern of Inequality and Mobility

TABLE 7.9

171

Total Land Purchased and Sold by Malegaon Sample

Decade

Purchases

Sales

1920s 1930s 1940s 1950s 1960s 1970s

35.33 240.44 92.43 58.65 79.42 25.65

87.37 162.79 15.25 24.84 27.95 75.26

(Areas in acres.) Sample survey in Malegaon.

~:

the reasons why some big landowners lost land while smaller ones (and landless laborers) gained. First, let us review the general pattern of crisis. As we have seen in Chapter 4, sugar andgul prices entered a steep decline in the middle 1920s, even before the onset of the depression. Sugar prices began to recover in the 1930s, due to the protective tariff, but gut prices remained low and erratic. Both shot up, however, during World War II. For cane growers in the Bombay Deccan, the crisis was severe. The Agriculture Department's annual Season and Crop Report testified to the problems the growers were facing, as did its occasional field-study reports (e.g. Patil 1932). The government's alarm was also evident in the work of the Deccan Canals Financial Improvement Committee (1932), which was concerned with the decline in sugarcane acreage and canal revenues. On a national scale, the work of the Indian Tariff Board (1931, 1932) testified to the perceived scale of the crisis in sugarcane production. The crisis reverberated throughout the family histories collected in Malegaon. Table 7.9 shows the total lands purchased and sold by the sample from 1920 to 1980. Two points are worth noting: First, the sum of both columns is more than 100 acres in most decades; in other words, there was an active market in land throughout this period. Second, the 1930s show an unusual degree of activity, with purchases and sales totalling more than 400 acres. In other words, the depression created an exceptional number of opportunities for mobility in either direction. The preceding sections have shown that initial inequalities in land distribution were not magnified by this crisis, contrary to standard predictions. Why did these predictions fail? Among other mistakes, writers on this subject tend to assume that the various enterprises (large, middle and small farmers) were all doing the same thing with varying degrees of efficiency. If that were the case, then the large farmers, with their greater resources, ought to have had a permanent advantage. However, the reality is that there existed a variety of options and limitations concerning the organization of production, and these were only loosely linked to the size of landholding.

172

The Pattern of Inequality and Mobility

Before elaborating these points, we should review the sources of efficiency which worked in favor of large cane growers and helped some to survive the depression. One advantage in their favor was the possibility of using capital-intensive, technically advanced inputs, which could be substituted for hired labor, as we have seen in Chapter 4. Big growers also benefited from credit on easier terms. A remarkable paradox of the depression was that there was no lack of credit for the creditworthy. The Bombay Central Cooperative Bank and the Poona District Central Cooperative Bank received a huge influx of deposits,· because opportunities for investment in trade were shrinking (Registrar 1927:6). The banks thus found it harder to locate good borrowers than depositors (Registrar 1929:9; 1935-36:28; 1937-38:30,33). Consequently, big cane growers were given direct access to the Baramati branch of the BCC Bank. The bank's original purpose had been to finance crop loans through village co-ops. However, temporary measures were adopted during the depression to provide loans directly to individuals against real security, such as land, gold or produce in a warehouse (Registrar 1935-36:28). Big farmers were also well treated by private moneylenders. Gut brokers who survived the crash were eager to find trustworthy clients, and the Baramati magnates, for example, received credit on favorable terms. Three magnates (Shembekar, Date and Jachak) did business with a single broker, and their combined sales were said to have been worth 50 percent of the Baramati gul market. Some idea of how credit terms varied according to the assets of the borrower can be gleaned from a study by the Department of Agriculture: With the exception of perhaps two, all the farmers selected were those who represent the average Deccan farmer. The average rate of interest charged in finding the costs in all the villages under study was twelve per cent per annum. In the case of the two substantial farmers the interest rate was nine per cent. (Patil 1927:4)

The largest cane plot in Patil's study was six acres. Since the magnates were growing cane on a much larger scale, they surely obtained credit on even better terms than the "substantial farmers" studied by Patil. Magnates and other big bagaitdars also made good use of personal connections with administrative and business elites. Their links with the Rajas of Malegaon and Phaltan, and with the industrialists who established sugar factories in the area, have been described in Chapter 5. Given all these advantages, it is surprising that some of the big landowners and cane growers went broke in the depression and that others rose in their place. Let us now review those factors which made life complicated for big growers and sometimes helped their smaller competitors. In addition to the amount of land owned, several factors made a difference in how production was organized and how efficient it would be.

The Pattern of Inequality and Mobility

173

Risks and Opportunities Outside Aariculture Elite connections smoothed the path for big bagaitdars and provided opportunities for nonagricultural investment. However, the latter also carried an element of risk. One reason the Chawres failed was due to their investment in the Nira Valley Sugar Company, which only operated for a couple of seasons before closing. The Raja of Malegaon and some of his relatives also lost on this venture. Other eminent families in Malegaon lost their wealth in a different way, in village politics. There were several leaders who paid little attention to farming, leasing their lands to tenants and putting their money and energy into politics. Debts accumulated from granting favors, pursuing litigation, consuming conspicuously, and indulging vices such as gambling and alcohol. When agricultural prices dropped, there was no way out of debt except by the strictest management of production. However, the habits of power, including relatively negligent management of the land, made it impossible for some to recover from the spiral of debt. Tenancy deprived these landowners of the skills needed to manage a farm out of debt.

Skills and Experience Villagers did not have to own any land at all, initially, to become successful cane growers. The Saswad Malis were small-scale market gardeners who migrated to the canal villages and became wealthy by growing cane on rented land. Dhangar immigrants to Malegaon did the same, though they started as laborers. The Malis made use of their traditional skill with garden crops, while the Dhangars (formerly shepherds and dry farmers) applied skills they had acquired while working as laborers and foremen for established cane growers. Referring back to Chapter 6, we might recall that several cases involved a classic sequence of occupations for immigrant laborers. As boys, they started out herding cattle and doing casual labor: weeding and helping with irrigation. Later, when they grew stronger and more experienced, they joined contract teams performing heavy, specialized operations, such as ditching, planting, and earthing-up. It took up to two years to learn to do these operations efficiently. Later still, they became annual laborers for individual cane growers, doing the more continuous work of irrigation, tending crops, and working bullocks. After this, they became overseers or foremen for their employers; and when opportunities arose, they became tenants and eventually landowners. Contrary to the common assumption that agricultural labor is mindless, repetitive, and simple in organization, we find here evidence of a complex progression of work experience which eventually gave some laborers the ability to grow sugarcane on their own. Maclachlan (1983) has described the complex decisions which farmers must make as they try to allocate family labor and other scarce resources to competing "windows of opportunity" for different crops. From this and other data, he has reasoned that the most efficient decision-makers are those who have learned to do the physical tasks of agriculture and thus have learned how these can be varied and combined under different circumstances.

174

The Pattern of Inequality and Mobility

In the Malegaon area, not all successful cane growers had this kind of handson experience. (Some of the magnates were Brahmans, who preferred to supervise labor rather than do it themselves.) Nevertheless, hands-on experience was an asset, particularly in the absence of other resources. There is little else to explain the rise of some immigrants from landless laborers to big bagaitdars. This argument is consistent with Schultz's (1980) summation of a lifetime of research: according to him, human capital is the most important factor in raising agricultural production. He argues that land is usually overrated, and human capital underrrated, as resources for production. Farmers' skills and knowledge are the key to their productivity.

Specialization in Different Crops At the broadest level, some farmers specialized in subsistence crops while others concentrated more on sugarcane. The former lost opportunities during commercial booms but were safer during busts. Some farmers survived the depression by avoiding commercial crops, purchasing safety at the price of opportunity. Evidence to support this inference may be found in Table 7.2, which shows that small farmers gained much less land than landless laborers. If upward mobility were based simply on physical assets, this difference would make no sense at all. It does make sense, however, when we recall that immigrant laborers were hired by big cane growers. These laborers became specialists in cash cropping and in working with new equipment, whereas small farmers apparently specialized more in subsistence cropping, which was safer for those with limited assets and limited experience in

sugarcane.

Of course, small farmers also worked as hired laborers, but mostly on an intermittent basis, when they could spare time from their own fields. Thus they missed the continuous, year-round experience with sugarcane which was gained by the landless laborers. Small farmers were protecting their families' semi-independent subsistence base by specializing in foodgrains, while landless laborers had no such base to protect. However, the comparative safety and independence of the small farmer was purchased at a price: his commercial expertise was limited, reducing his ability to buy more land when it came up for grabsY

Hazards of the Family Cycle For enterprising peasants, success is contingent on the size and composition of the family. In some phases of the cycle, families could earn more wage income while having fewer dependents to feed; they could also increase the labor on their land, thus saving and investing in future production. On the other hand, if the head fell ill or died at the wrong time, the family enterprise could easily fail. This happened to the wealthy Chawre family at the worst possible time, in the early years of the depression. Their decline provided opportunities for some of their laborers, who became tenants and bought

The Pattern of Inequality and Mobility

175

them out. Much depended on the coincidence of commercial opportunity with unpredictable variations in the family cycle.

Efficiency of F11mily L11bor Big cane growers had to face serious problems in terms of the cost and efficiency of hired labor. After the onset of the depression, officials in the Department of Agriculture observed that big growers were operating at a disadvantage. A careful study of production costs in the late 1920s concluded that "Business farmers renting land and borrowing capital have little scope for sugarcane growing under the present circumstances" (Patil 1927:9). A later study by the same author (Patil 1932:3) showed these farmers caught in a price scissors: wages were still fairly high but gut prices remained low. As other officials noted, "the big cultivator employing hired labor on a large scale," that is, "the Deccan Canal irrigator and the cultivator of commercial crops" were suffering most from the fall in prices, while the small farmer did not suffer as seriously (Agriculture Dept. 1930-31:24). This was partly because smaller farmers were not as heavily involved in cane production. Another factor was that small and middle farmers relied more on family labor. Family labor was cheaper because family members did not have to be paid a cash wage. It was also more efficient, because family members were self-supervising and their skills improved through long-term experience and cooperation with each other. This point should not be confused with the notion of "self-exploitation," which Chayanov (1966) and others have used to explain how small farmers survive on limited resources-the idea being that family members contribute far more labor to their own small plot than a commercial farmer could afford to hire. In other words, family members will work for themselves at below the going wage, in order to secure an adequate supply of food from their own land. While this is certainly true, especially for farmers barely able to survive on small holdings, it does not explain how some middle farmers, tenants, and even laborers managed to do much more than survive and feed their families. Virtually every form of self-employment also involves self-exploitation, in the sense that more labor is contributed than could possibly be replaced by hiring. (The authoring of this book is an example.) What makes family labor valuable is not only that it can be obtained without paying a cash wage but also that it is obtained at lower transaction costs, provides a cumulative payback on cooperation and experience, and embodies a managerial component-a willing hand and mind, not just a strong back. Chayanov's idea is essentially quantitative: small farmers work longer hours (and presumably, at lower returns than they might get if they worked as hired laborers) in order to survive independently. However, the point here is essentially qualitative: family labor is more efficient than hired labor. There are several reasons why this is so. First, hired labor entails transaction costs: either there is a risk that it will not be available from the right persons

176

The Pattern of Inequality and Mobility

at the right times in the right amounts, if hired on an intermittent basis; or else one pays a reliability premium (wages in advance, extra food, clothing and shelter) to keep laborers on a yearly contract. Family labor is not only more reliable in terms of availability but also more flexible: it can be applied in small amounts over several days, in coordination with other tasks, instead of all at once. In addition, family labor is self-supervising, reducing managerial costswhich are real costs in terms of opportunities foregone to use the farmer's time on other tasks. Moreover, the quality of family labor is more predictable, based on the continuity of past cooperation and experience. Thus family labor embodies a managerial component, not only in terms of self-supervision but also in terms of continuity, coordination, and learning. People who have worked together with the same bullocks and equipment on the same piece of land can work more efficiently with each other than with temporary hired labor. For all these reasons, Schultz (1964, 1980) and Paige (1975) have stressed that family-operated farms are usually the most efficient, because decisions are made by those who know how to perform specific tasks in a specific micro-environment. Maclachlan (1983) has grounded this general argument in a detailed village study, arguing that the better agricultural managers are those who have experience working their own lands in coordination with other family members. Thus larger cane growers were obliged to use labor which was not only more expensive (in cash and transaction costs) but also qualitatively less efficient than family labor. During the depression, moreover, low crop prices made the cost of wages even higher in real terms. In general, then, there were a variety of limitations and options concerning how much family vs. hired labor to employ, how much supervision to provide, how much capital to substitute for labor, how much land to rent, how much effort to put into which crops, how much education and outside employment to pursue-and these options were only partially determined by size of landholding. Because families were constantly facing new risks and opportunities, both internally and externally, it is simplistic to assume that all families in a given size-category pursued essentially the same strategies of production and survival. If there were only one uniform way to combine the factors of production, then one size-category (big farmers, perhaps) might win out all the time. However, efficiency could be achieved in more than one way, and commercial cycles of boom and bust may have rewarded different factor combinations at different times. The commercial economy brought prosperity to those who combined access to land with experience in farming and managerial skills. Those who had land but lacked the other requirements gradually lost ground in the competition, or else they cashed in on rising land values and moved toward urban occupations. Conversely, those with no land had opportunities to acquire commercial farming skills, particularly if they came from castes which specialized in farming. These opportunities for upward and downward mobility came to a peak in the 1930s.

The Pattern of Inequality and Mobility

177

Political Economy So Far It is time now to assess what all this means in terms of more general interpretations of the last half century of British rule. In certain respects, it is difficult to make a summing-up, since the fruits of this period would become evident only after independence, in the course of events which are discussed in later chapters. Thus our assessment must be incomplete at this stage. Nevertheless, it will be helpful to consider what we have learned so far and to compare this with widely held opinions about commercial agriculture in the late colonial period. The conventional wisdom has been summed up by Chithelen (1985), who argues that, in the first decade of this century, a stratum of rich peasants "had crystallised at the apex of rural Western Maharashtra." Canal irrigation was a crucial factor in bringing about this "rise of the rich peasants." Because sugarcane was such an expensive crop, it was only natural that those with the necessary resources moved in to make use of such opportunities at the cost of the meagre landholdings of the local small peasants. Thus, introduction of canal irrigation itself contributed substantially to the differentiation process. (Chithelen 1985:606)

From this point of view, differentiation in the canal villages was first caused by the Saswad Malis, who immigrated and took over land for commercial production-land which mostly belonged to "local small peasants." Later, to the extent that local villagers became more actively involved in the production of sugarcane, "it was mostly the rich peasantry who took to canal irrigated farming," according to Chithelen (1985:606). Moreover, the differentiation which started at the very beginning of the sugar economy continued to grow: rich peasants invested in mechanical cane crushers; established cooperative purchase and sale unions (in addition to the credit co-ops which they already controlled); consolidated their political interests by joining the non-Brahman movement and the Congress party; and after independence, established a host of cooperative sugar factories, which served their interests almost exclusively (Chithelen 1985:607-11). This scenario, which simply restates the standard theory in terms suitable to the canal villages, is inconsistent with a number of points discussed in the last few chapters. • First, it suggests a naive interpretation of the social order into which the sugar entrepreneurs moved. • Second, it offers a reductionist view of who these "rich peasants" were, where they came from, and where they were going-as if those who are rich and powerful today must have come from rich and powerful families in the past. (This is known as reading history backwards.) • Third, it assumes that initial advantages in the commercial economy simply multiplied, at no risk to the "rich peasants."

178

The Pattern of Inequality and Mobility

• Fourth, it oversimplifies the rise of rural leaders to political power in the region. • Fifth, it overlooks the coalition (between large and small cane growers) which would make the cooperative sugar factories economically viable and politically powerful. The fourth point will be discussed in the next chapter, while the fifth must wait until Chapter 10. Meanwhile, the first three deserve attention.

What Was the Social Order Before Sugar? Chithelen evidently believes the villages were inhabited by a mass of small peasants, plus a few who were already rich in terms of their control over land and their ability to act as moneylenders (1985:604-06). It was small peasants, in this view, who rented or sold land to the entrepreneurs. The data from earlier chapters support this view in some respects but suggest it is oversimplified. Villages were already highly differentiated in the 18th century, and this differentiation was based not only on land and credit, but also on caste and kinship, family prestige, administrative and military power, literacy and religion. Malegaon serves as just one example of how 18th-century differentiation affected 20th-century inequality. Malegaon had a Jagirdar in residence, who attracted many specialists to his service. Among these were Brahmans, who were given large landholdings for their religious and administrative services. Descendants of this old elite continued to own large landholdings and exert influence in village affairs-including the cooperatives-at least until the 1930s. Their wealth was based partly on sugar enterprises but owed its origins to processes of differentiation which were rooted in the 18th century. Was Malegaon unusual in this respect? Not every village had a Jagirdar, but many had inamdars, deshmukhs, etc., with big landholdings in the precolonial period. Close to Malegaon, for example, is a village called Wadgaon, which was an inam village (exempt from paying revenue) and the seat of a prestigious family, connected by marriage to the Jagirdar. In Baramati, there was formerly a powerful Brahman family known as Naik Baramatikar. This family was active in banking and finance in the 17th century; and in the 18th century, Bapuji Naik temporarily joined the inner circle involved in power struggles at the Peshwa's court in Pune (Gordon n.d.:ch.7). Just across the Nira river, the Rajas of Phaltan ruled a sizeable princely state that included, by the end of the 19th century, 73 villages yielding an annual revenue of Rs.174,228 (Bombay Government 1910:281). Many historians have by now given up the notion that pre-colonial villages were economically and politically homogeneous. Consequently, the bare facts are not necessarily in dispute. However, their significance has been missed by Chithelen, who does not even mention the possibility that some of the old elites were shoved aside by new entrepreneurs. We know that some old elites managed to ride the wave of the new commercial economy, at least for a while. Until the 1930s, the Jagirdar and

The Pattern of Inequality and Mobility

179

his Brahman employees did so with considerable success. Across the river, the Raja of Phaltan encouraged the Phaltan Sugar Works and immigrant sugar magnates. However, the Jagirdar and his Brahmans were hurt badly both by the failure of the Nira Valley Sugar Company and the depression. Although old elites and new entrepreneurs collaborated in various ways, the former were gradually eclipsed. For the Jagirdar's family, the eclipse was completed in 1952, whenjagirs were abolished, thus terminating their revenue rights and distinct legal status. In the 19th century, the old landed elites had put up a stiff and largely successful resistance against British efforts to investigate and reduce their revenue rights. As a result, "the Bombay government, during the early 1860s, abandoned the ideal of bringing all alienated lands within its full scrutiny" (Charlesworth 1985:55; also Preston 1988). However, in the 20th century, the old elites melted away in the face of the commercial economy and the rise of new rural leaders. Had this not been the case, it would have been difficult for the post-independence government to abolish revenue rights which the British could not curtail a century earlier; but the bill abolishing jagirs and most inams was passed with virtually no organized opposition. The fact that some Rajas had earlier helped to usher in new educational institutions, sugar factories, and cooperatives (and were, in a few cases, active in post-independence politics), should not obscure their ultimate fate, which was to become politically and economically marginal. Thus, to return to Chithelen's argument, it is an oversimplification to suggest that commercial entrepreneurs expanded at the expense primarily of small peasants. This reflects a naive view of the old order and completely overlooks the fate of the old elites. It is clear that, as entrepreneurs expanded, they sought out bigger landlords. In the 1930s, magnates in the Baramati area moved across the river to Phaltan state, where they were able to rent large holdings from state officials and relatives of the Raja. Moreover, much of this land was being brought under cultivation for the first time: it had to be cleared of rocks, scrub forest, foxes and snakes. Thus much of this new land was not owned or cultivated by small peasants. Of course, the entrepreneurs, or at least those with luck and skill, were getting richer than small peasants: on this point there is no argument. But Chithelen goes much farther than this, stating that rich peasants must have expanded by taking land primarily from small ones: "migrant farmers must have leased in lands largely from poor cultivators for cultivating irrigated crops in the Ahmednagar canal tracts" (Chithelen 1985:606). It is hard to believe that entrepreneurs could assemble large, consolidated areas under sugarcane, as they often did, simply by renting or buying land from small peasants. The transaction costs would have been prohibitive. Logic suggests that many followed the procedures which we know were used near Baramati: they looked for big landlords. And they found them, because this was already a highly differentiated society. Much of the land owned by old elites was not being transformed into commercial production by its

180

The Pattern of Inequality and Mobility

owners, since the latter often preferred to let tenants make the effort and take the risks.

Who Were the Rich PetUants? According to Chithelen, successful cane growers were either Mali immigrants or local villagers who were already big landowners. In view of our data from the Baramati area, this seems highly reductionist. In Malegaon, one early entrepreneur who fitted the formula was Keshavrao Taware, a Maratha from the dominant lineage whose father had plenty of land. Rather similar in economic terms, but markedly different in terms of caste and culture, was Rangnath Chawre, who inherited a large landholding that his ancestors had acquired in the Jagirdar's service. Chawre, a Brahman administrator, was hardly a peasant of any kind-neither in his origins nor in his enterprising cultivation of sugarcane and his management of the gut shop in Baramati. However, in building a commercial venture on a large inherited base, Chawre at least corresponded to one element in Chithelen's formula. The same cannot be said of the three sugar magnates-Shembekar, Date, and Jachak-since they all started out poor. Moreover, Shembekar and Date were Brahmans, and neither they nor their fathers could be described as peasants. There is no denying that these were special cases, and in the postindependence period, Brahmans faded away as sugar entrepreneurs. Nor can it be denied that some of the old Maratha elites, in Malegaon and elsewhere, established themselves as successful cane growers. Nevertheless, the last two chapters reveal numerous cases of upward mobility, including some nonMaratha migrant laborers who were able to buy large landholdings. These cases contradict the assumption that rich peasants were homogeneous in their origin, composition, or destiny. The rich of one generation, including some Maratha patils, might become the poor of the next.

Did Lllml Distribution Become E11er More Polarized? Chithelen suggests that irrigated cash-cropping led to a fairly rigid polarization of classes: he uses the term "crystallised," for example. The idea is that those who achieve an initial advantage in the commercial economy will hold on or expand it. However, if we look at the patterns of migration and mobility outlined in the last two chapters, we find no evidence of progressive polarization, and much contrary evidence of continuous mobility. Instead of solidifying into rigid classes, the villagers in Malegaon followed paths of "multidirectional mobility" (Shanin 1972). Thanks to the irrigation frontier, it was possible for entrepreneurs to obtain cheap land without crowding out small farmers growing food crops. Later on, thanks to the depression, those who got a head start in the sugar economy did not necessarily stay ahead, while those who came later found numerous opportunities.18 As we saw in Chapter 2, recent historians, looking at the Bombay Deccan as a whole, have found no sharp increase in land concentration during the late 19th and early 20th centuries (e.g., Charlesworth 1985:293-94). Guha

181

The Pattern of Inequality and Mobility

concludes that small peasants clung to their land, that rich farmers were handicapped by their need to employ expensive wage labor, and that there was a decline in the total share of land concentrated in large holdings (1985:140-41, 153-58). In western India, some localities were able to take up cash cropping, while others were not. In the Deccan districts, this depended on the availability of irrigation. The result of the general shortage of water was that canal villages became prosperous while dry ones remained poor. The outcome, as we have seen, was seasonal and permanent migration to the canal villages. This did result in the creation of a landless proletariat, but not by the mechanism assumed in the standard theory: not, that is, by the wholesale impoverishment of small farmers in the commercialized villages. Immigrant laborers came to Malegaon because the subsistence economy of the dry villages could not sustain them as farmers-not in the face of growing population pressure and recurrent droughts. It would be a distortion to try to save the differentiation thesis by arguing that irrigated villages exploited dry ones, that this was the malign social mechanism which created the proletariat. To exploit people, one needs to deprive them of essential resources, threaten them, constrict their choices, or diminish their subsistence. The cane growers did none of these things to the dry farmers. As we have seen in earlier chapters, the canal villages sustained many dry farmers by giving them intermittent refuge and a source of employment. Thus the dry villages were relieved of a growing population which could not survive at home. Notes 1. A total of 109 valid cases made up the sample used in the previous chapter. However, the sample had to be reduced to 107 here, due to the nature of the statistical procedures used. In Chapter 6, we simply counted the number of families who moved between the landless and landowning categories. Here we are concerned with how much land each family gained or lost, and this puts different constraints on the selection of cases. Specifically, in the linear regression analysis, it was found that two cases were extreme outliers: that is, they exerted undue influence on the slope of the regression line (as in Figure 7.1). Checking the data revealed that the two cases were not invalid, just unusual. Both were families with large landholdings that underwent big changes between 1930 and 1950. (See cases [108] and [109) in Appendix Table A.l.) As a result, they would appear in the extreme right-hand portion of Figure 7.1 but rather far from the regression line. According to the norms adopted by statistical experts, they should be removed from a sample used to calculate regressions. For the sake of consistency, they have also been removed from the other statistical procedures used in this chapter. As an anthropologist, I hate to give up two cases (especially from a sample which is not very large to begin with} when the data have been cross-checked and appear to be as valid as for the more typical cases. However, I accede to the norms of the statisticians for two reasons. First, the two cases have been included in the qualitative analysis. Second, if these cases were kept in the sample, they would strengthen my

182

The Pattern of Inequality and Mobility

overall argument. Faced with a choice, the argument must find what support it can in the more typical cases. It may be recalled from note 12 in the previous chapter that the Jagirdar's immediate family was also removed from the sample: the reasons are exactly the same. 2. Ratios between different types of land were established from data on land sale prices and from the norms used in land reform legislation (Maharashtra 1974). (See also Table 4.1 for earlier data which is pertinent.) Because the landholding data have been standardized by division, the resulting numbers look more precise than they actually are. For the earlier decades (1920-40), I estimate that the data come to within ten percent of the real values. (See the Appendix for details.) 3. There are two exceptions to the general observation that the correlations decline over time: these are for 1980, but only in columns one and two (baseline 1920 and 1930). As we shall see later, these hardly imply a reversal of the dominant trend. Chapter 9, with a larger sample (N = 121), shows results conforming to the trend. 4. The data could have been handled in other ways. However, with the procedures used overall, it is much simpler to treat each family as a single case, not as severalwhich would have been the result if collateral branches were also included in the analysis. In any case, data on long-gone collaterals were usually not as detailed as those on the lineal family. 5. The weights were explained in Chapter 6, and more details are given in the Appendix. Wei.qhts apply to the same families throughout the period: they attach only to those who were landless in 1970, not to those who happened to be landless at other times but owned land in 1970. 6. It would be possible for all big farmers to exchange positions with small ones without showing any change in the aggregates, when standard procedures are used. Such procedures are generally followed because the two sets of cross-sectional data are based on independent samples (e.g., Sarma 1982:34); but even where this is not the case, the standard and less revealing procedure is often followed (e.g., Dandekar 1986:299). 7. We must be careful not to over-interpret correlations and other measures of association based on the weighted sample. When the sample size is increased, the computer tends to regard more correlations as significant (and at higher levels)because significance tests are strongly influenced by sample size. However, the weighted N is expanded merely by repeating information on certain cases. Thus it is prudent to compare patterns of correlations (weighted vs. unweighted) but not to rely heavily on significance tests from the weighted sample. 8. The data in Figure 7.1 do not fit the "normal" distribution: that is, the land values for 1930, 1950, or any other decade do not fit the standard bell-shaped curve in terms of frequency distribution. As a result, the data do not strictly conform to the assumptions required for linear regression analysis. (Efforts were made to normalize the data by converting them to logarithms, etc., but these did not resolve the problem.) More generally, it can be said that all statistical procedures used here were created for other purposes and are not ideally suited for the present analysis. However, they are, at least, based on straightforward and widely used assumptions. (Even nonstatisticians commonly make arguments in terms of linear relationships.) Examining the data from a number of angles with a variety of methods is, I believe, the best way to avoid any problems which might arise from inherent limitations in the procedures. 9. Lorenz curves showing the distribution of land or wealth commonly display inequalities which are less extreme than those in Figure 7.3. The reason is that I have graphed the entire sample, including landless families, whereas the usual procedure

The Pattern of Inequality and Mobility

183

is to exclude such families. That would make little sense here, since we have been measuring, among other things, the amount of mobility across the landless boundary. 10. The lessening of inequality would appear more significant if the figure excluded all families which were landless in 1930 and 1950. (See note 9 above.) These landless families control the shape of both curves for nearly half their length, pulling them close together regardless of changes in their middle and upper portions. 11. Those who know the caste system more in terms of scriptural theory than in terms of daily life in Maharashtra may be surprised at this lumping together of Brahman and merchant castes. According to ancient scriptures, Hindu society is divided into four ranked categories, or varnas: Brahman, Kshatriya, Vaishya, and Shudra. Vaishyas, who are generally considered to be merchant castes, are thus ranked well below Brahmans. This model has salience in some contexts, though it misrepresents the everyday order of village life (see Srinivas 1966). For example, ordinary village Marathas, according to orthodox theory, are Shudras; but in their own eyes, the Marathas are Kshatriyas, and no one in the village would openly suggest otherwise. Another problem with the varna model is that it has no explicit place for the low castes, who rank below Shudras. In the categories used here, I have taken economic factors into account, lumping Brahmans and merchant castes together. This also takes account of the way the caste system is perceived by the villagers. 12. I have lumped the Muslims in this category, though strictly speaking they are a religious group, not a caste. However, in terms of everyday status, wealth and power, they belong here. 13. Two other families were also deleted from the sample, for reasons explained in note 1 above. One was a wealthy Brahman family [109], who lost most of their land in the depression. The other was a wealthy Taware family [108], who increased their landholdings through purchases and added irrigation. If these two were also included in the sample, the latter case would partially offset the losses incurred by the Jagirdar's family. However, including all three cases would, on balance, strengthen my argument that the top two categories (high castes and major Marathas) were the big losers in this period. 14. A simple linear correlation would not be suitable, since we are testing the relationship between variables of different types. Caste status is an ordinal variable without any precise measure of how much higher one category is than anotherunlike economic variables, which lend themselves to more precise statistics. A nonparametric correlation of land ownership with the status of individual castes could be used. However, the results might be: (a) obscured by the overlapping status of the smaller castes; and (b) distorted by the small number of families in these castes, creating spurious associations (or non-associations) between status and wealth. 15. Not much significance should be read into the higher F ratios for the weighted sample, because the F ratio is strongly affected by sample size. Thus the difference between weighted and unweighted ratios is mostly an artifact of the weighting procedure. 16. The strength of this prediction can be inferred by squaring R and Eta. For 1920, R squared is 0.26 and Eta squared is 0.43, suggesting that caste explains up to 43 percent of the variance in landholdings. By 1950, Eta squared declines to 0.13, suggesting that caste then explains only 13 percent of the variance in landholdings. 17. As we shall see in Chapter 9, the conservatism of the small farmer diminished after 1950, thanks to the security provided by the cooperative sugar factory. 18. Although at variance with the standard theory, these results are not unique. According to one specialist on Latin America, "Peasants have not regularly and rapidly differentiated into capitalist and proletarian classes when they came into contact with capitalism" (Cancian 1989:159).

PART FOUR

Cooperative Sugar, 1950-1985

8 The Politics of Sugar Independence brought new classes to power (the urban and rural middle classes) and led to a new policy framework for development. Prime Minister Nehru, who was ideologically committed to a peaceful path toward socialism, used his personal authority to persuade the Congress party and the government to adopt a program consisting of: (a) land reforms and state-sponsored development projects in the countryside; (b) tight controls over private industry; (c) a massive build-up of public-sector enterprises; and (d) an overall framework of five-year plans (Nayar 1989). One general result of Nehru's program was a massive expansion of government bureaucracy at every level and the intrusion of bureaucracy into almost every facet of economic life. This bureaucratic expansion proved to be a vital source of employment, power and privilege for the urban middle class. Thus it was not long before this class began to project its own values and interests as virtually identical with Nehru's version of socialist development. Meanwhile, something of a counter-trend was taking root. in the countryside: this was the rise to political power of rural leaders, at the expense of urban, middle-class Congressmen. The 1952 general election was the first ever to be based on a universal adult franchise. As a result, the voting power swung massively toward the countryside, and rural politicians began to be elected to office at every level. This trend was reflected in the occupations of Members of Parliament: between 1947 and 1962, MPs from the urban professions declined in number, while those based in the rural areas multiplied (Rosenthal 1970:182). The result of this counter-trend was competition and tension between the intermediate classes: that is, between the urban middle class and the "rich and middle peasantry." This tension was temporarily resolved, in the 1960s and 1970s, by a tacit division of power: rural leaders came to dominate some of the state legislatures, while the urban middle class held onto power in the central cabinet. Although the urban middle class was being forced to accept a more limited role in politics, its power in the bureaucracy was expanding exponentially. The result was a system so favorable to both classes that each tolerated the other's sphere of influence (Nayar 1989:215). 187

188

The Politics of Sugar

Sugar Interests at the National Level This confluence of political forces led to major changes in the sugar industry. As we saw in Chapter 4, the industry achieved its first rapid growth in the north, after the tariff was imposed in 1932. During the 1930s and 1940s, the industry was dominated by private factories located in U.P. and Bihar. A significant number of these factories were controlled by a few large firms, which succeeded in forming a cartel, the Indian Sugar Syndicate [ISS], in 1937. The ISS was promptly given official recognition by the provincial governments of U.P. and Bihar (Baru 1990:53,58). With help from official committees, the Sugar Syndicate strove to reduce over-production and limit competition. It was intermittently successful. Wartime price controls limited its freedom of action, but in the first two years after independence, it managed to push up sugar prices (Baru 1990:131). A variety of forces were ranged against the ISS before independence, and they all increased in strength after 1947. First of all, there were the cane growers, who were becoming more vocal and effective in their demand for fair cane prices from the private mills. The Sugar Factories Control Act of 1938, which gave official recognition to the Sugar Syndicate, also created cooperative cane supply unions in U.P. (Baru 1990:58-59). These unions, which were managed by the provincial government, were intended to introduce a more balanced relationship between the growers and the factories. Immediately after independence, representatives of the northern cane growers began to demand a restructuring of official committees and policies dealing with the industry. They were able to press their demands effectively because they now had the power of the rural vote behind them. As a result, they played an active role in the shaping of sugar policies. Northern growers and factory owners clashed over two issues concerning cane supply: cane prices and gur production. In years when gur prices rose relative to cane prices, the farmers used much of their cane to manufacture gur. This, of course, disrupted cane supplies to the factories. Thus in 1947, the factory owners tendered a resolution to an official committee, calling on the government to prohibit gur production inside the cane-supply areas of the factories. This resolution provoked vigorous opposition from the growers' representative (Baru 1990:77-80), and it never became government policy. Other interests also had to be taken into account. Urban consumers demanded that food prices be held in check. And finally, there were the interests of the private factory owners in Bombay and southern India, which were often at odds with the policies of the Sugar Syndicate. Sugarcane grows better in the tropics than in sub-tropical regions like northern India. Though the industry got off to a slow start in Bombay and southern India, it was clearly ready to expand by the late 1930s or early 1940s. At least in its home region, the Bombay industry could easily undersell sugar which was produced in the north. Thus the Bombay factory owners were opposed to any efforts by the Sugar Syndicate to limit production and competition. Such restraint would prevent the Bombay factories from ex-

The Politics of Sugar

189

panding and taking a larger share of the market (Baru 1990:62). This conflict of interests between north and south was played out on a larger scale after independence. After 1947, the central government gradually set up a complex array of regulatory systems, which were designed to balance these competing interests and promote the expansion of the industry. The first major step was a decision in 1954 to restrict all new licenses to cooperative factories. This policy was clearly in the interests of the growers, particularly those in the Bombay Deccan, who had already started a successful cooperative. The new licensing policy was opposed by private factory owners-especially those in Bombay, who saw great opportunities being snatched away-but to no avail (Baru 1990:85-86). With the help of the new licensing policy, the cooperative sugar industry expanded rapidly in the western and southern regions. The second set of regulatory mechanisms concerned cane prices; and here again the government's goal was to protect the interests of the growers, particularly those in the north. Using various formulas, which changed over the years, the government set minimum sugarcane prices for the factories to pay the growers (Baru 1990:146-51). As a result, many of the northern factories found themselves in a no-win situation. Cane and sugar prices were regulated but gur prices were not. Gur was produced by many thousands of small-scale units, so the government did not attempt to regulate its price. There were times when the gur-making units could offer higher prices for sugarcane than the factories, if gur prices rose more rapidly than sugar prices. Thus the private factories found it increasingly difficult to purchase a reliable supply of cane at a price they could afford. The factories were boxed in by another set of regulatory mechanisms: those concerned with the price of sugar. Here the interests of consumers entered in. The government learned from experience that a rapid rise in the price of sugar or other basic foodstuffs generated protests and unfavorable returns at the ballot box. Consequently, it sought to balance the interests of cane growers and factories, on the one hand, with those of consumers on the other. In particular, it sought to hold down any rise in sugar prices. Sugar price controls were also used to balance another conflict of interests: that between the high-cost factories in the north and the lower-cost factories in the west and south. The expansion of the latter threatened to drive many of the former out of business, and this would hurt not only the factory owners but also the many growers who supplied them cane. The central government's way of dealing with this problem involved, most of the time, a system of partial control over sugar prices. (There have been periods of total control and total decontrol, but partial control has been the preferred mechanism over the long run.) What this meant was that a certain proportion of the sugar produced by any factory (usually between 35 and 55 percent) could be sold on the open market, subject to certain regulations. The remainder had to be sold to the government as "levy sugar" at prices which the government fixed in advance.

