PUTTING F I1RS]T At#ibw ~~~ ~~~~ *I I ' I~~~~~~E AI RW Sociological Variables in Rural Development S)ecOafd Edition Revised, and EvpandAed Edited by Mtchatel M Cernea A WoArIld, BaBn k Pubil iiation - Putting People First A WORLD BANK PUBLICATION Putting People First Sociological Variables in Rural Development edited by Michael M. Cemea SECOND EDITION Revised and Expanded PUBLISHED FOR THE WORLD BANK Oxford University Press Oxford University Press NEW YORK OXFORD CORBY LONDON TORONTO NEW DELHI BOMBAY CALCUTTA MADRAS SELANGOR SINGAPORE HONG KONG TAIPEI TOKYO BANGKOK KARACHI LAHORE MELBOURNE AUCKLAND CAPE TOWN JOHANNESBURG DURBAN NAIROBI DAR ES SALAAM KAMPALA JAKARTA IBADAN Copyright i 1985,1991 by the International Bank for Reconstruction and Development I The World Bank 1818 H Street, N.W., Washington, D.C. 20433, U.SA. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of Oxford University Press. Manufactured in the United States of America. The findings, interpretations, and conclusions expressed in this publication are those of the authors and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. Published October 1985 First printing of second edition, August 1991 Library of Congress Cataloging in Publication Data The original edition of this title was cataloged as follows: Main entry under title: Putting people first. Includes index. 1. Agricultural development projects-Social aspects. 2. Rural development-Sociological aspects. 3. Sociology, rural. 4. Rural poor. L Cemea, Michael M. LI. International Bank for Reconstruction and Development. ILt. Title: Sociological variables in rural development. HD1433.P87 1985 303.4'4 85-1 1574 ISBN 0-19-520465-4 Contents Preface to the Second Edition xi Preface to the First Edition xiv Acknowledgments for the Second Edition xviii Acknowledgments for the First Edition xx 1 Knowledge from Social Science for Development Policies and Projects Michael M. Cernea 1 Increased Recognition of Social Science Knowledge 1 Financially Induced Development and the Sociological Variables S Entrance Points for Sociological Knowledge 12 Policy Formation Grounded in Social Knowledge 21 Methodologies for Social Action 24 Two Models: "Enlightenment" and "Social Engineering" 28 Institutionalizing Development Social Science 32 Notes 37 v vi CONTENTS PART I. IRRIGATION PROJECrS Editor's Note 43 2 Planning Technical and Social Change in Irrigated Areas E. Walter Coward, Jr. 46 Conceptualizing the Sociology of Ihigation 48 Studying hrrigation Institutions and Organizations 50 Ignoring Local Irrigation Institutions and Organization 53 Regional Field Studies of Local Irigation: Two Examples 57 Baskets of Stones: Local Irrigation in a District in Northern Sumatra 58 The Book of Irrigation Customs: Irigation Rights in Palampur 63 Regional Field Studies and Project Planning 68 Some Suggestions for Action 70 Notes 71 3 Developing Irrigators' Organizations: A Learning Process Approach Benjamin U. Bagadion and Frances F. Korten 73 The Learning Process in the Participatory Irrigation Program 76 Development of the Learning Process 78 Agency Change 81 Benefits of the Participatory Approach: Key Findings 91 Costs of the Participatory Approach 97 Comparison of Costs and Benefits 98 Developing Local Capacity 100 Developing Implementing Agency Capacity 103 Prject Cycle Issues 104 Notes 107 4 Middle-level Farmer Organizations as Links between Farms and Central Irrigation Systems David M. Freeman and Max K. Lowdermilk 113 The Dry and the Wet: Sociology and Irrigation 114. The Problem: Farmer Water Control 118 Organization Building for Irrigation 121 Design for Precarious Partnerships: The Charter 122 Organizational Behavior and Deviance 137 Conclusion 138 Notes 140 CONTENTS vii PART II. SETTLEMENT AND INVOLUNTARY RESETTLEMENT PROJECTS Editor's Note 145 5 A Sociological Framework for the Analysis of New Land Settlements Thayer Scudder 148 Sponsored Settlements and Social Science 149 A Dynamic Model of Settlement Process: Four Stages 159 Development Issues Associated with Stage One 168 Development Issues Associated with Stage Two 178 Development Issues Associated with Stages Three and Four 182 Conclusions and Policy Implications 183 Notes 185 6 Involuntary Resettlement: Social Research, Policy, and Planning Michael M. Cernea 188 The Unraveling of Social Organization during Displacement 190 Formulating a Sociologically Informed Policy 198 The Content of the Resettlement Policy 201 Formulating Institutional Procedures 203 From Policy to Practice: Reestablishing the Displaced 204 Social Skills in Resettlement Projects 209 Notes 213 PART III. LIVESTOCK PROJECTS Editor's Note 217 7 Pastoral Production Systems and Livestock Development Projects: An East African Perspective Neville Dyson-Hudson 219 Pastoral Production Systems 223 Social Organization and Pastoral Production 225 Pastoral Strategies 233 The Twentieth-century Context of Pastoral Strategy 236 Livestock Development Projects 237 Incorporating Anthropology in the Project Process 252 Addenda 254 Notes 255 Viii CONThNTS PART IV. FisHERY PRoJcETs Editor's Note 257 8 Social and Cultural Characteristics in Small-scale Fishery Development Richard B. Pollnac 259 Sociocultural Characteristics of Fishing Communities 260 Sociocultural Information Needed for Developing Fishery Programs 267 Cooperatives for Small-scale Fishermen 273 Sociocultural Aspects of Developing Fishermen's Cooperatives 279 Designing and Implementing Projects 286 Conclusions 294 Notes 295 PART V. FOREsTRY PROJECTS Editor's Note 301 9 Sociological and Envirommental Dimensions of Social Forestry Projects Scott Guggenheim and John Spears 304 Policy: Forestry Lending and the Environment 306 The Social Organization of Agroforestry 314 Sociology and Social Forestry Projects 319 Social Forestry and Project Cycle Issues 329 The Road Ahead 335 Notes 336 References 336 10 The Social Actors of Participatory Afforestation Strategies Michael M. Cernea 340 The Social Premises of Reforestation Programs 341 Land Tenure and the Anatomy of Trial and Error 346 The Privatization of the Commons 350 Designing Strategies around Social Actors 357 Community Woodlots: Programs without Participating Actors 362 The Family as Actor in Social Forestry 370 Alternative Units of Social Organization 376 Conclusions 386 Notes 388 CONTENTS LX PART VI. RuRAL RoAD PRojEcrs Editor's Note 395 11 Social Analysis in Rural Road Projects Cynthia C. Cook 397 The Role of Infrastructure in Rural Development 398 Sociological Perspectives on Rural Road Projects 402 Identification: How Planners Plan Roads 405 Social Analysis in Project Design 408 Social Issues in Project Appraisal 413 Social Sensitivity in Project Supervision 419 Evaluating Project Performance 422 Conclusion 424 Notes 424 PART VII. EVALUATION Editor's Note 429 12 When People Don't Come First: Some Sociological Lessons from Completed Projects Conrad Phillip Kottak 431 Secondary Analysis of Social Evaluation Findings 433 Social Design for Innovation 437 When People Don't Come First: Development Incompatibilities 443 People-oriented Strategies for Implementation 448 Overview 457 Information Gathering and the Culture of the Development Agency 459 Notes 463 PART vyi. PARTICIPATION Editor's Note 465 13 Fitting Projects to People Norman Uphoff 467 Problems with a Nonparticipatory Approach 468 Different Kinds of Participation 477 Project Design and Implementation 485 Information Needs 491 x CONTENTS Working with Local Organizations 495 Conclusions 502 Notes 505 PART IX. COLLECTION OF SOCIAL DATA Editor's Note 513 14 Shortcut and Participatory Methods for Gaining Social Information for Projects Robert Chambers 515 Information: What and Whose 515 The Two Traps of Outsiders 518 In Search of Cost-effectiveness: Rapid Rural Appraisal 521 The Menu of Methods 523 Participatory Rural Appraisal 527 Potentials and Dangers 529 Conclusion 533 Notes 534 Notes on the Contributors 538 Index 545 Preface to the Second Edition The second edition of Putting People First owes its publication to the spread and success of the first edition: both far surpassed the expectations of the authors and publishers. The volume found a wide audience in academia, in addition to the development community itself. It has been used in classes on development anthropology, development sociology, applied social research, social forestry, evaluation methodology, and other aspects of social change. Putting People First has been published in full translation into Bahasa-Indonesia, and partial translations have circulated in many other countries. Soon after its publication in hard cover, demands for a paperback edition started to arrive. We have resisted, however, the temptation of a quick and simple reprinting of the initial volume, published in 1985. This second edition of Putting People First is a much richer book. In fact, to a considerable extent, it is a new volume. In preparing it, the contributors attempted to weigh their earlier con- clusions against development experiences accumulated during the sec- ond half of the 1980s. Their chapters seek to capture the newly emerging trends in development thinking and practice that are likely to characterize the 1990s. Some authors even went back to the field to update cases investigated earlier and report their new findings. As a result, most of the chapters have been rewritten and expanded, one chapter has been dropped, and two new chapters have been included. The volume's index xi PREFACE xii has been made much more detailed to facilitate research, operational use, and classroom use of the volume. Experience over the past five years has confirmed and reinforced the main themes of this volume. "Putting people first" in development programs is no less an imperative now than before: in fact, it is even more readily recognized as the crucial requirement for inducing accelerated development. The volume takes a firm stand against the technocratic and econocratic biases in development work. It criticizes explicitly or implic- itly the neglect of social or cultural dimensions, the rigidity of blueprint thinking in project design, the focus on commodities rather than on the social actor, the disregard for farmers' knowledge, and the indifference toward people's grassroots institutions and organizations. Any interven- tion that does not recognize the centrality of the social actors in develop- ment programs is bound to clash, rather than to fit, with the natural dynamic of socioeconomic processes. And understanding the basic con- cept that unifies the volume-the concept of social organization-pro- vides the main key for translating the prerequisites of people's centrality and participation into specific project strategies and approaches. In preparing the second edition, it was gratifying to realize the great extent to which the "new" themes emerging on the development agenda and gradually modifying the development paradigm of the 1990s were already embedded in the volume's core. Emphasizing them further in the new edition was thus within the logic of our argument. This edition highlights more explicitly the issues related (a) to natural resources management-particularly water, forests, and fisheries; (b) to the environmental implications of development programs; and (c) to the development of human capital