The Politics of Sugar

190

The most salient features of the levy-sugar prices were (a) that they varied from region to region, and (b) that they were calculated on a cost-plus basis. In other words, the government calculated the average cost of production for a given region, added a certain amount of acceptable profit, and fixed the resulting amount as the levy-sugar price for that region. Thus factories in the north, which is a high-cost region, were paid more for their levy sugar than factories in the west and south. The objective of this system was to protect the high-cost northern factories against competition from other regions (Baru 1990:140). In other words, the northern factories (and their cane suppliers) were granted the equivalent of extra tariff protection. Overall, then, we have an industry in which there is some room for growth and competition but also a high degree of regulation and protection. Sugar policies and sugar prices reflect to some extent the impact of market forces, but they also reflect a complex political process, in which a variety of competing interests are held more or less in balance. No one set of political interests or market forces can be used to explain sugar policies or the behavior of the industry overall. Sugar Interests at the State Level

After independence, Marathas and other rural leaders came into their own in state politics, since they represented the rural majority. By 1957, Marathas already constituted a majority on the Pradesh (State) Congress Committee (Rosenthal 1977:32). By the mid-1960s, they had risen to undisputed command of state and local politics; and in the irrigated areas of western Maharashtra, they were using sugar cooperatives as a major power base. Consequently, it is tempting to regard the successful establishment of the cooperatives, and the support given them by the government, as a result of Maratha domination. However, there is a danger of reading history backwards, of assuming that the early cooperative factories succeeded because of political power at the state level. It would be more accurate to view these factories as experiments in which the commitment and performance of local leaders were all-important. The state supported these experiments, but not because it was under the control of big cane growers or cooperative leaders. Before independence, Maratha leaders had already acquired political experience and an articulate sense of rural interests through participation in the cooperative and non-Brahman movements. After 1930, militant leaders like Keshavrao Jedhe joined the Congress party. However, this did not mean that Marathas moved smoothly into control of the Congress, even though they had won elected seats on district local boards and the provincial legislature. In fact, "Jedhe was for a long time the only non-Brahman member of the Pradesh Congress Committee. The 1937 elections resulted in a Congress government dominated by Gujaratis and Maharashtrian Brahmans" (Omvedt 1976:281). Shortly after independence, Jedhe and other non-Brahman leaders broke away from the Congress to form the Peasants and Workers Party (PWP),

191

The Politics of Sugar

• Nasik

('-

f\S

• Ahmednagar (

Pune.

• Boramati

•Pholtan Sholapur •

( ~"'

f

/

s

J

.J

'\_..... ..... ---....

J

t..

l

c-

'-

,..r.l

~) '-/

FIGURE 8.1

0

I

~

\ ..... r . ..S

.... .I

N

f\-J

\"'" "'-,.....

Kolhapur. ,.,. /

....

100

200

Km.

l

}

Maharashtra State in 1960

since they felt that Congress was still dominated by urban, high-caste interests. The PWP was an important opposition party in state elections during the 1950s, culminating with a strong showing in 1957, due to its participation in the movement demanding a united Maharashtra. In 1960, Bombay state was, indeed, divided and reorganized to form a united Maharashtra, with boundaries based on the distribution of Marathi-language speakers. 1 (See Figure 8.1.) Y.B. Chavan, a Maratha, was the first Chief Minister of the new state, and he brought many Maratha leaders from the PWP back into the Congress fold (Kamat 1983:69-76). The formation of Maharashtra as a linguistic state was the culmination of a long upswell of Marathi nationalism. (The term "Marathi" refers not only to the language but also, in a loose sense, to native speakers of the language.) In the early 20th century, Tilak and other urban Congress leaders had incited Marathi nationalism as a basis of support for Indian nationalism (Cashman 1975). By the 1950s, the new rural leaders were using Marathi nationalism as a means of opposing urban interests, particularly the nonMaharashtrian businessmen who were influential in Bombay city. During the 1950s, Marathas were pressing claims to political power in various arenas, including the PWP, the Congress party, the state legislature, and the movement for a united Maharashtra. Meanwhile, other rural leaders were establishing the first cooperative sugar factories. However, it would be

192

The Politics of Sugar

incorrect to assume that the "sugar barons" were already powerful in regional politics when the state government decided to support the nascent factories. For one thing, power at the state level was shared with Gujaratis until1960. Even in the Marathi-speaking districts, urban high-caste Freedom Fighters were still influential, particularly at setting the ideological agenda for the government. For example, in the late 1950s and early 1960s, the state government enacted land reform legislation, giving more rights to tenants and placing a ceiling on the amount of agricultural land which could be owned by an individual. These were high-priority legislative acts, crucial steps in Prime Minister Nehru's strategy of agricultural development. These reforms were definitely not in the interest of the big cane growers, who tried to organize a lobby to defeat them. The Congress government was able to enact the reforms anyway, not because there was an organized demand for them from the countryside, but because Nehru backed them with all his personal authority (Nayar 1989). The fact that the big cane growers could not block passage of this legislation tells us much about their lack of power in the state government at the time, as well as their perception of the government's unfriendliness. The government prevailed, at least in passing the reforms, because the big cane growers did not have a mass political base in the countryside. 2 Middle and small farmers were crucial for establishing such a base; and having no interest in protecting big farmers on this issue, they remained quiet. Besides, the big cane growers had no experience, and little interest, in politics. Prior to the emergence of Maharashtra as a state, two decisions by the Bombay government were crucial to the establishment of the cooperative factories. The first decision, taken around 1949, committed the government to buy shares in the Pravara Cooperative Sugar Factory and to guarantee its debts (Gadgil1955:203-05). The second, in 1954, was a recommendation to the central government that licences for new sugar factories be given preferentially to cooperative enterprises. At the same time, the state government drew up a list of twelve sites where the supply of sugarcane would be sufficient for new factories and where cooperatives would be encouraged (Baviskar 1980:35). One man who occupied a pivotal role in making these decisions was the finance minister, Vaikunth Mehta. Mehta belonged to a wealthy Gujarati business family in Bombay and had managed the Bombay Central Cooperative Bank for decades. It was his devotion to the cooperative movement that led him to commit the government to buying shares in the Pravara factory. Mehta was only one of several non-Maratha leaders who played a significant role in bringing the cooperative factories to life. D.R. Gadgil, the eminent Brahman economist, was another. V.P. Varde, the honorary managing director of the Central Cooperative Bank (another Brahman), "was largely responsible for the active involvement of the bank in encouraging sugar cooperatives" (Baviskar 1980:117-18). Their ability to mobilize official support for the nascent sugar cooperatives was based, not on Maratha ambitions for power,

The Politics of Sugar

193

but on a vision of rural development through local initiative-a vision similar to that of Gandhi and Nehru. In the 1950s, a sugar lobby was only starting to emerge, and it did not yet have much strength in the Congress party. For example, in 1957, when the Malegaon Cooperative Sugar Factory began its first season of operation, the Baramati constituency elected a Maratha sugar magnate from the Peasants and Workers Party to the state legislature. A decade later, this constituency became a stronghold of the Congress party, but in 1957 it was not at all obvious that this would be the case. The election of a PWP candidate demonstrateq that many leaders in the area, including a number of big cane growers, did not feel that their interests were adequately represented by the Congress. At that time, big cane growers were politically weak. Most had avoided active involvement in the nationalist movement, so they had no credentials as Freedom Fighters. They had little experience in government and party politics, and none at all in the politics of agitation. Moreover, a number of the biggest bagaitdars were from minority castes, like the Saswad Malis in Ahmednagar district or the Brahmans near Baramati. They were in no sense established leaders in party and state politics. Even the Maratha leaders who organized some of the first cooperative factories were not, at that time, personally influential in the Congress party. In fact, Vitthalrao Vikhe and G.R. Autade, who organized the first two factories, were more sympathetic to the PWP (Baviskar 1980). Other Marathas were rising to power in the party and the government, and it took time for a sugar lobby to make its presence felt among many competing interests. In the 1950s, no one expected the sugar factories to become important power centers in the state. The factory founders were committed to social and economic, rather than political, goals. As we shall see in the next few chapters, sugar cooperatives eventually became enmeshed in state politics in all sorts of self-seeking ways; but they did not owe their origin merely to a crass scramble for political power, nor were they the offspring of Maratha domination. In fact, as we shall see later, it was beneficilll for the early factories that their founders were not particularly influential in state politics. This helped them avoid the trap of excessive state support and state control, which would have been fatal in terms of economic efficiency and innovation. The First Cooperative Factories

As we saw in Chapter 4, the perennial problem of the Indian sugar industry was cane supply. In 1934, the Saswad Malis worked out a solution for themselves: a farmer-owned factory, which served as a model for other cane growers after independence. 3 During World War Two, commodity prices soared, private sugar factories boomed, and village cane growers prospered. However, as the latter expanded production, they yearned for access to the growing white sugar market. In

194

The Politics of Sugar

Ahmednagar district, just north of Pune district, cane growers (mostly of Maratha caste) began organizing in opposition to other interests (Baviskar 1980:24-34). First, they wanted more access to canal irrigation, a large share of which was already taken by the Saswad Malis and private sugar factories. Second, they wanted greater control over marketing than they could get as individual clients of gul brokers. Third, they were fed up with the private factories, which preferred not to depend on independent growers in the first place and were not considered prompt or fair in paying for their cane in the second. Cane growers in Ahmednagar thus tried as early as 1945 to organize a cooperative factory. Gul brokers and private factory owners lobbied the state government to discourage this effort, which would compete with their businesses. Saswad Malis, the biggest cane growers, initially declined to participate in the new venture, fearing the political power of the Marathas in a cooperative organization (Baviskar 1980). This meant the organizers had to struggle along, at first, without the capital and expertise which the Malis could have contributed. Despite these obstacles, Vitthalrao Vikhe, a Maratha farmer with experience in the cooperative movement, began working in consultation with the Bombay State Cooperative Bank to raise share capital for the Pravara factory. Under his leadership, the farmers contributed Rs.600,000 in share capital. The state government also agreed to invest Rs.600,000 in shares. To finance equipment and construction, an additional Rs.2 million was borrowed from the Industrial Finance Corporation, plus short-term loans from the State Bank, making a total investment of Rs.4.1 million (Gadgil 1955:205-06). The Pravara Cooperative Sugar Factory went into production in 1950-51 and was immediately successful. Much of this success was due to the dedication of the leaders and the loyalty of the members: The greatest test came in the last quarter of 1950 whengul prices soared very high, the erection of the factory was not yet complete, and nobody could say definitely how soon it would begin to manufacture sugar. At this time the vast bulk of the members refrained from turning their sugarcane intogul and showed the most remarkable restraint and patience, even though it meant definite monetary loss to many, which amounted in the case of some to substantial sums. (Gadgil 1955:210)

The Pravara factory was soon followed by another in nearby Kopargaon. This one was organized by G.R. Autade: In 1932 he played a prominent part in starting the Cooperative Sale and Purchase Union in Kopargaon ... [and) he took the lead in organizing the Taluka Cooperative Supervising Union. . . . His active participation in cooperation brought him into close contact with such leading co-operators as V.L. Mehta and D.R. Gadgil, as well as with the officials of the State Cooperative Bank and the Department of Co-operation of the State government. (Baviskar 1980:27-28)

The Politics of Sugar

195

Autade thus exemplified the rising village leaders who were experienced with growing cane and marketing gul and also with the problems and potentials of the cooperative movement. These rural leaders obtained vital assistance from highly educated, urban leaders. D.R. Gadgil was a Brahman economist who had taken a keen interest in rural development for many years, as the first director of the Gokhale Institute of Politics and Economics. Toward the end of a long and distinguished career, he served as Deputy Chairman of the Planning Commission in Delhi. Among many other achievements, he conducted a study (1948) on the impact of canal irrigation in Ahmednagar district. Gadgil thus had intimate knowledge of the economic condition of the villages and good contacts with rural leaders-based partly on his active participation in the leadership of the Rayat Shikshan Saunstha (the Farmers Education Institute, founded by Bhaurao Patil), plus his close connections with leaders of the Bombay State Cooperative Bank. He was called on to guide the Pravara factory at its inception and served as its first chairman for nearly ten years (Baviskar 1980:31). In fact, meetings of the board of directors used to be held at the Gokhale Institute, in Pune, to accommodate his busy schedule. Vaikunth Mehta was the son of Lallubhai Samaldas, who founded the Bombay Central (later State) Cooperative Bank. Mehta worked as the Bank's manager for decades, travelling throughout the countryside to supervise its operations and promote the ideals of the cooperative movement (Catanach 1970; Shaikh 1978). Like Gadgil, then, he had wide practical knowledge of conditions in the villages and longstanding contacts with their leaders. At the time of independence, he was asked to serve as Minister for Finance and Co-operation in the state government. (Mehta had never been active in politics, though he had been sympathetic toward the nationalist movement.) As such, he took an interest in the proposed sugar factories and ensured that the government would support them. The founders of the first cooperative factories were thus backed up by some of the best intellectual and financial minds in the state, men with a commanding overview of the history and problems of the cooperative movement. Autade, Vikhe, and other rural leaders had experience in dealing with these same problems, and they enjoyed the confidence of the local farmers. One reason the cooperative sugar factories got off to a strong start was due to this extraordinary combination of experienced rural leaders allied with brilliant and committed men from the intellectual and business communities. In addition, Autade and Vikhe managed to win over the big Mali cane growers to their cause. Since cooperative shareholders each have one vote, regardless of the number of shares they own, the Malis were afraid their interests would be swamped by the Maratha majority. Autade protected their interests within the Kopargaon factory by including two Malis on the first board of directors, which was nominated by the state government (Baviskar 1980:36). Thus the capital and entrepreneurial abilities of the Malis were brought on board.

196

The Politics of Sugar

When the Pravara factory had proven successful at raising capital, procuring cane, and producing sugar, the central government decided that licenses for future sugar factories would go preferentially to cooperatives. As a result, dozens of new factories sprang up along the Deccan Canals and wherever lift irrigation systems could be built along the rivers. Although Pravara depended on Gadgil as its first chairman, local organizers soon learned how to manage their own factories; and the state government wisely decided not to assume direct control. There were at least two reasons for this policy of restraint: the principles of the cooperative movement, and the interests of the factory organizers. For decades, influential cooperative organizers, like Vaikunth Mehta, had been inspired by the ideals of democratic autonomy and had sought to limit government control of the cooperative movement-particularly when the government in question was British. Thus experienced officials in the State Cooperative Bank and the Cooperative Department were not driving toward ever-increasing bureaucratic control over the movement-a miracle of restraint in the Indian context. On the other side, the factory organizers were interested in building institutions which they could control, not in strengthening the state bureaucracy, which was heavily dominated by urban Brahmans. In the 1950s, as we have seen, they had reason to regard the bureaucracy and the legislature as not particularly friendly to their interests. The result was unusual for India, where government bureaucracy is usually considered the proper instrument to establish new institutions, and where a higher education, preferably at an English-medium college, is normally considered essential for managing large-scale enterprises. Some of the most innovative and enterprising factory organizers, including Vikhe, Autade, and many others, lacked even a high school education. They were real "sons of the soil," at home among their constituents, and they kept the cooperative factories under local control. As we shall see in later chapters, this degree of autonomy is rare for cooperatives in other parts of India.

Factory Organizers Near Baramati The factory founders who arose in the Baramati area did not quite conform to the pattern in Ahmednagar district. The Malegaon Cooperative Sugar Factory was organized by Dadasaheb Shembekar, while its sister factory near Baramati (Shri Chhatrapati) was organized by nephews of Sopanrao Jachak. The Jachaks were Marathas with prior experience in credit and marketing cooperatives, similar to Autade and Vikhe in Ahmednagar. Shembekar, however, was a Brahman with quite different experience as a public leader. After independence, the Baramati magnates were drawn into the burlyburly of state politics in order to protect and advance their interests. For example, they joined with associations of big ba.ga.itdars in Ahmednagar district, trying unsuccessfully to lobby against land reforms proposed by the new Congress government. 4 Of more local interest, the magnates also tried to block the reorganization of the cooperative banking system in Baramati. In 1911, when the Bombay

The Politics of Sugar

197

Central Cooperative Bank was organized, it opened a branch in Baramati to facilitate supervision of the gul shop and the village credit societies. Some years later, a District Central Cooperative Bank (DCCB) was established in Pune to finance and supervise credit societies outside the Nira Canal tract. The Nira Canal societies continued to use the Baramati branch, linked directly to the Bombay Bank, to meet the heavy financial needs of the cane growers. After independence, however, the state government proposed to bring the Nira Canal societies under the jurisdiction of the Pune DCCB and thus abolish the direct link to the Bombay Bank. The magnates opposed this move because it would reduce their access to credit. During the depression, the Bombay Bank adopted a policy of lending to individuals as well as credit societies. The magnates thus had direct access, through the Baramati branch, to the assets of the Bombay Bank. The District Bank, which was financing mainly dry-village societies, had assets and lending policies that were comparatively paltry, so the magnates did not want the Baramati branch taken over by the DCCB. They attempted to organize a "central" bank in Baramati-one which would be on a par with the District Bank. Shembekar, Jachak and others (including Govindrao Pawar, manager of the Sale-Purchase Union) raised Rs.100,000 in share capital. However, the Cooperative Department ruled against a "central" bank at this level. The magnates could organize an urban cooperative bank, if they wished, but this would not be for financing agriculture. They were finally forced to give up and return the share capital to the contributors. The lesson from this episode was similar to that from the land reforms: political power and influence over government policy would depend on a broader mobilization of support. This lesson made cooperative factories a more attractive option for the magnates, despite the effort and risk they entailed. As mentioned, the central government announced in 1954 that future sugar factory licenses would be reserved for cooperatives. Without this policy, the magnates might have started private factories of their own. (In fact, one magnate from the Baramati area had earlier joined a partnership to build a private factory in Ahmednagar district.) Shembekar was quite dubious about the feasibility of cooperative factories and only changed his mind after visiting the Pravara factory in Ahmednagar. What weighed on his mind was the memory of the Nira Valley Sugar Company, which had failed due to lack of loyalty among the cane suppliers. The new licensing policy convinced the magnates to give cooperatives a try, since private factories were no longer an option. An additional incentive was a new limit on the area of sugarcane which a farmer could irrigate by canal: the limit would be six acres for nonmembers and 25 acres for members of cooperative sugar factories (Baviskar 1980:36). In 1954, then, the magnates applied themselves to organizing the first two cooperative factories in Pune district. The Jachaks were the leading organizers of the Shri Chhatrapati factory, east of Baramati, while Sliembekar organized the Malegaon factory to the west. The Jachaks were assisted in

198

The Politics of Sugar

their efforts by Govindrao Pawar, former managing director of the SalePurchase Union and father of Sharad Pawar, who was later to become Baramati's most notable leader. Shembekar, for his part, was assisted by Freedom Fighters from the independence movement, who lent symbolic legitimacy to the new enterprise. Aside from business expertise, the assets which the magnates applied to these projects included their experience at using administrative contacts to expedite licenses and credit, plus their reputations as solid businessmen-as leaders who would look after business rather than pursue power and glory for their own sake. Having learned a lesson from the failure of the old Nira Valley Sugar Company, Shembekar was careful to build up the confidence of the village farmers who were to be the shareholders and cane suppliers. He held public meetings in all the villages surrounding Malegaon, taking with him village leaders, successful cane growers, Freedom Fighters, and the Pune Collector, who happened to be a close associate of Vaikunth Mehta and was deeply committed to cooperative goals (Pawar 1978:53). Shembekar found leaders in every village to help the cause. (The proposed factory area included 13 villages within a radius of about ten miles from Malegaon.) In Malegaon, two of the foremost leaders were Madhavrao and Bhalchandra Taware, both big bagaitdars. They had already helped establish new institutions, including a village high school. Along with leaders from other castes, they had organized the Shikshan Seva Mandal (Education Service Society} for this purpose. The same village leaders, with a few exceptions, were also involved in establishing a new cooperative credit society in 1951, to replace the one liquidated in 1932. These were the men who helped Shembekar raise share capital for the factory. Shembekar raised Rs.l.88 million in share capital, the great majority of shares belonging to thousands of farmers holding two shares or less. The state government also bought shares worth Rs.1 million and thus helped to guarantee repayment of capital loans, worth Rs.7.38 million, which were borrowed from the Industrial Finance Corporation and the District Central Cooperative Bank. Dadasaheb Shembekar died in 1955, before the factory came into operation. He is commemorated by a large bust in the factory guest house, near one commemorating Raja Shambhusingh. The chairmanship was then taken by Dadasaheb's son, who did not continue in the factory leadership after 1960, since he was occupied with an industrial enterprise in Pune city. Because Dadasaheb could not continue as chairman, the Malegaon factory had a somewhat unusual political history. In many factories, leadership tends to be controlled by the founder or his family. (Shri Chhatrapati is an example, as is the Someshwar factory not far from Baramati.) In other cases, factory politics is governed by competition between powerful and ambitious leaders who involve themselves in district and state politics as well: Kopargaon factory is an example (Baviskar 1980). However, after Shembekar, the Malegaon chairmanship has shifted among a number of leaders, none of whom has become powerful beyond the local area. This was due not only to the death

199

The Politics of Sugar

of the founder and to his caste (which might have precluded a lengthy tenure in any case), but also to the rise of a powerful Congress leader, Sharad Pawar, who eventually controlled the factory through his supporters. Cooperative Organization Before exammmg the further evolution of leadership in the Malegaon factory, it is important to understand how the factory is organized. What makes a cooperative distinct from a private factory? First, the co-op is owned by village cane growers, not urban shareholders. Each share, worth Rs.SOO, commits a farmer to supply the cane from one-half acre annually and commits the factory to buy that cane. (In some factories, each share costs Rs.l,OOO and is equivalent to one acre of cane.) One member can own a maximum of 50 shares, equivalent to 25 acres of cane. 5 A co-op is managed by a board of directors elected from among the shareholders, each member having only one vote, regardless of how many shares he owns. The elected directors make all policy decisions concerning hiring and firing, purchasing equipment, selling sugar, diversifying operations, and so forth. These decisions are implemented by managerial and technical staff under the board's control. Elected members of the board in turn elect their chairman and vicechairman. Ex-officio members of the board, representing the District Central Cooperative Bank and the state government-the factory's creditors and public shareholders-are not allowed to hold the chairmanship nor even vote for the chairman. A convention has been established, in fact, that these outside representatives do not interfere in board decisions and rarely attend its meetings (Baviskar 1980:54). The board answers for its decisions to the general body of shareholders, which must meet at least once a year. The general body meetings which I witnessed at the Malegaon factory were well-attended and included vigorous comments and questions from the floor-in sharp contrast to the apathy and non-participation which characterize cooperative meetings elsewhere in India and the Third World. The board appoints, among its own members, an executive committee, purchase committee, sugar sale committee, lift irrigation committee, and central harvesting committee, among others. Of these, the executive committee is the most important. The board and its committees meet regularly, usually once a month, sometimes more often. The administrator who oversees daily operations is the Managing Director (MD}, a salaried employee and ex-officio member of the board. After the MD come the chief engineer, chief chemist, agricultural officer, chief accountant, and secretary, followed by hundreds of clerks, skilled workers, and laborers. The MD is usually a technocrat from within the sugar industry: a former chief chemist, chief engineer, agricultural officer, secretary, chief accountant, office superintendent, or labor welfare officer, but occasionally a government

200

The Politics of Sugar

official on deputation. The MD is selected by a committee consisting of the factory chairman, the chairman of the Maharashtra State Federation of Cooperative Sugar Factories, and the Director of Sugar, an official in the state government (Baviskar 1980:56). Even in cases where he comes on deputation from government service, the MD answers only to the board, which answers to the cane growers. The primary economic goal of the directors is to pay the highest possible cane price to the growers. This is, of course, opposite to the goal of private factories buying cane from independent growers. Cooperative chairmen and directors also have strong political goals. They use their local influence to compete in state politics, and many have become Members of the Legislative Assembly (MLAs) or Members of Parliament (MPs). Consequently, there is keen competition for election to the factory boards. In these elections, the directors are judged, among other things, by how well the members have been paid for their sugarcane. Cooperative sugar factories have been supported by government in several ways (Inamdar 1965; Mohite 1974). The central government grants licenses for starting new units and expanding old ones. The state government helps to raise capital by buying shares. (This is considered temporary assistance, and the better co-ops have bought back most or all of their government shares.) The government also audits the co-ops and regulates them in various ways: a state committee regulates cane prices, for example. However, the government does not intervene directly in management decisions. Even though there is at least one ex-officio representative of the government on the board of directors, he keeps a low profile. Long-term investment credit is obtained from public agencies like the Industrial Finance Corporation, which lends to private industry on equivalent terms. Many of the older coops, like Malegaon, have long since paid back their original loans and taken more credit for expansion. Another vital and distinctive feature of the cooperatives is their harvest and transport system. The agriculture department in each factory has a staff which keeps track of the planting dates of every cane field, scheduling the harvest accordingly. The factory hires harvest and transport teams through contractors, and the teams work according to a schedule laid down in advance and supervised by the factory staff. Harvest and transport teams typically consist of about a dozen bullock carts, each owned and operated by a dryvillage farmer, along with his brother or other male kinsman, plus their wives. (Another kind of team is hired to cut cane and load it onto trucks, but cart teams predominate.) Teams migrate for the harvest season (usually lasting five to six months) from villages in other districts, where the dry rabi or winter-season crops are of minor importance. As for the labor which grows the crop in the first place, this is supplied by the families of shareholders and the hired labor which they employ. Larger farmers hire laborers on annual contracts; and nearly all cane growers hire teams intermittently for certain operations, such as planting, earthing-up, and weeding. These laborers are local residents, either landless or marginal

The Politics of Sugar

201

landholders. They and the migrant harvest laborers come from the middle and lower castes, the majority from the same farming castes (Maratha, Dhangar, Mali) which own and cultivate most of the land. Credit to finance sugarcane cultivation is linked to membership in the factory. Each village has one or more credit or multipurpose co-ops, to which the farmers belong. Repayment of crop loans for sugarcane is arranged by the factory: advances are deducted from its cane payments and paid to the credit societies. In turn, all the credit societies in a factory's area of operation are entitled to become "B class members" and to elect among themselves one member to the board of directors. Another distinctive feature of the cooperatives is their tendency to invest in social infrastructure in their areas of operation. Many have built schools, high schools, polytechnic institutes, colleges, clinics and even hospitals, which serve not poly their employees but also the surrounding communities, including non-members. Quite a few have been active in promoting other kinds of cooperative enterprises: lift irrigation societies, cooperative banks and dairies, poultry and consumer co-ops, etc. Sometimes, as in the case of two outstanding factories in Kolhapur district, they try to match each other in the promotion of ancillary institutions. In a political system where cooperative leaders compete for other public offices, these social investments help expand and consolidate local support among non-members as well as shareholders. During the crushing season, the cooperatives hum with activity. Outside the wall enclosing the factory proper, there are long lines of bullock carts and other vehicles waiting to deliver the sugarcane. In Malegaon, at any time of day, one will find a large number of farmers going to and fro in the administrative offices. They come to collect their cane payments, to confirm harvest schedules, to consult the administrative staff, to meet the Managing Director or the Chairman. Every office is open to them, and there is usually a group of shareholders gathered in the Chairman's and MD's offices. By simply observing the demeanor of the farmers, one can see that they regard the factory as their own. The demeanor of the Chairman, MD, and administrative staff also indicates that they are available to answer questions, deal with complaints, discuss the season in progress, and generally ensure that the shareholders are satisfied. Both groups, the farmers and the office people, share a keen interest in the technical performance of the factory; and the current results in Malegaon and other nearby factories are constantly under discussion. This, as much as anything else, demonstrates the strong sense of members' participation in these co-ops. Apart from these characteristics, the cooperative sugar factories are constructed and managed just like the nearby private ones. The typical cooperative now crushes from 1500 to 2500 tonnes of cane daily, though some crush as much as 5000 tonnes. They use vacuum-pan machinery, which is in the realm of modern, large-scale industrial technology. In the 1950s, the early co-ops had to buy machinery and parts manufactured abroad, but today their equipment is made in India.

202

The Politics of Sugar

Evolution of Factory Politics The first board of directors (1956-60) in the Malegaon Cooperative Sugar Factory was nominated by the state government, through its Registrar of Cooperative Societies. The board included Shembekar's son as chairman, plus a Brahman lawyer from Baramati with experience in the cooperative movement. One Maratha Freedom Fighter, who had been elected to the state assembly in 1952, was also included. The remaining local directors were village leaders like Madhavrao and Bhalchandra Taware, the representatives from Malegaon. Five of the ten local directors were Marathas, two were Dhangars, two were Brahmans, and one was a Mali. The board also included the District Collector, an official from the Irrigation Department, and the Joint Registrar of Cooperative Societies (Sugar), as representatives from the state government. There was also a nominee from the District Central Cooperative Bank. Since the factory had lost its founder, there was a problem of succession in 1960, when the first election took place. However, Shembekar's absence only hastened an inevitable process of sorting out the political strengths of various groups among the shareholders. Not only was there an internal balance of power to sort out, there were also growing opportunities to connect leverage in one institution with leverage in others. A system of patronage politics began to emerge, connecting the factory with other arenas, such as district cooperative banks, district and state Congress committees, the state legislature, and district and local development councils (thepanchayati raj institutions, which were just being established). The chief patronage resources in the hands of the factory directors were jobs and promotions. By 1970, an estimated two-fifths or more of the Malegaon ·employees belonged to shareholders' families. (Higher technical posts tended to be filled by outsiders, veterans of the sugar industry; but less-skilled jobs were open to locals.) There was tremendous pressure on directors, from members linked to them by caste, kinship, or village residence, to provide jobs. The directors were eager to comply as best they could, since this was a sure way of securing votes for re-election. Moreover, their clientemployees provided volunteer labor in election campaigns, including campaigns for state offices. Almost the first order of business of the board elected in 1970 (an election which I observed) was to hire about 50 new workers. The essence of patronage politics is the exchange of votes for favors. This style of politics is competitive and non-ideological. Although individual competition is vigorous, group interests also come to the fore. It is assumed that a person who belongs to the same lineage, caste, or village as a particular director is more likely to be able to get favors from him. The result is individual and factional competition within groups, coupled with competition between groups for better access to more goodies. The processes are similar to ethnic "machine" politics in a number of American cities. In this competition, the Marathas are faced with potential opposition from a variety of small castes which ordinarily have no unity among themselves. Thus the Maratha majority pyramids into even greater control over top

The Politics of Sugar

203

positions. This happens for two reasons: first, Maratha voters will often unite if a single Maratha candidate is opposed by a non-Maratha; and second, minority candidates are often not supported by voters from other minorities. Nevertheless, Marathas are always competing with each other, and a minority caste can sometimes wield power by voting in a solid block and demanding to be included in a slate of candidates. In these circumstances, an astute minority can control the swing vote between Maratha factions. In the Malegaon factory, caste competition was brewing even on the first (nominated) board, because it seemed to non-Marathas that the Marathas, who had five directors out of ten, wanted to hog all the employment patronage. 6 When the time for the first election came in 1960, the nonMaratha leaders collectively worked out a strategy for trimming the Maratha power. The brains behind this strategy, it was said later, were two Brahmans and a Lingayat from Baramati-men who were embroiled in urban politics, where Marathas had less voting strength. Their method was to assemble a slate of candidates representing the non-Maratha vote blocs and to split the Maratha vote. The first objective was achieved through selecting one Brahman, one Lingayat, two Hatkar Dhangars, two Shegar Dhangars, and one Khatik Dhangar (a Freedom Fighter) for their slate. Since some villages in the factory area are dominated by Dhangar lineages, the Dhangars have important pockets of strength (see Orenstein 1965). The second objective, that of splitting the Maratha vote, was accomplished by including V.C. "Bapusaheb" Jadhavrao, the grandson of Raja Shambhusingh, on the slate. Bapusaheb was respected for his college education, fluency in English, wealth, and family prestige. Moreover, he was not part of a large Maratha lineage, so potentially he could attract votes from any group, since he did not have a lot of kinsmen to satisfy. The essence of the non-Maratha strategy was to persuade other nonMaratha leaders to support them, even though not all could be included on their slate. This made it hard for the other two slates, which were dominated by Marathas, to find good non-Maratha candidates. The non-Maratha slate was so well constructed that it won nine seats out of eleven. The resulting board had four Marathas and seven non-Marathas, an historic high point for the latter. Three board members, including two from Malegaon (Bapusaheb Jadhavrao and Gangadhar Taware), had some college education. The variety of groups on the non-Maratha slate became a problem when it came time to elect the new chairman. Unlike the board as a whole, the chairman could belong to only one caste, and it was difficult for one minority to give the nod to another. Oddly enough, then, they settled on a Maratha: on Bapusaheb, the Jagirdar's grandson. This was a logical compromise, since Bapusaheb did not have a hungry lineage in tow, whereas other directors, except the Brahman and Lingayat, did. Thus it was difficult for an otherwise successful coalition of minorities to share the top seat-an outcome which essentially doomed such coalitions in the future. The pattern of succession after the 1960 election was complicated. At that time, the by-laws required that every two years, half the board would

204

The Politics of Sugar

resign and open their seats for election. The first resigners, in 1962, were determined by drawing lots; and Bapusaheb, the Chairman, was among them. It happened that most of those who were compelled to resign were Marathas, and four of the six who remained for another two years were Dhangars. This, of course, spurred the Marathas' urge for revenge. They were particularly worried, it was said, that the Dhangar Freedom Fighter, whom they regarded as a hot-headed agitator, might be elected chairman by the Dhangars. Bapusaheb was induced to join a coalition of four Marathas and one Mali, a slate which excluded the other three resigners who were running for reelection. Bapusaheb's panel won four of the five empty seats, and the board then had five Marathas and six non-Marathas, with no new Dhangars. The four Dhangars and one other director absented themselves from the first board meeting, thus blocking a quorum and the election of a chairman. In order to regain the chair, then, Bapusaheb had to negotiate with the Dhangars. He and two members ofhis new slate split from the other Marathas, and Bapusaheb became chairman again. However, in 1964, after the other half of the original board resigned, another Maratha was elected chairman. Bapusaheb had been a businesslike chairman, and he obviously had abilities as a coalition-builder. His problem, over the long run, was that he did not inspire durable loyalties among other leaders or among the shareholders. His strength was that he was not attached to a large body of kinsmen; but this was also a weakness, since he had no automatic vote bloc behind him. Also, he was too aware of his family status to suit the rising politicians. Bapusaheb had some later successes in factory politics, but he never became a durable chairman. Given that his competitors were all fairly ordinary village leaders, his limited success says much about the shift of power away from hereditary elites-even when such elites were enterprising enough to compete in the new system. Thus it happened that the first two chairmen did not become durable strongmen: Shembekar's son because he was a Brahman and had business elsewhere; and Bapusaheb because he did not command lasting support among any particular group. Consequently, the factory entered a period of uncertain leadership. In four elections, from 1962 through 1967, the following conditions always obtained: there were at least three slates of candidates (which meant that potentially powerful vote blocs were split among two or more slates); and no single slate won an outright majority on the board. Consequently, there was always a reshuffling of coalitions before a chairman could be elected; and after Bapusaheb, no chairman was re-elected until 1970. In fact, only one director was re-elected continuously from 1960 to 1970; and he was, of all things,·a Brahman. This was Shankarrao Date, son of the magnate discussed in Chapter 5. A businesslike leader, Date did not favor patronage politics, and he was popular with big farmers whose interest lay in efficiency and high cane prices rather than jobs in the factory. By the same token, he was popular with shareholders who did not have a kinsman as director, since he was not trying to steer favors to his own group. (Brahmans

The Politics of Sugar

205

did not have large lineages of shareholders.) Date had good contacts with villagers throughout the factory area, since he could not rely on a single lineage or village for support. He built up his following by helping those who had no patron-director to call on, seeing to it that those without patrons were treated fairly. He kept the shareholders informed of what the board was doing-which might have had an inhibiting effect on other directors, none of whom wanted to acquire a reputation for excessive venality. In two board elections, Date got the highest number of votes among all candidates, but he did not become chairman. Whatever slate he was on, Date was usually the only member of that slate to be elected. Obviously, the stronger Maratha leaders wanted nothing to do with him, since they were playing a different game. During most of the 1960s, other leaders relied essentially on votes from their own lineage, caste, and village. The voters responded in kind, voting mainly for those with whom they had some connection, regardless of their slates. Indeed, slates were never elected as such. It was the ambition of most leaders to deny votes to all other candidates, including those with whom they were supposedly allied. The theory behind slate-making was that partners would share their bases of support, since each shareholder could vote for as many candidates as there were seats to be filled. 7 However, there was not much trust in these arrangements and no means of enforcing them. Consequently, each candidate tended to tell his close supporters, in private, not to vote for anyone else at all. Since the elected candidates would be those with the most votes overall, votes not given to others were as good as extra votes for the one who got this kind of support. The general results for coalitions were thus unpredictable and disintegrative. Only Date was able to collect votes far and wide, and he was one of the few leaders consistently re-elected in the 1960s. However, the 1967 election heralded the emergence of a new political order. The by-laws were changed in 1967: the entire board was vacated and a new board was elected for three years. As before, there were three slates, each with candidates from Malegaon, one of the largest villages in the factory area. Madhavrao Taware was on one slate and Bhalchandra's nephew, Murlidhar, was on another. The two families had been rivals in various spheres: Bhalchandra's family held the office of sRrpRnch (head of the village pRnchRyRt) for more than 20 years. Madhavrao, on the other hand, had more influence in the village credit societies and had been elected to the Zilla Parishad (District Council). Both Madhavrao and Gangadhar, another of Bhalchandra's nephews, had been elected at various times to the factory board. Madhavrao's slate won five seats in 1967 and Bhalchandra's nephew's slate won six. (The third slate won none.) Madhavrao's slate included Bhimdevrao Gophane, a bright young Malegaon leader from the Dhangar caste. Although Bhalchandra's nephew's slate had a majority, it was not enough for a quorum of seven directors. Madhavrao's slate played the usual trick of not attending the first board meeting, so there was no chairman elected.

206

The Politics of Sugar

Now it happened that Madhavrao and Gophane had been supporters of a young Congress party worker, Sharad Pawar, in his first election to the state assembly that same year. 8 Another member of their slate, a Maratha patil from Baramati, was also a supporter of the young MLA. In addition, Shankarrao Date, who belonged to the opposing slate, was a Pawar supporter, and Date had two allies on his slate. The remaining three members of the second slate (including Bhalchandra's nephew) had no particular leader. Thus an unusual compromise was worked out under the guidance of the MLA. Madhavrao Taware, Shankarrao Date, and the patil from Baramati were brought together by agreeing that each would serve one year as chairman. (The vice chairmanship also rotated among their supporters.) Thus they brought together a total of eight votes, enough for a quorum. When it came time for Madhavrao Taware to take up his chairmanship, he made a touching gesture, handing over the chair for half a year to his young ally, Bhimdevrao Gophane. (This was in return for Gophane's help in getting Madhavrao elected to the district council in 1962.) Gophane was the first Dhangar chairman in the Malegaon factory and one of very few anywhere. The same coalition set up a new election slate in 1970, with the help of the MLA. Their opponents taunted them over the makeshift arrangement for rotating the chair in the preceding three years and predicted that the coalition would fall apart. But the MLA's influence was so strong that every member of the slate (except one) was elected, and Madhavrao Taware became chairman for a full three years. The ballot counts showed that many shareholders voted for the MLA's slate as a whole, contrary to previous custom. There were 2,428 ballots cast, representing 88 percent of the shareholders. On these ballots, there were a total of 18,033 votes, equivalent to 77 percent of the total possible, since each shareholder could vote for up to ten candidates. Clearly, many shareholders had used most of their votes. The winning slate was constructed with skill. Careful representation had been given to minority castes (Dhangars and Malis) who were inclined to vote in solid blocs, since they otherwise had few ways of asserting themselves against the Marathas. In fact, three of the top four vote-getters were Dhangars, with Gophane in the lead, confirming that their supporters had voted in blocs. One other Dhangar, one Mali and five Marathas completed the list of winners from this slate. Apart from caste and lineage affiliation (they were almost all from large lineages), the candidates on this slate were selected according to the size of their villages: two candidates each from Baramati, Malegaon and another large village, and one each from five smaller villages. The art in constructing such a slate was not only in knowing where the votes lay-even a visiting anthropologist could know that. The real art was in convincing dozens of prospective candidates (a total of 75 had filed nomination papers) to withdraw in favor of others. Those who do not withdraw, even if they run as independents without a coalition, are potential spoilers: that is, they can split the vote blocs which the slate-makers are trying to weld together. Every village and lineage has its internal factions, so slate-makers invariably have more candidates than they can possibly use.

The Politics of Sugar

207

Winnowing out the redundant candidates depended on the influence of the young MLA, Sharad Pawar. He had to choose candidates likely to unite their own lineages and factions and draw the widest support elsewhere; and he had to meet personally with all the others-with those who might be called "nuisance candidates." The essence of these meetings was that he had to indicate the need for a personal favor from the nuisance candidates: that is, he needed them to withdraw so that his slate of supporters could win. In most cases, there was no specific quid pro quo promised in these meetings, though many of the nuisance candidates had specific requests in mind. (These could take myriad forms: help in speeding approval of a cooperative bank loan; or support in some other election, such as the Baramati municipal council, ttzluka panchayat samiti, zilla parishad, or district cooperative bank elections.) In most cases, the MLA simply offered his good will, which might be worth a lot later on. Inevitably, there were candidates who did not succumb to these blandishments, and they formed the opposition slate. (There were no independents left over after the winnowing was done.) Unfortunately for the opposition, they were united only in not wanting to cooperate with one section or another of the MLA's supporters. For example, there was no room on the same slate for Bhalchandra's son and his family's rival, Madhavrao Taware, who was close to the MLA. Some opposition candidates were personal supporters of the MLA's own rivals, the Kakade family. The Kakades were powerful in cooperative sale-purchase unions, at the taluka and district levels, and dominated the nearby Someshwar Cooperative Sugar Factory. The opposition slate had some good candidates, including one ex-chairman of the factory and three former directors. However, they had no central leader and they failed to entice into their ranks candidates from two mediumsized, Dhangar-dominated villages which provided solid vote blocs for the MLA's slate. In fact, the opposition slate was not well balanced: it had only three Dhangars and eight Marathas. It included three candidates from a single large village which was, however, rife with factionalism, and it had two candidates from one of the smallest villages. Malegaon itself had only one candidate, Bhalchandra's son; and the same was true for Baramati. The only winner from this slate was from a medium-sized village dominated by a single Maratha lineage. Conversely, the only loser from the MLA's slate was Shankarrao Date, the Brahman. As always in such cases, it was said that Date was betrayed at the last minute by his allies, who spread the rumor that the MLA did not want Date to win. Whether the rumor was true or not, Date's fate was consistent: he had previously won alone, and this time he lost alone. His appeal had been to those who wanted good government and those who wanted insurance in case their favorite sons did not get elected. However, the latter appeal was effective only when there was no strong central leader to balance competing claims from other castes, lineages, factions and villages. When such a leader arrived, Date was no longer needed. The MLA put an end to political confusion in the Malegaon factory, and he continued to hold sway over factory elections through the 1980s. Subsequent

208

The Politics of Sugar

elections were arranged as in 1970, with supporters and nuisance candidates competing for his nod. The chairmanship rotated among competent but not very independent leaders: Bapusaheb Jadhavrao, the Jagirdar's grandson, and S.S. Hiremath, for example. Hiremath was a Lingayat by caste, a lawyer by training, and a businessman and big bagaitdar by profession. He was also a technocrat, having served at one period as Managing Director of the Malegaon factory and later as MD of another. He and Bapusaheb were similar, in that both were better educated than most of the factory leaders (they were both fluent in English) and neither belonged to a large Maratha or Dhangar lineage. In other words, they could be counted on to make sound business decisions, to give a fair hearing to leaders from all vote blocs, and to follow the wishes of the MLA. The latter's wishes surely included a preference that nobody stay in the chair too long, since he needed to keep his many supporters happy by rotating favors among them and did not need a strong deputy who might become his rival. One problem with this p9wer arrangement, over the long run, was that ultimate control rested with a leader who was not involved in the daily working of the factory. Sharad Pawar, on his way to becoming Chief Minister of Maharashtra, was preoccupied with other matters. 9 One criticism of current factory leaders is that they are said to be unable to make important decisions without consulting him. This delays decisions and undoubtedly weakens the sense of personal responsibility which is vital to healthy management. Other factories have arrived at a similar condition via different routes: for example, when a powerful chairman gets elected to parliament and spends more time in Delhi than in managing his factory. Cooperatives and State Politics In 1984 there was an event which dramatized Malegaon's connections with the outside world and with the historical past. An equestrian statue of Shivaji, the founder of the Maratha kingdom, was dedicated on the grounds of the sugar factory. Presiding over the ceremony was the President of India, along with the Chief Minister and Governor of Maharashtra. (The Chief Minister at that time was Vasantdada Patil, founder of an outstanding cooperative sugar factory in Sangli district.) According to the annual report of the Malegaon factory, the occasion was also graced by the presence of several Members of Parliament and other political leaders, literary figures, chairmen and directors of various sugar factories, and senior government officers. These distinguished guests came to honor Shivaji and thus, by implication, "the rise of Maratha power," which first occurred in the 17th century. By further implication, they were there to honor the renewal of Maratha power, which came after independence. Malegaon is a good enough factory, as we shall see in the next chapter, but not terribly innovative or distinguished when compared with the best in the region. What brought these eminent guests to Malegaon was the

The Politics of Sugar

209

influence and negotiating skills of Sharad Pawar. At that time, Pawar was president of a regional party that he had formed in opposition to the ruling Congress party led by Indira Gandhi. In retrospect, inviting the President of India to Malegaon may have been a signal of Pawar's intention to eventually rejoin Mrs. Gandhi's Congress. After her assassination, the two parties were merged, and Sharad Pawar became (for the second time) Chief Minister of Maharashtra. Following the electoral defeat and recent assassination of Rajiv Gandhi, Pawar is one of a handful of leaders who must attempt to revive and reconstitute the Congress party under new leadership. Thus events in Malegaon reflect changes in the wider world; but they also do more. Local events are an integral part of what happens in that world. These events are broadly consistent with other studies of rural politics in Maharashtra. (See, e.g., Sirsikar 1970, Carter 1974, Rosenthal1977, Baviskar 1980, Lele 1981, 1984.) The general pattern of patronage politics can be summarized as follows: First, factory leaders and their competitors form among themselves a set of shifting coalitions. Successful coalition partners are often, but not always, members of dominant Maratha lineages in the larger villages. (A few strong chairmen have come from other castes, particularly in southern Maharashtra.) Coalitions are assembled to include those who command vote blocs in these lineages; but other castes, lineages and villages are also represented, roughly in proportion to their numerical strength. Second, leaders retain the loyalty of their supporters by providing favors. Particularly esteemed is the favor of a job in the factory, and many shareholders or their kinsmen hold such jobs. Such favors cannot be provided for all, so the leaders also project an image of general goodwill by supporting religious and cultural events, irrigation projects, drought relief, educational and medical institutions, and so forth. Third, local leaders are also judged by their access to higher patrons: to state party leaders, cabinet members, and the like. In return for political support, the latter can encourage development projects, such as schools, roads, banks, clinics or irrigation projects in the local area. Thus the interaction between local and regional leaders consists of intense bargaining over the exchange of votes for patronage. Four~h, the most effective factory leaders are those who can maintain the support of a solid vote bloc, despite competition from other leaders in the same communities, plus access to powerful state leaders. Since state politics is highly competitive and non-ideological, patronage alliances sometimes change with bewildering rapidity, tending to crystallize during important elections and then shifting as competitors adjust to the results of one election and prepare for the next. This system creates a network of alliances and information which reaches from the highest offices down to the villages. In 1969-71, I found that village leaders and their friends (belonging to several castes) gathered every evening in a credit society office to discuss current events. These were not

210

The Politics of Sugar

just local events: they had a keen and informed interest in political struggles within the state and central cabinets. One reason for their interest was that the local economy was governed in part by Delhi's decisions on sugar policy. Another reason was that their immediate patron, Sharad Pawar, was linked directly_ to Y.B. Chavan, a top minister in the central government and the most powerful Maratha leader in India. I have no doubt that such discussions still go on today, and that the level of interest and information regarding Pawar's career is even stronger. This is an important feature of rural politics in Maharashtra, particularly in the irrigated areas where farmers are increasingly educated and can afford to be concerned with more than just their immediate subsistence. These villagers are actively concerned with what their governments are doing at the local, state, and central levels, and they also have, through cooperatives and other institutions, the possibility of negotiating with these governments over policies affecting their interests. (For example, there are state and national federations of cooperative sugar factories which collect information, negotiate, and lobby with the government over sugar policies.) This is an impressive outcome of the "rise of Maratha power." At this point, however, we need to raise some issues of interpretation concerning the political impact of the cooperatives-issues which are followed up in later chapters. The cooperative sugar factories are often described as components of a regional system of domination by "elite Marathas." For example, Lele (1981:27) writes of the post-independence rise of a "hegemonic rural class" resulting in nearly total Maratha domination of rural politics. He stresses the power of elite Marathas at every level, ranging from the state cabinet down to the villages. There is little disagreement over the basic facts. Maratha leaders have risen to powerful positions in most rural institutions and in the state legislature. Other castes have not shared equally in the power of the Marathas. The countryside is still filled with poor people: with landless laborers and dry farmers belonging to the middle and lower castes (including Marathas). The fruits of political and commercial success have not been distributed evenly. However, Marathas are the majority in the countryside. In a democratic system, assuming that caste, religion, ethnicity, tribe and nationality will continue to influence political loyalties (as they do almost everywhere), should the majority not prevail? (Of course, minorities must have equal rights to vote, lobby, petition and protest.) As Lele notes, the size of the Maratha caste is partly the result of political processes (1981:46-54). The caste has created itself by various means, including unification of kunbis (peasants) with high-status Marathas, via consciousness-raising in the non-Brahman movement. (This process was outlined in Chapter 5.) Such an expansive base for Maratha power would seem like a good thing in a democratic system. However, Lele argues that the Maratha majority is, in effect, a contrivance: "A look into history shows that the numerical strength of the Marathas has itself been a function of

The Politics of Sugar

211

the rulership strategy of the elite" (1981:54). Majority rule, in this case, is not a system in which potential leaders are required to build broad, popular bases of support. To Lele, it is instead the outcome of manipulation by an elite. In essence, Lele says: this is the same old Maratha elite which always dominated the region. Lele is aware of the flexible boundaries of Maratha identity and the upward mobility which is possible within the caste (1981:55). However, he views this as just more evidence of deliberate contrivance by the elite to maintain their power. Some questions must be raised regarding this interpretive stance-a stance which is fairly typical for critics of the cooperative factories. • • • •

First, does it over-interpret a transient assortment of leaders? Second, does it obscure the position of minority castes and lineages? Third, does it minimize the significance of urban elites? Fourth, does it help to understand caste politics in other regions?

Over-interpretation. Does the concept of a permanent, hegemonic elite create a grand abstraction out of an assortment of leaders who happen to be in power at a given time? Going back to the pre-colonial period, I have stressed the competitive and mobile character of Maratha leadership. Taking this leadership at one point in time and assigning it an historical mission which extends back over centuries, and which invariably serves to keep the "same" elite in power, seems a bit overdone. Many Marathas who occupied the highest positions in state politics after independence came from humble backgrounds. Y.B. Chavan was one example, Sharad Pawar another. Their rise to power was analogous to the rise of the sugar entrepreneurs, many of whom came from poor, or at least undistinguished, families. Some of the individual leaders described in this chapter could be cited in support of Lele's position. In Malegaon, the three Maratha leaders who were most active and successful in post-independence sugar politics were Madhavrao Taware, Bhalchandra Taware (through his nephews), and Bapusaheb Jadhavrao. The latter comes from an aristocratic family. Likewise, Madhavrao is related to former patils (headmen); and Bhalchandra's family has been wealthy and influential at the village level since at least the early part of this century. However, the other leaders mentioned above do not fit so well. S.S. Hiremath comes from a well-to-do family, but he is Lingayat, not Maratha. Bhimdevrao Gophane is a Dhangar, whose father was an immigrant laborer. Dadasaheb Shembekar, the founder of the Malegaon factory, was a Brahman who started out poor; and Shankarrao Date's father, another Brahman, also started out poor. In rebuttal, one might note that these non-Maratha leaders could only attain limited positions in local and regional politics. However, this can be

212

The Politics of Sugar

explained by the demography of caste and kinship as well as by elite domination. Minority ciiStes and lineages. As we have seen, minority castes participate actively in sugar politics, and they sometimes take advantage of Maratha factionalism to win access to power positions. If the minority castes (and even minor Maratha lineages) find it difficult to compete with the major Marathas, it is because they usually find it impossible to unite behind a single leader. So long as caste and kinship remain a consideration in voting, the largest castes and lineages will tend to prevail. This works to the advantage of Dhangars in villages where they are in the majority; but at higher levels, they cannot out-vote the Marathas. Nevertheless, since Marathas compete so vigorously among themselves, they actively seek the support of minority castes-as we saw in the analysis of slate formation for Malegaon factory elections. (See also Baviskar 1980.) Minorities thus gain some access to positions of power and to the patronage which flows in return for electoral support. Other things being equal, Maratha voters tend to support Maratha leaders. However, voters also respond pragmatically to non-Maratha leaders who show exceptional capacity for institution building. Hence a few of the most impressive sugar co-ops have been chaired by non-Marathas (Baviskar 1991). The low castes, in comparison with the minority peasant castes, are more burdened with poverty, landlessness, and low status. Consequently, they are in a weaker political position. Neo-Buddhists (formerly Mahars) are one of the larger minority castes and have sometimes been successful in gaining political representation and making demands on the state government. If we assume that caste will continue to influence political behavior, then the NeoBuddhists can only increase their power through alliance with other castes, or through mobilizing on broad economic issues. In a democratic system, it would seem to be a healthy sign that the majority caste is in power. As we shall see in Chapter 11, there are major regions in India in which high-status minority castes are politically dominant. Surely this is less desirable than domination by a broad, middle-status majority? Well, perhaps not. When urban intellectuals are speaking privately, they sometimes state what they see as the better alternative: the political system should be controlled by those with the most education. The self-interested character of this view need hardly be emphasized. Given the demography of caste and lineage and the problems of unification among minority groups, it is not necessary to invoke a hegemonic elite to explain why Marathas so often prevail. Of course, the concept of a permanent elite may be useful for other reasons. Urban elites. Lele discusses the rise of powerful urban elites under colonial rule, and he .describes how "the Maratha elite" sought to reassert itself against these new power-holders (1981:26-27, 50-54}. What he does not discuss is the possibility that Maratha leadership was itself transformed in the struggle with urban interests. The rise of cash-crop entrepreneurs, cooperative organizers, and agro-industrial pioneers, as a new breed of rural

The Politics of Sugar

213

leaders, is not considered a real change in the character of Maratha domination over the countryside. Likewise, the relative decline in power of the urban high castes is considered merely a victory for elite Marathas, not for rural interests more broadly construed. Other regions. Lele argues that elite Marathas broadened the boundaries of their caste in order to continue their long-term domination of the region (1981:55). If this is such a convenient and successful "rulership strategy," why has it not been imitated elsewhere? There are, after all, several heavily populated regions in which minority castes are dominant. Why should the Maratha elite be any different? Why not dominate as a minority and leave the kunbis to their own devices? My answer, of course, is that the broadening and unification of the Marathas came about not so much from a conscious elite strategy but as a result of factors over which individual leaders had little control. These factors included the historical flexibility of caste boundaries, the competitiveness of Maratha alliances, and the mobility of Maratha leadership-combined with the new demands and opportunities posed by mass politics under late colonial rule. Thus we agree on the basic fact of Maratha domination, but not on how to characterize its underlying dynamics. Lele's conception of a permanent elite closely parallels the standard theory of commercialization: that is, the powerful just get more powerful. Any evidence of broader participation and competition is characterized as a rulership strategy of the elite. Comparing modern Maharashtra with earlier periods and with other regions of the world, I see the current political system as highly competitive and dynamic, resulting in ever-increasing participation by a wide variety of groups. The permanent elite is an attractive concept because intellectually tidy. If I had to advance a phrase to summarize what I see as a more complex set of processes in 20th century Maharashtra, I would offer the following: The main structural change, both economic and political, has been the rise of the rum/ middle class. And who lost power relative to this class? The old aristocratic elite, like the Jagirdar's family, and the urban intellectual, political, administrative, commercial and industrial elites. (This argument will be elaborated in Chapter 11.) The rise of Maratha leaders since independence has, of course, benefited some villagers more than others: big bagaitdars more than dry farmers, Marathas more than lower castes. We may certainly hope that the expansion of power and participation will spread more evenly. Nevertheless, the recent transition is as extraordinary as it is unprecedented. It is difficult to think of any post-colonial regime where peasants have risen so decisively to power. This was, in fact, a revolutionary transition. Since 1947, Marathas have seized a host of new opportunities to exercise the skills that made them famous long ago: mobilizing followers, building support networks, negotiating alliances, maneuvering for position, distracting adversaries, suborning their commanders, raiding their lines of supply, and

214

The Politics of Sugar

hammering out terms for (temporary) truces. It has all been very exciting. The new leaders were not only pursuing great personal opportunities, they were making a social revolution-a revolution which brought peasants into power in a large, rapidly industrializing state. (Maharashtra, with more than 70 million people and an aggregate state income of more than $10 billion, has a larger economy than most Third World countries [Muthiah 1987:228).) This was not only a political revolution, it was a chance for rural leaders to usher in a period of remarkable growth and innovation in the countryside. However, perhaps because it involved so little violence, this revolution has been ignored by today's shell-shocked journalists and historians. Notes l. When Maharashtra state was formed, it combined the Deccan and Konkan regions and part of the Southern Maratha Country, from the old Bombay Presidency, along with Vidarbha, which was formerly part of the Central Provinces, and Marathwada, which was part of the Nizam's state of Hyderabad (Kamat 1983:67). 2. The land ceiling legislation was widely evaded in practice, with the result that little "surplus" land was acquired by the government for redistribution. Successful evasion does not, however, mean that the land ceiling was in the interests of big bagaitdars. Some would have rented or purchased even more land, if the reforms had not impeded them. Moreover, entrepreneurial tenants were compelled to give up their rented lands as an indirect consequence of the tenancy reforms (see note 4 below). 3. The Malis had tried to register their factory as a cooperative, but the laws and regulations did not encompass this possibility in the 1930s. Consequently, their factory was registered as a joint stock company. Nevertheless, the basic concept of a factory owned by village cane growers provided the model for cooperative experiments after independence. 4. Paradoxically, the magnates, who were tenants on a grand scale, were threatened by the tenancy acts of 1948 and 1957, which were designed to transfer land from non-cultivating landlords to their tenants. What plagued the magnates was that their landlords, fearing the permanent loss of their lands, wanted to cancel their leases. In many cases, the magnates agreed-partly, perhaps, because the government was also preparing legislation to limit the maximum acreage owned by any individual, which would have made it hard for the magnates to buy out their landlords. 5. Under the land reforms, in fact, the maximum area of cane land per individual owner is now much less. The state's Land Ceiling Act of 1961 specifies that one person may own a maximum of 18 acres under perennial canal irrigation or 27 acres under perennial well or lift irrigation (Maharashtra 1974:183-86). Under normal crop rotations and irrigation rules, farmers could devote a maximum of one-third of these areas to a new crop of sugarcane each year. Thus the maximum annual cane area would be nine acres per individual owner. (Some families, of course, jointly own much more.) 6. The two Brahmans on the board would have seen this a bit differently from the other non-Marathas. Their view was undoubtedly that patronage politics was a threat to good business management, and that Maratha leaders, with the largest numbers of kinsmen to satisfy, were the greatest threat. 7. One director was elected by representatives from the village credit co-ops (the "B-class members"), so the producer-members did not vote for this seat directly.

The Politics of Sugar

215

8. Sharad Pawar was the youngest son of Govindrao Pawar, who had managed the Shahu Maratha Boarding in Raja Shambhusingh's time and later managed the Nira Canal Sales-Purchase Union. Govindrao also helped the Jachaks establish the Shri Chhatrapati Cooperative Sugar Factory near Baramati. One of Sharad's brothers had been an active political worker (for the PWP), and another became Managing Director in a cooperative sugar factory. 9. Sharad Pawar's career, which encompasses sugar politics on a grand scale, would require a chapter in itself.

9 Performance and Impact of a Sugar Co-op Because the sugar cooperatives are successful and powerful, they provoke a certain amount of resentment. The Indian press delights in printing stories of conflict among the sakhar samrat, or sugar barons, and there have been plays and films written about them. The temptations of power are manifold, and the barons have earned some of the criticism they receive. But some portrayals of their misdeeds (on film, for example) have been gross and implausible caricatures. Much criticism is a political reflex, since urban intellectuals are uncomfortable with a regional power system in which less-educated rural leaders have a strong voice. Intellectuals believe they are the natural leaders of Indian society; and the mainstream ideology of the political system, which favors state control of the economy, strongly reinforces that view. Cooperative sugar factories would appeal more to the intellectuals if they were managed by state bureaucrats. Since they are both independent and powerful, however, the factories are the object of intense suspicion and criticism. As an American, my own cultural bias inclines me to admire the sugar barons for precisely those qualities which other people deplore-for I see them as competitive hustlers, as pragmatic organizers who value results over rhetoric, as countrymen who outmaneuver the city slickers. The critics see several negative sides to this: First, the factories are dominated by a rich peasant elite. Second, these rich peasants monopolize all benefits for themselves (Chithelen 1985:612). Third, the factories exploit their laborers (Matson 1983; cf. Breman 1978-79). Fourth, the factories are heavily subsidized by the state (Lele 1984:173)-and thus, by implication are inefficient and parasitic. Fifth, the factories are rife with patronage, factionalism and political corruption-which again suggests they are inefficient and parasitic. Some of this is true, some is half-true, and some is wildly implausible, as we shall see. Some criticism reflects a lack of comparative perspective. Aie the co-ops more subsidized, less efficient, and more corrupt than private industries or public-sector enterprises? The critics have not, in fact, tried to demonstrate this empirically. It is difficult to compare levels of corruption, except to 216

Performt~nce

and Impact of a Sugar Co-op

217

observe that the co-ops are probably no more corrupt than the private firms and government agencies with which they do business. Regarding efficiency, there is much to be said in these next two chapters. We may note, for purposes of comparison, that the public sector is massively subsidized and wasteful, and that private industries are heavily subsidized by tariffs and regulations, which drastically limit competition (Maddison 1971; Bardhan 1984; Rudolph and Rudolph 1987; Nayar 1990). This chapter and the one following give a detailed analysis of efficiency in the sugar co-ops and draw comparisons with other types of sugar enterprises, public and private. There is an element of truth in some criticisms, but the reality is more complex and interesting than they suggest. The critics ignore the remarkable achievements of the cooperatives: not only their spectacular growth in sugar production but also their technical efficiency and innovation; their mobilization of resources to set up ancillary enterprises and meet basic needs (like education); and their mobilization of village farmers to give them an active voice in the state government. Dogmatic views blind some critics to the most interesting questions about these cooperatives. The basic dogma is that the cooperatives are playthings of the rural rich, that the benefits all accrue to big bagaitda.rs. What the critics fail to ask is why the cooperatives have expanded so remarkably, why they are technically efficient and innovative, and why they have not been imitated successfully in other regions. Every region in India has big farmers involved in politics; but in no other region, except nearby Gujarat, have they built a huge and complex system of agro-industrial cooperatives under local control. This suggests that we must look beyond dogma to understand how "rich peasants" have accomplished this in Maharashtra but not elsewhere. In this chapter and the next, we shall review the performance and efficiency of the cooperative factories, concentrating first on Malegaon as an example. We shall also consider the distribution of benefits-to see whether "it is still the rich peasants who mostly cultivate sugarcane and almost exclusively enjoy the benefits accruing from cooperative sugar enterprises" (Chithelen 1985:612). This question is continued in the next chapter, which considers why benefits must, on the contrary, reach a broad range of middle and small farmers in order for the factories to function efficiently. Economic Rationale

As we saw in the preceding chapter, the Malegaon factory was organized by a Brahman; thus it was not started as a scheme to advance Maratha political ambitions. Was there a real economic need, a need felt by cane growers of all types, for a cooperative factory? Shembekar and his fellow organizers had several reasons for wanting to move from gul to sugar production. First, because gul contains moisture, it is more perishable. Gul is about 75 percent sucrose, while white sugar is nearly 100 percent sucrose. The other ingredients in gul, including glucose, absorb water in damp weather, causing the gul to ferment and dissolve

218

Performance and Impact of a Sugar Co-op

(Knight 1905:13). Gut stays in peak condition for only a few months and is riskier to store in the expectation of better prices after the harvest season. Second, sm;tll-scale equipment cannot extract all the juice from sugarcane, which has a tough stem. Bullock-driven cane crushers are particularly inefficient, extracting only about 65 percent of the sucrose (Padhye 1924:1617). As a result, one-third of the sucrose literally goes up in smoke, when the crushed fiber is used as fuel to boil the juice. Considering what it costs to grow sugarcane, this is a serious loss. Engine-driven crushers are better, but they still cannot do as well as large-scale, industrial machinery. A modern sugar factory can keep total sugar losses to less than two percent. Third, a cooperative factory resolves the basic problem of the Indian sugar industry, which is cane supply. Since the growers own and manage the factory, they are paid the highest possible cane prices. This removes the conflict and uncertainty that emerge when private factories depend on a multitude of small suppliers. Moreover, since the growers are paid well, they willingly submit to a centralized harvest schedule. Such a harvest system greatly enhances the efficiency of the cooperatives. Fourth, working as an organized group, the cane growers can expect to get better prices from selling sugar in bulk. The same idea applies to all processing or marketing cooperatives-though it is insufficient, by itself, to make them all operate effectively. In the case of sugar, however, the technical factors just mentioned made this expectation more plausible. Finally, cooperatives can go even further in promoting the vertical and horizontal integration of the industry. As we shall see, they have financed new irrigation facilities, created integrated credit and marketing systems, procured fertilizers and improved seeds in bulk, and provided soil-testing and extension services-as well as hiring harvest workers, organizing harvest and transport operations, processing the cane, and selling the sugar. For all these technical and organizational reasons, then, sugar cooperatives have the potential to work cohesively and efficiently, providing a better outlet than the gul market. Growth of the Malegaon Factory The Malegaon Cooperative Sugar Factory was registered under the Cooperative Societies Act in 1955, the year that Dadasaheb Shembekar died. The plant was erected during the next two years and went into operation in 1957-58. At that time, the factory had 1,160 members and a paid-up share capital of Rs.2.88 million. The total share capital contributed by producermembers (cane growers) was Rs.l.84 million, and shares worth Rs.41,000 were purchased by cooperative credit societies (B-class members) in the factory's area of operation. Thus 65 percent of the share capital was raised locally, while 35 percent was provided by the government (see Table 9.1). 1 In order to erect the plant, the factory also had to obtain long-term finance from the Industrial Finance Corporation (IFC) and the Poona District Central Cooperative Bank (DCCB). It purchased West German equipment

219

Performance and Impact of a Sugar Co-op TABLE 9.1

Sources of Initial and Expansion Capital, Malegaon Factory

Source of Capital Share capital from producer-members: Share capital from local cooperative societies: Factory funds (deposits deducted from members' cane prices): Share capital from state government: Loan from Industrial Finance Corporation: Loan from Poona District Central Co-op Bank: Loan from Industrial Development Bank: Total

Initial (1957) Investment

Two Expansions (1968 & 1973)

Rs. 1,836,000

Rs. 1,500,000

41,000 5,000,000 1,000,000 5,250,000

8,500,000

2,125,000

6,000,000 5,000,000

Rs.l0,252,000

Rs.26,000,000

Sources: Malegaon 1969, 1977, 1986.

with a crushing capacity of 1000 tonnes of cane per day. The total capital cost of the plant was Rs.10.25 million. In its first year of operation, the factory achieved a recovery rate of 11.63 percent, and in subsequent years the rate was over 12 percent, which is quite good. (The recovery rate is a measure of technical efficiency: it is the ratio of tonnes of sugar produced divided by tonnes of cane crushed, indicating what proportion of the raw material was converted into the final product.) Within a decade, the factory had paid back its initial capital loans and was preparing for expansion. In 1968-69, the factory expanded its crushing capacity from 1000 to 1500 tonnes per day. The total cost of this expansion was Rs.6.1 million, which was financed by loans from the IFC and DCCB, plus capital raised from the members by issuing new shares and deducting non-refundable deposits from their cane payments. More than half the new capital was raised from producer-members by these methods. (By 1969, the paid-up share capital was Rs.3.72 million, of which 73 percent came from producer members and local cooperatives.) The factory expanded again, to a capacity of 2500 tonnes, in 1973-74. The total amount invested in these expansions was Rs.26 million, derived from the sources listed in Table 9.1. One factor which accounted for the efficiency of the factory was its compact area of operation, which in 1969 covered only 13 villages with more than

220

Performance and Impact of a Sugar Co-op

TABLE 9.2

Shares Owned by Producer-Members, Malegaon Factory, 1969

Area of Cane Pledged to Factory

Equivalent Shares

0.5 or 1 acre 1. 5 to 5 acres 5.5 to 10 acres 10.5 to 15 acres 15.5 to 25 acres

1 or 2 3 to 10 11 to 20 21 to 30 31 to 50

Total

Number of Members 2269

424

70

7 2

2772

Source: Malegaon 1969.