through investments in forming grassroots organizations and promoting participation. After a decade of structural adjustment to financial crises, it is also appropriate to draw attention in this volume to certain adverse consequences of development, such as the risk of greater impoverishment for some marginal groups, the forced displacement and involuntary resettlement of populations, and the dete- rioration and dissipation of common property income-generating assets. All of the authors are concerned with understanding the conditions for long-term sustainability of development investments. For those interested in how sociological and anthropological knowl- edge can be professionally applied in all stages of development work, this edition brings new experiences and methods. It analyzes specific cases in which the difficult transition from social science knowledge to xiii PREFACE the formulation of policy principles and institutional planning procedures was made successfully; it explores ways to overcome disciplinary limi- tations in social sciences; and it discusses new, recently invented or tested, rapid and participatory procedures for generating information. Altogether, the new edition strengthens the argument that sociological analysis brings an increment of professional precision to the thinking and practice of induced development. The authors of this volume look forward to receiving as rich a feedback to this second edition as we did to the first. As editor, I full-heartedly encourage our readers-development practitioners and social scientists, teachers and students-to write and share their views, questions and comments about this volume with us. Michael M. Cemea Preface to the First Edition This volume is the joint product of a group of people much more numerous than the set of social scientists who wrote its chapters. Many managers, beneficiaries, and even victims of development programs, as well as technical experts, economists, and policymakers, have shared their experiences with the social scientists who carried out the research and operational work reported here. Benefiting from these various per- spectives, the volume's contributors have undertaken to give sociologi- cal-anthropological analysis of issues that are central to induced, or planned, rural development. This volume thus critically discusses lessons of development experience, brings into the limelight both the obvious and the hidden (or the recognized and the denied) sociological variables of development activities, and presents new concepts, models, and methodologies for addressing these variables. The basic tenet of the book is that people are-and should be-the starting point, the center, and the end goal of each development interven- tion. In sociological terms, "putting people first" is more than an ideo- logical appeal. It means making social organization the explicit concern of development policies and programs and constructing development projects around the mode of production, cultural patterns, needs, and potential of the populations in the project area. Several themes and concepts have consistently guided the authors of this volume toward unity of treatment in their chapters. Briefly, these refer to: xiv xv PREFACE * The social organization ofproduction systems. The authors started from the shared conviction that generalities about the processes of social change did not need to be rehashed, but that it was necessary to identify the type of social organization and the relevant sociocul- tural characteristics at work in the process of induced rural change. Although certain characteristics are common to the entire agricul- tural sector, others differ from one subsector to the other, depending on the technical production process and the type of population. For instance, the characteristics of pastoral populations differ pro- foundly from those of fishermen, which are different from those of farming communities, and so on. In the same vein, the use of irrigation systems requires a pattern of social organization that does not exist (and is not necessary) in rainfed agricultural areas. These sociocultural characteristics are intertwined in the social organiza- tion of the respective pastoral or farming communities and their mode of production. The authors of this volume believe that soci- ology and social anthropology can contribute more to planned development by supplying the understanding of these patterns of social organization, their dynamics and manner of change, rather than by offering Band-Aids, or piecemeal "solutions," to the prob- lems. * Typology of development projects. The generally accepted taxon- omy of agricultural projects (based on subsectors such as irrigation, settlement, forestry, and livestock) is used here in addressing the various areas of development activities. The authors considered that development projects in any one subsector usually have more or less similar prerequisites for their preparation, and, apart from many local differences, they need the same kind of social information, general organizational arrangements, and basic implementation procedures and approaches. The challenge was to define, in con- nection with the overall social organization of production activities, the social and cultural issues most likely to emerge in each category of projects. Such an identification might help the practitioner and the reader understand the cultural issues underlying various types of agricultural development. * The stages of the project cycle. To increase the operational rele- vance of the book, we have used the project cycle paradigm and discussed the sociologist's task and potential contributions within the broad framework of the main project stages: identification, PREFACE xvi preparation, appraisal, implementation and supervision, and ex post evaluation. Just as the economist's or agronomist's project-related tasks differ at each stage, so should the contributions and methods of the social analyst be tailored to the specific needs of each phase of the cycle. The discussion of the sociologist's contributions can gain in specificity and relevance if critical cultural issues can be pinpointed for each phase. Of course, not every author in this volume has taken his or her recommendations through all stages of the project cycle: some address only one or a few stages. But since the project cycle is a structuring device for the work of many development practitioners, the authors have tried to sharpen the precision of their operational recommendations and improve the communication with practitioners by following this paradigm. Lessons from many World Bank-assisted development projects are discussed in this volume, and the contributors reflect on these varied experiences, both positive and negative. Although this is the first book to discuss extensively World Bank experience with sociological work in rural development projects, the contributors also analyze development experiences of other international, bilateral, or national development agencies. This, we hope, will help give the authors' points relevance in various operational and national contexts. One of the difficulties in writing this book was caused by the authors' desire to address two audiences simultaneously, both crucial. On the one hand, there is the large audience of development practitioners (project designers, managers, policymakers, technical experts, and others) for whom the authors wished to point out the sociological content that inevitably is present in the problems these developers handle daily. On the other hand, there is the audience of professional social scientists from developed and developing countries with whom the authors wanted to share their experiences in project work and information about the type of sociological "products" they found were needed and feasible for development work. The authors hope that by suggesting practical approaches for address- ing the social issues of several distinct categories of projects, pointing out typical pitfalls or fallacies, and recommending constructive and feasible alternatives, they will give the volume both practical and aca- demic value. Although the book is not intended to be a simplified how-to-do-it manual, nonsociologist readers should be able to under- xvii PREFACE stand and use it to enhance the social quality of the design and strategies of their projects. Social scientists themselves, learning how to work in development interventions, will, we hope, find some guidance for mak- ing their efforts more relevant to the concrete needs of development-ori- ented research and operational work. Michael M. Cernea Acknowledgments for the Second Edition In preparing this second edition, considerably revised and enlarged, the editorand contributors toPuttingPeopleFirsthave benefited from many public and personal reactions to the first edition, received through book reviews, discussion articles, letters, and face to face encounters. Particular thanks are due to Theodor Bergmann, Subhash Chandra Sarker, Michael Schulman, Shiv Mehta, Mary Tiffen, Tim Allen, and others, who reviewed and discussed the volume in print and helped spread its message. David Brokensha, Ted Downing, Alan Hoben, Michael Horowitz, Gerald Murray, Michel Petit, V. Rajagopalan, G. Edward Schuh, H. Schwartzweller, Vijay Vyas, Thomas Weaver, and others generously shared their comments with us directly, and we are grateful for their valuable insights. A special word of thanks is due to Scott Guggenheim, research assistant for the first edition, who this time joined as a new chapter coauthor and gave a hand in some of the chores that are the editor's toil. It was especially gratifying and encouraging to receive comments from numerous colleagues who used Putting People First in their classrooms and who conveyed their and their students' reactions. We hope that they will recognize their suggestions in the enriched substance of the new text. The editor would like to acknowledge that a good part of his work on the new edition was carried out during the year he spent at Harvard University as Visiting Scholar, at the invitation of the Department of Anthropology and Harvard Institute for International Development xviii xix ACKNOWLEDGMENTS (1989-90). Thanks are expressed to David Maybury-Lewis, Nur Yal- man, Dwight Perkins, Robert Hunt, Steve Gudeman, and to many other colleagues and graduate students, whose intellectual companionship during that year was of benefit for my work on this volume. The word processing of the second edition demanded a substantial investment of skills and patience from Gracie Ochieng: authors and editor alike are thankful to her. M. M. C. Acknowledgments for the First Edition This volume has evolved over several years, and we are grateful to those who have assisted and helped in its preparation. The volume's chapters were reviewed by many readers-both social scientists and technicians with experience in the specific subsectors-and their comments greatly helped improve the manuscript; a general expression of thanks is offered to them. In particular, Leif E. Christoffersen, Donald C. Pickering, and Montague Yudelman were unwavering in their support and strong en- couragement throughout the successive phases of generating and