4,000 acres of sugarcane. Nearly 90 percent of the cane supplied to the factory came from within a radius of seven miles, which facilitated supervision of harvesting by the agriculture department and reduced the cost of transporting the cane. At this time, the factory had 2,772 producer-members. As seen in Table 9.2, the vast majority (82 percent) held only one or two shares. Two shares were equivalent to one acre of cane pledged annually to the co-op. Since, at that time, the crop was planted on a three-year rotation, two shares corresponded to one acre of cane, which in turn was one-third of a threeacre canal block. Thus the vast majority of members were small farmers, with three irrigated acres or less, as Table 9.2 indicates. 2 In 1988-89, the most recent year for which audited reports are available, Malegaon produced 45,770 metric tonnes of sugar and sold it for Rs.340 million. Total income (including the sale of alcohol derived from molasses) came to Rs.507 million. All loans for the expansion of fixed assets had been repaid with interest by 1980. In 1989, the paid-up share capital was Rs.6.8 million, of which 97 percent was owned by producer-members. The factory had redeemed most of the shares purchased by the state government and was essentially owned outright by the cane growers. 3 Because Malegaon is a cooperative, and because it disburses net income in the form of higher cane prices, it pays no corporate income tax. Nevertheless, it generates significant revenue for the state and central governments. In 1988-89, the state government collected cane purchase tax worth Rs.12.3 million. In addition, the central government collected excise duty of Rs.40 million on sugar and alcohol. Thus direct taxes took slightly more than 10 percent of total income. (Hidden taxes, which cost nearly as much, are discussed in the next chapter.) Turning to Table 9.3, let us consider the factory's long-term performance. Here we note a strong upward trend in membership-hardly consistent with the belief that a small elite are the sole beneficiaries. We also find an upward

167,007 166,461 180,861 172,503 183,780

244,262 425,892 449,100 357,604 346,114

371,723 348,964 388,411 425,628 603,285

l, 714 1,792 1,832 1,856 1,898

1,963 2,035 2,518 2,772 2,811

3,183 3,297 3,480 3,572 3,831

3,973 4,442 4,590 4,687 4,783

4,985 5,607 6,021 6,259 6,485

6,771 7,059 7,274 7,430 7,863

60-61 61-62 62-63 63-64 64-65

65-66 66-67 67-68 68-69 69-70

70-71 71-72 72-73 73-74 74-75

75-76 76-77 77-78 78-79 79-80

80-81 81-82 82-83 83-84 84-85

85-86 86-87 87-88 88-89 89-90

419,777 376,021 405,322 457,700 640,030

273,085 477,666 493,841 407,325 391,177

326,600 443,627 403,931 393,526 188,089

276,557 225,360 171,600 216,637 345,749

239,170 209,100 195,485 269,640 245,850

210,171 213,205 217,202 216,229 220,930

117,962 157,761 166,845

Sugar bags Produced (quintals)

11. 31 10.78 10.44 10.72 10.61

11.18 11.20 11.01 11.25 11.27

11.24 11.45 10.81 10.78 9.63

11.37 10.92 10.92 10.49 11.47

11.62 11.06 11.16 10.95 10.48

12.59 12.82 12.00 12.54 12.01

11.63 12.43 12.46

Sugar Recovery (\ cane)

• Average yield in the factory's area of operation. b Cane price before deductions. �: Malegaon 1969, 1977, 1986, and Annual Reports.

289,889 387,125 374,335 363,617 195,567

240,808 205,649 157,120 206,510 302,230

206,312 189,252 174,719 247,068 234,268

104,605 130,333 137,362

1,160 1,313 1,410

57-58 58-59 59-60

(tonnes)

Producer Members

Cane Crushed

162 147 173 198 247

128 180 195 160 152

163 187 188 178 106

207 130 181 181 176

217 183 177 215 207

191 186 207 181 182

185 159 186

Total Crushing Days

92.9 87.2 97.1 106.4 150.8

61.2 106.5 112.3 89.4 86.5

72.5 96.8 93.6 90.9 48.9

100.3 85.7 65.5 64.5 75.6

128.9 118.3 87.4 102.9 97.6

104.4 104.0 113.0 107.8 114. 9

65.4 81.5 85.9

Capacity Utilization (\) 42.00 52.00 58.00 59.00 58.00 55.00 65.25 65.00

46.09 45.05 47.52 47.07 45.76 50.89 52.02 52.10

tonne)b

cane price (Rs. per

Price drop Price impact 260.00 233.00 219.00 250.00 300.00 316.00 301.00 290.00 320.00

39.00 43.77 40.86 35.17 36.46 37.38 33.88 33.30 26.86 37.26

n.a.

Expansion 129.00 120.00 115.00 137.50 175.00

42.90 41.70 48.05 47.03 39.15

Drought Drought Drought

Drought Drought

Expansion

93.65 134.00 132.25 125.00 143.00

51.30 45.03 42.00 39.21 30.00

Expansion

64.00 63.00 173.00 112.50 75.50

Drought

Conwnenta

55.08 54.02 49.03 49.88 46.10

(tonnes per acre)•

Avg. yield

Performance of Malegaon Cooperative Sugar Factory, 1957-58 to 1989-90

Year

TABLE 9.3

....

N

222

Performance and Impact of a Sugar Co-op

but irregular trend in total cane crushed and sugar produced. There are several reasons for the irregularities. First, variations in the weather affect the supply of irrigation water, which determines the yield of sugarcane. Another factor affecting cane supply is fluctuation in sugar prices. Part of this fluctuation is due to a "sugar cycle," which goes up and down about every four or five years (Baru 1990:204-08). When prices fall, sugarcane production declines. 4 In addition, production is sometimes affected by spasmodic changes in price-control policy by the central government. Normally the government maintains partial control of the market, meaning that it buys a certain share of the output from each factory at fixed prices. Sometimes, however, it makes abrupt changes which unhinge the market, as happened in 1978-79. Finding that sugar production was at a high level, the government dropped all price controls, including monthly quotas on sugar released onto the open market. Sugar prices dropped sharply, down to less than Rs.2.00 per kilo (wholesale) from a high of more than Rs.4.00 the previous year (Cooperative Sugar 1985-86:661-62). There was a glut on the market, chaos reigned among sugar sellers, and the farmers experienced great hardship. The next year, predictably, sugarcane and sugar production fell off sharply throughout the country, whereupon prices rose steeply to Rs.8.50 per kilo. At this point, the government tried to impose a ceiling on sugar prices, creating a huge black market and causing traders to offer sub rosa payments on an unprecedented scale to the sugar factories. The results of this roller-coaster ride for Malegaon can be seen in Table 9.3, where production fell catastrophically in 1979-80 and next year had still not recovered to the level reached five years earlier. This was a period of hardship for the whole industry. Consumers also suffered from the sudden rebound in sugar prices after production dropped, and the government was forced to spend scarce foreign exchange importing sugar into one of the world's largest sugar-producing nations. 5 Variations in weather and prices also affect the other performance measures listed in Table 9.3. However, the table shows not only variation but decline in yields over the long run. This is not unusual, as data on other factories indicate (Baviskar 1980:38; Maharashtra 1976). The main reason for this decline is probably the increased demand for irrigation, combined with effects of the factory's expansion. When the factory started in 1957, there were about 2200 acres of cane in its command area, while today there are about 7,000 acres. Meanwhile, the capacity of the Nira canal has not increased, in fact it has decreased as far as sugarcane is concerned. The rules governing cane blocks on the Nira Left Bank Canal were changed in 1975, so that the area under sugarcane was reduced from one-third to one-fourth of each block. (This change was in response to repeated droughts: it was an effort to make water available for seasonal crops over a wider area.) Less canal water for cane meant lower yields. Farmers have countered by deepening old wells and constructing new ones, but the available irrigation per acre of cane is generally less than it used to be. One indicator of the

Performance and Impact of a Sugar Co-op

223

problem is that the Nira canal formerly provided water every 10 to 15 days, whereas now it provides water only every 30 days during much of the year. This has become a widespread problem: in Ahmednagar district, for example, the older canals now serve far more villages than they were designed for. Meanwhile, the factory's latest expansion was not well-timed. The basic reason for expanding is to reduce average costs per bag of sugar. For example, a factory with a daily crushing capacity of 1,250 tonnes employs about 800 workers, while the Malegaon factory, with twice that capacity, employs about a thousand. Thus the average cost of labor is reduced by investment in equipment. However, Malegaon needed more sugarcane after expansion, just when the irrigation supply became more problematic. It was not until 1981-82 that the factory managed to operate at full capacity for the season, and the problem persisted in subsequent years. 6 In order to solve this problem, Malegaon had to buy more cane from growers beyond the local area. In 1988-89, for example, 66 percent of the cane supply came from outside the factory's area of operation. This added to the cost of transport and lowered recovery rates, since cane dries out more when brought over longer distances. (As we shall see in the next chapter, this geographic expansion of cane procurement has occurred in many factories, greatly intensifying the competition for cane supplies.) Due to these problems, sugar recovery rates in Malegaon have never returned to the high levels achieved in the late 1950s and early 1960s. Overall, the performance measures listed in Table 9.3 depend as much on agricultural conditions as on the efficiency of the factory itself. The next-to-last column in the Table shows cane prices paid by the factory. From the point of view of the cane growers, this is the single most important fact about the factory's performance. All net income, except that which is held in reserve for investment, is distributed in the form of higher cane prices. Thus the overall growth in prices indicates a healthy organization as long as interest payments on external debts and depreciation schedules are being met. The long-term rise in cane prices is impressive, especially given that sugar prices were held tightly in check by the central government. According to the National Federation of Cooperative Sugar Factories [NFCSF] (1988a), sugar prices have risen slower than the prices of all other basic consumer goods, including foodgrains, and slower than the cost of purchased inputs, such as fertilizers. Malegaon has been able to raise its cane prices only because it engaged in a program of expansion, technical improvement, and diversification, reducing the average cost per bag of sugar. Another way to judge a factory's performance is to compare it with nearby factories, especially in terms of results which are under the direct control of the factory management. It is pertinent to compare a local cluster of factories, since variations between agro-climatic zones lead to significant differences in cane quality across the state. Table 9.4 compares factories located in or near the Nira valley, within a radius of about 50 kilometers from Malegaon. 7

224

TABLE 9.4

Performance and Impact of a Sugar Co-op

Performance of Local Cluster of Sugar Factories, 1987-88

District & Factory

Recovery Rate (%)

Capacity Util.(%)

Time Lost(%)

Sugar Lost(%)

Pun~ Qistri£t Malegaon Someshwar Chhatrapati Yeshwant Walchandnagar

(C) (C) (C) (C) (P)

10.44 11.07 10.40 10.42 9.80

97.10 129.64 96.30 113.43 90.90

8.23 13.57 5.85 12.08 14.39

2.14 2.48 2.20 1.98 2.29

§atj!ra Dhtrict Shriram New Phaltan

(C) (P)

10.49 10.42

120.28 102.07

14.55 19.96

2.01 2.22

Shola11ur Dist, Shankar Mohite Saswad Mali

(C) (C) (P)

10.66 10.22 10.13

125.36 108.50 105.55

13.85 11.73 11.87

2.15 2.01 2.23

(C - Cooperative, P- Private) Source: Deccan Sugar Institute 1988:37-38, 89-90.

The general finding in this table is that Malegaon performs about as well as other factories, both cooperative and private. On recovery rates, Malegaon ranks in the middle. Regarding capacity utilization, Malegaon ranks among the lowest-a legacy of over-expansion. However, the factory has been taking steps to increase the supply of irrigation and sugarcane, and the results look better by the end of the 1980s (see Table 9.3). The amount of time lost during the crushing season (due to cane shortages, mechanical breakdowns, cleaning and maintenance) is an important indicator of managerial efficiency. On this count, Malegaon has been doing well (Table 9.4). Finally, one of the best overall measures of performance is the percentage of sugar lost in production. Unlike recovery rates, this measure is not affected by the quality of the raw material: it measures sugar lost in proportion to the sucrose actually present in the cane. Total losses of under two percent are considered quite good for the industry. In this respect, Malegaon did about average for the duster, with losses slightly above two percent. On the whole, it can be said that the technical performance of this duster is good and that Malegaon is keeping up with the average. The table also indicates that cooperative and private units are at about the same level of performance, though the latter have been faced with special problems in recent years, as we shall see in Chapter 10. It is interesting to compare the same cluster of cooperative factories in terms of their financial standing. (Similar data are not available on the private

225

Performance and Impact of a Sugar Co-op TABLE 9.5

Financial Performance of Local Cluster of Cooperatives, 1987-88

District & Factory fl!DI!

1:!1UJ:1!Ct

Malegaon Someshwar Chhatrapati ~!UU

l:!htJ:1!Ct

Shriram

Sb!!lllli!YJ: l:!ht,

Shankar Mohite

Paid-up Share Capital

Reserve & Other Funds

Borrowings

Working Capital

Total

Govt.

6.5 4.0 6.7

0.1 0.3 0.1

49.2 38.4 40.9

4.5

129.1 78.5 130.5

7.4

0.2

28.7

17.6

n.a.

5.2 8.5

2.1 0.7

34.5 12.7

14.5

88.6 145.6

(Amounts in millions of rupees.) Maharashtra State Cooperative Sugar Factory Federation 1988.

~:

factories.) Table 9.5 indicates that most are healthy and that Malegaon is among those which have redeemed nearly all their shares from the state government. Malegaon has also accumulated a healthy supply of capital (Rs.49 million) in reserves and other funds. Relative to the size of its reserves and share capital, its borrowings are not heavy. Malegaon may also be compared with the whole cooperative industry in Maharashtra. A government-appointed committee evaluated the health of all 59 factories which were in operation as of 1979-80. It found that 19 factories, or 32 percent, were "sick" in terms of accumulated losses and other problems (Patil Committee 1983:18-20). The central agro-climatic zone, in which Malegaon is located, was generally healthy: only five of the sick units were located in this zone, and all except one of these had been started after 1975. (New factories were more vulnerable to the sharp fall in prices which occurred in 1978-79.) In the central zone, 22 out of 27 factories (81 percent) were found to be healthy, or at least not sick; and Malegaon came out in the middle of the ranking on the committee's "health index." Malegaon's Other Investments Malegaon has recently diversified, building a distillery to produce alcohol from molasses, a by-product of sugar. (In this venture, it has followed the example of other factories.) Capital was raised from bank loans and by deducting Rs.S per tonne from the cane price paid to members over a period of several years. (In 1985-86, for example, this method generated Rs.l.18 million in capital.) The distillery went into operation in 1987-88, reducing the cost of molasses storage and bringing additional revenues. In 1988-89,

226

Performance and Impact of a Sugar Co-op

alcohol sales came to Rs.21 million, adding nearly Rs.19 (net) to the price which the factory could pay per tonne of cane. Malegaon has also been investing in irrigation, to help improve capacity utilization. It procured government permission for lift irrigation systems along the Nira river, serving farmers who did not have canal blocks for sugarcane. The factory also took over the operation of twelve lift irrigation cooperatives which were in financial trouble. In order to make the lift schemes more effective, the factory organized the construction of medium-scale irrigation works in collaboration with the Shriram cooperative factory, just across the Nira river. Four Kolhapur-type weirs (so called because this system of irrigation was first perfected in Kolhapur district) were built. The weirs impound river water during the monsoon season and canal water seeping down through the soil. Bank loans worth Rs.ll million were used to finance construction. The weirs were built by the state's irrigation department, which will eventually pay back the loans. However, the two factories paid all the interest charges, which cost Malegaon about Rs.500,000 annually. The four weirs, in conjunction with lift irrigation schemes, brought about 8,300 acres under irrigation on Malegaon's side of the river. Malegaon has also increased its cane supply by subsidizing loans taken by members for deepening wells, constructing new ones, or installing pumps and pipelines. Another scheme, organized with the state government, helps small farmers from the lower castes (Scheduled Castes and Neo-Buddhists) reduce the cost of cane cultivation. (About 150 farmers with 100 acres of cane were subsidized in the mid-1980s, at a cost to the factory of about Rs.150,000 per year.) A more general program aimed at raising cane yields is the subsidized distribution of fertilizers and seeds for green-manure crops. Another investment in the cane supply system has to do with transport equipment. Much of the factory's cane supply is transported by bullock carts, operated by migrant teams of harvest workers. The factory has replaced the traditional wooden carts with more than 800 steel-frame carts with rubber tires, which it rents at Rs.8 per day to the harvest workers. These carts have at least two advantages: they can carry a much heavier load of cane each trip, with no greater effort by the bullocks; and they do not destroy the roads, which the wooden carts did to a remarkable extent. (The factory has built and maintained roads to almost every farm in its area of operation.) The factory has also encouraged its members to invest in trucks and tractors, for even greater efficiency in the cane transport system. The factory guarantees loans taken for this purpose and also guarantees to put the vehicles to use during the crushing season. As of 1985-86, the purchase of 72 trucks and 65 tractors had been arranged for producer-members. Like other cooperative sugar factories, Malegaon has also made a series of social investments which have no direct pay-off in terms of its own profits. It has built primary and secondary schools which are open to the children of employees and villagers. A total of 1,600 students attend the schools,

227

Performance and Impact of a Sugar Co-op TABLE 9.6

Deductions from Cane Price Paid by Malegaon Factory, 1986-87

Type of Deduction 1. 2. 3. 4. 5. 6. 7. 8. 9.

Non-refundable deposit Fund for Polytechnic Institute 5-year deposit for distillery Area development fund Sugarcane development fund Chief Minister's fund Small housing scheme Small savings scheme Drought relief fund

Amount per Tonne Rs.

5.00 2.00 5.00 1.00 4.00 0.50 0.50 1.00 2.00

Total deductions

Rs. 21.00

Cane price after deductions

Rs.280.00

Source: Malegaon 1986-87:7.

costing the factory about Rs.300,000 per year. The factory has also financed the construction of a polytechnic institute, which offers diploma courses in civil and automotive engineering, and industrial electronics. As of 1988-89, there were a total of 526 students in the polytechnic. Funds for the institute were collected by deducting Rs.2 per tonne from the price paid for sugarcane. Deductions from the cane price for this and sirililar purposes must be approved at annual general body meetings. Table 9.6lists deductions approved for 1986-87, including those for the distillery and polytechnic. Most of the others are standard practice, similar to those made by other cooperative factories. This is a convenient and effective way to raise capital: each onerupee deduction raised at least Rs.225,755 in 1986-87, since the members provided that many tonnes of cane. (Non-members who supply cane are not charged for all deductions.) The factory has also undertaken projects for the benefit of less fortunate communities. It constructed 50 community wells for irrigation in nearby dry villages. In addition, like other factories, Malegaon makes regular contributions to funds for drought relief, housing for the poor, and other charitable works. In 1986-87, contributions under ~hese headings came to more than Rs.700,000. Another investment by the factory has been in the realm of livestock and dairy production. Milk is a fast-growing cash "crop" in the region-with the advantage, from an equity standpoint, that dairy production is laborintensive. Small farmers and dry farmers can take advantage of a growing network of cooperatives for collecting, processing, and marketing milk. A veterinary center has been established at the Malegaon factory, where artificial inseminations are performed and advice given on health care for dairy animals. The factory also offers loan guarantees for producer-members purchasing high-quality dairy cattle.

228

Performance and Impact of a Sugar Co-op

The local dairy industry has boomed since the early 1970s. The ability to organize and manage the new dairy co-ops results partly from experience in the sugar industry. By the same token, it can be said that other agroindustries (for example, in poultry and wine production) have sprung up in the Baramati area as a result of experience gained in the sugar industry. Malegaon's primary responsibility, of course, is to its shareholders. It has not made major investments in the neediest villagers, those who have no land. (However, many farmers with dry or seasonally-irrigated land have been helped by its investments in irrigation.) The factory is not, and cannot be, a welfare agency for the whole area. Yet there are cooperative factories which have done more in terms of social investment. Investments by Other Sugar Co-ops Many cooperative sugar factories have made extra contributions to the economic and social development of their areas, and some have been especially innovative in this respect. The following paragraphs give a few examples. 8 Different factories have concentrated on different investments, according to the needs and interests of their areas. In groundnut-producing districts, they have helped set up oil-processing cooperatives and solvent extraction plants. In cotton-growing areas, they have helped establish cooperative ginning and pressing units. Some factories have established pulp and paper mills; other have started distilleries and chemical plants; still others have set up printing presses, cattle feed plants, cooperative banks, department stores, and cooperative power grids for electricity (which is vital for well irrigation). Once there is a cooperative factory in an area, it becomes a focal point of further growth. By 1980, the sugar co-ops had generated a total capital investment of Rs.1,620 million from members' share capital, deposits, and deductions (Maharashtra State Cooperative Sugar Factories Federation [MSCSFF] 1980:97). For 57 factories, this came to an average of Rs.28.4 million apiece. Sugar factories have mobilized capital and expertise to provide what farmers need most: that is, irrigation. In the southern districts, which have little canal irrigation, cooperative sugar factories have organized numerous lift irrigation schemes for the benefit of their members. For example, the Shetkari factory, started in 1956 by Vasantdada Patil (later Chief Minister of Maharashtra), had by 1983 invested Rs.12.5 million in lift irrigation schemes, irrigating 20,000 acres. When the factory was started, it had a crushing capacity of 1,000 tonnes per day and only 1,500 acres of sugarcane in its area. By 1983, it had a capacity of 6,000 tonnes per day, one of the largest in the country, and there were 40,000 irrigated acres nearby. Meanwhile, membership had expanded from 1,100 to 28,000 growers (Abraham 1983). Factories in Ahmednagar district have also made substantial investments in irrigation, since the available canal water is stretched too thin. For example, Sanjivani factory had to invest in well irrigation for the majority of its

Performance and Impact of a Sugar Co-op

229

members, so that they could grow sugarcane. By 1980 the Pravara factory had constructed 55 percolation tanks, 40 lift irrigation schemes, and other irrigation facilities; and the Rahuri factory had spent about Rs.2.6 million for lift irrigation (MSCSFF 1980:107-17). An estimated total of 125,000 acres have been brought under irrigation schemes promoted by cooperative factories (Director of Sugar [DS] 1984:4). Many factories have veterinary centers, and several have raised capital for cooperative dairies. Warna factory in Kolhapur has built an ambitious dairy complex with a marketing branch in Bombay (Apte 1991). Warna and others have also established poultry cooperatives. Many factories have built and maintained roads in their areas, facilitating the transport of sugarcane and also contributing to the general development of the villages. The factories have jointly created the Deccan Sugar Institute, to train technologists and conduct research. Maharashtra's sugar co-ops have also jointly established a heavy engineering corporation to manufacture equipment for sugar factories. The corporation has completed several contracts on a turnkey basis, competing with established private firms. Most factories have also contributed significantly to welfare facilities, particularly in the fields of health and education. When the first factories were established, there were few secondary schools in the rural areas. These co-ops started primary and secondary schools for the children of employees and nearby villagers. A number of factories have established colleges of arts, science, commerce, engineering, and medicine. All factories have established dispensaries, and most have organized family planning camps (Rane 1983:AS25). The Pravara factory has gone further, creating a medical trust of Rs.5 million, and establishing a hospital with 150 beds (MSCSFF 1980:108). In 1976 the state government launched a program to construct housing for the homeless. Shareholders in every factory contributed Rs.0.75 per tonne of cane, and by 1980 a total of 45,000 houses had been constructed (DS 1984:7). Other projects have been started for the benefit of the lower castes. The Panchaganga factory has taken over the debts of lower-caste small farmers and assumed cultivation of their lands until their debts are repaid. A credit scheme for small farmers from the Scheduled Castes has been implemented by a number of factories, including Malegaon, with financial help from the state government (Rane 1983:AS23). During natural calamities, the resources of the factories have been useful for organizing relief and other activities. In 1970-73, for example, most parts of Maharashtra suffered from serious drought. There was a shortage of foodgrain and fodder, and in some places even drinking water was not available. The Kopargaon factory contributed over Rs.l.6 million to the Chief Minister's drought relief fund, provided fodder and molasses worth Rs.147,000 to the cattle camps in the district, and spent over Rs.35,000 on the construction of percolation tanks. In total, Kopargaon spent over Rs.2 million on relief work during this critical period, and many other factories did likewise. The

230

Performance and Impact of a Sugar Co-op

state government's success in organizing relief work owed a great deal to the initiative, resources, and infrastructure created by the sugar cooperatives. Impact of the Malegaon Factory: Employment

Having reviewed the performance of the factory and compared it with others, let us weigh its impact on Malegaon village. The direct employment effect is quickly summarized. The factory started in 1957 with several hundred workers and today employs over a thousand. More than half are year-round employees; seasonal workers are employed for seven months each year, and the same ones are rehired next year. Their positions, duties, qualifications, and wages are defined by the central government's wage board, which periodically revises wages and allowances. Several hundred workers (mostly skilled workers, hired from outside the local area) live in the factory's staff colony, provided with facilities for education, health care, recreation, and worship. Other employees live in the surrounding villages. All these may be called "local workers," to distinguish them from migrant harvest workers. The factory stimulated a new wave of permanent migration to Malegaon. Table 9.7 lists immigrants after 1950-those captured in my 1970 sample. Eight immigrants became factory workers and settled in the staff colony. Of these, two had attended high school (both were Brahman) and one had attended college: all three took up white-collar jobs. The others were largely from the middle peasant castes, with at most a bit of primary-school education: they got jobs as semi-skilled or unskilled laborers in the factory. (Because they are not part of Malegaon proper, these eight are excluded from the village sample analyzed below.) Many factory jobs were filled by residents of Malegaon and surrounding villages. A total of 30 sample households (or 25 percent of the sample, excluding residents in the factory colony) had one or more members working in the factory in 1980. Of these, six were landless, nine small farmers, nine middle, and six large (see Table 9.8). It is clear that factory jobs, which are avidly sought after, were not monopolized by big farmers. There was little variation according to size of landholding: the rate of employment varied between 24 and 36 percent. The landless were the exception: because they were not shareholders, they were less favored in the competition for jobs. Table 9.8 also shows the distribution of sugar factory jobs by caste. Only the low castes had a low rate of employment-primarily because most were landless and thus not shareholders. The mid-low castes, on the other hand, were strongly represented; and the middle peasant castes, which included most of the shareholders in the village, were also well represented. It is interesting that the major Marathas (mostly members of the Taware lineage) did not have an exceptionally high rate of employment in the sugar factory, even though they had the strongest leaders and largest vote blocs in the village. It is also interesting that, among the Tawares, factory jobs went mostly to small and middle farmers. There may have been a tacit

231

Performance and Impact of a Sugar Co-op TABLE 9.7

Immigrants to Malegaon after 1950

Type of Immigrant

Year of Arrival

Caste

Education

First Occupation in Malegaon

Sugar factory employees

1952 1956 1957 1957 1957 1957 1958 1960

Maratha Maratha Dhangar Mali Brahman Muslim Brahman Dhangar

None None 4th None 11th 4th 11th College

Construction, laborer Watchman Asst. mechanic Workshop laborer Switchboard operator Asst. electrician Clerk Asst. storekeeper

Skilled workers

1959 1967 1968

Maratha Maratha Gujar

3rd 11th 3rd

Cloth-shop owner High-school teacher Skilled mason

Agricultural laborers

1956 1956 1957 1959 1964 1965 1967

Dhangar Dhangar Mali Nav-Buddha• Mali Wadar Maratha

4th None None 6th None None None

Laborer, sharecropper Laborer Laborer Mason, laborer Annual laborer (saldar) Laborer Laborer

Landowners

1957 1959 1964

Wadar Maratha Maratha

None 11th 4th

Construction contractor Government officer Farmer

• Mahars converted to Buddhism in 1956. ~: Sample survey of village 4DQ factory staff colony (N - 130).

understanding that Taware leaders would provide jobs to their less-fortunate lineage mates and also to leaders of the minor Marathas and other peasant castes, as a way of consolidating political influence among these groups. All local workers, below the managerial level, belong to state-wide sugar factory unions. There are several of these unions, affiliated with different political parties, ranging from centrist to militant leftist. Unions are granted recognition as official bargaining agents on a territorial basis, by taJuka (Matson 1983). The unions compete for recognition, to some extent, by offering different degrees of militancy. Because many local workers belong to shareholders' families, they generally enjoy cordial relations with the management. In fact, most are protected by three sets of institutions: by the government wage board, the unions, and the informal network between workers, shareholders and directors. An additional workforce, consisting of about 6,500 seasonal migrants, provides harvest and transport labor. These laborers are recruited in contract teams from dry villages, mostly in other districts. Most of the harvest workers are farmers from the middle castes, seasonally unemployed because the rabi or winter crops are unimportant in the dry areas.

Performance and Impact of a Sugar Co-op

232 TABLE 9.8

Sample Families with Jobs in the Cooperative Sugar Factory, 1980

Caste & Lineage Category Low Hid-Low Other Peasant Minor Maratha Major Haratha High Total Percent of Sample"

Size of Landholding Landless

3 2 1

Small

Middle

1 3

1 1 2

1 4

6

9

3 2 9

14

32

36

Large

Total 2

7

2 2 1 1 6

6 4 8 3 30

24

25

Percent of Sample• 12 32 25 14 36 38 25

• The number of households with factory employees divided by the number of sample households in the same category. Source: Sample survey of Halegaon, excluding the factory colony (N- 121).

About half are recruited in family units with bullock carts, the typical unit consisting of two men, one woman, two bullocks, and a cart. (Until recently, each unit used its own cart; now most hire rubber-tired carts from the factory.) These workers are known as gadivale, gadi meaning cart and vale, people. The men cut the cane, the women dean, bundle, and load it in the cart, and the men drive the cart to the factory. Each cart unit is paid in two ways: first, for the weight of the sugarcane, and second, for the distance it is transported. In addition, they are allowed to keep most of the green tops from the cane, which are excellent fodder for bullocks. (The significance of this part of their payment will be appreciated if we note that, in northern India, cane harvesters are paid only in green tops, not in cash [Batra 1988, 1991).) In 1989-90, each cart unit earned an estimated Rs.120 to 135 per day from the Malegaon factory. (Out of this amount, contractors who brought cart teams to the factory deducted 12 percent commission, plus interest on earnings they had advanced to the harvesters.) In addition, the workers often sold green tops to the farmers whose fields they were harvesting, bringing an additional income of perhaps Rs.15 per day. Another category of harvesters works with trucks and tractors. The vehicles often belong to local contractors, including factory shareholders, but the workers come mostly from the dry villages. These workers are known as koytavale (from koyta, or knife). They cut the cane, bundle it, carry it, and load it in the trucks. Since Malegaon has been procuring more cane from distant areas, requiring rapid transport by truck, the proportion of koytavate in the labor force has increased. (They numbered about 3,500 out of 6,500 in 1989-90.) As compared with the gadivate, they are paid slightly more

Performance and Impact of a Sugar Co-op

233

for harvesting and loading: Rs.31.15 vs. Rs.28.40 per tonne. Transport payments, on the other hand, go to the truck and tractor owners. Gadivale are recruited in teams consisting, on average, of about a dozen cart units apiece. Before the crushing season starts, the factory's agricultural officer tours the dry districts, meeting contractors at local market towns. In exchange for advance payments, the contractor arranges to bring so many cart units to the factory area, starting on a certain day. Often he recruits these units from among his kinfolk and fellow villagers. 9 (Detailed studies of harvest workers may be found in Matson 1983; Breman 1978-79.) A number of factors have combined to force up harvest payments in recent years. First, the number of sugar factories has multiplied, so the factories are bidding against each other for available workers. Second, the state has been investing in medium and large-scale irrigation works in the dry areas. As soon as farmers have irrigation, they can grow rabi-season crops and have no need to work as migrant harvesters. Third, the state's Employment Guarantee Scheme (EGS), started in 1974, provides employment on public works on a routine basis in all the dry areas. The combined effect of these changes is to make migrant labor scarcer and more expensive for the sugar factories. Large-scale public investment in irrigation schemes and the EGS shows that the "sugar lobby" does not control ~:tate policy to the exclusion of other rural interests. This point is illustrated by yet another set of factors which exert upward pressure on harvest wages: unions and state-level negotiations. In the 1970s, sugar factory unions began to organize harvest workers (Matson 1983). As a bargaining tactic, some unions organized strikes at the start of the harvest season. These strikes were effective because the factories depended on meticulous scheduling to harvest the cane at peak maturity and deliver it rapidly to the factory gate. The state government stepped in, convening tripartite negotiations among the unions, the State Federation of Cooperative Sugar Factories, and the Chief Minister. Once an agreement was reached, it applied to all factories. In the latter part of the 1980s, there were two sharp increases in harvest wages, both negotiated under Chief Ministers unfriendly to the sugar lobby. One gave an increase of 62 percent and the next an additional 35 percent. From 1985-86 to 1989-90, harvest wages rose by a total of 119 percent. These decisions show that the interests of the sugar factories are overridden, at times, in a political system which depends on broad-based support throughout the countryside. The migrant harvest workers are still not paid on the same scale as local workers (who now earn a minimum of Rs.1,100 per month), but they are better paid than other agricultural laborers. A cart unit of two men plus a woman now earns about Rs.120 per day, as compared with the legal minimum wage for agricultural laborers, which is Rs.22 per day. Moreover, the minimum wage is unenforceable for laborers employed by village farmers: nowadays, these workers actually get a maximum of about Rs.12 (for men) or Rs.9 (for women), depending on the season. Thus a cart unit earns more than three times what its members would earn if they worked as casual laborers.