com- pleting the manuscript. Clive Collins, Fred Hotes, and Francis Lethem were always ready to read yet another version and to share their advice. Michael Bamberger, Gloria Davis, Ted Davis, John Duloy, Hussein Fahim, Francine Frankel, Peter Hammond, and Maritta Koch-Weser contributed especially helpful comments and insight. I owe a special debt of gratitude to my assistants-first Deborah S. Rubin, whose exceptional skills were not spared during the editorial preparation of the manuscripts, and then Scott Guggenheim, who took over from her in the final stretch and ran fast under the editor's prodding. Both were at the time graduate students in social anthropology at The Johns Hopkins University and both managed to complete and obtain their Ph.D. just before book went to print. Joy Vendryes, Mae Gahl, Sonia Moral, and Arda Kehyaian diligently typed their way through the many different handwritings that we inflicted on them. xx .rd ACKNOWLEDGMENTS Last but not least, to the volume's coauthors, too, I owe a special debt, for the forbearance with which they responded to my repeated rounds of demands and comments on their draft chapters. To my children, Dana and Andrei, I owe gratitude for their warm overall support, and also an apology for the overtime that went into this volume that otherwise would have belonged largely to them. I hope they feel it was all worthwhile. M. M. C. 1 Knowledge from Social Science for Development Policies and Projects Michael M. Cemea Several scientific disciplines, particularly economics, preside over the processes of planned development. But until rather recently, sociology and social anthropology have not been called on to serve extensively.1 Though planning is a knowledge-based exercise, the storehouse of knowledge and knowledge-generating methods amassed within these social sciences has been largely overlooked. However, this imbalance is gradually (though too slowly) changing. Why is this change occurring? Increased Recognition of Social Science Knowledge A key factor in this change is the increasingly widespread recognition that repeated failures have plagued many development programs, which were-and largely because they were-sociologically ill-informed and ill-conceived. Although uncomfortable for development agencies, this recognition heightens interest in identifying and addressing the socio- cultural variables of projects. The heavy influence of such variables upon the success or failure of projects is recognized more readily than before. Consequently, more efforts go into providing projects' "hardware" with indispensable "software." A second process conducive to using more socioanthropological knowledge in project planning began in the mid-1970s. At that time, a I 2 SOCIAL SCIENCE KNOWLEDGE reconceptualization of development policies-moving away from "trickle down" theories to poverty alleviation through development- was proposed, and it started to modify development interventions in third world countries. Perhaps for the first time the concept of "target group"-the poor, those with an income below the absolute or rela- tive threshold of poverty-was brought to bear upon project strategies. With it came the need to define the target group, the social actor, the beneficiaries (and sometimes the victims, too) of development. The shift from a virtually exclusive emphasis on physical infrastructures to a recognition of social structures, and from free-swinging ethnocentricity in development interventions to recognizing indigenous cultures, was part of the shift in the direction of actor (people)-centered development projects. The third factor is the work performed by development-oriented social scientists themselves who are demonstrating their usefulness and effec- tiveness for induced (planned) development programs. The prior absence of a critical mass of anthropologists and sociologists within the develop- ment community at large has long prevented them from asserting them- selves systematically and has undercut the cumulative effect of their contributions. The 1980s have registered a change in this respect too. The former slow growth seems to have picked up speed, and although in my view such social science professionals are still too few, their steadily increasing number makes the intrinsic value of social analysis more visible to the rest of the development community. The 1980s saw a tremendous increase in the number of published (and gray) papers and reports on development anthropology and sociology. Furthermore, other factors are the rising public concern for environmental protection, sus- tainable development, and participation and institution building, as well as the sharpened public criticism of development allocations wasted on half-baked programs and backfiring inept interventions; these factors contribute powerfully to employing more behavioral scientists to imn- prove the quality of development work. Thus, at the beginning of the 1990s the noneconomic and social sciences are better positioned to influence development interventions. Their pleas, before dismissed as unproven claims, have become more "respectable." The practical uses of sociology and anthropology are regarded with a bit more confidence. Much of past beliefs about the exercise of noneconomic social sciences becomes now obsolete or obsolescent. SOCIAL SCIENCE KNOVLEDGE 3 When the orientation to poverty alleviation started, there was little experience about how to do a socioanthropological analysis that would fit into the project process. In hindsight, it can be said that the social science community didn't then know how to use this opportunity and largely underused it. Today, almost two decades later, there is consider- ably more-but still not enough-experience. Factors internal to the discipline also constrain the use of noneconomic social sciences in planning and policymaking. Most damaging is the fact that the accumulated experiences of applied social scientists have not yet been systematized, conceptualized, and codified either by these practi- tioners themselves or by academic social scientists; therefore, theory formation is sporadic and the epistemological and methodological con- cerns of applied science are insufficiently stated.2 Hidden under an apparent but ambiguous truce, the split within these professional com- munities between those in academic teaching and basic research and those in development and applied work continues to exist and weakens both sides. There is limited mutual intellectual empowerment, and little deliberate building of the development-oriented enterprise, although this new enterprise is steadily laboring to expand the frontier of its parental disciplines. Such inner dysfunctions reduce the actual impact of the disciplines upon society. If, nevertheless, development sociologists and anthropologists are entitled to rejoice in the growing demand for their expertise, then I believe that they are also under the obligation to define more clearly the types of contributions they propose to make. Will they simply supply baseline social information and background knowledge? Design social models? Formulate theories or strategies for social action? Or just offer methods for data gathering and evaluation? Furthermore, since some of these contributions must be context specific, how will they vary from one culture or location to another? The contributors to the present volume attempt precisely to answer such questions in practice and to identify the entrance points to introduce knowledge emerging from noneconomic social sciences to planned development programs. They carried out these social analyses in the context of real development programs, and they have examined the social organizational and cultural variables, either to propose models, define methodologies for social action, or translate the lessons from past failures into improved approaches. Overall, this represents an effort to sharpen the tools of social science for inducing development. 4 SOCIAL SCIENCE KNOWLEDGE Sociologists and anthropologists dealing with different production processes (farming, animal husbandry, fishing, forestry, and so on) not only have to address different patterns of social organization; they are also faced with several common issues that result from the nature of their operational involvement and are distinct from those encountered in academic research. This opening chapter deals with some of these common issues that I regard as essential, yet are little addressed in the current literature: the nature of financially induced development; the development project as a framework for socioanthropological endeavor; the entrance points for sociological knowledge in development planning; policy-permeated so- cial inquiry; the methodologies for social action; the role of social engineering; and the institutionalization of social science in development work. In discussing planning, this chapter is concerned primarily with rural and agricultural change in developing (rather than developed) countries. Today, more than ever, this vast area is open to and in demand of sociological and anthropological inquiry. The role played by the state (as well as the state-sponsored international and bilateral aid agencies), in inducing and directing accelerated development in developing countries, has generated a demand for empirical knowledge and theoretical models about the structures and processes that such interventions aim to affect. The social scientists from developing countries are best positioned to respond to this demand. The noneconomic social sciences have a long tradition of commitment to the betterment of society. Some of the founders of sociology and anthropology, from Comte to Lester Ward and from Lewis Morgan to Margaret Mead, have pioneered this respectable tradition, striving also to tackle the new ethical and methodological issues involved in such work. Their interests have been continued and refned by subsequent generations of scientists focusing on specific social problems that de- mand research-grounded solutions. Therefore, the call for social science contributions to development is hardly a new appeal. The challenge now, however, is to get involved in the extraordinary expansion of planning and of state interventions for inducing directed development. There is a further intradisciplinary challenge as well. Social scientists must overcome the "disaffection for public service and governmental involvement," as Goldschmidt described it,3 and the "disenchantment with . .. the validity of their involvement in official activities,"4 in Hall's SOCIAL SCIENCE KNOWLEDGE S words, which was pervasive in the social science community until quite recently. If systematic interest in these issues increases among scientists from both developed and developing countries, these disciplines and development planning, will benefit substantially. As a social scientist working for seventeen years within a large international development organization, I am involved in many World Bank-assisted development projects in various countries. I have ob- served the work of many sociologists and anthropologists in these projects and in programs assisted by other agencies, as well as the position that socioanthropological knowledge holds relative to