234

Performance and Impact of a Sugar Co-op

Migrant harvest workers live in rude huts during the crushing season and work at a back-breaking pace, as described vividly by Breman (1978-79).1° Still, they are better paid than other agricultural workers because thegadivale are not purely laborers. They are also suppliers of vital capital equipment: their bullocks, the most valuable equipment that most farmers own. As suppliers of capital equipment, the gadivale are in a strong bargaining position. They cannot be replaced by any old laborers who happen to be around. In the irrigated villages, most local laborers are landless, which means they do not have adequate fodder supplies to keep bullocks. The factories need workers with bullocks, so what they really need are seasonally unemployed farmers, who can only be found in the dry villages.u Indirect Employment Effects The indirect employment effects of the Malegaon factory are extensive and only partially captured in my sample data. The nearby town of Baramati, located between two cooperative sugar factories, has boomed in the last few decades, with a rapid growth of employment in construction, trade, transport, repairs and other services. Likewise, occupational diversity has grown rapidly in Malegaon and other irrigated villages. Malegaon's population nearly doubled, from 6,528 to 12,741, in the 1950s; and it rose to 17,020 by 1981 (Bombay Government 1954; Maharashtra Government 1966, 1986). 12 Surveys in nearby Ahmednagar district show the employment effects of the Pravara Cooperative Sugar Factory. Expansion in cane cultivation, stimulated by the factory, caused a threefold increase in the employment of annual contract laborers (saldars) in just five years. Most of the newly employed saldars immigrated from outside the local area. There was also a great expansion in the employment of contract labor teams (Padki 1974:2627). Some of these effects were captured by my 1970 census and sample survey in Malegaon. Table 9.9 shows the distribution of principal occupations of heads of households for the whole village, including the factory colony. Farming households accounted for only 24 percent of the total, with agricultural and other unskilled laborers accounting for 45 percent and nonagricultural workers accounting for the remaining 31 percent. The Table actually under-represents nonagricultural workers, since all farming households were categorized as such, even though many contained other kinds of workers. For example in 1980, 27 percent of sample farming households contained sugar factory employees, and others had workers employed in Baramati. Laborers were drawn to the village by the expansion of cane cultivation, and self-employed traders and craftsmen also immigrated to meet the rising demand for new services. Table 9.7 includes these other immigrants, who fall into three categories (besides factory employees). The first consists of skilled immigrants who responded to the rising demand for specialized services. The second category consists of immigrants who came to Malegaon as agricultural laborers, attracted by the rapid expansion of cane acreage and

235

Performance and Impact of a Sugar Co-op TABLE 9.9

Principal Occupation of Heads of Households, Malegaon, 1970

Households Occupational Category Farmers• Landless agricultural workers Other unskilled workers Total Agricultural Households Self-employed traders, craftsmen Sugar factory employees Other non-agricultural wage workers Total Households

N

Percent

535 762 238

24.0 34.1 10.7

1535

68.8

234 348 116

10.5 15.6 5.2

2233

100.0

• All other categories are landless. Landowning households are counted only once, although many also had other occupations. Source: Attwood 1974:317 (village census including the factory colony).

employment on this crop. The majority of these left small landholdings with their kinsmen back in the dry villages. In terms of caste status and education, this category was less advantaged than the others. The final category consists of immigrants who acquired land shortly before or after settling in Malegaon. In one case, a Wadar family had already purchased land in Malegaon while they lived in a nearby village. In two other cases, Marathas came to Malegaon because their wives' fathers had good landholdings and no sons. Except for the latter three, who came to Malegaon because they already had established land claims, those who migrated to the village after 1950 did not become land-gainers. In this respect, they differed from the pre1950 immigrants. Many immigrants came to practice specialized occupations, not work their way up from laborer to tenant to farmer. The factory stimulated occupational diversification in two ways: by attracting immigrants with specialized skills; and by encouraging local people to seek education and skilled employment, or to set up new businesses, often as adjuncts to their agricultural occupations. (Surveys show similar effects from the Pravara Cooperative Sugar Factory [Padki 1974].) In Malegaon, the number and diversity of shops expanded in the village proper and near the factory colony. This expansion was stimulated by the growing incomes of farmers and laborers, and the greater variety of their needs. For example, the maintenance and repair of bicycles, motorcycles, jeeps, tractors, trucks and irrigation pumps became essential services. Likewise for the supply of motor fuel and parts. By 1970, one big farmer had established a petrol pump on the main road to Baramati.

236

Performance and Impact of a Sugar Co-op

In providing jobs for educated villagers, the factory gave further impetus to the rising demand for education. By 1961, Malegaon's high school had the second-largest enrollment in the taluka. There were three high schools in nearby Baramati and three others in canal villages near Malegaon, while only one had been established in a dry village (District Census Handbook 1966:271). In other words, canal villagers had both the incentive and the means to increase their educational opportunities, and they acted with dispatch. In the 1970s, a college was established in Baramati and, in the 1980s, a polytechnic institute in Malegaon. We shall see how important education has been as a means of economic mobility. Diversification and Mobility Migration and mobility continued in the post-1950 period as they had earlier. There were less dramatic shifts in land ownership, which I attribute to three factors: first, there was no great depression forcing large owners to sell out; second, there was ever-increasing population pressure, which meant that the land market became tighter; and third, there was greater economic security for most farmers, due to the reliable market for sugarcane provided by the factory. On the other hand, there was a much greater diversity of opportunities as a result of two factors: the spread of education, and the impact of the sugar factory. Let us consider some examples of mobility between the landless and landowning strata. Subsequently, we shall review the statistical pattern of mobility. In the 1940s, Ramsevak was a landless Maratha tenant, renting about 35 acres of irrigated land, which he cultivated with his sons, on a one-half share. (The owners of the land were two Taware widows.) Shortly before 1950, Ramsevak and his two elder sons bought eight acres of seasonally-irrigated land. The sons began to divide the land in 1957 (by which time there were another three sons growing up), but they jointly invested in two wells, constructed around 1960. One son got a job in the sugar factory, which provided savings for this investment. [38)

This is a case where the joint family cycle encouraged upward mobility. The two elder sons made it possible for Ramsevak to sharecrop a large holding and then to buy land and irrigate it. Their investments were facilitated by the assured market for cane as well as the wages provided by the factory. Table 9.10 lists land-gainers and land-losers, that is, those who crossed the boundary between landed and landless, since 1950. (The time period is divided into two segments, before and after 1970, since the sample frame was drawn in that year.) Both gainers and losers come from the same caste categories: mainly from the middle-peasant, mid-low and low castes. Gainers outnumber losers twelve to eight, but the losers seem to be more numerous than they were before 1950. (In the weighted sample, gainers number 26 and losers 28.) This is partly due to increasing population, which

237

Performance and Impact of a Sugar Co-op TABLE 9.10

Land-Gainers and Land-Losers in Malegaon, 1950-1980

1970 to 1980

1950 to 1970

l&Jlg;:

.l.!2.U.n

Caste

N

Caste

N

Major Maratha Minor Maratha Mali Dhangar• Nhavi Ramoshi• Chambhar• Neo-Buddhist•

2 1 1 1 1 2

Brahman

1

1

Mangb

3

1 -2-lb

12 -5•

l&Jlg;: ~

Net 1950-80: Weighted Net:

-7(27)

Lingayat Wani Minor Maratha

1 1

Dhangar Gopal Bhoi Muslim

1 1 1 1

Mangb

2

+ +

-7(7)

-1(1)

Minor Maratha Mali Dhangar•

4

Ramoshi•

1

Chambhar• Neo-Buddhis t•

2 1

-8-lb Net 195G-80: Weighted Net:

Net Gainers and Losers, 1950 to 1980

8 (28)

1 1

10 -5•

+ +

-5(19)

12 (26)

• Five families (one Dhangar, one Ramoshi, two Chambhar, and one NeoBuddhist) lost land between 1950 and 1970, then gained in the 1970s. b One Mang family gained land 1950-70, then lost it 1970-80. ~: Sample survey, excluding factory colony (N- 121).

forces more sons to give up shares in tiny family holdings. Most of the land-losers were of this type. That is, they separated from their fathers or brothers without taking a share of the family land. In two cases, this was made easier by holding a job in the sugar factory: one branch of the family kept using the land, while the other relied on the factory job. Two other land-losers were remarkable because they were Tawares, that is, members of the dominant lineage. In both these cases, sons separated from their fathers and became landless laborers. (Their own sons have since completed high school and found work in Bombay or in the army.) In two cases, there was a cyclical process: men who had been landless laborers became landholders again after their fathers died. That is, they inherited the land from which they had been separated when the family

238

Performance and Impact of a Sugar Co-op

split. In a similar case, a brother regained a share of his brother's land after losing his job in the sugar factory. This points to a second reason why landlosers appear to be more common in recent decades: the data are more precise. Some of those classified as landless in 1970 (a crucial step in the sampling process) turned out to be only temporarily landless. If these cases had been viewed retrospectively decades later; the landless interregnum might have slipped through the mesh. One case illustrates a more fundamental process missed by the retrospective data. This was a Brahman, who retired from a clerical job at the sugar factory in 1978 and, because he had no sons, sold off his remaining land and moved to Pune. He had entered the sample, of course, only because he was still farming in 1970. If I had started the research in 1980, I would have missed him. Similar but earlier land-losers-those who left the village before 1970-may have gone undetected by the sampling procedure. It is reassuring, then, to find only one such case during the 1970s. This suggests that the rate of disappearance was not so high as to vitiate the patterns found among those who remained. On the other side of the ledger, there were extra land-gainers in the 1970s, due to a new round of land-reform implementation which provided small bits of dry land to three landless families in the sample. Bajaba, a Maratha laborer, came to Malegaon before 1920 and died when his three sons were quite young. The sons remained landless laborers throughout their lives; but one grandson, Shankarrao, after a long life as a member of a gul-making team and then as an annual contract laborer, acquired two acres of dry land under the land reform scheme in 1975. [36]

State land-ceiling legislation was enacted in the early 1960s, and again in the 1970s. Most large farmers evaded the law by dividing ownership of their holdings among several family members. However, some land was acquired by the government and gradually distributed to landless families. Shankarrao was one beneficiary, perhaps because he had a son with a highschool diploma, who was able to handle the paperwork. After completing his education, the son became a gram sevakJ or village development worker, in a dry village about 15 miles from Malegaon. In the long run, this job may be more significant than the two acres, since the income is more secure. This case shows the importance of educational opportunities, which expanded in the Malegaon area thanks to the pioneering efforts of men like Raja Shambhusingh, plus the subsequent work of the Shikshan Seva Mandal and the sugar factory. Recently, even some of the poorest families have used education as a means of mobility. Maloji was the son of a migrant laborer, a Mali who came to Malegaon around 1920 [as mentioned in Chapter 6). When Maloji grew up, he worked as an annual laborer and member of a contract team. He and his wife also sharecropped on a one-fifth basis. In 1970, Maloji said it had been possible to educate his children because of the sharecropping arrangement, which gave them enough

Performance and Impact of a Sugar Co-op

239

food to eat without the sons having to work as hired laborers. By 1979 he had sent two sons to college, and one was employed as a city clerk in Pune. Another son managed to get 1.75 acres of dry land in 1975, as part of the land reform scheme. [34)

It would be noteworthy in any society, perhaps, for the illiterate son of a migrant laborer to send his sons to college. The expansion of educational opportunities has made a real difference to families like Maloji's. Balaku is a Mang whose father and grandfather were landless laborers in Malegaon. In the late 1920s, Balaku completed primary education and a training course, becoming a primary-school teacher. In the 1960s, near retirement, he bought seven acres of dry land and joined a cooperative lift irrigation society. (Three of his five sons were still living at home then.) By 1972 he had four acres under seasonal irrigation; but a couple of years later he stopped cultivating the land, since the soil was poor, the water supply was inadequate, and four of his sons had moved away. (His youngest son, with a year of college, got a job as a primary-school teacher in another part of the district.) [1]

Here we find a low-caste family which has moved up and down at the border of landlessness. Except for one son, the others moved away from Malegaon for nonagricultural employment, following the strategy of many low-caste men. By 1980 Balaku was impoverished, since the family cycle (as in many landless or near-landless families) led to a dispersion rather than concentration of human capital. Even so, education was of prime importance both to Balaku and his sons. Occupational diversity, like education, also provided an expanding array of opportunities for the poor. Small farmers used diverse sources of nonagricultural income for investment in agriculture-often with more favorable results than in Balaku's case: For example, Malhari and his two brothers got jobs in the sugar factory, starting in 1956. Their family, which was Neo-Buddhist, 13 already owned 5.7 acres under seasonal canal irrigation. In the 1970s, thanks to the savings accumulated from their jobs, they bought three acres of dry land and added well irrigation to four acres of the canal land, enabling them to grow sugarcane for the first time. The cost of these capital investments was about Rs.25,500. [68) Vasantrao, another Neo-Buddhist, became a government clerk and then a highschool teacher after completing a B.A. and B.Ed. (His younger brother became a primary-school teacher.) Starting in the late 1950s, then, Vasantrao and his father were able to enlarge their small holding and add well irrigation-allowing them to grow sugarcane. [61) Another case was Buwasaheb, a Dhangar whose grandfather migrated to Malegaon as a laborer in 1935. Buwasaheb's father rented seven acres from one of the Jagirdar's sons and in the 1960s managed to purchase three acres with the help of tenancy reform legislation. Meanwhile, Buwasaheb got a job as a seasonal laborer in the sugar factory. With the help of his wages, the family bought

240

Performance and Impact of a Sugar Co-op

more land and added lift and well irrigation, giving them a holding of 9.75 perennially irrigated acres by 1976. Their total investment in land and irrigation, over two decades, was about Rs.55,000. [29]

This case parallels the upward mobility of the Dhangars discussed in Chapter 6. If it was less spectacular, this was probably because there has been no crisis like the great depression to shake up the land market since 1950. In general, as we shall see, small farmers and landless laborers increased their average landholdings more than middle and large farmers. Due to the tight land market, the total acreage which they gained was not huge; but such gains as the poorer families did make were helped considerably by the diversification of other employment opportunities. Changes in Relations of Production

Before discussing recent changes in land and income distribution, let us consider how the organization of farm labor has evolved since the period before independence. Perhaps the biggest change since 1950 is that, if small farmers have irrigation, they are now consistently sugarcane growers. Thus, as compared with the 1920s, they devote more labor to their own fields and probably work less for other farmers. Even so, marginal farmers (those with less than one acre) still seek employment as agricultural laborers. But even here, there has been change, since there are now other options. Many small and marginal farmers have sons or brothers with education and formal-sector jobs. (This will be discussed further when we turn to the distribution of nonagricultural income.) Thus small farmers are more viable, thanks to the secure market for cane provided by the factory, and also thanks to the variety of new employment opportunities. Consequently, small farmers and landless laborers are perhaps more distinct from each other than they were two or three generations ago. The corps of immigrant laborers has swollen apace since 1957; and as we shall see, the land market has been much less volatile than it was between 1920 and 1950. The result is that immigrant laborers have fewer opportunities to become farmers. Instead, they seek upward mobility through education and formal-sector employment, as the farmers do. Middle and small farmers still depend more on family than on hired labor. The amount of labor hired depends as much on the number of household members working in agriculture as on the size of the farm. On small farms, annual contract laborers are rare. (The number of saldars tends to be directly proportional to the size of holding.) Middle and small farmers do employ contract teams for specialized operations, plus casual workers on an occasional basis. As in the 1920s, large farmers depend more on hired than on family labor. Some hire permanent foremen, and some even employ managers. Large farmers devote most of their time to supervision, to negotiating loans and

Performance and Impact of a Sugar Co-op

241

equipment purchases, to experimenting with new crops and technologies, to setting up and running ancillary businesses, and sometimes to politicking. Many of their sons now have college educations, which are utilized in farm and business management or as passports into the urban middle class. As earlier, there are a variety of labor relations used to produce sugarcane and other crops. Large farmers employ saldR.rs on annual contracts. Saldars get the security of a fixed monthly wage, plus a variety of extras, such as food, housing, and clothing once a year at the diwali festival. (These extras are all subject to negotiation.) Annual laborers and their employers sometimes stay together for long periods, though the arrangement can be terminated at year's end by either party. Laborers sometimes ask for, and receive, a year's wages in advance-in order to get married, for example. They do all types of farm work but particularly tasks, such as irrigation or tending bullocks, which must be repeated on a daily or weekly basis. On some big farms, there may be five, ten, or even more annual laborers, and different ones will specialize in different tasks. Annual laborers commonly sharecrop certain plots belonging to their employers. As explained in Chapter 6, much of the labor is provided by the sharecropper's wife. The employer tills the soil, sows the crop, and then turns the plot over to the sharecropper, who will weed it, water it, guard it as it ripens, and then harvest it. For this contribution, the sharecropper usually gets one-fifth of the harvest. Since female laborers are hired only intermittently for cash wages, sharecropping provides a way for women in landless households to increase their income. Only certain crops are cultivated on shares: wheat, sorghum, millet, sometimes cotton, but never sugarcane. Cane is too expensive, too labor-and capital-intensive, to be managed this way. Large farmers use several other types of labor arrangement as well. Women (including the wives and daughters of their annual laborers) are hired in teams for certain operations-particularly weeding and harvesting. Men are also hired as casual, daily laborers (rojandars) whenever needed. Large farmers also employ contract labor teams (toli), which are organized by the laborers themselves. Only young men with skill in certain heavy operations (such as planting cane and earthing it up) can work in these teams, since they agree to a fixed payment per acre. The sooner they complete one job, the sooner they can start another, so farmers and laborers both benefit from payment by the acre, instead of by the day. Each team is led by a mukkadam, who arranges the contract and collects the payment. The mukkada.m does the same work as the other team members and collects the same share of the payment. His benefit, if he is a good organizer, is that he is more sure of getting employment by arranging it himself. (Other members may not always be included when the team is hired on smaller jobs.) Farmers no longer hire migrant contract teams to harvest their cane and process it into gul, since this part of the production process is now organized by the factory. Likewise, equipment repair and other ancillary services are

242

Performance and Impact of a Sugar Co-op

provided by specialists in return for cash payments. Blacksmiths, carpenters and cobblers no longer provide baluta services on a regular basis to most farmers. Those who repair motorcycles, bicycles, jeeps, tractors, trucks, irrigation pumps, oil engines, and electric motors come from a variety of castes and charge what the market will bear. Women contribute to crop production in the ways just mentioned: that is, they work as sharecroppers, usually on a one-fifth share, and they also work as casual laborers. In addition, if their families rent or own land, they help farm it. They particularly help with weeding, guarding and harvesting the crops. (Most work with bullock teams or tractors is done by men.) In general, the economic security of a family is reflected in how its women work. On large farms, women as well as men are hired as laborers, but the farmer's own wives and daughters do not work in the fields. On mediumsize farms, the farmer and his sons do much of the manual work themselves, and their wives and daughters often work alongside them. On smallholdings, the women nearly always work on their own farms. When the farm is less than an acre or so, both men and women also work as hired laborers. However, if a family has a slightly larger holding, the women may not work for other farmers, though the men may do so. In contrast with other regions of India, family honor is not demeaned if women work in their own fields. Whether they do so or not is mainly an economic decision, reflecting whether a family can afford to hire others to do the heavy work. However, on small farms, when women might need to seek employment from others, questions of family honor do come into play. If they can afford to make the distinction, families will send only their men, not their women, to work for others. In other words, there are many smallfarm families, in which the women could earn extra cash but do not, while their men do. This double standard seems designed to protect family honorto keep the women out of harm's way-if the family can subsist without their extra wages. So far as I know, there have been no major changes in the sexual division of agricultural labor since the 1920s; but I have to admit that I did not pursue this question systematically in the family-history interviews. At the time of my field work, there were no women employed in the cooperative sugar factory. A small number of educated local women are employed in the schools. More and more daughters are being sent to high school and college, but women's education still lags behind men's. Young women (like men) invariably expect to be married, though the average age of marriage appears to be increasing as women's education is increasing. In nearly every case, village parents still select their children's marriage partners, though the son or daughter may nowadays be given the option of meeting a potential partner and exercising a veto. Maharashtra resembles south India, in that women have more personal autonomy and respect than in the north (see Miller 1981; Dyson and Moore 1983; Maclachlan 1983; Jeffery, Jeffery and Lyon 1988). Nevertheless, women and men maintain largely separate spheres of activity and social interaction.

Performance and Impact of a Sugar Co-op

243

If not working in the fields, women will generally be found at home. Women do not expect to be routinely insulted or harassed when out in public, not even in cities like Pune or Bombay, but it is considered seemlier for them to work at home or to associate in public with other women.

Changes in Land Distribution In this section, we consider the overall changes in land distribution from 1950 to 1980. As compared with Chapter 7, we are using an expanded sample of 121 families, including 13 who immigrated after 1950. (Eight immigrants who settled as factory workers in the factory colony are excluded.) 14 In general, we shall find that, compared with the period from 1920 to 1950, the land market was less active' and there were fewer opportunities for dramatic upward mobility. As a result, there was no great change in the pattern of inequality. This can be seen as both good and bad. On the one hand, it is much better than the outcome predicted by the standard theory, which holds that small farmers will become impoverished and lose their lands under commercialization. On the other hand, the inequalities which persist from before the rise of the sugar economy are softened only slightly, at least in terms of land distribution. The story does not end there, however, for we shall see that nonagricultural sources of income do help to weaken these longstanding inequalities. Figure 9.1 shows the relationship between landholdings in 1950 and those in 1980. As in Figure 7.1, points on the line of stasis indicate families with no net gain or loss of land. Those above the line show cases of upward mobility, while points below indicate the opposite. We see, first of all, a fair amount of scattering. The correlation (R) is 0.76, and R squared is 0.58, meaning that 58 percent of the variance in 1980 can be explained by the 1950 values. Conversely, 42 percent of the variance is due to changes which disrupted continuity with the past. As compared with the period before 1950, however, there is far more continuity (see Tables 7.1 and 7.4). This reflects the absence of a great depression and the stability provided to commercial farming by the sugar factory. The figure also shows who benefited from these changes. Of the 12 largest landholders (those with more than 30 acres in 1950), two moved up, three stayed the same, and seven moved down. In the next-largest category (between 15 and 30 acres), seven moved up, none stayed the same, and six moved down-including one who became landless. Among middle and small landowners (those with less than 15 acres), 15 moved up and 20 down, while 18 stayed the same. Among those who were landless, 12 moved up and 31 stayed landless. As in the period before 1950, then, there were many exceptions to the predicted process of class polarization. What was the pattern overall? The regression line for Figure 9.1 has a slope of 0.61. As mentioned in Chapter 7, a slope of less than 1.0 indicates a decline in inequality. This can be tested by summing the cases on Lorenz curves, as in Figure 9.2. Here we see that the 1980 curve has risen slightly

244

Performance and Impact of a Sugar Co-op

0

co

0\

0

20

40

60

80

100

Landholdings in 1950 (Acres) FIGURE 9.1

Scatterplot of Malegaon Sample Landholdings, 1980 vs. 1950

closer to the diagonal, meaning that inequality was reduced. In 1950, farmers with more than 15 acres owned 82 percent of the total land, while in 1980 their share had declined to 75 percent. Table 9.11 shows correlations between landholdings in 1950 and those in later decades, along with the slopes of the regression lines. The latter show a trend toward less inequality, whether the sample is weighted or unweighted. Table 9.12 pinpoints the kinds of changes which created this mobility. As in the period before 1950, the fate of the large farmers is crucial. Here we find that they bought and sold about the same amount of land, obtaining no overall advantage from the land market (contrary to standard assumptions). However, the other size categories came out ahead on purchases and sales. One process which gave the large farmers an advantage was adding irrigation, due to heavy investments in new wells and pumpsets. However, these gains were more than offset by losses due to partitions.

245

100

ic ~

ij

Solid line - 1980 Broken line - 1950

80

..:I

""'0 Q)

60

1:11

"'c

+J

Q)

u

I.. Q)

a.

40

Q)

>

•1"'1

,"'

+J

......

20

§

u

0

20

40

60

80

100

Cumulative Percentage of Households FIGURE 9.2 Land Distribution in Malegaon Sample, 1950 and 1980 (Lorenz Curves)

TABLE 9.11

Later Decade 1960 1970 1980

Correlation of Landholdings in 1950 with Landholdings in Later Decades

R

R2

Slope

Weighted Slope

0.86 0.81 0.76

0. 75 0.65 0.58

0.79 0.71 0.61

(0.80) (0.73) (0.64)

(All Rs are significant at 0.01.) Sample survey of Malegaon, excluding factory colony (N- 121).

~:

246

TABLE 9.12

Changes in Land Distribution, Malegaon Sample, 1950-80

12~Q ~ndholdings

Sum 0 35.66 171.82 956.84

Mean 0 1. 27 6.87 38.27

1164.32

9":62

:i1;e !;;ltegQr:£ Landless Small Farmers (15A) Total

LL SF MF LF

Purchases 1950-80 Sum Mean W, Mean 20.05 0.47 (0.17) 21.35 0.76 (0.55) 25.82 1.03 (0.76) 113.00 4.52 (4.52)

Total

180.22

LL SF MF LF

Added Irrigation 1950-80 Sum Mean W, Mean 37.96 0.88 (0. 27) 0.84 (0.48) 23.55 50.15 2.01 (1. 54) 200.77 8.03 (8.03)

Total

312.43

---r:-49

2Ts

(0.76)

( 1. 25)

LL SF MF LF

Net Changes 1950-80• Sum Mean w, HeAn (0.42) 52.96 1.23 12.62 0.45 (-0.17) -26.81 -1.07 (-2.52) -218.43 -8.74 (-8.74)

Total

-179.66

-1.48

(-1. 00)

(area§ 1n Wtd. N H 43 (141) 28 (53) 25 (34) 25 (25)

121

(253)

I~I~§)

Weighted M~!D

(0) (1. 33) (6.80) (38.27) (4.98)

Sales Sum -6.00 -10.13 -12.59 -118.75

1950-80 Mean W, Mean (-0.04) -0.14 -0.36 (-0.46) -0.50 (-0.49) (-4.75) -4.75

-147.47

-1.22

(-0.66)

Partitions 1950-80 Sum Mean W, Mean -5.00 -0.12 (-0.04) (-0.87) -28.68 -1.02 -101.53 -4.06 (-4.88) -401.01 -16.04 (-16.04) -536.22

-4.43

(-2.45)

1980 Distributionb He§n W, Mean (0.42) 52.96 1. 23 (1.16) 48.29 1. 72 (4.29) 145.01 5.80 738.45 29.54 (29.54)

:i!.!l!!

984.71

--s.I4

(3.98)

• Including other transactions not shown: gifts, reduced irrigation, acquisitions under the tenancy act, etc. b 1980 distribution broken down by 1950 size categories. SQurce: Sample survey, excluding factory colony.

Performance and Impact of a Sugar Co-op

247

Small farmers gained a bit more, and middle farmers less, than in the period before 1950. The net gains of the landless were smaller than before. Even so, net gains by the landless and small farmers exceeded the overall average. The analysis with weighted cases yields the same basic pattern, except that middle farmers did worse than the overall average. In general, the period after 1950 shows greater stability than the earlier period because there was no depression and the cooperative factory reduced the risks inherent in cash cropping. The normal assumption is that risk reduction is most beneficial to small farmers, and this is probably true. We saw evidence in the earlier period that small farmers were sticking to subsistence crops, rather than incuring the risks of cane cultivation. After 1950, virtually all those with irrigable land became factory shareholders and started growing cane. As for the large farmers, they too were protected against risk, and we find them selling off less land than they did before 1950. The result was a less volatile land market and fewer opportunities for the upwardly mobile to acquire large holdings. What about changes in land distribution by caste? Table 9.13 shows the cross-tabulation of land-size with caste categories for 1950 and 1980. In contrast with Table 7.6, there is not much change overall: indeed, the measure of association (Spearman's Rho) changes only slightly. The continued success of the middle peasant castes is not surprising, but the continuity of landed wealth among the high castes is. Despite their political vulnerability, the high castes have held their own at the very top of the economic scale. Their farms are often supported by good sources of nonagricultural income, such as technocratic jobs and family businesses. We shall consider this pattern later, in terms of income distribution. Table 9.14 gives the aggregate changes from 1950 to 1980, broken down by caste and lineage categories. On average, other peasant castes and major Marathas gained the most from purchases, while high castes lost the most from sales. The net outcome of purchases and sales was most favorable to the other peasant castes and least favorable to the high castes. On the other hand, high castes added more irrigated area, on average, than the other castes. Taking all transactions into account, net changes were unfavorable to high castes and Marathas. Only the low castes actually increased their holdings, and that by a tiny amount. However, simply by not losing land, they raise further doubts about the assumption that the poor inevitably get poorer. The pattern with weighted cases is similar. What about the association between ungrouped land-size data and caste? Table 9.15 presents the analysis of variance statistics. The F ratio (weighted and unweighted) increases moderately after 1950, though it was falling sharply before that date. The first impression, then, is that caste and land were becoming more strongly associated. However, it is interesting that the R coefficients do not increase. Since R measures linear association while F does not, the increase in F may not imply a stronger linear relationship. In

248

Performance and Impact of a Sugar Co-op

TABLE 9.13 Cross-Tabulation of Caste & Landownership Categories, 1950 & 1980

Caste Category Size Category (acres)

Low

Other Peasant

Minor Maratha

12 2 3 1 2 2

10 3 2 1 2 3 3

12 2 1 5 5

22 (56)

24 (55)

28 (62)

9 4 1 2 2 3 3

8 7 5

HidLow

Major Maratha

High

Row Total

Wtd. Row Total

A. 1930 Landless 0.01-1.0 1.01-2.5 2.51-5.0 5.01-15 15.01-30 >30

8

2 6 1

Column Ttl. 17 Wtd. Total (45)

3

1

43 14 14

2 3 2

14 13 12

22 (28)

8 (8)

121

2 2 3 2 7 1 5

1

39 18 14 9

5 2 3 3 5 4

11

(141) (22) (31) (17) (17) (13) (12) (253) 0

B. 1950 Landless 0.01-1.0 1.01-2.5 2.51-5.0 5.01-15 15.01-30 >30

8 4 2 2 1

Column Ttl. 17 Wtd. Total (45)

11

1 3 3 4

22 (56)

24 (55)

4 2 2 28 (62)

22 (28)

3 2 2 8 (8)

21

8 12 121

(143) (40) (20) (9) (21) (8) (12) (253) 0

(For 1950, Spearman's Rho- 0.43 [unweighted], significant at 0.01.) (For 1980, Spearman's Rho- 0.41 [unweighted], significant at 0.01.) • Weighted Ns are rounded to the nearest integer; hence the total N is slightly less than the sum of the partial Ns. ~: Sample survey in Halegaon, excluding factory colony.

other words, the association of caste with size of landholding may have become stronger, but not to the benefit of the higher castes. 15 Another approach to the question is to test the relationship between caste and net changes in landholdings from 1950 to 1980. The results are simple: F = 0.24, which is not significant at any level. In other words, there is no association, linear or otherwise, between caste category and net change in landholdings, 1950-1980. In short, the long-term shifts in land distribution after 1950 run counter to the standard model of change under rapid commercialization. The large farmers end up with a smaller share of the land, as do the high castes and Marathas. Small farmers, landless and low-caste families are relatively better off than they were, though their increases are hardly impressive. Overall, it is safe to say that the poor have not lost their lands due to increased

249 TABLE 9.14

Changes in Land Distribution by Caste Categories, 1950-80

(![!i§ in I~[!§) Wtd. Weighted MilD H H (45} (0.66} 17 (1. 73} (56} 22 (5.38} (55} 24 (4.19} (62} 28 (14.19} (28} 22 (23.50} (8} 8

l2~Q ~D~2ldiDK§ ~IIU ~J:2W!

Low Castes Mid-Low Castes Other Peasant Marathas: Minor Major High Castes Total

Low Mid-Low Other P. Minor M. Major M. High

15.39 75.50 277.08 225.19 383.15 188.01

Hun 0.91 3.43 11.55 8.04 17.42 23.50

1164.32

9.62

~!.!!I!

Purchases 1950-80 M!an W, Me!n 0.36 6.06 (0.29} 1.22 26.91 (0.50} 2.91 69.75 (1. 27} 24.00 (0.46} 0.86 47.53 2.16 (1. 72} 0. 75 5.97 (0.75}

liYill

1':49

Total

180.22

Low Mid-Low Other P. Minor M. Major M. High

Added Irrigation 1950-80 Mean w, Hun ~!.!!I! 0.43 7.33 (0.16} 0.90 19.71 (0.39} 46.54 1.94 (0.85} 34.56 1.23 (0.60} 129.60 5.89 (4.70} 74.69 9.34 (9.34}

Total

312.43

Low Mid-Low Other P. Minor M. Major M. High

Net Changes 1950-80b Mean W, Mean 2.53 (-Q.05} 0.15 -29.52 -1.34 (-0.84} -0.04 -1.01 (-0.32} -82.51 -2.95 (-1. 78} -47.07 -2.14 (-2.01} -22.08 -2.76 (-2.76}

Total

2'Ts

(0.76}

(1.25}

~urn

-179.66

-1.48

(-1.00}

121

(253} 8

(4.98}

Sales 1950-80 Hun w, Mein -4.76 (-0.31} -0.28 -32.00 (-0.58} -1.45 (-0.23} -0.30 -7.12 (-0.91} -1.85 -51.75 (-0.75} -0.94 -20.75 -31.09 (-3.89} -3.89

liYill

-147.47

-1.22

(-0.66}

Partitions 1950-80 Hun w, M!!n -6.05 (-0.28} -0.36 -56.02 (-1.30} -2.55 -111.15 -4.63 (-2.26} -102.21 -3.65 (-2.27} -191.46 (-7.24} -8.70 -69.33 -8.67 (-8.67} ~um

-536.22

-4.43

(-2.45}

1980 Distribution Sum Hun W, Milan 17.92 (0.61} 1.05 46.01 2.09 (0.89} 276.07 11.50 (5.07} 142.69 5.10 (2.41} 336.09 15.28 (12.18} 165.93 20.74 (20.74} 984.71

8':14

(3.98}

• Weighted Ns are rounded to the nearest integer; hence the total N is slightly less than the sum of the partial Ns. b Including other transactions not shown: gifts, reduced irrigation, acquisitions under the tenancy act, etc. Source: Sample survey in Malegaon, excluding factory colony.

250

Performance and Impact of a Sugar Co-op TABLE 9.15

Analysis of Variance for Landholdings,Broken Down by Caste Categories

Correlations

Year

F Ratio

Degs. of Freedom•

R

Eta

Weighted F Ratio

1950 1960 1970 1980

3.30 3.51 4.75 4. 77

115 115 115 115

0.31 0.32 0.33 0.32

0.35 0.36 0.41 0.42

(8.06) (8.40) (11.10) (10.45)

(All F ratios are significant at 0.01.) • Degrees of freedom within caste categories. (N - 121) Source: Sample survey in Malegaon, excluding factory colony.

commercialization, nor have they gained dramatically. As regards land distribution, the stratification system seems to have stabilized in the wake of the great depression. Nonagricultural Incomes However, the preceding section gives a limited assessment of the relationship between economic change and poverty. It ignores two fundamental changes which have had positive effects on the poor. One is increased agricultural production, which, as we have already seen, generated a huge increase in employment and immigration. The other is occupational diversification, which opened up more sources of nonagricultural employment. In my surveys, I did not attempt to quantify family incomes before 1970, because there were too many problems of methodology and conceptualization. However, I did record estimates of nonagricultural income in 1969-71 and again in 1978-79. Because data from the latter survey are more complete, they are used here. Generally speaking, there are two kinds of nonagricultural income: income from wages earned in enterprises owned by others, and income from familyowned enterprises. The first I refer to as "job income" and the second as "business income." Job income derives from employment in the formal sector-that is, in sugar factories, government offices, private business enterprises, transport companies, and so on. People in Maharashtra refer to this type of occupation as nokri (in Marathi) or as employment "in service" (in English). A striking feature of the regional culture is that everybody, whether low-caste laborer, rich cane grower, or sophisticated urbanite, wants some of his or her children to be employed "in service." All believe that greater security and opportunity are found in the formal sector, particularly in government service. This reflects, of course, the massive growth of bureaucracy since independence.

Performance and Impact of a Sugar Co-op TABLE 9.16

251

Distribution of Business Income by Land-Size Categories, 1980

Mean 1980 Business Income (Rs.)

W'td. Mean Business Income (Rs.)

Households with Business Income

Rate of Participat ion•

Landless Small Farmers Middle Farmers Large Farmers

16 8 7 8

37\ 29\ 28\ 32\

1950 2175 3314 7325

(1950) (1732) (3314) (7325)

Total

39

32\

3344

(2492)

Land-Size Category

• The number of households with business income, divided by the total number of sample households in each land-size category. ~: Sample survey, excluding factory colony (N- 121).

My survey data on job incomes are probably quite accurate, since everybody is aware of how much income derives from local job categories. In other words, respondents could not be too inventive on this subject. Business income, on the other hand, derives from family-owned enterprises such as running a cloth or tea shop, peddling milk, eggs, or vegetables, selling petrol, renting bicycles or repairing motorcycles. My data represent estimates of annual net income (based on costs and turnover) for all the business activities in each household. (Quite a few households had more than one business, depending on the number of adults and the amount of land which they also managed.) The estimates of net income are necessarily quite rough. One can see at a glance the contrast between a flourishing shopkeeper and a marginal hawker, and the income data reflect these gross differences. Beyond that, they are not very precise. The distribution of business income by size of landholding is given in Table 9.16. (This table includes only households that were earning business income in 1980.) The first point to note is who was participating in the world of business. Sixteen out of 43 landless households (or 37 percent} were engaged in business activities-which is hardly surprising, given that their main alternative was agricultural labor. (Of course, many families combined both sources of income.) Small, middle and large farmers participated at somewhat lower rates, from 28 to 32 percent. Business investments varied roughly according to assets in land: some landless people hawked eggs and vegetables, while one large farmer bought a truck and became a transport contractor, another built a movie house, and another established a petrol station. Hence the mean business incomes of the different land-size categories show a linear progression, whether weighted or unweighted. This progression is confirmed by the correlation of ungrouped land-size data with business income: R = 0.47 (weighted) or 0.46 (unweighted}, significant at one percent. In other words, there is a fairly strong association between landholding

252 TABLE 9.17

Performance and Impact of a Sugar Co-op Distribution of Business Income by Caste Categories, 1980

Caste & Lineage Category

Households with Business Income

Rate of Participat ion•

Low Mid-Low Other Peasant Minor Maratha Major Maratha High

7 12 6 6 4 4

41% 55\ 25\ 21\ 18\ 50\

1157 2850 4600 2900 4725 6050

{906) {2190) {3705) {2705) {4725) {6050)

Total

39

32\

3344

{2492)

Mean 1980 Business Income {Rs.)

\ltd. Mean Business Income {Rs.)

• The number of households with business income, divided by the total number of sample households in each caste category. ~: Sample survey, excluding factory colony {N • 121).

and business income, though there is plenty of variation which is not related to land size. It is interesting to consider the distribution of business income by caste category, as in Table 9.17. Here we find a three-tiered pattern of participation: the low and mid-low castes have high rates of participation in business enterprises, the middle peasant castes have much lower rates, and the high castes are high. This pattern derives partly from skills and specializations established long before the sugar economy. The middle castes tend to devote their efforts to the land, while the high and low castes use a variety of skills in other occupations. However, there is no clear association between caste category and business income: F = 1.04 (unweighted), which is not significant at any level. 16 In other words, there is so much variation within caste categories that they overlap, and the differences between their means are not very significant. Nevertheless, there seem to be exceptions at the extremes. The low castes have a distinctly lower mean and the high castes a higher one. Turning to job incomes in 1980, the distribution according to land-size categories is given in Table 9.18. Participation rates are high for small and middle farmers, much lower for large farmers and landless families. The latter are at a disadvantage because they are not shareholders in the sugar factory. On the other hand, low participation by large farmers is presumably a result of their own preference. In any case, we find that small and middle farmers are the main beneficiaries in terms of job participation. In terms of mean job income, we find a two-tiered distribution, with the middle and large farmers receiving much higher incomes than the landless and small farmers. If one size category stands out in terms of both participation and income, it is the middle farmers. However, the differences between the means are not very significant, since the correlation of job income with

253

Performance and Impact of a Sugar Co-op TABLE 9.18

Distribution of Job Income by Land-Size Categories, 1980

Land-Size Category

Households with Job Income

Rate of Participat ion•

Mean 1980 Job Income (Rs.)

Wtd. Mean Job Income (Rs.)

Landless Small Farmer ~Iiddle Farmer Large Farmer

15 14 7

11

26\ 54\ 56\ 28t

3800 3600 7614 6971

(3800) (4510) (7614) (6971)

Total

47

39t

5345

(4899)

• The number of households with job income, divided by the total number of sample households in each land-size category. ~: Sample survey, excluding factory colony (N- 121).

ungrouped landholding data is 0.23 (unweighted) or 0.25 (weighted}, and

only the latter is significant at five percent. The distribution of job income is decidedly less skewed than the distribution of land or business income. In 1980, the top 20 percent of landowners (excluding the landless) owned 68 percent of the total land in the sample; and the top 20 percent of families active in business earned 60 percent of the total business income. On the other hand, the top 20 percent of families with jobs earned only 45 percent of the total job income. As elsewhere, the distribution of income is less skewed than the distribution of wealth. Since agricultural and business incomes are based partly on the ownership of physical assets, incomes from these sources tend to be more skewed than those from jobs. Turning to the distribution of job incomes in terms of caste (Table 9.19), we find a remarkable pattern, in which all categories except one are roughly the same. The exception is the highest category, which has an average income more than twice the overall mean, coupled with a high rate of participation. High castes stand out, of course, due to their longstanding tradition of literacy and specialization in administrative occupations. Equally remarkable is the similarity of the other categories to each otherthough the lowest castes have the highest mean among them. Their rates of participation are also similar. If the high castes were removed from this table, the F ratio would be less than one (not significant at any level). That is, there would be no relationship at all between job income and caste category. In the table as it stands, the F ratio is 2.49, significant at 5 percent. However, in the analysis of variance, the deviation from linearity is stronger and more significant than the linearity. (This is even more true for the weighted sample.) This means that job income does not tend to increase with caste status overall. If we removed the high castes from this table, the low castes would be outstanding in terms of their average job income, which is about 1.5 times

Performance and Impact of a Sugar Co-op

254 TABLE 9.19

Distribution of Job Income by Caste Categories, 1980

Caste & Lineage Category

Households with Job Income

Rate of Participat ion•

Mean 1980 Job Income (Rs.)

Wtd. Mean Job Income (Rs.)

Low Mid-Low Other Peasant Minor Maratha Major Maratha High

6 8 8 11 9 5

35% 36% 33% 39% 41% 63%

6200 4025 3400 4864 4889 11420

(6344) (4098) (3276) (4660) (4583) (11420)

Total

47

39%

5345

(4899)

• The number of households with job income, divided by the total number of sample households in each caste category. Source: Sample survey, excluding factory colony (N- 121).

higher than the average for all the others. This inverted disparity is probably due to two factors. First, the low castes have stressed education as the path toward individual mobility and group emancipation, and education is the main qualification for employment in the formal sector. Second, the low castes have benefited from quotas in public-sector employment. In 1980, four low-caste sample households had members employed as primary and high-school teachers and in the state electricity board. Job and business incomes are only weakly related to caste, then. In this respect, the nonagricultural sector differs from the agricultural. (Recall that in Tables 9.13 and 9.15 we found moderately strong associations between caste and landholding.) In other words, as the economy grows and diversifies, caste exerts less influence over income distribution. This is clearly an advantage for the lower castes, who find paths to mobility (locally and in nearby cities) through education and jobs in the formal sector. These opportunities also have a positive effect on land distribution. Let us consider sample families earning job or business incomes, or both. (They number 73, unweighted.) An interesting pattern emerges when their 1970 landholdings are compared with those in 1980. Among those with business incomes, three out of 19 landless families acquired land by 1980; and four out of six small farmers acquired more land. Among those with job incomes, four out of IS landless families became landowners; and four out of 12 small farmers expanded. (A total of four middle farmers also moved up, but no big ones.) Only one small farmer with nonagricultural income lost part of his holding during this decade. In other words, these families were bucking a trend, since the average progression was downward rather than upward (see Figure 9.1 and Table 9.12). Their access to nonagricultural income was probably the key factor which helped them to acquire land. 17

Performance and Impact of a Sugar Co-op

255

Seven out of 17 families moving upward in this way belonged to the low and mid-low castes; the rest were from the peasant castes. The expansion of education and nonagricultural employment made a difference to the poor, even in terms of opportunities to acquire land in a tight market. Poverty, Inequality, and Economic Expansion Before the sugar economy, high-caste villagers obtained large landholdings • through political, bureaucratic, and religious mechanisms. Their control of land was weakened, however, by the growth of the commercial economy. High-caste owners tended to rent out their lands, so commercial farming skills were acquired by their middle-caste tenants; and the latter were sometimes able to buy out the former. 18 In 1930, the middle peasant castes had 69 percent of the sample land; but this rose to 77 percent by 1980, thus accentuating an old pattern of specialization in farming. These results are consistent with a broader-based study on the transfer of lands in 28 villages of Maharashtra and Gujarat during the period from 1956 to 1965 (Rao 1972). In most localities, more than 40 or 50 percent of the total area purchased was bought by smallholders, landless tenants, and other landless families. Likewise, the greatest portion of lands sold (from 58 to 80 percent in most talukas) belonged to the large landholders. Rao's study confirms that lands did not necessarily flow from the poor to the rich as villages became more commercialized, and that even the landless could acquire land. Since inherited status (in terms of caste, lineage, and ancestral landholdings) only partly determined a family's economic fate, we were forced to look for other factors to explain mobility. Skill and experience in commercial farming turned out to be crucial factors, and this conclusion is supported by the caste analysis. Over the whole period from 1920 to 1980, the middle peasant castes achieved the greatest upward mobility in terms of landholding. This was partly a result of their work ethic and cultural identity as tillers of the soil. These changes in land distribution in Malegaon suggest one reason why intellectuals tend to believe that the "sugar barons" have risen to power at the expense of others. (This is in addition to the fairly obvious reason that urban, educated, high-caste leaders have been pushed aside in state politics.) At least in this village, rising commercial entrepreneurs from the middle peasant castes did take land away from others: but the "others" were not small farmers. Instead, they were rich, high-caste landowners, who had the largest average holdings, by far, in 1930 and who also had the best access to job and business incomes, just as they do today. If Malegaon reflects a broader pattern throughout the region, then changes in economic power under commercialization have paralleled changes in political power, with the greatest losses incurred by the old high-caste elite. We should note that these were relative losses only, since the high castes continue to own the largest mean landholdings in 1980 and also have (as

256

Performance and Impact of a Sugar Co-op

they surely had before) the largest job and business incomes. Nevertheless, it is understandable that intellectuals, who mostly belong to the high castes, feel resentment concerning the rise of Maratha power. This is the resentment of an old and still-privileged elite against aggressive newcomers. What is particularly interesting in the Malegaon data is that the high castes retain their privileged economic position, even in terms of landholdings and even where Maratha political power is overwhelming. Everyday experience in Pune and Bombay cities suggests that the level of high-caste privilege in the urban economy is even greater. Of course, employment income is connected with individual achievement and thus conforms with some principles of an egalitarian society. However, English-medium schooling, leading to training in the best universities, is a tradition going back generations in many urban, high-caste families; and it is this cultural inheritance which perpetuates their privileged access to employment in the formal sector. This old pattern of privilege is seldom discussed by intellectuals, in contrast to their intense concern with land reform-that is, with redistributing the assets of their rivals. India has run through a wide array of policies, projects and proposals intended to create more equitable opportunities for the poor. However, little attention has been given to this entrenched pattern of high-caste privilege, which grows ever stronger under most plans for "socialist" (that is, bureaucratic) development. Little wonder that the highest aspiration for many villagers, including many "rich" farmers from the dominant lineages, is to obtain English-medium schooling for their children and move them into the urban middle class. If commercialization of the agricultural sector favors the middle peasant castes, and if bureaucratization of the formal sector favors the high castes, then what happens to the low and mid-low castes? In general, nonagricultural opportunities, particularly jobs in the formal sector, have been more open to the lower castes than opportunities to buy land. However, nonagricultural employment also reacts back on agricultural investment, helping some lowercaste families to buy land in a tight market. In general, the pattern of economic inequality has moved radically in neither direction: neither toward ever-increasing polarization nor toward a truly egalitarian society. The limited positive results are partly a consequence of migration: as opportunities expand in Malegaon, they are rapidly filled by migrants from the surrounding dry villages. This prevents the real wages of agricultural laborers from rising significantly, since there is an almost unlimited pool of labor outside the canal tracts. The solution to this problem is to expand the commercial economy by irrigating the dry villages-as the Government of Maharashtra is doing. Pervasive mobility, ignored in most accounts of rural India, has at least loosened the old pattern in which opportunity was strongly wedded to caste status. Opportunities have become more evenly distributed across caste categories, and the privileged status of high-caste landowners and hereditary Maratha elites has declined. Moreover, regardless of caste, there is now less land concentrated in large holdings than there was in 1930 or 1950.

Performance and Impact of a Sugar Co-op

257

This is an important conclusion, not simply because it conflicts with widely held assumptions. If it were true that economic disparities were becoming more polarized in the canal villages, then we would need to ask why, in such active centers of political competition, there is so little evidence of class-based political action. In the 1980s, there were large-scale agitations in western Maharashtra; but these were focused on agricultural prices, since farmers believed the government had allowed other prices (including those of inputs, such as fertilizers) to rise much faster than the prices of their crops. Farmers showed their ability to organize and agitate around this issue-an issue which clearly appeals to a broad spectrum of middle and small farmers, as well as big ones. Indeed, this was the first issue since independence to engage a mass rural constituency in the politics of agitation. From the farmers' point of view, then, the crucial class issues are those that divide them from the urban middle class. This would be puzzling if the distribution of wealth and opportunity were becoming ever more polarized within the villages. However if, instead of polarization, there is economic expansion and mobility (even though inequalities remain sharp), then incipient pressures toward rural class conflict will be drained off in opposition to urban interests, or in efforts to move up the ladder, or in both directions. The result will be competitive, pragmatic and non-ideological politics, such as one finds in rural Maharashtra over the last few decades. Discussions of rural poverty in the Third World tend to assume that wealth must be redistributed in order to rescue the poor. This may be true under some circumstances. However, particularly in the work of non-economists, the positive effects of economic expansion are generally ignored. Yet countries like South Korea, Taiwan, Hong Kong, Singapore, and perhaps a few others, have shown not only that they can generate high long-term per-capita growth rates but also reduce poverty to much lower levels than might have seemed possible twenty years ago (World Bank 1990). Thus the anti-poverty effects of the cooperative sugar factories should not be judged simply in terms of land redistribution. The cooperatives are not designed or suitable for promoting land reforms, and their indirect effects on land distribution are minor (though positive). What is important, instead, is that they have expanded the scale and complexity of the rural economy. This provides opportunities for all sorts of new skills and occupations to emerge; and these build on each other to promote newer enterprises, such as dairies and distilleries. Parts of Maharashtra, like parts of Gujarat and Punjab, may thus come to resemble the "Mini-Dragons" of East Asia, with favorable results for the employment opportunities and living standards of the rural poor. Notes 1. The data in this section come from pamphlets published in English by the factory (Malegaon 1969, 1977, 1986), from its annual reports in Marathi (e.g. Malegaon 1987-89), and from interviews with factory officers and shareholders.

258

Performance and Impact of a Sugar Co-op

2. However, this table cannot read as simply showing the distribution of members' landholdings. In the first place, the same farmers had other kinds of land as well: grazing land, dry crop land, and land under seasonal irrigation. In the second place, the family holdings of large farmers were divided among individual members, enabling them to comply with reforms which set a ceiling on the ownership of agricultural land. Thus many of the individual shareholders listed in this table are actually clustered in larger family units, making them bigger farmers than they appear. However, it may be assumed that most of the really big farmers (the ones who were threatened by land reforms) did not divide their farms into dozens of tiny holdings, corresponding to just one or two shares apiece. Thus the great majority of small shareholders on the first line of the table were, in fact, small farmers. 3. The state government now holds shares worth only Rs.100,000. It is prudent for the factory not to redeem all the government's shares. So long as the state holds shares, it can guarantee capital loans to the factory and regulate certain management decisions, particularly regarding cane prices (see Chapter 10). 4. When the weather happens to be good, there may be overproduction, driving down sugar prices. Low prices signal farmers to plant less cane or to economize on fertilizer and other inputs. But because the crop is in the ground for 12 to 18 months, there is a long lag before the effect of their decisions reaches the sugar market. When it does, there may be a shortage of sugar, which drives prices back up again. Because of the long lags between planting and marketing, there is a tendency for farmers to over- or under-shoot in response to changes in sugar prices. 5. In the early 1980s, sugar policy held steadier, with the government purchasing 65 percent of the output from each factory as levy sugar. The proportion was reduced to 55 percent in 1985-86, 50 percent in 1986-87, and 45 percent in 1988-89. Since levy-sugar prices are lower than open market prices, this reduction in the proportion of levy sugar benefited cane growers and encouraged production. 6. The seasonal capacity of a factory is calculated by multiplying its daily crushing capacity times the number of days in a standard season (which varies according to region). In Maharashtra, the standard season is 160 days. In Malegaon at present, 160 days x 2,500 tonnes per day = 400,000 tonnes per season. This amount is divided into the actual tonnage of cane crushed during a given season and multiplied by 100, giving the percentage of capacity utilized. 7. This table includes only the older factories in the area, those established before 1980. The newer ones face greater problems (due to high capital costs) in coping with bad weather, cane shortages, and price fluctuations. Including them would bias the comparison in favor of Malegaon. 8. I have drawn here on examples provided by B.S. Baviskar (in Baviskar 1980; Attwood and Baviskar 1987). 9. The cooperatives did not create this migrant harvest system. It was first started by the Saswad Malis and others, who recruited migrant teams for harvesting and makinggul (Patil 1927:10; 1928b; Knight 1905:9). In the 1920s and 1930s, it was used by the private factories (Deodhar 1950) and adapted to fit their requirements. 10. However, much of Breman's analysis, based on fieldwork in south Gujarat, does not apply to the situation in Maharashtra. In south Gujarat, most of the cane is transported by truck, possibly because there is a better road network and because one huge factory has an area of operation covering 170 villages. Hence the great majority of harvest workers are koytavale, those who harvest cane and load it into trucks. These migrants come from poorer backgrounds and sometimes lower castes than the gadivale (Breman 1978-79:51-61; Matson 1983:25). Thus the bargaining position of migrant workers in Gujarat may not be as strong as in Maharashtra, where the gadivale have led the struggle for higher wages.

Performance and Impact of a Sugar CrJ-op

259

ll. At least one cooperative factory (Shriram, near Malegaon) tried to rely on local contractors for harvest labor. (The intention was to direct more income opportunities toward shareholders acting as contractors.) However, during the rabi harvest of crops such as wheat and sorghum, local bullock carts and laborers were in heavy demand by the farmers themselves. Consequently, Shriram's cane supply was disrupted, at serious cost to the factory. The same problem has been encountered in northern India by cooperative factories relying on the local supply of labor and carts (Batra 1988, 1991). Thus, contrary to what Breman (1978-79) believes, migrant harvesters cannot be replaced by local ones without a serious loss in efficiency-not so long as their bullocks are needed. 12. Census data must be used with caution, however, since occupational categories and village boundaries have been changed from one census to the next. 13. In 1956, the Mahars ofMaharashtra, under the leadership of Dr. B.R. Ambedkar, converted to Buddhism. Unlike Hinduism, Buddhism does not legitimize caste hierarchy and untouchability. 14. The statistical analysis in Chapter 7 excluded two extreme outliers, cases [108] and [109]. In the post-1950 data, only one of these [108] remains an extreme outlier; the other is included in the calculations. 15. In the analysis of variance, the deviation from linearity, which was not significant in 1950, became significant at the five percent level by 1980. Compared with the unchanging R, the increase in Eta (another non-linear measure of association) also suggests that the association grew non-linearly. 16. The F ratio from the weighted sample is 4.02, significant at one percent. However, much of this apparent significance is an artifact of the weighting, since F tends to get stronger when N is increased. 17. Job and business incomes are reported here as of 1980. However, in most cases, families with jobs or businesses in 1980 also had these income sources in 1970. Thus it is reasonable to infer that nonagricultural incomes helped them to retain or increase their landholdings from 1970 to 1980. 18. There were some Brahmans, like Shembekar and Date, who got their land through commercial enterprise; but these were exceptions.

10 Why Do Some Cooperatives Work? Why have the cooperative sugar factories been so successful? Cooperatives have been promoted in many forms to solve the problems of rural development, and on the whole they have not worked very well. Are the cooperative sugar factories in Maharashtra really different, and if so, why? It is necessary to compare them with other kinds of sugar factories and other kinds of cooperatives in order to answer these questions. 1 Starting with just one factory in 1950, the cooperatives multiplied to 85 operating in 1988 (see Table 10.1). Production of industrial white sugar has risen equally rapidly: as Table 10.2 indicates, Maharashtra's output multiplied nearly 25 times in the period from 1950-51 to 1987-88, due to the increase in cooperative production. Sugar production in north India has grown at a much slower rate. In Uttar Pradesh, the leading state in sugar production at independence, the output from private factories hardly grew at all between 1960 and 1985, and the output from cooperatives grew very slowly. Maharashtra, which at independence produced only ten percent of the nation's white sugar, now produces more than 30 percent. Each cooperative factory is a major enterprise in itself. The annual turnover of a medium-size plant, crushing 2,500 tonnes of cane per day, is on the order of Rs.350 million, equivalent to about US $20 million. The cost of building such a plant today ranges from Rs.210 to 250 million (National Federation of Cooperative Sugar Factories 1988b). In this chapter, the cooperatives are compared with private sugar mills in northern India, to determine whether the former are really more efficient. We shall then consider a number of reasons why the factories in Maharashtra, unlike many cooperative enterprises, manage to remain cohesive and efficient. Regional Differences in Sugar Production

The sugar industry first arose on a large scale in Bihar and U.P., where cane was grown without great expense by village farmers. However, transaction costs in getting a regular supply to the mills were high, since each mill had to deal with thousands of widely scattered growers. In addition, the growers 260

261

Why Do Some Cooperatives Wtwk?

TABLE 10.1

Number of Sugar Factories in Operation, by State & Sector

State & Region Maharashtra

Sector

North India•

1970-71

1980-81

1987-88

1 13

14 27

30 41

67 77

85 90

c

0 68

2 71

4 71

17 91

30 104

c

0 100

3 103

ll 107

25 132

46 152

c

2 139

30 174

73 215

149 315

196 357

T T

All India

1960-61

c

T

Uttar Pradesh

1950-51

T

(C - Cooperative, T - Total) • North India- Bihar, U.P., Haryana & Punjab. Source: Cooperative Sugar Directory & Year Book 1986-88:500-01. TABLE 10.2

White (Vacuum-Pan) Sugar Production by Sugar Factories

State & Region Maharashtra Uttar Pradesh North India• All India

Sector

1950-51

1960-61

1970-71

1980-81

1987-88

c

4 ll3

264 523

773 1055

1877 2085

2691 2795

c

0 584

28 1427

74 1299

170 1224

654 2666

c

0 826

75 1934

137 1723

243 1572

946 3460

c

5 llOO

450 3021

1262 3740

2903 5148

5270 9ll0

T

T T T

(Production measured in thousand metric tonnes.) (C- Cooperative, t- Total.) • Bihar, U.P., Haryana & Punjab. Source: Cooperative Sugar Directory &Ye1r Book 1986-88:506-07.

looked to alternative markets whenever possible, selling their cane to gur (gut) or khandsari makers, or manufacturing gur themselves. One result of these chronic problems is that about half of India's total cane crop still goes to the production of artisan or open-pan sugars, especially in the north (see Table 10.3). Because gur-making units are small and numerous, their labor is not unionized and they need not pay minimum

262 TABLE 10.3

Why Do Some Cooperatives Work? Utilization of Sugarcane for Different Products, by State

1983-84

1984-85

1985-86

1986-87

1987-88

ws

GK

WS

GK

WS

GK

WS

GK

WS

GK

Maharashtra U.P. Bihar Haryana Punjab

67.4 23.4 49.3 31.9 26.8

20.6 64.8 24.9 56.3 61.4

79.0 21.8 40.7 28.0 27.2

11.5 64.4 45.1 59.5 60.3

89.9 23.6 66.9 33.0 29.7

0.6 62.6 18.9 54.5 57.8

90.2 32.0 na 43.3 39.5

0.3 54.2 na 44.2 48.0

na 32.1 na 52.3 36.3

na 54.1 na 35.2 50.2

All India

33.9

54.0

35.3

52.9

39.9

48.2

45.8

42.3

47.8

40.2

State

(WS - Percentage of (GK- Percentage of (Remainder used for Source: Cooperative

sugarcane used to produce white [vacuum-pan] sugar.) sugarcane used to produce~ and khandsari.) seed, chewing, etc.) Sugar Directory & Year Book 1985-86:616-17, 1986-88:496.

wages, nor minimum cane prices. Moreover, their product is sold on the open market, untaxed. They also benefit from schemes promoting smallscale industries. For these reasons, the sugar industry in north India is in a strange condition: from time to time, the factories are unable to compete with artisan units. This is particularly true during some phases of the sugar cycle (the oscillations in production and prices which occur every four or five years). In northern India, some private mill owners have responded by not investing in new equipment. Politicians have threatened to nationalize these "sick" units, and some have been taken over by the state governments of U.P. and Bihar (Franda 1979), while others have been auctioned off for tax arrears (Brass 1965:122). The result is a form of economic stagnation which may be described as a deadlocked class conflict, involving cane growers, mill owners, and the government. The roots of these problems are set in the awkward cane supply relationship, with its underlying conflict of interests (as described in Chapter 4). Meanwhile, the older sugar interests of northern India have been faced with the rise of vigorous competition from the south and west, especially from the new cooperatives. The latter are successful because they have resolved the problem of cane supply: the interests of factory and cane growers are no longer antagonistic. As a result, about 80 percent of the cane grown in Maharashtra is processed by sugar factories, as compared with about 25 percent in U.P. (see Table 10.3). Cooperative cane prices in Maharashtra are so attractive that few growers prefer to make gut, whereas the opposite is true in the north. The cooperatives in Maharashtra have done more than reduce the antagonism between themselves and their cane suppliers: they have established a harvest and transport system that contributes greatly to efficiency. Contract teams of migrant laborers, hired and directed by each factory, work on a

263

Why Do Some Cooperatives Work? TABLE 10.4

Production of Sugarcane, by State and Region

State & Region

1950-51

1960-61

1970-71

1980-81

1987-88

Maharashtra Uttar Pradesh North India•

5,302 28,153 35,517

12,090 54,516 71,693

14,770 54,672 73' 131

23,591 64,205 76,265

24,984 93,350 108,856

All India

54,823

110,001

126,368

154,248

196,723

(Production measured in thousand metric tonnes.) Bihar, U.P., Haryana & Punjab. Source: Cooperative Sugar Directory & Year Book 1986-88:486-87.

a

schedule laid down by the factory's agricultural department. The cane harvest is planned like a military operation: the Agricultural Officer directs teams of harvesters from one field to the next, with operations occurring simultaneously in dozens of locations scattered over many villages. He has a jeep at his disposal. Under him, there are perhaps a dozen overseers, each with a motorcycle, to supervise a particular set of villages. Under each overseer, there are mukkadams (foremen) with bicycles; and under each mukkadam, there are slip boys, who observe every field and make a record of every cartload: when and where the cane was cut, and by whom. Thus a steady supply of fresh cane is obtained throughout the season. The efficiency of this system pays off in terms of net income, which is passed on to the members in the form of high cane prices. One result is that more farmers want to become members, grow more cane, and organize more factories. This accounts for the rapid multiplication of sugar factories and sugarcane production in Maharashtra (see Table 10.4). As seen in Table 10.2, cooperatives in Maharashtra, whose output was less than one-fifth that of U.P. in 1960-61, had caught up with the latter by 1987-88. Moreover, prior to 1979-80 (a year of severe crisis for the whole industry), the cooperatives showed remarkable steadiness in their growth (Mohite 1974:71-73). In 1971-72 and 1975-76, when sugar production dropped sharply in the country as a whole, the co-ops seemed relatively immune to the industry's boom and bust cycles. This stability was brought about by the willingness of the co-ops to go on paying high prices for their members' cane, even during slumps in the sugar market. Cooperative factories on the Maharashtrian pattern may not represent the only solution to the problems of the sugar industry, but they seem to represent the best solution discovered so far. (Other kinds of cooperative and state-owned enterprises are compared below.) Cooperative organization not only resolves the cane supply problem, it protects the interests of small growers. In a co-op, each member has just one vote-that is, his vote is not proportional to the shares he owns. This arrangement gives some security

264

Why Do Some Cooperatives Work?

and influence to the small and medium shareholders who make up the great majority of members; and as we shall see, this is vital for efficiency. Technical and Economic Efficiency Claims for the efficiency of cooperative factories were tested, in Chapter 9, only by comparisons within Maharashtra. Now we shall compare the coops with private factories in general. In India, any large-scale comparison between cooperative and private sugar factories becomes essentially a comparison between regions. Private factories are concentrated in northern India, a region with a subtropical climate where the yields from sugarcane (a tropical species) are lower. Only about 25 percent of northern sugar is produced by cooperatives, nearly all the rest coming from private factories. (There are also some state-owned mills.) Thus data on the north as a whole, or on U.P., its largest state, approximate the operations of the private sector in that region. Likewise, data on Maharashtra as a whole reflect mainly the performance of cooperatives: in 1970-71, cooperatives accounted for 73 percent of the state's sugar production, rising to 96 percent by 1987-88. If cooperatives outperform private factories on a regional basis, this is due to some combination of the following reasons: (1) superior organization of cane supply; (2) regional differences between northern and western India; and/or (3) political leverage crystallized in the form of public subsidies. Before we examine the latter issue, we need to compare technical efficiency. One common measure of technical performance is the recovery rate: that is, the weight of sugar produced from a tonne of cane, expressed as a percentage. As indicated in Table 10.5, average recovery rates for the cooperative factories in Maharashtra are the highest in India, while factories in the north are well below average. 2 Recovery rates are influenced, however, by the quality of sugarcane, which varies from region to region. For comparison across regions, a better measure of technical efficiency is total sugar losses, which shows how much sucrose actually present in the cane is lost in the production process. This measures the efficiency of the factory itself more than the quality of the cane. Table 10.6 compares regional sugar losses, showing that losses are higher in the north than in Maharashtra. (Other measures of efficiency, such as reduced overall extraction, tell the same story.) Another measure of efficiency is capacity utilization. Table 10.7 shows again that Maharashtra surpasses the national average, while the northern states (except for Punjab and Haryana) are generally below average. 3 Good capacity utilization reflects two achievements: first, avoidance of breakdowns; and second, efficiency of cane supply. Capacity utilization tends to be strongly related to overall economic performance. Turning now to a more direct discussion of financial efficiency, this can be measured in terms of the average cost of converting a tonne of cane into sugar. It is difficult, however, to make comparisons between the cooperative

265 Sugar Recovery Rates, by State and Sector

TABLE 10.5

State

Sector

1950-51

1960-61

1970-71

1980-81

1987-88•

c

ll.46 ll.65

ll.71 ll.64

ll.32 ll.28

11.07 11.04

10.85 10.82

9.81

9.22 9.53

9.27 9.16

9.43 9.47

8.82 8.90

10.26

9.21

9.. 00 9.01

9.17

8.97

9. 77

9.14 9.01

8.80 8.69

8.85 8.43

9.81 9.94

T

7.76

8.65 8.78

8.69 8.57

8.90 8.67

10.19 10.13

c

ll.l7 9.99

10.64 9.74

10.65 9.79

10.51 9.98

lO.ll 9.70

Maharashtra

T

c

Uttar Pradesh

T

c

Bihar

T

c

Haryana

T

c

Punjab All India

T

(Recovery rate measured in sugar produced as percent [by weight) of sugarcane crushed.) (C- Cooperative, T- Total.) • Provisional. Source: ~Qo~eritive ~ugar Dire~tOII &Y~ii BQQk 1986-88:508-09

TABLE 10.6

Total Sugar Losses by Sugar Factories

State

1982-83

1983-84

1984-85

1985-86

1986-87

Maharashtra Uttar Pradesh• Biharb Haryana Punjab

2.44 2.86 3.20 2.62 2.55

2.32 2.86 3.03 2.57 2.55

2.44 2.75 3.23 3.29 2.51

2.33 2.69 2.86 2.53 2. 72

2.36 2. 70 3.36 2.54 2.76

All India

2.64

2.69

2.76

2.66

2.66

(Sugar losses measured as percent of sucrose in sugarcane which is not recovered as white sugar.) • Simple average for three zones (East, Central & West U.P.). b Simple average for two zones (North & South Bihar). Source: Indian Sugar Year Book 1986-87, v.I, pt.II:52.