technical or economic expertise. This chapter is rooted in some of my personal experiences that have shown what sociology and andtropology can or cannot do in such contexts. Financially Induced Development and the Sociological Variables How are projects instruments of intervention in development? And how propitious is the framework offered by a project for social inquiry and sociologically informed planning? Projects arepurposive interventions used foraccelerating and targeting economic growth and social development. Notwithstanding the inherent limitations of their current format, I venture to say that development projects offer broad scope for intensive, applied socioanthropological work. In fact, they offer the context to employ a much wider array of social science "products" and approaches to inducing change than the professional social science community has invented to date. At the same time, the sociology of development offers the theoretical argument for an alternative model for projects, a model in which the social actors are the pivotal element, the central core around which all other resources should be marshalled for action, as will be explained further in more detail. Social scientists make the distinction between spontaneous develop- ment, on the one hand, and induced orplanned development, on the other. Yet it is worth remembering that an overall theory of induced social development has not yet been articulated. With measures (policies, programs, international aid, and state interventions of all types) for inducing development being multiplied and expanded enormously, prac- 6 SOCIAL SCIENCE KNOWLEDGE tice has moved ahead of still absent comprehensive theory, and it suffers badly from this absence. An absence of concern with social dynamics has been inherent and endemic in the econocratic or technocratic models guiding planned interventions. Economists, as the professional body presiding most often over the rites of project maldng, have done little to incorporate cultural variables into project models.5 The neglect of social dimensions in intervention-caused development always takes revenge on the outcome. Contrary to myth, it is a grievous misunderstanding to imagine that project interventions are a simple linear unfolding of a well-reasoned, time-bound sequence of preprogrammed activities with all but predefmed outcomes.6 Beyond what is being planned, and often despite it, development interventions occur as proc- esses subjected to political pressures, social bargaining, administrative inadequacy, and circumstantial distortions. A host of necessary or un- warranted reinterpretations modify the intended outcome. It is true that development interventions do not take place through projects alone, but projects are the most widely used vehicle. Their underlying model contains a greater potential for case by case fine-tuning than planners typically use. The debate over whether current projects are an adequate vehicle for development interventions has pointed out both strengths and weak- nesses inherent in the project model of intervention. On the one hand, projects concentrate resources on selected priorities, focus on a circum- scribed geographic area, and can address specific population groups. They zero in on identifiable constraints on development. Projects can also be social laboratories that use innovative approaches on a limited scale to gain experience for attempting large-scale interventions (for example, national plans). As a form of development investment, Hirsch- man noted, the very term project "connotes purposefulness, some mini- mum size, a specific location, the introduction of something qualitatively new, and the expectation that a sequence of further development moves will be set in motion."7 Nonetheless, development projects have been criticized because they are only segmented units of intervention; they often bypass overall structures, are subject to the hothouse overnurturing syndrome, and thus may develop atypically. Projects are also criticized because they tend to create enclaves and to siphon resources from nonproject activities, while sustained development at the same pace beyond their limited time frame may be doubtful. The flow of project allocations is also prone to SOCIAL SCIENCE KNOWLEDGE 7 diversionary pressures that often channel resources away from intended beneficiaries. These and other shortcomings of the project approach are further highlighted from a sociocultural perspective in several chapters of this volume. Despite the recurrent debates on the merits and disadvantages of projects as instruments of development intervention, no effective alter- natives have emerged, and projects are likely to remain a basic means for translating policies into action programs.8 Therefore, as long as the project approach is routinely being used, it is both legitimate and neces- sary to identify and address the sociological requirements intrinsic to this model of development intervention. I contend that social scientists could gain from further exploring project potential and from overcoming its limitations, in order fully to use-and also broaden-the opportunities available for inquiry and action-oriented sociology. Since development anthropology and sociology are often defined as a "sociology of inter- ventions"9 or a "sociology of strategies," it is fitting to examine what in the nature of interventionist state policies and programs creates a demand for social knowledge. T-he systematic use of social knowledge, as a complement to economic and technical knowledge, is indispensable for "putting people first" in planned development interventions. Putting people first in projects is not just a goodwill appeal to the humanitarian feelings of project planners, a mere ethical advocacy. It is a concept for constructing programs for inducing development and an imperative for their effectiveness. I submit that "putting people first" in development programs must be read as a scientifically grounded request to policymakers, planners, and technical experts to explicitly recognize the centrality of what is the primary factor in development processes. This interpretation implies a call for changing the approach to planning. When the requirement to admit the centrality of people in projects is addressed to those who currently design projects-primarily, to the technical and economic planners of development programs-it becomes tantamount to asking for reversal of the conventional approach to project making. This is not to say that people are totally out of sight in conven- tional approaches. But many approaches are so overwhelmingly domi- nated by the priority given to technical factors or economic models stripped of the flesh and blood of real life that the characteristics of the given social organization and the very actors of development are dealt with as an afterthought. 8 SOCIAL SCIENCE KNOWLEDGE The argument of this chapter, indeed of this volume in its entirety, is that the model adopted in projects that do not put people first clashes with the model intrinsic to the real social process of development, at the core of which are-simply-its actors. This clash seriously undercuts the effectiveness of projects that attempt to induce and accelerate devel- opment. Putting people first is a reversal because it proposes another starting point in the planning and design of projects than that taken by current technology-centered approaches. This specific "reversal" de- mands to identify-in every single technical, financial, or administrative intervention-4he sociological angle and the variables pertinent to the social organization affected or targeted by intervention.10 This is why putting people first is not a simple metaphor, but rather a tall demand to restructure the approach to planning. Lately, support to "people-centered" projects is pledged even in the official planning rhetoric. But this is done in a manner that often reduces the concept's applicability to projects in the so-called "social sectors"- education, health, family planning, nutrition, and so on. The point is that "putting people first"-in other terms, the concern for social organization as the central issue-should be recognized as paramount for all projects, not only for projects in social service sectors such as those listed above. It should be recognized explicitly as a paramount objective for projects supporting productive activities, where the apparent primacy of techno- logical or economic variables still serves as a fallacious justification for neglecting social organization variables. Development projects are in essence vehicles for financing induced growth and change. Often financial resources are a project's single most massive input injected into an area to accelerate growth. The financial investments-resulting either from loans or from central budget alloca- tions, in other words, from sources exogenous to the project area and its own capacity for capital formation-are used as the lever apt to eliminate constraints and set development in motion. By their nature as financially driven interventions, such projects must-but often do not-provide explicitly for complementary social or institutional reorganization. This is so because the abrupt large infusion of external resources into arural society modifies the internal and gradual processes by which resources for development are created, saved, and accumulated. When these resources are generated internally and grad- ually, they more or less commensurate with the capacity of the socioeco- SOCIAL SCIENCE KNOWLEDGE 9 nomic structure to absorb and use surplus. However, if the financial resources alone grow suddenly by external injection, while the patterns of social organization or the institutional structures remain the same and no matching change in the nonfinancial factors of development is sought, then serious discrepancies set in. People's economic activities are embedded in a structure of social relations; therefore, for the theory and practice of induced development, defining accurately the levels of such embeddedness in different-mar- ket or nonmarket-societies becomes operationally relevant."1 The de- gree of embeddedness and level of congruence between accelerated technoeconomic activities and the existing structure of social relations is not just an academic problem, but a practical one, inasmuch as the interventionist stand of the state should beware of causing desquilibria and discontinuities. Indeed, when investments through state-sponsored programs are made only in the technological infrastructure, they are highly likely to be an implicit disinvestment in the social infrastructures of the given society. Often such disinvestment is only relative, but sometimes it is also an absolute disinvestment. Furthermore, the very know-how for investing financial resources in institutional structures is much less refined than the know-how for investing in technical and physical assets. Thus, when public investment in technical infrastructures proceeds alongside disin- vestment in the social, cultural, and institutional structures within which the former are embedded, the sustainability of the technical advancement itself is undermined. Massive financial resources may trigger a short-term