266

Why Do Some Cooperatives Work?

TABLE 10.7

State

Capacity Utilization in the Sugar Industry, by State

Sector

1981-82

1982-83

1983-84

1984-85

1985-86

c

146.7 145.8

147.2 145.7

96.3 94.2

103.4 100.7

100.9 97.5

c

131.5 131.3

111.9 125.1

94.2 102.3

76.1 83.6

81.2 90.5

T

100.1

106.5

64.7

41.4

71.2

c

123.8 148.3

133.3 156.9

119.1 148.3

53.9 88.4

67.1 103.9

Punjab

c

172.9 169.3

177.1 171.6

177.1 165.9

177.1 161.4

124.5 132.6

All India

T

c

136.5 130.7

130.2 125.4

86.4 86.0

89.4 84.5

96.1 92.4

Maharashtra

T

Uttar Pradesh

T

Bihar Haryana

c

T T

(Capacity utilization measured as percent of seasonal crushing capacity.) (C- Cooperative, T- Total.) ~: Cooperative Suiar (monthly) Jan. 1985:293, Aug. 1988:893.

and private sectors because they are organized differently and comparable data are hard to find. The cooperatives in Maharashtra include a large office staff and field labor force devoted to the harvest and transport system, which costs an additional Rs.85 or more per tonne of cane. On the other hand, the private factories in northern India do not organize the harvest and transport of their cane. In the jargon of the trade, their cane is purchased "ex-gate" (at the factory gate), whereas the cooperatives in Maharashtra purchase their cane "ex-field" from the grower and pay all harvest and transport costs themselves. Consequently, when costs are compared, the extra harvest and transport charges incurred by the cooperatives should be excluded. Table 10.8 provides one set of regional cost estimates, calculated from obscure formulas by the central government. These estimates show that the average cost of producing a bag of sugar in Maharashtra is generally lower than in the north. 4 (Similar cost estimates are used by the central government to calculate its levy sugar prices.) Data also exist on the number of factories which have run at a loss. These figures must be used with caution, however, since some of the cooperatives deliberately run at a loss, at times, in order to continue paying high prices for their members' cane. In the early 1980s, nearly one-third of the cooperatives in Maharashtra were running at heavy losses and categorized as "sick"

Why Do Some Cooperatives Work?

TABLE 10.8

267

Average Cost of a Bag of Sugar, by State

State

1983-84

1984-85

1985-86

1986-87

1987-88

Maharashtra Uttar Pradeshb Bihar 0 Haryana Punjab

314.60 338.06 392.50 347.42 335.60

329.52 370.94 414.52 355.73 329.51

373.16 420.00 456.66 402.64 395.99

405.13 8 407.80 470.96 370.52 371.78

407.60 8 465.29 515.11 422.59 422.96

All India

328.86

339.75

384.93

393.70

432.90

(Cost measured as the unit cost 'L' in rupees per quintal of sugar, announced by the Government of India.) • Simple average for two zones (North & South Maharashtra). b Simple average for three zones (East, Central & West U.P.). c Simple average for two zones (North & South Bihar). Source: Indian Sugar Year Book 1986-87, v.I, pt.II:l5; Cooperative Sugar Directory & Year Book 1986-88:523.

(Director of Sugar (DS] 1984:20-21). In the same period, half of the northern factories sampled were running at a loss (Indian Sugar 1986-87, v.I pt.I:210). In 1980-83, a government-appointed committee investigated the problem of sick factories in Maharashtra. (This was just after a series of abrupt changes in the central government's sugar-pricing policy caused widespread losses across the whole industry.) The committee found that 19 out of 59 factories were sick in terms of accumulated losses and other problems (Patil Committee [PC] 1983:20). However, 12 of the 19 were new factories which had been started in 1975 or later. These new units faced rapidly rising capital costs and shortages of cane, which made them more vulnerable to price fluctuations. In addition, most of the sick factories were located in relatively backward areas, outside the classic sugar zone of western Maharashtra. In 1969, the state government had decided to encourage new factories in the northern and eastern zones, which had not yet benefited from the industry. (This was, of course, in response to political pressure from these zones.) By 1980 there were 18 factories (mostly new) located in these relatively backward areas, and a total of l3 (or 72 percent) were classified as sick (PC 1983:20). On the other hand, only five out of 27 factories (19 percent) in the central zone were sick, and only one out of 14 (seven percent) in the southern zone. According to these figures, then, the cooperative industry was basically sound, except where it was implanted in new zones, where the supply of sugarcane was inadequate. The points discussed so far indicate that cooperatives in Maharashtra are at least as efficient as private factories nearby and more efficient than those in the north. However, the possibility still remains that co-ops are so heavily subsidized that their efficiency is artificial, not a result of fair competition.

268

Why Do Some Cooperatives Work?

This is a subject which has engendered little published data or systematic analysis. Consequently, we are here embarking into the unknown with the use of very limited information. We begin by considering the various ways in which the state and central governments subsidize the cooperative factories. The central government gives preference to cooperative units in the granting of licenses for new sugar factories. This does not, however, prevent private factories from expanding; and the northern factories have faced little competition from co-ops in their own region. The state government of Maharashtra subsidizes the sugar co-ops in several ways. Some critics regard sugarcane itself as a subsidized crop, since irrigation charges do not pay the full cost of canal construction and maintenance. However, low irrigation charges are the rule throughout India, regardless of the crop grown (see Stone 1984; Ramamurthy 1989); and sugarcane, as we saw in Chapter 3, was the one crop which helped pay the costs of the Deccan Canals. Moreover, irrigation subsidies go to the farmer, not the factory. This is an important distinction, since after all, the cane growers of northern India pay even lower rates for irrigation; but they have not established successful cooperatives on a large scale. From a comparative perspective, then, irrigation subsidies cannot explain the success of cooperative factories in Maharashtra. There are other subsidies from the state government, however. It buys shares in the cooperative factories; it guarantees their debts to the Industrial Finance Corporation and similar agencies; it sometimes suspends the cane purchase tax (as, for example, during 1979-80, when the industry was in crisis); and it sometimes allows sick factories to postpone depreciation. As noted, sick units are in a minority, concentrated in the relatively backward zones of Maharashtra, so the latter subsidies do not go to the cooperative sector as a whole. Likewise, loan guarantees are not costly for the majority of factories, which are healthy. The one subsidy which does appear to be a major cost is the purchase of shares by the state government, since no dividends are paid on these shares. It is possible to make a rough estimate of the income foregone by the government as a result of unpaid dividends. In 1977-78, for example, the government held shares worth Rs.390 million in the cooperative sugar factories (DS 1978a:7). If this investment had been in the form of a loan, it would have fetched perhaps Rs.39 million (at 10 percent per year) or Rs.58.5 million (at 15 percent) in annual income for the government. Let us say that the interest foregone was equivalent to about Rs.SO million per year. Had this amount been charged to the cooperative factories, their total costs would have increased by less than three percent. (In 1977-78 the Maharashtra cooperatives produced 1.82 million metric tonnes of sugar at an average cost of Rs.927 per ton, for a total cost of Rs.1,689 million [DS 1978a:20; 1978b:5].) The subsidy does make a difference, then, but hardly a decisive one for the cooperatives as a whole. Unless withdrawal of the subsidy coincided with one of the periodic slumps in the sugar cycle, the additional cost could be absorbed with little difficulty by many factories. In

Why Do Some Cooperatives Work?

269

fact by 1988, nine of the older factories had redeemed most of their shares from the government (Maharashtra State Cooperative Sugar Factories Federation 1988). As far as I know, this is the largest subsidy which the cooperative factories received. Or at least the largest positive subsidy. To put this in perspective, we should note that cooperative (and private) factories in Maharashtra are the recipients of a large neg11tive subsidy from the central government: in other words, they pay a heavy subsidy to that government. This subsidy is extracted in the form of differential sugar prices. The central government normally compels all sugar factories, both private and cooperative, to sell about half their output for distribution at low prices through government ration shops. However, this "levy sugar" is not purchased at a uniform rate: different prices are set for different regions of the country (for the same quality of sugar). Levy sugar prices are calculated on a cost-plus basis, so the high-cost sugar factories of the north are paid more than the low-cost factories in Maharashtra. Thus, in selling levy sugar to the central government, the cooperative factories are subsidizing the government, the northern factories, and urban consumers. It is possible to make a rough estimate of the size of this subsidy in the following way. First we note the difference between levy sugar prices in the north and in Maharashtra for a given year, then we multiply this difference times the amount of levy sugar produced by Maharashtra's co-ops. In 1985-86, these co-ops produced 2.21 million tonnes of sugar, of which 55 percent was requisitioned as levy sugar at a rate of Rs.3,728 per tonne. That year, the factories in U.P. were paid, on average, Rs.4,324 per tonne of levy sugar-a bonus of Rs.596 per tonne (Cooperative Sugar 198586:658-63). Had the cooperatives in Maharashtra received the same price, their income would have increased by Rs. 724 million. In other words, the cooperative factories in Maharashtra were forced to pay roughly this amount as a subsidy to the central government. This far outweighs the positive subsidy received from the state government. Had Maharashtra's co-ops been paid at the same rate as the northern factories, their revenues from levy sales would have increased by 16 percent. In other words, they paid a hidden tax of at least 16 percent on these sales. In fact, the tax was even higher since the levy price, even for the north, was well below the free-market price of sugar. Some critics have claimed that the government's price policies are beneficial to the sugar co-ops. An example is the following statement: "The rapid growth of state-subsidized cooperative sugar mills has provided a guaranteed and lucrative market for cane. The mills have thrived as a result of statecontrolled, favorable prices for sugar" (Lele 1984:173). It is difficult to infer what this means. Possibly it alludes to the protective tariff-which helps the private factories even more because they are competitively weaker. Alternatively, Lele may be suggesting that the government's purchase of levy sugar helps to stabilize the market. This is true, at least when the government follows a consistent set of policies. Again, however, the prime beneficiaries must be

270

Why Do Some Cooperatives Work?

the weaker, private factories. So far as I know, there is nothing in the sugar pricing system which could be construed as more favorable toward cooperatives. The levy-price system places a hidden tax on sugar production in Maharashtra; but there are regular taxes as well. If all cooperative factories in Maharashtra were taxed at the same rate as the Malegaon factory in 198687, they collectively paid Rs.342 million in cane purchase tax to the state and about Rs.1,292 million in excise duty to the central government. 5 To summarize, the average technical efficiency of the cooperatives in Maharashtra is higher than that of the northern private factories, in part due to locational advantages, but primarily due to better organization. A number of cooperative factories have run at a loss, but a higher proportion of northern private factories have done the same. Moreover, loss-making is common only in co-ops that have been newly sited in relatively backward zones. 6 In the case of well-established co-ops in western Maharashtra, most subsidies are not a significant cost to the public. One is costly, in the form of shares for which the state government does not receive dividends. However, this turns out to be much smaller than the negative subsidy which flows out of the cooperatives to the central government, due to the pricing system for levy sugar. The Basis for a Class Alliance The key to the cooperatives' success has been resolution of the cane supply problem. However, cane supply is only one aspect of cooperative organization. We must also ask another basic question, which pertains not to sugar factories specifically but to all cooperatives generally: how is it possible for the members to cooperate? For example, what prevents big farmers from monopolizing all the benefits? What prevents the members from promoting their personal, short-term interests at the expense of the organization? Why are the co-ops not burdened with "free riders"? (Olson 1965) These questions do not even occur to critics of the sugar co-ops, who assume that benefits are monopolized by big farmers and that cohesion is maintained by their domination of the other members. Likewise, these questions do not occur to government planners and cooperative promoters, who believe that anything labelled "cooperative" should be considered beneficial to the poor. Experience shows, however, that villagers do not automatically cooperate with one another, and that many cooperatives have been bled to death by corrupt leaders. Why do things work out differently in the sugar co-ops? The cohesion of the cooperative sugar factories appears to be due in large part to a tacit alliance between large and small cane growers. The explanation for the success of this alliance is twofold: First, the political history of the region and its structure of agrarian relations are favorable toward pragmatic coalitions and a sense of common purpose among farmers in general. Second,

Why Do Some Cooperatives lVtJrk?

271

there are technical and economic factors operating inside the co-ops which compel large farmers to promote the participation of smaller ones. The structure of agrarian relations in Maharashtra is relatively open and flexible as compared, say, with northern India. The dominant caste (Maratha in most villages) includes the majority of the rural population and owns most of the land. Large and small landholders belong to the same caste and are linked by patrilineal descent and marital connections, by geographic and economic mobility. Most Maratha farmers rely primarily on family labor (including female labor) to cultivate their lands, and there is no disdain for doing heavy labor. Other castes are small in size; and when farmers hire laborers, the latter come from a wide range of middle and lower castes, the majority being Marathas and other peasant castes. Thus differences of caste status are, to some extent, weakened by shared economic roles; and class differences within the majority caste are softened by connections of kinship, marriage, political patronage, and cultural identity-including a work ethic that does not stress contempt for those who labor in the fields. While these factors do not necessarily promote widespread solidarity among Marathas, they at least make possible pragmatic alliances wherever the situation warrants. All this stands in contrast to northern India: particularly to the classic sugar-producing region of Bihar and eastern U.P., where dominant castes such as Rajputs and Bhumihar Brahmans are often minorities in their own villages; where these castes often refuse to touch the plow and have contempt for those who work in the fields; and where much of the heavy work is done by dependent, Scheduled-Caste laborers, who are at least as numerous as the dominant-caste landowners (see Schwartzberg 1968; Sharma 1978). All this makes for a more rigid and polarized stratification system, with a stronger cultural divide between those who are political masters of the village and those who do much of the actual farm work. In semi-arid regions like Maharashtra, the main political and cultural divide is often between the middle peasant castes and the educated urban castes who control commerce and administration. In humid regions of dense population like eastern U.P., Bihar and Bengal, on the other hand, urban lifestyles and values have long been emulated and absorbed by village elites, creating a cultural divide which is centered more within the villages than between rural and urban interests (Ludden 1988a). 7 In north India, the dominant landowning castes are not only smaller in proportion to the rural population, they are also less unified, politically and culturally, than the Marathas. In U.P., there are several dominant castessuch as Ahir, Brahman, Jat, and Rajput-which have overlapping territories and compete with each other for regional power. This leads to "unstable intercaste competition and alliances" (Lele 1981:xii}, making cooperation difficult beyond the village or village cluster. Willingness to cooperate across class lines may also be influenced by opportunities for economic mobility. As we have seen in Chapter 4, Maharashtra's irrigation frontier encouraged migration and mobility, which contributed to a sense of optimism and innovation among farmers in the

Why Do Some Cooperatives l*Jrk?

272 TABLE 10.9

Annual Cane Acreage of Members 0.5 1. 5 2.5 3.5 4. 5

or 1.0 or 2. 0 or 3.0 or 4.0 & More

Total

Distribution of Shares Owned by Members of 57 Cooperative Sugar Factories in Maharashtra, 1977-78

Number of Members

Percent of Total Members

Estimated Avg. Cane Acreage

Estimated Total Cane Acreage

Percent of Total Acreage

111,045 61,310 25,260 11,704 14,569

49.60 27.38 11.28 5.23 6.51

0.75 1. 75 2.75 3.75 6.75

83,284 107,293 69,465 43,890 98,341

20.70 26.67 17.27 10.91 24.45

223,888

100.00

1.80

402,273

100.00

Source: Maharashtra State Cooperative Sugar Factory Federation 1980:94, 102-03 0

irrigated villages. Many big bagaitdars are known to have come from humble backgrounds, and this knowledge reinforces a sense of identity and opportunity among large and small cane growers. These are reasons why one might expect better cooperation between large and small farmers in Maharashtra than in the north-central region. This is not to suggest that class differences or potential conflicts are unimportant in rural Maharashtra, only that they are less extreme than in other parts of the country. Next, let us consider internal factors, specifically the interests of large and small cane growers inside the sugar co-ops. In 1977-78, half the farmermembers in 57 factories owned shares equivalent to one acre of cane or less (see Table 10.9). Sugarcane is usually grown on a three-or four-year rotation with other crops, so one acre of annual cane corresponds to a holding of three or four acres of irrigated land. However, not all the small shareholders are small or middle farmers, since shares are sometimes divided among several family members, and many families have other land besides cane fields. Even so, the great majority of shareholders are certainly small and middle farmers, owning less than 15 acres of irrigated land per household. Individually, the small and medium growers supply minor amounts of cane to the factories; but collectively, they supply a significant share of the total. This can be demonstrated with estimates from Table 10.9. The second column shows the distribution of members according to how much cane they have pledged to their factories (that is, according to how many shares they own). According to the third column (derived from the second), 88 percent of the members annually grow three acres or less. Their average cane acreage is not given in the source, but it is possible to estimate this for the shareholders grouped in each row. 8 These estimates are given in the fourth column, making it easy to calculate the total and percentage distributions of sugarcane acreage. The sixth column shows that 21 percent of the cane

Why Do Some Cooperatives Work?

273

acreage is owned by small farmers with at most one acre of annual cane. Likewise, 65 percent of the acreage is owned by small and middle farmers with a maximum of three acres of cane, equivalent to nine or twelve acres of irrigated land. The data and estimates in Table 10.9 must be treated with caution, for the effect of the state's land ceiling act was to cause families with large joint holdings to divide up their lands, which were previously registered in the name of the eldest male. Thus, while the table gives the approximate distribution of cane acreage by individual shareholders, it underestimates the proportion of land held jointly by families with large farms. Let us suppose that the wealthier families (those with more than four acres of cane) actually hold twice the amount of land suggested by the table. That is, suppose that these families hold 48 percent instead of 24 percent of the annual cane area. Even so, a significant share of the total cane land still remains with small and middle farmers. The point of this estimate is to show that the small and middle farmers supply a significant share of the total sugarcane (and share capital). Their participation is essential for the factories. Contrary to what the critics assume, there is no question that small and middle farmers benefit from the services provided by the factories. If they did not, the factories would not have been able to enrol so many new members as they expanded. (All the older factories, like Malegaon, have expanded considerably since they were established.) Smaller farmers would never agree to join these enterprises if they saw that others like them were not benefiting. Small and middle farmers participate actively in factory politics. A large share of the members attended annual general body meetings and nearly 90 percent voted in elections of the board of directors, in cases observed by B.S. Baviskar and myself. (This is in contrast to the apathy and nonparticipation which characterize cooperatives elsewhere.) This level of participation suggests that smaller farmers have a real stake in the factory's performance. Each large farmer certainly grows more sugarcane and earns a larger total income from it; but he often earned big profits even before the factories were organized. A small grower earns less total profit, but the cooperatives help make him viable. Since big farmers would profit with or without the cooperatives, while small farmers might not survive at all, it can be argued that the latter benefit most. Large farmers dominate the elected boards of directors and use these positions to manoeuvre for greater power in state politics. The "sugar barons" are wealthy and powerful figures in the countryside, but they are certainly not a closed or reactionary elite. One observer describes them as part of an "expansive elite" (Rosenthal 1977). What holds these leaders in check? What prevents an oligarchy of large farmers from exploiting the cooperatives to the detriment of other members? What ensures that small farmers' cane will be harvested at the right time,

274

Why Do Some Cooperatives Work?

weighed correctly, and paid for promptly? Part of the answer lies in the technical character of the production process. Sugarcane, especially cane of the hard, high-yielding varieties, is more efficiently processed by heavy industrial machinery. Cooperative factories process from one thousand to several thousand tonnes of cane every working day. Such machinery cannot be run at a profit unless it is used at full capacity throughout the crushing season. Now, what would happen, given these constraints, if the leaders decided to manipulate the factory system primarily for their own advantage, to the exclusion of small and medium growers? If the latter felt they were being treated unfairly, they would either reduce cane production; manufacture gul (as in the north); or not join the factory in the first place. If a large number of small and middle farmers stopped supplying cane to a factory, the results would be fatal. Capacity would be under-utilized, profits would fall, and so would cane prices. If prices fell, cane supplies would also fall, and capacity utilization would go down even further, in a vicious circle. In other words, the leaders would be cutting their own throats if they tampered with the supply system to the disadvantage of many smaller farmers. The leaders face a simple choice: run an efficient factory, with a fair supply system, or operate an unfair system which leads to financial ruin for the factory. Most members of a cooperative factory know that these are the operative conditions, stemming from the basic technical constraints just mentioned. Critics assume that big growers can manipulate the co-ops any way they please; but this view ignores the technical and organizational constraints on operating a sugar factory. Critics also assume that small growers are not benefiting and are too intimidated to complain. This is a patronizing view of people who are not, in fact, easy to intimidate. In talking with dozens of Malegaon shareholders, large and small, over the years, I heard many complaints about specific leaders (for personal and political reasons) but seldom complaints about the factory's overall performance. What is interesting is that this is not just an equity issue. The cooperatives could not function well, and certainly could not expand, if they discriminated against the vast majority of their shareholders. Comparisons with Other Types of Co-ops To what extent do these factors explain the performance of other types of cooperatives? Specifically, under what conditions will an effective alliance be created between large and small farmers? There are various types of cooperatives which provide opportunities for comparative testing of the hypothesis outlined above. The first comparison will be with cooperative cane supply unions in northern India. As mentioned in Chapter 4, these unions were organized by government officials in the late 1930s, for selling cane to the private factories. Though each union had elected leaders, these had little real authority

Why Do Some Co(Jjleratives Work?

275

or responsibility and were more concerned with sub rosa manipulations for their personal advantage (Amin 1984:272-77). Hirsch (1961:111-116) mentions widespread accusations of favoritism in the allocation of cane purchase slips; and he goes on to describe an unsuccessful attempt by thousands of cane growers to boycott the unions and sell cane directly to the factories. Batra (1991) describes massive corruption in the allocation of purchase slips; and Verma (1983:56-77) shows how large farmers conspired with mill owners and union officials to monopolize the benefits from these "cooperatives." Why should cane supply co-ops, supposedly dedicated to the common interests of the growers, work so poorly and inequitably in northern India, when cooperative sugar factories work so well in Maharashtra? Perhaps the most basic problem with the cane supply unions is that they assume common interests among the cane growers but do nothing to create or strengthen them. Large farmers who manage cooperative factories in Maharashtra must encourage a steady supply of cane from smaller farmers, or else the factories will run at a loss, and nobody will get good cane prices. In the cane supply unions, on the other hand, there is no such complementarity of interests. Minimum cane prices are set by the government, and the private factories usually will not pay more. Large farmers are not injured if smaller ones are unable to sell their cane; on the contrary, they benefit from restricted competition. Since the large farmers have no direct stake in the profitability of the private factories, they have no stake in an efficient or equitable cane supply system. Their rational goal is to seek to dispose of their own cane on the best terms, regardless of the consequences for the factory or the other cane growers. This comparison between cane supply unions and cooperative sugar factories reveals the frailty of much cooperative planning. It is, of course, well understood by planners that large and small farmers often have divergent interests. Much planning assumes, however, that this problem can be overcome through bureaucratic controls. Unfortunately, such controls tend to be subverted by large farmers through bribery. In fact, as I shall argue more fully in the final chapter, increased bureaucratic control simply stimulates corruption and removes responsibility from local leaders. If an organization does not motivate large farmers to encourage smaller ones to participate, it will not work equitably and perhaps not at all. If cane supply unions are doomed to failure, according to this analysis, perhaps the north Indian sugar industry could be revived through the spread of cooperative factories? We shall review this possibility toward the end of the chapter. For another comparison, let us consider the famous Amul dairy cooperatives of Kheda District, Gujarat. As in the case of sugar, dairy processing for distant markets needs heavy industrial equipment, requiring full capacity utilization in order to run at a profit. (Amul was created to market dairy products in Bombay, about 400 kilometers to the south.) Milk is perishable, like cane, and must be processed quickly, meaning that suppliers and processors

276

Why Do Some Cooperatives Work?

must have a stable and reliable relationship. Consequently, as in the case of sugar, milk processing provides the technical basis for an alliance between large and small farmers (cf. Esman and Uphoff 1984:213). Whether this alliance actually succeeds will depend on external factors such as the agrarian structure of the region and the character of government intervention. In Kheda district, these factors were clearly favorable: Amul, like the cooperative sugar factories, was established not through a government program but on the initiative of local farmers. Amul was organized by leaders from the enterprising caste of Patidars, who were often big farmers, as compared with Barias and other castes of small farmers (Baviskar 1988; Patel 1988). Patidars encouraged other castes to join the dairy co-ops, because more milk ensured greater profits. When the capacity of Amul dairy was expanded, more villages were included from outside the Patidar-dominated Charotar area. Continuous expansion has thus meant that a majority of the village dairy co-ops are now controlled by Barias. Nevertheless, it was in the interest of Patidar leaders to encourage participation by small farmers from other castes because milk, like sugar, is perishable and requires heavy equipment to process and sell in distant markets. Thus good capacity utilization is vital, and this can only be achieved through a stable, concentrated and coordinated supply system. Competition and Efficiency Returning to the sugar cooperatives in Maharashtra, there are still some basic questions about how they manage to operate efficiently. One seeming paradox is that all of them-including those with the best performanceare rife with political competition and factionalism. Moreover, the elected chairmen and directors use factory politics as a base for competition in district and state politics. This is contrary to standard cooperative ideals, which stress political neutrality, the absence of factionalism, and the primacy of cooperation over competition. For many people, competitive politics in the sugar factories seems incongruous or improper: after all, a cooperative should presumably function on a basis of cooperation. Factional competition carries the threat of favoritism, corruption, and institutional paralysis-ills that undermine many co-ops throughout the world. A few sugar factories may have succumbed to these ills, but not many. The combination of economic efficiency and political competition seems paradoxical. How can any enterprise be technically wellmanaged while at the same time serving as an arena for factional politics? Twenty years ago, Baviskar (1968) raised this question and offered some answers. He pointed out that political competition inside the factories actually restrains corruption and mismanagement. Elected leaders know they will be judged by the economic performance of their factory. If performance declines, they may be voted out in favor of eager competitors. In addition, because the factories are publicly audited and because their cane prices are both (a) direct reflections of economic performance and (b)

Why Do Some Cooperatives Work?

277

vital outcomes as far as members are concerned, there is little chance that gross incompetence or corruption will escape public scrutiny. As a result, leaders must be sensitive to the well-being of their factories, not merely to opportunities for personal enrichment. Contrary to standard expectations, then, internal political competition keeps leaders' attention on economic performance. However, there are some problems with this explanation. As Baviskar himself (1991) has recently pointed out, many factories have been chaired by solidly entrenched leaders over long periods-suggesting that, in these cases, leaders are not so vulnerable to the discipline of internal competition. In some factories (Malegaon has been an example), power shifts back and forth between coalitions; but in other cases a dominant coalition, centered on one powerful leader or family, tends to stay in power for a long time. (Two factories near Malegaon provide examples.) One might expect these entrenched leaders to be more arbitrary and irresponsible; and a recent state law limiting factory chairmen to ten years in office is designed to restrict that very possibility. However, the technical performance of factories managed by long-established chairmen is often as good as, or better than, those with more internal competition. Moreover, factory chairmen are sometimes far more competent and innovative than one would expect simply on the basis of internal checks and balances. Such leaders expand and diversify their factories, adopt new technologies, and experiment with innovations. What leads them to take such risks? There is more to be understood here than just restraints on mismanagement. At this point, we might ask whether political competition, particularly extern~~l competition, may be an incentive to improve technical and financial performance? One noteworthy feature of the sugar co-ops is that the better factories are improving their performance in ways that would be familiar in private industry: buying new and larger-scale machinery, installing more efficient equipment, expanding production, and even diversifying into new product lines. Most of these strategies are inherently risky, demanding substantial investment and experimentation with new technologies and production processes. Consequently, the results are somewhat unpredictable, at least in terms of short-term profitability. What drives factory leaders to take such risks? The answer is external competition-competition between sugar factories. It has long been evident that factories compete politically with one another. The dominant faction in one is always allied with its counterparts elsewhere and opposed to those in still others. There is constant maneuvering and realignment as coalitions compete for control not only of nearby factories but also seats in the state assembly and directorships of district cooperative banks. (See Baviskar [1980] for a vivid portrayal of these processes.) As a result, there is always at least one nearby factory whose current' leaders are in direct competition with, and ready to support dissident factions in, your own.

278

Why Do Some Cooperatives Work?

However, this is merely the political side of the competition, the side which attracts most attention in the press. There is also an economic side which is, in some ways, more fundamental. Regardless of whether two nearby factories are political allies at a given moment, they will always be economic competitors. This competition has several sides. In the first place, the factories compete more or less openly for their supplies of sugarcane (and thus also for irrigation water and labor teams for harvesting the crop). Second, they compete in terms of technical performance. Third and most important, they compete in terms of cane price. All these competitions are intertwined, and factory leaders who can consistently offer higher cane prices are more likely to carry political weight.

Competition for Cane Factories compete to expand their membership and cane supply. Within certain limits, more members enhance the political clout of their leaders, and more tonnes of cane reduce the average cost of production. Hungry for cane, the co-ops sometimes encroach on each other's territories. The state government has tried to regulate this competition by passing a "Zonal Order" act, reserving certain areas of operation for each factory. However, nonmembers are allowed to sell their cane to any factory; and the bidding for non-members' cane has been fierce in recent years due to three years of continuous drought. As a result, many factories have been sending their trucks and harvest teams over long distances (often more than 150 km. and sometimes into neighboring states) to obtain non-members' cane in huge quantities and in direct competition not only with adjacent factories but also with those in other districts. Many factories have also engaged in irrigation development as an alternative to territorial encroachment. Others, as we shall see, have enticed the State Fal'llling Corporation to sell them cane which was intended for the private factories.

Competition o11er Technical Performance The leaders and members of one factory are constantly aware of how their technical performance compares with their neighbors, and they actively seek detailed information on the operations of other factories. This concern is often summed up in terms of the recovery rate: that is, how much sugar is produced from a tonne of cane. If costs of equipment, labor, and transport are similar, the factory with the higher recovery rate will earn a higher net income, to be passed on to the shareholders. Baviskar (1980:220,227) has described how one factory appointed a study team, consisting of shareholders and technical experts, to visit other factories and prepare recommendations for improving their own sugar recovery. Comparisons are also commonly made in terms of the number of days a factory shuts down for repairs durin_g the crushing season. If this number is much higher in one factory than its neighbors, the members become concerned and angry. Every factory arranges a boiler puja (a ritual by a

Why Do Some Cooperatives Work?

279

Brahman priest) at the start of the crushing season, in order to avert breakdowns. If these occur anyway, the members are likely to say that the leaders of another factory picked a more auspicious date for the ritual, thus implying superior management in general. Whatever the specific points of comparison, shareholders will ask tough questions about a factory's performance at annual general-body meetings. I have seen this myself, and Baviskar describes it thus: For nearly two hours the shareholders grilled the Chairman and other directors, asking them uncomfortable questions about the working of the factory. This indicated that they were thoroughly conversant with every aspect of the working of the factory. The questions ranged from delays in the harvesting of members' cane to losses in the manufacturing process resulting in lower recovery of sugar. (1980:228)

Comparative information on other factories supplies the premises on which these questions are based.

Competition over Ctme Prices The ultimate force behind these comparisons is the price paid per tonne of sugarcane. Cooperative sugar factories do not pay dividends to their shareholders; instead, net income is distributed in the form of higher cane prices. The popularity of chairmen and directors is directly related to the cane prices which their shareholders receive. Intense competition for election to the board thus translates into intense competition to offer higher cane prices: higher than before in the same factory and higher than others nearby. A state-appointed cane-price committee serves to regulate this competition, which might otherwise spiral out of control and induce some leaders to run down their corporate assets in pursuit of short-term political gains. The committee consists of several cabinet ministers, assisted by the state's Director of Sugar, plus the chairman of the Maharashtra State Cooperative Sugar Factory Federation. This committee inspects the annual accounts of each factory after the close of the crushing season and renders a decision on the highest cane price which it may pay, given its current financial obligations as well as its obligations for future maintenance and replacement of equipment. 9 Even within these regulatory limits, directors and shareholders are keenly aware of which nearby factories have paid more and which ones less. For example, when I visited the Malegaon factory in 1989, a former chairman expressed his misgivings about the current board by pointing out that a nearby factory had paid Rs.30 more per tonne of cane than Malegaon, as compared with a normal differential of Rs.S to 10. Price differentials affect the cane supply, particularly from non-members outside the area of operation, but over the long run even from members. Price differences are a major factor used by shareholders to evaluate the performance of their elected leaders. As a result, leaders are usually careful not to waste their corporate resources, at least not in a flagrant manrier, since the reckoning will surely come after state auditors have checked their

280

Why Do Some Cooperatives "WtJrk?

books and the cane-price committee has made its decisions. Only wellmanaged factories can expect to go on raising their cane prices year after year, and leaders must keep pace with their rivals in other factories. Thus political rivalry is a spur to economic performance, while the demands of economic competition serve to keep political misbehavior within reasonable limits. Competition and Innovation

Competition not only serves to maintain standards of performance, it also prods directors into making big investments with unforeseeable consequences. When nearby factories are adopting new technologies, expanding capacity, seeking out more distant sources of sugarcane, and diversifying into by-products, others feel pressure to do likewise. As in private industry, competition stimulates imitation and innovation, and some changes prove more profitable than others. A case in point is provided by the Malegaon factory, which expanded in 1973-74 and then found the cane supply inadequate, due to the growing scarcity of canal irrigation. As a result, Malegaon had to buy more cane from distant suppliers, which meant greater transport costs and lower sugar recovery. Despite such risks, there are strong motives for expansion. Directors seek to expand their constituencies and thus their chances of election to other offices. They also seek to impress their allies and rivals in other factories. Most of all, they seek to reduce the average cost of production so they can afford to raise cane prices. Higher prices not only make shareholders happy, they also protect the cane supply. A dramatic object lesson in this regard is provided by the recent fate of the private sugar factories in Maharashtra. In the 1930s, these factories created their own estates along the newer canals. However, they lost control of these estates under post-independence land reforms, which turned the land over to the Maharashtra State Farming Corporation (MSFC). The MSFC was created to sell cane to the private factories, but the cooperatives eventually undermined this arrangement. Driven by the expansionist logic just outlined, they offered to buy cane from the MSFC at higher prices than the private factories were willing to pay. The latter protested that they had an exclusive right to the cane, but the MSFC decided to sell to the highest bidder. Thus several private factories have recently shut down, and others are converting into cooperatives. The point is not that this was a fair fight, since the political cards were stacked against the private factories. (The state government, which ultimately controls the MSFC, is influenced by the "sugar lobby" led by cooperative factory chairmen.) The point is that the private factories faced fierce competition for sugarcane, competition which the cooperatives also face from each other. Since private shareholders expect to be paid dividends, their managers could not distribute net income in the form of ever-higher cane prices. But the cooperatives do precisely that.