development spurt, but without institutional and social scaffoldings built in at the same pace, the new edifice is not durably constructed. The long-term positive effects of financially induced changes will remain at risk. Such risks may initially stay hidden, but eventually they will surface into unanticipated and undesired outcomes. It is important to emphasize that induced development depends ulti- mately on the overall quality of the program, rather than on the absolute amount of its financial inflows. In other words, the developmental impact of aid resources could be increased even if the financial flow levels are kept constant or at the limit reduced, provided the overall quality of project design and its effective implementation are enhanced. For in- stance, participatory project implementation and improved management 10 SOCIAL SCIENCE KNOWLEDGE are likely to extract greater "development mileage" from the financial investments of a project than what can be obtained by increasing the financial allocations while disregarding participatory implementation. While the need for financial resources is indisputable, it should be understood that internationally assisted rural development programs have often languished not because of lack of finance, but because of either the inability of the given rural society to use external finance effectively or the planners' inability to formulate an efficient social construction strategy for absorbing those new financial resources. Money is not everything. In certain situations money may be the least important contribution to processes of change. The financial levers of development cannot soundly substitute for the nonfinancial ones.12 The financial inflows of the project may temporarily create the appearance that anything can be done, but this appearance is misleading and transi- tional. Salient sociocultural factors continue to work under the surface. If the social variables remain unaddressed ormishandled, then the project will be unsustainable and fail, no matter which governmental or interna- tional agency promotes it. Sociological knowledge-and the social analyst-can help identify, conceptualize, and deal with the social and cultural variables involved in fmancially induced programs. In doing so, the sociologist's contribution consists not just of uncovering social variables overlooked in the planner's approach; it often amounts, as Robert Merton pointed out, to a reformulation of the problem that requires solving. "Perhaps the most striking role of conceptualization in applied social research is its trans- formation of practical problems by introducing concepts which refer to variables overlooked in the common sense view of the policymakers. At times the concept leads to a statement of the problem that is diametrically opposed to that of the policymaker.",13 Or, we can add, is diametrically opposed to that of the project planner or manager. Specifically, in situations like those discussed here, the social analyst participating in designing a project intervention will ask basic sociolog- ical questions: Can the existing social and institutional structures func- tion effectively at the accelerated pace triggered by a large financial influx? What social adjustments are needed to keep step with the other elements of the intervention? The sociologist must help chart the opera- tional steps for creating the institutional changes necessary for both the social and cultural sustainability of the financially launched develop- ment. As a result, the entire course of practical action can be changed. SOCIAL SCIENCE KNOWLEDGE 11 The penalty for not carrying out the social analysis and not incorpo- rating social knowledge into financially induced growth programs is costly and swift. An anthropological secondary study of fifty-seven World Bank-financed projects, which examined the association between the sociocultural fit (or misfit) of project design and the estimated economic rate of return at project completion (audit) time, found that attention to issues of sociocultural compatibility paid off tangibly in economic terms. Specifically, thirty out of the fifty-seven projects were judged to have a project design compatible with traditional cultural and local socioeconomic conditions, while in the other twenty-seven projects serious sociocultural incompatibilities were identified. The most signif- icant findig was that the compatible set of projects had an average rate of return at audit of 18.3 percent, which was two times higher than the economic rate of return (only 8.6 percent) of the other twenty-seven projects in the second group (see chapter 12). Another significant finding comes from analyzing twenty-five Bank- financed projects that were reevaluated several years after the financial flows channeled through the projects were termiinated. The purpose of this analysis was to assess the long-term sustainability of those projects. Thirteen out of twenty-five projects were found to be nonsustainable; among the primary reasons for their nonsustainability was not insuffi- cient financing, but factors of a sociocultural nature (mainly the lack of farmer organizations and participation) neglected during project formu- lation and implementation.14 Such examples only confirm with economic facts that financially induced growth interventions stand a high risk of being less effective than planned or of failing altogether, if they neglect to build up the sociocultural structures for development. It is said sometimes that doing social analysis for each project would increase the cost of project design and waste scarce resources. In hind- sight, one learns, however, that the cost of not doing social analysis is much higher. Projects should not bear the unaffordable expense of not addressing the sociocultural variables knowledgeably. In fact, as a cul- turally sensitive development economist argued, the "advances in social science knowledge reduce the cost of institutional change,"1" the same way advances in the natural sciences reduce the cost of technical change. Development-oriented sociologists and anthropologists have become not only more involved with projects in the past decades but also more able to insert their contributions effectively. Although sometimes still 12 SOCIAL SCIENCE KNOWLEDGE uncomfortable with projects and straitjacketed by blueprints, tech- nocratic biases, short time frames for fieldwork, and other restrictions, many social scientists have gradually been learning, as the following chapters in this volume demonstrate, how to make operational contribu- tions within this planned approach to development. Applied social sci- entists have also developed new research procedures (see chapter 14) and discovered that the project format opens up not one, but multiple points of entrance for contributions of social knowledge. Entrance Points for Sociological Knowledge Where are the entrance points for effective incorporation of sociolog- ical knowledge into rural development planning? Past errors and lingering misperceptions have clouded the answers to this question. My argument is that the habitual entrance points used in the past-"social impact assessment" or "ex post evaluation"-have been few and not the most effective ones. They must be broadened. Therefore, I argue (a) that entrance points should be multiplied and opened up in every important juncture of the planning and execution of projects and (b) that the single most important entrance area where sociological knowledge can and must contribute is the design for pur- posive social action. A quick retrospective look reveals several fallacies and lessons of experience in this respect. Conventional Entrance Points Historically, and until recently, the first and main entrance point for social scientists was the ex post evaluation of development results. Sociologists were sometimes called in to assess whether a certain project had indeed accomplished its overall objectives and triggered the desired consequences or some unanticipated ones. Unfortunately, this was the wrong end of the cycle: it was then too late to affect the project process. While the use of a sociologist or anthropologist for this task was positive, it was not a substitute for multidisciplinary planning. At issue here, of course, is the role of the discipline as a body of knowledge, not the task given to an individual sociologist. Although an individual expert can correctly perform a segmented role such as evalu- ation, the noneconomic social sciences as disciplines should not be SOCIAL SCIENCE KNOWLEDGE 13 pigeonholed into only one segment of the project or planning process. If used only as evaluators, sociologists arrive late, long after other experts have made their contributions. They appear wise after the fact and are seen as those who only complain about what others have actually done. Their skills are not brought to bear on ongoing social action; since the social process has taken place before the evaluation study, it cannot be improved or redirected in retrospect. It is often said (with a consoling undertone) that the lessons drawn from evaluation may, of course, be useful for the next program. However, many sociologists, the author included, have been generating evaluation findings that should have led to the modification of subsequent programs, only to find themselves in the unenviable position of having no part or influence in the actual formulation of the follow-up programs. Instead they watched new projects without sociological inputs being designed again by econocrats oblivious to earlier findings and repeating the same mistakes. Even when the sociological evaluation findings are correct and relevant, whether or not they will materially affect new programs or policies depends on the decision of others. Incorporation of past lessons is never automatic. Moreover, skills similar to those used for identifying the social lessons of past programs are best suited for incorporating such lessons into the new design. That is why it is necessary to involve professional sociologists in the preparation process for the new project. There is no legitimacy in relegating them to the function of ex post evaluators only. Another role for the social sciences has emerged in what is called the social impact assessment (sIA). This is a kind of ex ante evaluation. In this role, the sociologist examines a development project prepared by a group of other experts and is asked to make a desk assessment about whether or not it will have positive or adverse social repercussions. Here again the social analyst is not called on to participate constructively in shaping the intimate structure and sequence of actions in development projects; rather, he or she is used simply to validate or partly modify a ready-made "package." Sometimes the social impact assessment carried out by the social analyst is genuinely taken into account in modifying the project's plans, and this service is undoubtedly worthwhile. Some anthropologists have rendered important contributions this way. My point, however, is that the social analyst is not to be employed merely to anticipate the effects of plans conceived without him, for damage control and mitigation. The 14 SOCLAL SCIENCE KNOWLEDGE point is to involve