Why Do Some Cooperatives ffiJrk?

281

An even more dramatic lesson is the recent shut-down of the Ganesh cooperative factory in Ahmednagar district. Ganesh happened to be located near two cooperatives which were aggressively innovative. Moreover, like other factories, Ganesh was hurt by the recent three-year drought and could not improve its operations. Though not supposed to do so, Ganesh shareholders began selling cane to other factories when the differences in cane price became too great for too long. This caused a downward spiral of decreasing cane supply and increasing losses. There are two vital lessons in this story. First, the competition is real: it matters not only in terms of votes but even in terms of survival. Second, Ganesh was allowed to fail: it was not automatically propped up by the political system. This reinforces a point made earlier: the factories have managerial autonomy, they are not wards of the state with unlimited subsidies. (As we shall see, there are government-managed "cooperatives" in other regions which are not allowed to fail and are the worse for it.) A closer look at Ganesh's neighbors will illustrate some points about competition and innovation. The Kopargaon and Sanjivani factories are particularly relevant here because they have overlapping areas of operation and are intimate rivals in terms of efficiency, innovation, and cane price. This makes them convenient examples, though the same general points would apply, albeit with less intensity, to other neighboring factories. 10 The Kopargaon Cooperative Sugar Factory was established in 1953, the second in Maharashtra. (See Baviskar [1980] for a detailed study.) It expanded capacity in two phases, going from 750 to 2,200 tonnes crushed daily. In 1985 Kopargaon started a distillery producing alcohol from molasses. (This was a joint venture with its rival, Sanjivani.) By 1986 Kopargaon also had a paper plant nearly ready to come on line. (Paper is produced from bagasse, the fiber remaining after the juice is extracted from sugarcane.) A former Managing Director was particularly proud that during his tenure, Kopargaon achieved its highest recovery rate and for two years had the lowest percentage of hours lost to shutdowns in the state. In addition, Kopargaon received an award for outstanding technical performance from the National Federation of Cooperative Sugar Factories. (These awards are based on criteria covering all aspects of a factory's operation: sugar losses, labor productivity, fuel efficiency, time losses, capacity utilization.) In discussing these achievements, the former M.D. went out of his way to compare Kopargaon with its rival, Sanjivani. It was clear that, in his mind and the minds of his employers, high performance was verified by comparison with neighboring factories. Kopargaon's nearest rival, Sanjivani, was started in 1964, when it faced two serious problems. First, there were already five sugar factories (three private units, plus Kopargaon and Ganesh) in the immediate area. Second, there was no more canal irrigation available, so most of Sanjivani's cane had to be grown on wells. This meant that the factory and its members had to develop their own irrigation sources, which took time. For both reasons, Sanjivani had trouble getting an adequate cane supply in its early years. After 1972, however, accumulated losses were wiped out.

282

Why Do Some Cooperatives Work?

From 1974 on, the management paid particular attention to new technologies for saving energy and increasing sugar recovery. By 1985, Sanjivani achieved a recovery rate of 11.98 percent, the highest ever in the district and third highest in the state. A wide variety of new equipment made this possible, including electronic monitoring and control equipment. In 1986, Sanjivani was installing a huge new sugar cooling column and planning to purchase high-pressure boilers at Rs.25 to 30 million apiece. We need to ask ourselves why we find here a bunch of farmers and their employees messing around with complicated new equipment, especially when the factory was already functioning efficiently. The question becomes even more urgent when we look at their latest project: diversification into industrial chemicals. Why diversify at all? The Managing Director of Sanjivani, who has to explain such things to his shareholders, had no trouble outlining the rationale to a visiting anthropologist. Sugar profits are inherently unstable for several reasons: the weather fluctuates, as do irrigation supplies, and the entire sugar economy goes through boom and bust cycles. On top of this, the central government keeps changing its policy on sugar prices, which have been partly controlled for most of the time. The prices of some by-products (molasses and alcohol, for example) are also controlled; but industrial and pharmaceutical chemicals, which can be produced from molasses and alcohol, are not controlled. Thus Sanjivani is hoping that chemical profits will smooth out fluctuations in sugar profits. (I use the term "profits" loosely, since net income is either held in reserve funds or distributed in the form of higher cane prices.) Lest anyone think this was an easy gamble to take, I should point out that Sanjivani borrowed Rs.28.5 million for its new chemical plants-more than the total borrowed to build and expand the original sugar factoryand that there were few, if any, other co-ops whose experience could guide them in the choice of products and technologies. In addition to alcohol, Sanjivani is now producing oxalic acid, diethyl oxalate, acetic acid, ethyl acetate and butyl acetate. Some of these are corrosive and dangerous. Partly for this reason, the first year of chemical production resulted in equipment failures and cost overruns. However, by 1986 the major bugs seemed to have been worked out. Expensive and risky are the words which spring to mind in contemplating this strategy of diversification. Yet many sugar factories across Maharashtra have already established distilleries, paper plants, and other by-product industries. Sanjivani is simply carrying this trend a step or two further. Sanjivani and Kopargaon have worked jointly on some projects while competing overall. Their first alcohol distillery was a joint venture, and they have jointly financed local colleges and other projects, such as a cooperative milk-chilling plant. Their areas of operation overlap and many farmers are members of both factories, so joint ventures make sense. In addition, their chairmen have often been allies in regional politics. With regard to cane prices, however, the two factories are intensely competitive. Every technical advance by one is carefully watched by the

Why Do Some Cooperatives Work?

283

other. In the late 1970s and early 1980s, their cane prices moved up and down with perfect consistency, within a few rupees of each other. In 1984, however, the chairman of Kopargaon lost his bid for re-election, it was said, because his factory paid seven rupees less than Sanjivani. (This was a difference of only three percent in the total cane price.) Rivalries and alliances between the two factories are complex (see Baviskar 1980), and the election cannot be reduced to one issue. Nevertheless, it is clear that price differences are important in sugar politics. In 1985-86, thanks to its new profits from chemicals, Sanjivani paid an unprecedented premium of Rs.28 per tonne over the cane price at Kopargaon. If the previous chairman of Kopargaon lost his seat over a much smaller price gap, what would happen now? Kopargaon's chairman was already prepared to meet the challenge: he had the paper plant almost on line. Unfortunately, paper plants had not yet been profitable for sugar co-ops: capital costs were higher than expected and other problems cropped up. As a result, some of the ablest chairmen thought paper might have been a mistake. In its first year of operation, Kopargaon's paper plant ran at a big loss; in its second year, losses were lower; and in the third year (1988-89), they were hoping to get out of the red. The chairman accomplished this by hiring a top-flight manager from private industry at an unprecedented salary and giving him carte blanche. If the strategy works, Kopargaon may soon have paper profits to match Sanjivani's chemical profits. Meanwhile, Kopargaon has been on the defensive. As the chairman himself admitted, the price differential had to be reduced to a maximum of Rs.10 to 15 per tonne of cane. I was told by an observer that, in order to match Sanjivani's prices at this stage, Kopargaon had been running at a loss. I also heard that the Sanjivani chairman had been building up his reserves as fast as possible, to be drawn down in future price wars if the Kopargaon paper plant starts to turn a profit. Kopargaon and Sanjivani are not presented as typical cases. While there are a number of factories which can match them, they rank among the best in Maharashtra. Moreover, their rivalry may be more intense than most, since their areas of operation overlap to an unusual degree. They are presented here to show precisely what happens on the cutting edge of competition. It is assumed that other factories respond to similar pressures, though often with less dramatic results. Kopargaon and Sanjivani are certainly not unique in their experiments with diversification. Their lesson is that political rivalry spurs economic innovation, even at considerable risk. Cooperative Factories in the North What happens to cooperative sugar factories outside this competlt1ve environment? For comparison we turn to northern India, where cooperative factories have grown very slowly. Why has cooperative sugar not flourished there?

284

Why Do Some Cooperatives liVOrk?

We are fortunate in having two case studies of cooperative sugar factories from this region (Batra 1988, 1991). Both factories are near Delhi, one in Haryana state and the other in western U.P. Even though the industry's performance is generally better here than in the north-central region, both factories have problems with unstable cane supply, mediocre technical performance, low capacity utilization, and apathetic leadership. There seem to be three serious problems which account for poor performance. First, the northern factories do not have centrally organized harvest and transport systems. Instead, they rely on farmers to harvest and deliver their own cane, which means the supply depends on local availability of labor and bullock carts. Local labor, however, is only available before the wheat harvest, which overlaps with the crushing season. Partly as a result of intermittent labor and cart shortages, then, the cooperative factories have weak cane supply systems. The situation is even worse in the U.P. case, because there are two distinct cooperatives, one governing cane supply and one managing the sugar factory. The cane supply union issues slips to the farmers, authorizing them to deliver cane to the factory; but there is massive corruption among union employees, since the farmers are all bidding to have their cane purchased at about the same time. 11 Thus the delivery schedule, which the union is supposed to plan, is essentially unregulated. It is impossible for the factory to improve its operations so long as the independently managed supply union performs so poorly. (This problem is inherited from the state's earlier approach to regulating cane supply for the private factories.) A second problem helps to explain the tradition of corruption in the cane supply unions. In the north, cane growers do not manage the "cooperatives." Instead, they are managed by government officials, who may not even be specialists in sugar and are often burdened with other responsibilities. For example, the chairman of the Baghpat cooperative factory in U.P. is the District Magistrate, who has responsibility for the administration of an entire district (Batra 1991). Day-to-day management at the factory site is in the hands of a General Manager, who is usually a civil servant. Files are sent back and forth from the factory to Meerut city (about 45 km. from the factory) for the chairman's approval. The weaknesses of this system are obvious. Thus the elected leaders of the factories and supply unions are not responsible for economic performance, and the real managers are not directly accountable to the shareholders. (They answer, of course, to higher officials.) Since the factory's board of directors has no real responsibility, corruption and negligence are unchecked by competition for election-and in fact, elections are not vigorously contested. Elected leaders are free to pursue their personal interests, which often run counter to the collective interest. For example, many big growers in the U.P. factory are also gur producers; and they organize annual cane-growers' strikes against the cooperative in order to divert cane to their gur-making units. In Maharashtra, the idea of a shareholders' strike against a cooperative factory would be ludicrous, like

Why Do Some Cooperatives Work?

285

shooting oneself in the foot. In the north, lack of accountability among elected leaders encourages lack of responsibility. Likewise, responsibility by state officials means virtually no responsibility: that is, unchecked corruption and negligence, particularly in the cane supply unions. The third problem is that the northern co-ops operate in a climate of heavy protection and official regulation. The private factories have always been small and inefficient by world standards, chiefly due to problems of cane supply. Consequently, they flourished only after a protective sugar tariff was enacted in 1932. Even more to the point, in recent decades the northern factories have been protected against competition from western and southern India by the government's sugar pricing policies. Thus, operating in a highly regulated and protected environment (one in which state governments, as a matter of course, assume direct management of "cooperative" institutions), cooperative factories in the north feel little pressure to operate more efficiently. In Maharashtra, competition engenders greater cooperation among directors and shareholders, because all have a stake in outperforming other factories, while the lack of competition in northern India encourages directors to pursue their personal interests at the expense of the organization. In Maharashtra, the factories are managed by local leaders, not by the state, and they are allowed to fail. In the long run, they are healthier and less wasteful because they are less protected. Competition and Accountability

The ideology of cooperation tends to stress that people should work together in order to better themselves and their communities. This principle is too general, however, to explain why people often fail to cooperate; and conversely, it does not explain why many cooperatives in Maharashtra have far surpassed their original mandate. If the crop was being processed and sold and the members were being paid, why make additional efforts? Why should factory leaders take on the heavy costs and risks of adopting new technologies, expanding capacity, or setting up complex and expensive byproduct plants? The answer can only be that these factories are competing with each other. It is not enough simply to offer an adequate cane price to your members: it must, if possible, be a better price than your neighbor's. Most economic theories of competition focus on costs, prices, and other monetary factors. However, innovation is propelled in this instance not simply by economic factors but in large part by politics. The leaders are held accountable for the performance of their factories, and at the same time they have wide-open opportunities in state politics. They use efficiency and innovation to raise cane prices to buy votes. Having introduced this image of a marketplace for political support, it may be wise to qualify it. No large-scale organization could take risks and plan for the future if the loyalty of its members were for sale for a few rupees at any time. On the contrary, it is clear that leaders can innovate and make big investments precisely because there is a certain amount of

286

Why Do Some Cooperatives Work?

trust and loyalty from the shareholders. However, this loyalty will stretch only so far. In the case ofGanesh factory (and presumably others), a prolonged decline in efficiency may lead to a loss of confidence and increased unwillingness to accept lower cane prices than those available nearby. As with other kinds of organizations, then, there are occasions, such as the crisis at Ganesh, when an organization loses credibility and individual interests come to the fore. On the other hand, there are collective projects, such as the diversification at Sanjivani, which are only feasible if members have some confidence in their leaders over the long term. Interpreting the dynamics of competition, innovation, and possible failure requires us to think about individual calculations of self-interest along with other factors which make for cohesion and a sense of common purpose. In some cases, competition promotes solidarity and growth, whereas in others, it may lead to disintegration. Political competition is often seen as wasteful-as indeed it is, in some respects. For one thing, the number of sugar factories in Maharashtra seems to have grown faster than the supply of irrigation. In every corner of the state, local leaders are eager to establish their own factories, but many of the new units have been sited in areas with inadequate irrigation and cane supplies. 12 In addition, there is a question of distributive justice: should a semi-arid region invest so much irrigation in sugarcane, when the same volume of water could be spread over wider areas to protect seasonal crops from drought? (See Rath and Mitra 1987.) Political ambitions in the sugar sector push in one direction, which may not be optimal for the state as a whole. Another problem caused by competition is that aspiring factory leaders spend lavishly on election campaigns, and some hope to recoup their expenses with under-the-counter payments from merchants, contractors and suppliers doing business with the factories. Such payments have become standard practice in some factories and constitute part of the cost of doing business. In a recent paper, Baviskar (1991) has discussed the reasons for the spread of this practice and the legislation which is intended to curb abuses. If leaders were only concerned with lining their pockets, however, the co-ops would be mired in corruption and unable to operate efficiently. This does not seem to be a serious problem in most cases, despite lurid stories in the press. Certainly, there is corruption: I have been told that contractors now expect to pay an additional ten percent under the counter to the chairmen of some factories; and this amount is returned in the payment which the factory officially makes to the contractor. Hence costs are raised and efficiency lowered, to an extent. What is remarkable, however, is that there seem to be limits to this kind of behavior, limits imposed by the competitive environment. Stories in the press tend to focus on high-profile politicians, giving the impression that all sugar co-ops are controlled by corrupt oligarchs. However, there is probably less corruption in the co-ops than in public- and privatesector enterprises. Different standards are sometimes applied to the coop-

Why Do Some Cooperatives Work?

287

eratives, in part because their founders were dedicated and self-sacrificing individuals concerned primarily with good management and community service. Today's leaders fall short of the standards set by these pioneers, but they still observe certain limits (Baviskar 1991). Bribes and kickbacks are universal, particularly where business interacts with politics. In this case, institutions survive and function efficiently because individual greed is held in check. One reason corruption is restrained is that the available resources extend beyond the sugar factories into other political arenas. Spoils are available, but they are not confined to the assets of the co-ops. Consequently, local leaders nurture their co-ops to build constituencies for competition at higher levels. Moreover, leaders sometimes feel they are compelled to acquire "black money" in order to meet the needs of their co-ops. I was told of one factory with an excellent technical and financial record which applied to the central government for a license to expand capacity. Nothing happened, despite various political strings being pulled. Then the Managing Director was sent to Delhi with Rs.20,000 cash in his pocket, and the license was granted forthwith. Such practices are universal in the business world, because government bureaucracies regulate virtually every aspect of the economy. In other words, some black money (there is no telling how much) has to be spent by leaders for the benefit of their organizations. This may or may not be good for the country as a whole, but the co-ops do not have much choice in the matter. Finally, we might also consider whether there is a less corrupt and more efficient model for the management of these enterprises. That there are worse alternatives has already been suggested by comparison with the northern sugar factories, both private and cooperative. But these are not ·the worst by any means. A recent news story concerns a public-sector enterprise, the Bihar State Sugar Corporation. This corporation, which runs at least 10 sugar factories, has recently lost over Rs.160 million. A committee of inquiry ruled that the mills had been run with such negligence that their losses should be recovered personally from the engineers and managers responsible (Times of India 1989). Expenditures in these factories were uncontrolled and unaccounted for, and technical results were ludicrous. Leadership in Maharashtra's cooperative sugar factories may be wasteful in some ways, but it does not approach the corruption and inefficiency found in some state-managed "cooperatives" or state-owned corporations. One vital reason for this difference is competition-and the accountability of elected leaders for the performance of their factories.

Notes l. Parts of this chapter were originally written and published in collaboration with B.S. Baviskar (see Attwood and Baviskar 1987). 2. Even within Maharashtra, co-ops outperform private factories-as seen in Tables 10.5 and 10.7, where the cooperative average usually exceeds the total average for Maharashtra.

288

Why Do Some Cooperatives Work?

3. Punjab and Haryana are minor producers, making just six and eight percent, respectively, of total northern sugar. U.P. produces more than 75 percent of the northern sugar, so weighted averages for the whole region would be close to those for U.P. 4. It is not clear from the sources for Table 10.8 whether harvest and transport charges are included in the cost estimates. I have been told by one expert that these estimates, based on official formulas, include part of the harvest and transport charges. If the latter were excluded, Maharashtra's average costs would be even lower. 5. The government offers tax rebates under various conditions, particularly to the newer factories. Consequently, this is a rough estimate. 6. It is difficult to compare the efficiency of Maharashtra's sugar industry with those outside India. The data cited above suggest the cooperatives meet reasonably high technical standards. Whether they are also economically efficient cannot readily be judged in the absence of competitive international markets. (World markets are heavily distorted by tariffs, quotas, bilateral and multilateral trade agreements, and so forth.) 7. These contrasts in agrarian structure between Maharashtra and the north apply with less force to the northwest (Punjab, Haryana and western U.P.). However, the northern plains in general, including the northwest, are inhabited by Scheduled Castes (mostly landless laborers) forming a larger share of the population than in Maharashtra. Conversely, the dominant landowning castes in the north form a smaller share of the population than the Marathas in Maharashtra (Schwartzberg 1968). This suggests that economic and social inequalities are more polarized throughout the northern plains. 8. For example, the 111,045 members in the first row hold shares equivalent to either one-half or one acre of cane. Taking the simple mean between these limits gives an estimated 0.75 acres per shareholder, which cannot be very far from the real value. Likewise for the next three rows. The only row which presents a problem is the fifth one, since the upper limit beyond 4.5 acres is not given in the source. This limit has been estimated at nine acres on the basis of other evidence: The state's land ceiling act specifies that each person may own a maximum of 18 acres of canalirrigated land or 27 acres of lift-or well-irrigated land suitable for cane growing (Maharashtra Government 1974:183-86). Under normal crop rotations and irrigation rules, farmers could devote a maximum of one-third of this area to a new crop of sugarcane each year. Thus the maximum annual cane area per shareholder would be nine acres. 9. Factories which have repaid all capital loans and redeemed shares owned by the state government can make their own decisions regarding cane prices. Another limit on cane prices is the minimum price set by the central government, which is intended to protect farmers selling cane to private factories. In Maharashtra, the minimum price is irrelevant, since the co-ops are competing to raise their prices. 10. My information on Kopargaon and Sanjivani comes from brief visits in 1986 and 1989. I am particularly indebted to C.S. Mujgule, who arranged these visits and whose long experience in the sugar industry has been extremely helpful. 11. This problem is avoided in Maharashtra because the factories plan their harvest schedule according to the sequence of planting dates. Since the farmers know who, among their neighbors, planted first, they could easily complain if the schedule were distorted by bribery. The factory administration could not show good results if the cane supply were disrupted, so there is limited tolerance for irregularities. 12. Even so, since some of the older factories, like Sanjivani, have created their own irrigation sources, it is a bit early to condemn the newer ones as unworkable.

PART FIVE

Conclusion

ll Revolution from the Middle We lmve found it of paramount importance that in order to progress we must recognize our ignorance and leave room for doubt. -Richard Feynman

It was mentioned in Chapter l that this book could be read as a critique of modernization as well as dependency (and world-system) theory. The reason is that these are simply variants of a single theory, which has been dominant in the West since the middle of the 19th century, if not earlier. They assume that the West acts on and transforms the East, or the Third World in general. The West is dynamic, almost omnipotent in some versions, while the East is passive and stagnant. · According to dependency and world-system theory, the West is activated by greed and other ugly motives; and in any case, the results of its actions are disastrous for ordinary people elsewhere. According to modernization theory, the results of Western expansion are more positive, at least in their potential. But whichever side you take, the source of new ideas, technologies and organizations is the West. Carried to extremes, this assumption leads one to believe that change in non-Western societies is brought about primarily by external influences (see, e.g., Moulder 1977). The world once had many small-scale societies with simple technologies. Most have been overwhelmed by Western and other imperial powers, such as China and Russia, or by post-colonial states, such as Brazil. Much of anthropology has been built around the study of these hapless societies, which explains why the anthropology of development is a pessimistic science. With regard to these smaller and simpler societies, the assumption of Western omnipotence is at least plausible. This book looks in a different direction: toward large and successful institutions inside a complex and powerful civilization. There is nothing commendable about this choice of topic per se, but it inevitably leads to a different view of Western omnipotence. Readers who have worked their way through the earlier chapters may be willing to concede that the sugar industry of western India was not a creation of external forces impinging on helpless villagers. Neither was it a product of purely endogenous forces. It was the outcome of a complex interaction 291

292

Revolution from the Middle

between East and West, internal and external forces; and the former have shown themselves to be vigorous and creative. Thus, what is needed is an interactive theory of commercialization in Indian agriculture: that is, a theoretical perspective which gives as much weight to inside as outside forces. Washbrook (1990), in assessing the strengths and weaknesses of world-system theory as applied to South Asia, suggests that we need a "multicentered" view of world history in order to reach this goal. This book attempts to meet that challenge in analyzing one small part of the world economy. Grass-Roots Development Reviewing the agrarian history and political economy of Maharashtra helps to explain how sugar cooperatives emerged and flourished. But this leads to further questions for research, and for revisions of development policy, in other parts of the world. Why did cooperative organizers and industrial entrepreneurs emerge in rural Maharashtra but not elsewhere? How could they be encouraged in other regions? The sugar co-ops are part of a broader pattern of creative rural leadership. Long before independence, there were Brahman social reformers, Maratha princes, and peasant-caste leaders who established an array of voluntary associations, cooperatives, and educational institutions. Peasant-caste activists organized the non-Brahman movement, which led after independence to the formation of the Peasants and Workers Party. Thus the sugar co-ops were part of a general awakening of leadership and institution building in the countryside. In recent years, much ink has been spilled on the problem of encouraging grass-roots development, or "development from below." (See, e.g., Chambers 1983; Esman and Uphoff 1984; Korten 1986; Salisbury 1986; Hyden 1988.) Grass-roots development means encouraging rural leaders to mobilize voluntary organizations and build institutions, as they have already done in Maharashtra. Thus we need to compare Maharashtra's political economy with others, to determine the characteristics which encourage such leadership. This question presses hard on the boundaries of my own doubt and ignorance. I have visited other regions in India, and I know other parts of the Third World through printed sources, but I lack intensive, first-hand experience with the political economies of other regions. A similar problem seems to be built into the practice of social science research in India. Social scientists generally specialize in one region and seldom attempt systematic, comparative research in other regions-particularly not research based on long-term fieldwork. Thus there are not enough attempts to develop systematic models of variation between regions. What I plan to do, then, is offer a speculative comparison of Maharashtra with other regions, highlighting features which may explain the sources of constructive leadership in the countryside. Such leadership seems to depend on two factors: how rural leaders (both old elites and new entrepreneurs)

Revolution from the Middle

293

interact with village people, and how they interact with urban elites. These factors depend, in turn, on caste politics and political cultures at the regional level. What seems distinctive about Maharashtra is that rural leaders tend to identify with their village roots and act on problems found in that environment. In many regions of the Third World, rural leaders seem to be primarily concerned with entering the urban elite, that is, with detaching themselves from their roots. In many regions (in Mrica, Latin America, and parts of India), this trend is encouraged by the failure of governments to invest in the countrysidethat is, to invest in transportation, electrification, irrigation, education, public health, and other infrastructure which would provide the base for local entrepreneurs to build new institutions. Much has been written on how and why governments neglect the rural poor (see, e.g., Lipton 1977; Hart 1982). The urban poor can be politically volatile and threatening, and urban elites prefer to concentrate their resources in a congenial environment. Thus it is not hard to explain the urban bias of many Third World governments. It is perhaps more difficult to explain why Maharashtra fails to fit the common pattern. It seems to me that some of the reasons go back to the region's distinctive geographic and historical roots. In the 18th century, Maharashtra was a semi-arid region of comparatively low population density, partially land-locked. The steep coastal mountains, crossed only by footpaths, cut the region off from much of the trade around the Arabian sea. Of course, luxury goods and military expeditions passed over the mountains, but not much in the way of ordinary, village commodities. There were extended trade routes leading across the Deccan plateau, but overland transport (before the railroads) was slow and expensive, suited only to high-value goods. One striking feature of the region's caste system bears out this historical assessment: Maharashtra lacks indigenous merchant castes of any significance. In the 19th century, as the countryside was being opened up with new roads and railroads, the big merchants in rural towns were mostly Marwaris and Gujaratis, that is, natives of Rajasthan and Gujarat. (There were also Lingayat traders from Karnataka.) The lack of indigenous merchant castes (except for a small number of local Wanis) points to the relative scarcity of trade before the 19th century. The land-locked character of the Deccan plateau has thus affected the skills of its inhabitants. Rural Maharashtrians are renowned for their political and agricultural, rather than commercial, abilities. In contrast, the rural people of nearby Gujarat are famous for their "business sense," both at home and abroad. It is not difficult to explain this, since Gujarat is largely a coastal region, and its people have had centuries of experience with trade around the Arabian sea. As a result, Gujarati entrepreneurs, even from rustic backgrounds, have spread across India and to East Mrica, England, Canada, and the United States. The city of Bombay, India's largest industrial center, is dominated economically by Gujaratis, Parsis, Marwaris and others from outside Maharashtra.

294

Revolution from the Middle

Business families from these same ethnic groups established most of the private sugar factories in the Bombay Deccan. Although Bombay is a part of Maharashtra, it is not really a Maharashtrian city in terms of the language and culture of its elites. 1 This is explained, at least in part, by the steep mountains which separate the inland plateau from the coast. Thus the pre-colonial elites of Maharashtra were renowned for their military, administrative, and political (not commercial) enterprise. This helps to explain why Marathas are known today for their expertise in cooperatives and politics, that is, in enterprises which depend on organizing public support. Unlike their counterparts in Gujarat, they are not known for their business skills, nor for migrating to other regions in pursuit of commercial profit. 2 Maharashtra is also characterized by a strong sense of regional identity, which binds people of different statuses and localities together. Semi-isolation, due to mountain ranges separating the Deccan plateau from Gujarat, Malwa, and the coast, is partly responsible for this distinctive identity. In addition, there is Maratha history. In the 17th century, the Marathas rose against the Mughal empire based in northern India, and they played a major role in bringing about the empire's decline (Pearson 1976). The huge, lumbering Mughal armies were nearly invincible, when faced with others of their own kind. However, they found it impossible to subdue the Marathas, who specialized in speedy, highly mobile attacks on Mughal supply lines and who persistently declined to meet their main force in pitched battles (Gordon n.d.). Many have seen in these events a parable of nationalist revolt against a foreign power, but I prefer to see another parable: that of local leadership out-maneuvering the imperial bureaucracy. In the 20th century, memories of earlier battles and military leaders strengthened the nationalist movement, the non-Brahman movement, and the movement for a united Maharashtra. This sense of regional and historical identity enabled kunbi peasants and Kshatriya princes to unite into a huge, politically conscious caste. The rise of the nationalist movement introduced a crucial process in the evolution of rural leadership. Early nationalist leaders were from Englisheducated, high-caste, urban backgrounds, and they spoke initially to urban interests. Later, under leaders like Gandhi, the movement began to mobilize broad public support in order to show that it actually represented a nation. Since the vast majority of Indian people were living in the countryside, it was inevitable that the movement try to address their concerns. In Maharashtra, the nationalists found that princes and peasant-caste reformers were already articulating rural interests, organizing voluntary associations, and building institutions. Moreover, young Maratha activists were organizing a non-Brahman movement and demanding political representation. Thus the nationalists had to come to terms with a rural leadership already conscious of its distinct interests and already in communication with a broad political base. At first, the nationalist and non-Brahman movements went their separate ways. Mter 1930, non-Brahman activists were absorbed into the Congress

Revolution from the Middle

295

party, but they did not lose touch with their rural identity and interests, as shown by their subsequent actions. In 1942, during the Quit India movement, they established a "parallel government" in Satara district, demonstrating the strength of their political roots and their ability to carry nationalist issues into the countryside. In 1948, they had the confidence and commitment to leave the Congress, the party which had won independence for the nation, and form the Peasants and Workers Party (PWP). They demonstrated their ability and determination to speak for a distinct set of interests, in opposition to the urban, high-caste leaders who still dominated the Congress. During the 1950s, the PWP picked up strength by supporting the movement for a united Maharashtra, a movement which was strongly opposed by Gujarati businessmen and Congress leaders in Bombay. After the state was formed, in 1960, Maratha leaders were lured back into the Congress and came to preside over the party and government at the state level. This whole progression of movements, trends, and events served to create a strong sense of political identity and potency in the countryside, in opposition to urban interests. By the same token, it muted consciousness of conflicting interests within rural society. Paradoxically, rising rural leaders benefited from their proximity to Pune and Bombay cities-that is, to dynamic intellectual, administrative, and business elites. We have seen in Chapters 5 and 8 how Brahman intellectuals and social reformers, along with British officials and Bombay businessmen, helped to establish agricultural cooperatives, educational institutions, sugar factories, and other organizations in the countryside. Thus rural leaders had expert help in creating new institutions and articulating new social goals. Moreover, the very strength of the political debates in Pune-debates among the intelligentsia concerning nationalism and social reform-stirred rural leaders to greater awareness of their own problems and interests. Finally, the strength of Pune and Bombay as intellectual and business centers, which could not conceivably fall under the sway of rural politicians, meant that the latter had to accommodate pragmatically with other elites. Rural populism became a strong theme in their campaign rhetoric, but they recognized clearly that the political system could only function through a blending of sectoral interests. The result has been a system of state politics dominated by competitive, pragmatic, non-ideological leaders from the countryside. This style of politics has both positive and negative aspects. As its detractors point out, it tends to dampen down any impulse to cure poverty at the expense of commercialpeasant interests. However, since few villages were dominated by oppressive landlords in the first place, there was not much to be gained from confrontational, class-based politics. Because the rural population consists mostly of small and middle farmers, not landless laborers or sharecroppers, the state government is more concerned with investing in employment and production than in redistributing assets. Critics would like the government to enforce minimum-wage laws and provide other support for agricultural laborers. The problem with this type

296

Revolution from the Middle

of demand, however, is that it cuts into the interests not of a narrow elite but a broad spectrum of large, middle, and small farmers who hire labor to grow cash crops. It is difficult for a democratic, centrist political party to move against such a broad segment of its own supporters. (Even so, the government has created an Employment Guarantee Scheme which reduces the demand for private employment and thus exerts upward pressure on rural wages.) The same problem arises with another basic redistributive issue. Many critics would like the state government to restrict or eliminate the canal irrigation of sugarcane, since the same volume of water could irrigate more farms if used extensively. But here again, the beneficiaries of the sugar economy cover such a broad segment of voters that only gradual, reformist measures are likely to be enacted. Thus the political system of Maharashtra has little to offer by way of radical rhetoric or redistribution. This does not mean it has neglected rural poverty, however. In fact, the pragmatic, non-ideological approach of patronage politics has some reformist virtues. The state government has encouraged many kinds of successful cooperatives: not only sugar factories, but dairies, banks, consumer co-ops, lift irrigation societies, and others. It has also brought primary and secondary education within reach of most villages. It has provided electrification to most villages, plus a dense network of rural roads, and an efficient famine-relief administration. It has also engaged in a huge program of constructing new irrigation canals. Now these programs are not specifically directed at the poorest of the poor: they have benefited a wide spectrum of villagers. But it needs to be kept in mind that the largest pool of poor people is in India's villages and that rural Maharashtra, outside the irrigated zones, is one of the poorest regions in the country. All these programs have benefited villagers belonging to the 40 or 50 percent of the population falling below the poverty line. 3 Because state politics is competitive and non-ideological, politicians build support through development patronage-that is, through ensuring that projects such as roads, banks, percolation tanks, schools, and clinics reach their constituents. Nobody believes that the benefits from these projects are distributed equally. Some constituencies are favored over others, and there are political hangers-on who manage to extract personal benefits from government contracts. Nevertheless, the competitive nature of the system assures continuous attention to meeting the demands of constituentsparticularly since the benefits of projects in nearby constituencies are quite visible. Moreover, another "virtue of a spoils democracy"4 is that pay-offs to constituents tend to expand in scope. An example is the Employment Guarantee Scheme (EGS}, a rural public-works program intended to create permanent assets, such as roads and irrigation works, as well as relieve unemployment. Maharashtra was the first to establish this program on a state-wide basis and is said to do better than other states at administering the scheme. Although the EGS inevitably has problems of inefficiency and

Revolution from the Middle

297

corruption, it brings significant benefits to the neediest areas (Echeverri-Gent 1988, n.d.). The EGS does more than alleviate an economic problem, however. It has created, perhaps inadvertently, a "right to employment" (Echeverri-Gent n.d.). Wherever there are local leaders or groups of seasonally unemployed farmers who want an EGS project, they know how to pressure political representatives and state officials to create one, forthwith. Thus the range of demands and responsibilities placed on the political system has expanded significantly. Furthermore, this demand is specifically one for the poor, since the non-poor are not interested in breaking rocks to build roads. This may not be the best political system for reducing poverty, but it is better than many which neglect the poor altogether. Although the system does not attempt to mobilize rural classes as such, it responds to the interests of the poor. Moreover, it does something which few regimes do anywhere:

it encourages constructf'Pe leadership at the grass roots.

The abolition of poverty must be an urgent goal of all governments. I believe, though I cannot prove, that grass-roots leadership is essential for fighting poverty and other social problems. (Esman and Uphoff [1984:286] note from their worldwide survey of development projects that these have tended to be more sn