the social analyst in the advance planning of develop- ment. This is a much broader endeavor than just assessing impact. A somewhat more promising "entrance" is offered to sociologists when they are invited to generate the basic social information necessary for a project. Such a contribution is often quite useful. But the role of the supplier of descriptive information still allows the sociologist little influence over what is done with the information; whether or not it is used at all; or whether it is incorporated into the design for development, into resource allocation decisions, and into the sequencing of planned actions. Thus, none of the entrance points noted above allow sociology to participate fully in the interdisciplinary modeling of planned rural devel- opment. The narrowness of the assignment-evaluation or data gather- ing-blocks out the crucial contribution that sociology should make to the actual content and design for purposive action in rural development programs. Going Outside the Cocoon of the Discipline In a powerful analysis of "why sociology does not apply" in public policy, Scott and Shore argue that sociologists have been largely ineffec- tual in policy-relevant work because they have remained captive to their disciplinary process and manner rather than interpolating themselves and their work into the policymaking process itself. "A main source of the present difficulty with applied sociology is that attempts to make sociol- ogy relevant to policy are conceived and executed with disciplinary and not with policy concerns in mind."'6 This thesis is important and conse- quential. It entails that the work to be done by sociologists, the methods used and their order of use should differ substantively in a policy perspective from what is habitual in a disciplinary perspective. When guided by an inward-looking disciplinary perspective, applied sociological work begins and ends with sociology and may not fully serve the specific purposes of policy (applied) work. Conversely, with afinality oriented to the public domain, a sociological perspective would begin and end with policy-not disciplinary-concems. The "significant other" for a social scientist writing up his field "product" in a policy or project perspective is the policymaker or the development practitioner, while the significant others for the one writing with a disciplinary perspective are his academic peers. To carry out policy-permeated social SOCIAL SCIENCE KNOWLEDGE 15 inquiry and analysis, one has to come out of the cocoon of the discipline in more than one way. When employing the policy perspective, Scott and Shore note, not only is the order of activities changed but their nature is different, too, since the purpose is "to adapt method to problems involv- ing questions and variables outside the ken of the discipline."17 Scott and Shore focus their argument on applying sociology to policymaking, but their critical analysis is valid also with respect to applying sociology to the planning process. The two areas are not identical, since planning is essentially in the realm of policy execution rather than policy fonnulation. Applied sociological work in one is different from the other, despite similarities or overlap. If the goal is to use sociological knowledge in projects, then this work should start from the needs intrinsic to the project model itself. The sociologist who decides to use his knowledge and skills in a project-related task needs to internalize the process of project making and to tailor his work so as to fit the structure of this process. He or she must then understand the project cycle, its specific stages (discussed in detail below), and how to relate to these stages as both starting points and intermediate ends of his or her activities. More than any other development professionals, sociologists must labor to bring the social actors, the people themselves, into the processes of project formulation, planning, and creative execution, and attune the other specialists-technical and economic experts-as well to the de- mands of putting people first. The two tasks are not opposed, they are complementary. The vital need to promote public participation goes hand in hand with recognizing that preparing a complex development pro- ject is a job that requires the cooperation of a number of professional experts, including the social scientist. Adopting the project model as a common denominator for applied sociological work also has the advantage of enabling sociologists to interact better with the other project professionals (planners, technical experts, and economists, for example) and to overcome their intimate biases, often self-paralyzing, against such otherprofessionals.'8 The better they grasp the "technical" criteria of the other professionals, the better able sociologists are to generate sociolog- ical answers (propose solutions) well-tailored to the project's circum- stances. Furthermore, this engagement enables the sociologist in turn to raise his challenging questions, compelling the other professionals to reflect on them and generate project-related (and project-funded) solu- tions. 16 SOCL4L SCIENCE KNOWLEDGE Rural development planning in many countries follows the format of the project cycle. Aside from some local variations, the essential stages of this cycle, to which sociological contributions can be matched, are: * Project identification * Project preparation (including design) * Project appraisal (including design correction) * Project implementation (including monitoring) * Project evaluation. Each stage of this cycle requires a different type of sociological contribution (either informational, analytical, or predictive) from sociol- ogists and anthropologists, in much the same way that the specific contributions of economists vary from one stage to another.19 Of course, the stage sequence of the project should not be fetishized or idealized as the ultimate embodiment of planning rationality or as an inflexible arrangement of time-bound activities. As shown before, the project model has its limits, and the project's social reality is much richer than its abstract model. It is the obligation of the social scientist to work so as both to adopt and adjust the model to the given social reality and to reject any procustian request to do the opposite. Not only challenges, but sometimes barriers as well are built into this very format. Control over the project cycle is generally in the hands of government officials, local politicians, planners, administrators, and technical managers, who decide whether and when to call in sociologists and what to ask them to do. In real life, such calls may be half-hearted or may be narrowly restricted to only one stage rather than to all of them consecutively; they may even misspecify the tasks of the social analyst. In practice, project planners and other officials often do not know precisely what to ask the sociologist. Nor do they know particularly what they are entitled to receive from the project's social analyst. To sum up, going "outside the ken of the discipline" means, in this case, adopting the stages of the project cycle listed above as the practical framework for organizing the activity of the social analyst. His or her contributions should be tailored to the specific purposes of each stage. To do so, most social scientists trained only in the ways of disciplinary academic research must go through a learning process. Numerous illus- trations of how sociological work can fit into each of these stages are contained further in this volume, and I will provide here only a brief characterization. SOCIAL SCIENCE KNOWLEDGE 17 Project Identification Project identification is the stage when the potential of a particular de- velopment intervention is approximated. On the social side, this re- quires a definition of the likely social actors in that development, such as the project area population and the place of its various subcategories in the social structure. The social and economic goals of that development must be delineated, and their likely consequences, both positive and neg- ative, must be subjected to an initial estimate-the first round in an iterative analytical process. During this stage, the social analyst has only little time for independent field investigation, but should be able to gather and assess all the available general social data (censuses, prior sur- veys, ethnographies, monographs, and so forth) relevant to the area and the type of intervention considered. For instance, if the potential of an irrigation dam is considered, at the identification stage the social analyst should broadly define not just the beneficiary groups in the command area but also the reservoir population certain to suffer the adverse impact-forced displacement. He or she should also consider the likely influx of construction manpower and the predictable boom- town effect at the construction site, and so on. At this preliminary stage, it is not possible, nor is it necessary to have definitive answers or the full set of social data. But it is necessary to identify the main risks and to set in motion the subsequent data gathering, social analysis, and social design processes to be carried out during the much longer, and more in-depth, preparation stage. Project Preparation This stage is the most important for the social construction of devel- opment interventions. At this point, the sociologist verifies the hypothe- ses about the development potential and translates them into planned sequences of actions. The social analyst, as a member of the preparation team, must use the full arsenal of relevant research tools, from surveys and case studies to piloting and other methods, together with his or her ability to anticipate changes in social arrangements and to design alter- native organizations, institutions, or strategies for participation. Using the same example of an irrigation dam project, the social analyst must plan in detail the process of involuntary resettlement, involving both the displaced groups and the host population; plan for the displaced to be 18 SOCAL SCINCE KNOWLEDGE reestablished on a viable socioeconomic basis,20 plan for the complex process of helping the future beneficiaries from irrigation organize themselves into a network of watercourse-based irrigators' associations (see chapters 2, 3, and 4). Project Appraisal The appraisal stage critically reexamines the preliminary version (feasibility study) of the project as produced during project preparation and, often, introduces significant corrections into the future project. Ideally, the specialist teams doing the preparation and the appraisal, including their social analysts, should not consist of the same individuals, so as to enable the second team to reassess freshly, independently, and critically the work of the first team. The social analyst on the appraisal team should reexamine the project's assumptions regarding the social actors, their current needs and expected behavior; he or she should also reevaluate the social arrangements recommended to accompany the induced technical changes and reshape the relations between people and their environment. In light of World Bank experiences with project appraisal, four main elements should be the focus of the sociological appraisal:21 * The sociocultural and demographic characteristics of local benefi- ciaries, including groups that may be adversely affected * The social organization of productive activities of the population in the project area * The cultural acceptability of the project and its compatibility with the needs of the intended beneficiaries * The social strategy forproject implementation and operationneeded to elicit and sustain beneficiaries' participation. A few details on each one follow. The social appraisal of projects should first verify that project design has taken into account sociocultural and demographic characteristics- the size and social structure of the population in the project area and its density and stratification patterns (including ethnic, tribal, and class composition). This is particularly important for project components affecting specific target groups (such as ethnic minorities, resettled populations, and women). It requires an understanding of the receptivity SOCIAL SCIENCE KNOWLEDGE 19 of the existing patterns of population settlement and community organi- zation to the proposed arrangements for supplying inputs, collecting outputs, facilitating access to education, health, or other social services, and for ensuring the desired distribution of project benefits. Appraisal should further ascertain that the project design is based on an accurate understanding of the social organization of productive activ- ities: (a) how the intended beneficiaries have access to, make use of, and exercise control over natural and other productive resources available in the area and what changes need to be promoted in the existing social arrangements; (b) how the characteristics of the household models and family systems prevalent in the area affect the development potential and constraints, labor availability, and ownership patterns; (c) whether small producers have reasonable access to and information on wider markets and regional economies; and (d) how land tenure systems and usage rights, as well as alternative employment opportunities, may affect intended beneficiaries' interest in the proposed project activities. In weighing the assumptions on which the project intervention is built, the social appraisal needs to ensure that the technological changes to be introduced will be complemented by supporting changes in local social organization patterns. Projects must be culturally acceptable, that is, understandable, agreed to, and capable of being operated and maintained by the local social actors and their institutions and organizations. For instance, projects for joining herders and cultivators in combined rangeland and farmland management schemes may not be workable if they ignore the history of relationships prevailing between the two given groups. Similarly, the health benefits of improved water supply or waste disposal systems may not materialize unless the intended beneficiaries appreciate the linkages. A judgment on the project's cultural acceptability and on the beneficiaries' willingness to contribute to its success must therefore take into account their values, customs, beliefs, and felt needs. The appraisal stage is not too late to carry out one more round of consultation and communication with the project area populations, thus facilitating various forms of public appraisal of the project as opposed to only expert appraisal. Moreover, the appraisal process must ascertain that the people likely to be affected-both positively or nega- tively-by the project were involved or consulted during the identifica- tion and preparation stages and that they will continue and expand their participation during project implementation, maintenance, operation, 20 SOCIAL SCIENCE KNOWLEDGE and monitoring. Whenever necessary, the project should contain explicit strategy provisions to help beneficiaries organize themselves to carry out these functions. If the social appraisal detenmines that the project is likely to be highly risky in social terms, but inadequate information is available to support a firm conclusion, consideration should be given to either a pilot project or postponement of the project until sufficient information is available. If certain technical, administrative, or other aspects of a project make it socially unfeasible, they will have to be modified or eliminated. Often it may be desirable to include in the project an information, motivation, and education component to help accelerate the necessary changes in social attitudes, behavior, and organization. The appraisal should ensure that the implementation process contains a realistic time frame and mechanisms for the expected behavioral responses to occur, and that there is enough built-in flexibility for making design changes in response to sociocultural information obtained during implementation. Project Implementation and Monitoring The actual implementation of the project opens up a new territory for applied social science activities-in fact the broadest area in which sociologists and anthropologists can and should work. They can bring their knowledge to bear on the organization, communication, and man- agerial realms of projects, on the shaping of project approaches to specific tasks, on mobilizing participation, and on daily problem solving. Implementation is precisely the stage when it becomes most obvious that projects never unfold linearly, exactly as planned, but involve changes, struggles between interested parties, and reinterpretations. How social scientists can work within this vast domain is illustrated in detail with actual project experiences throughout this volume. To summarize, the overview of the project cycle stages provides the development social scientist with an insider's key to open the entrance doors into the process of inducing social change through projects. I must emphasize, in light of my own and other researchers' experi- ences, that by adopting the project cycle model as a framework for conducting applied sociological project work, the social researcher does not have to and should not abdicate his critical thinking. He must not surrender any of his tools of trade-conceptual or methodological-or his ability to critique or reject one course of project action or another. To SOCLIL SCIENCE KNOWLEDGE 21 the contrary, anthropologists and sociologists can assert their views more influentially by reorganizing the conduct of their work according to the project cycle model and by becoming insiders to the project-making process-both intellectually and organizationally. Sociological knowl- edge can thus aspire to statutorily inhabit the project process, rather than be temporarily called in from the cold for an in-and-out, yo-yo-type contribution. Policy Formation Grounded in Social Knowledge The formulation of development policies is a substantively different terrain for the use of social science concepts and research tools, distinct from the project planning process and its key stages. The important principle to be stressed, as Solon Kimball reminded us, is that basic conceptualizations guide the gathering, organization, and analysis of data. Therefore, these conceptualizations must differ in the case of applied work in a project context, from, say, an ethnographic exercise, or from the conceptualization required for policy formulation. "The simple conceptualizations of traditional ethnographic descriptions pro- vide inventories of cultural items, but are inadequate for policy pur- poses."22 The demands and methodological options for using policy formulation as an "entrance point" have been discussed in detail, and arguments in its favor continue to be added from within both anthropol- ogy (see Weaver, 1985; Grillo and Rew, 1985)23 and sociology (see Hall and Midgley, 1988),24 obviating the need to repeat them here. In fact, the theoretical argument for putting people first in development, for designing strategies around the social actor rather than starting with technical factors, concerns primarily the shaping of developmentpolicy and secondarily the shaping of project design. The incorporation of socioanthropological knowledge into develop- ment policies is the most effective way of employing this body of knowledge, by far more effective than influencing one or another project piecemeal. It has a multiplier effect produced by the institutionalized recognition of sociocultural variables in the programs that follow policy formulation. Implicitly and explicitly, it causes multiple subsequent uses of social science knowledge in specific programs and projects. Two instances from the experience of sociological work in the World Bank prove this convincingly. 22 SOCLAL SCIENCE KNOWLEDGE The first is the case of the elaboration of a social science-informed policy for projects that cause involuntary population displacement and resettlement. Forced displacements imposed by the construction of dams, highways, or ports dismantle people's settlements and their prior mode of production, shatter community networks and patterns of social orga- nization, and cause homelessness, landlessness, and impoverishment. The seriousness of such negative consequences is compounded by the magnitude of forced displacement in certain projects: for instance, the Narmada Sardar Sarovar and Almatti dams in India will displace over 70,000 and 160,000 people respectively. The recently built Cirata and Saguling dams in Indonesia have displaced some 70,000 and 65,000 people, respectively; the Kayraktepe and Ataturk dams in Turkey, some 20,000 and 55,000 people, respectively; and the Sobradinho and Itaparica dams in Brazil, some 65,000 and 40,000 people, and so on. Over many years, anthropologists and sociologists have carried out research and generated knowledge on people's responses to, and conse- quences of, exogenously imposed displacement. Nevertheless, the avail- ability of knowledge-on-the-shelf about involuntary resettlement exercised hardly any influence on the governments and agencies engaged in the practice of forced displacement. Studies and books produced by social scientists kept accumulating, but they had little effect: they were by and large ignored by officials responsible for programs causing compulsory displacement. Development policies of most governments and major agencies, including the World Bank, did not formulate explicit demands that involuntary resettlement operations be carried out under more stringent criteria based on social science knowledge; displacement of people was usually dealt with last, as an afterthought. This lack of sociologically informed planning backfired during project implementa- tion stages, causing underfmanced and unplanned relocation to be exe- cuted disastrously as a last moment crash operation. What was primarily missing was not "more research for better knowledge." Already accu- mulated sociological knowledge was not incorporated into an institution- alized policy for involuntary resettlement, which could guide the agencies dealing with forced relocation. The significant turning point occurred when, for the first time, this body of knowledge was operationalized and translated into action guide- lines adopted by the World Bank as an explicit policy statement address- ing the "social issues associated with involuntary resettlement."25 This happened initially in 1979-80 (when the first statement was issued) and SOCIAL SCIENCE KNOWLEDGE 23 was followed in 1986, 1988, and 1990 with new, strengthened policy papers.26 The formulation of this policy was grounded in social science knowledge. The policy mandates that planning for projects that cause resettlement should start effectively with "putting people first," making the reestablishment of the living standards and productive capacity of those displaced a priority concern, favoring resettlement in groups, and protecting the interests of the host populations as well (these moves are treated in detail in chapter 6).27 Once a formal resettlementpolicy was instituted and enforced, the kind of social science knowledge that informs the policy became manifestly in demand. The policy itself explicitly prescribes the use of this special- ized knowledge and, in fact, was the lever that increasingly moved the "knowledge-on-the-shelf" into actual application in operational work. Knowledge publicly available for some three decades but largely ignored or underestimated was now suddenly put to use (thus confirming Zuckermann's observation that social science knowledge informs policy making in a "more diffused fashion" than do physical and biological sciences, "involving longer intervals and more complex chains of influ- ence"28). For instance, during 1985-90 more than 200 field missions that included one or more anthropologists or sociologists were sent by the World Bank to projects causing forced displacement, to assist in the preparation, appraisal, or imnplementation and supervision of the reloca- tion operations at standards defined by the policy. This involvement of sociologists or anthropologists in resettlement work in Bank- financed projects the world over represents to date the highest density of socio- logical presence in a single sector of World Bank lending. Even though formidable problems are always faced in resettlement, the practical results of the sociological contributions are visible and measurable in many projects. This does not mean that the social analysts were always free of mistakes in their work and certainly does not mean that it is sufficient to mobilize social knowledge to resolve all the major problems in such projects. Another comparable example of social science-informed policy is the adoption by the World Bank of its policy guidelines concerning tribal or indigenous populations in Bank-financed projects.29 Based directly on anthropological knowledge, these policy guidelines have directly con- tributed to a better protection of the cultural identity and socioeconomic rights-especially the demarcation of land rights-of vulnerable tribal groups and ethnic minorities inhabiting areas covered by projects. This 24 SOCIAL. SCIENCE KNOWLEDGE policy led to substantive corrections in Bank projects and in the treatment of such population segments by many borrowing agencies. Methodologies for Social Action The need to go beyond explanation to action and beyond assessments to recommendations, raises requirements that make the practice of applied sociology and anthropology a distinct enterprise. This enterprise is not science in the classic sense and often may not be linked to policy formulation either. It is "simply" applied sociological or anthropological work, bound to generate new and valuable "products" that are different from the typical research products of science. The concept of an "applied social scientist" who does not do "science" may at first seem paradoxical, but the contradiction is ontly apparent. The applied social science work often takes its practitioners far beyond what they have studied as trained social science professionals and beyond what has been conventionally regarded as the discipline's boundaries. Some practitioners of social and environmental impact assessment studies, for instance, have realized that important as such assessments are, "the assessment process is ... neither science nor policy, despite the conven- tional wisdom which views it as a scientific procedure that lays the groundwork for well-informed policy."30 And indeed, this is true not just for assessment studies, but for a much larger part of applied sociological and anthropological work. Moreover, much of the applied anthropolog- ical work cannot be called applied research either, as it often is labeled, simply because not every application of anthropological knowledge involves new research. This opinion, however, is not shared by all those who do applied sociology or anthropology, and a bit more consensus about the nature of this endeavor would much help this subfield. But if this view is accepted, it clearly ensues that new types of "products" of applied social science must be invented and defined. One such product is the methodologies for social action. Such meth- odologies may be seen as a kind of applied social science product that is intermediate between general policy work and individual project (or case) social analysis. While it is essential to contribute to defining broad or sectoral development policies that indicate directions and goals, the formulation of policy does not usually include the elaboration of detailed methodologies for action to reach those goals: articulating them requires SOCLAL SCIENCE KNOWLEDGE 25 a distinct professional effort. Project-focused analyses must be specific- ally tailored to the given set of circumstances. Yet many social processes, under many different programs, share similar basic characteristics, and consequently the types of social actions required are in essence similar as well. For situations that are more or less recurrent (save different particular conditions), the elaboration of social action methodologies can supply what is missing in both general policy analysis and piecemeal project-focused social analysis. Such methodologies then can (must) be adjusted to individual situations, thus making tested approaches opera- tional and avoiding repetitive work in each single instance. A case in point, which demonstrates the acute demand for social action methodologies, is the impasse with participatory planning for rural development. Among obstacles is also the absence of tested methodolo- gies for organizing people's participation. Now we often hear sudden declarations of fashionable support for participatory approaches from politicians, planners, economists, and technocrats. Social scientists should not confuse these statements with actual participatory planning because, under the cloud of cosmetic rhetoric, technocratic planning continues to rule.3' The rhetoric of intent is still far ahead of the design for action to promote participation. However, government officials or planners are not the only ones to blame for this gap. Anthropologists and sociologists have been busier advocating participation than working out social techniques for organiz- ing it. But without the know-how to organize it, participation will remain a hot ideology lacking a social technology. Many developing countries have authoritarian regimes that place structural and political restraints on grassroots participation, but often more participation is feasible even within existing political limits. We must ask and resolve pragmatic questions. Are the social sciences able to offer a methodology for organizing actual participation in different cultural contexts? Do social scientists have sets of procedures and methods transferable to planners and managers? What should be done during project preparation to shape the project so that it elicits and depends on participation? What should be done for organizing participation during implementation? An interesting initiative to design such a social methodology was taken by a group of anthropologists, sociologists, and planners under the PiDER development project in Mexico. Against many odds, the team developed a model for participatory investment planning at the community and municipality levels. The team wanted to replace the top-down, paternal- 26 SOCIAL SCIENCE KNOWLEDGE istic decisions imposed from on high with a methodology for eliciting farmers' own proposals and choices and for mobilizing their initiatives and material resources. The innovation was to develop this model itself not by a desk-bound effort but through action-research-actual tests in the field and community planning experiments-followed by iterative returns to the drawing boards.32 The "product," the new methodology, consisted of a conceptual framework and a set of procedures, rules, and approaches that instituted field assessments, information exchange, and structured interaction between local and outside experts. Application guidelines and field manuals for planners were prepared. This method- ology was then applied not only to many PiDER areas, but to planring for the entire Zacatecas state (over 1,000 municipalities). Such an approach can be replicated in other contexts where action research for this or other issues is needed, using similar social science-craftsmanship. The interest in generating such methodologies usable in development interventions, still incipient, is nevertheless growing. Based on field- work, the Institute for Development Anthropology (IDA) has designed a methodology for identifying optimal locations for the siting of deep wells to provide potable water to dispersed rural homesteads and small settle- ments in Central Tunisia. Represented as a series of thematic maps overlaid on satellite imagery, the methodology employs sociological, demographic, and landuse criteria, as well as the more customary hydro- logic and financial criteria, and takes into account projected population growth, political or administrative divisions, and environmental capaci- ties. The institute also advised the government on the creation of organ- ized, self-managed groups of potable water users, Unites d'Autogestion, to assume responsibility for the operation and maintenance of the new water points, fee collection, and the allocation of water surpluses. This social methodology appeared so effective that the government of Tunisia has invited IDA to participate in elaborating a national strategy for water user groups based on the Central Tunisia model.33 Another example comes from Sri Lanka's Gal Oya project area. Based on prior findings from the sociology of irrigation, and on action-research in the area, a team led by Uphoff developed, jointly with local water user associations, a social methodology for detennining small group capacity to undertake development tasks and for building up this capacity.34 A comparable methodology is described in chapter 3. A domain that urgently calls for such methodology creation is organi- zation building at the grassroots. All over the world, the degree of formal organization in rural communities lags far behind that of urban popula- SOCIAL SCIENCE KNOWLEDGE 27 tions. This is afundamental characteristic of rural underdevelopmentthat accounts largely for the vulnerability of rural societies. Many rural programs collapse for want of grassroots organizations able to foster collective actions (yet the same programs seldom attempt to establish organizations that aggregate and enhance individuals' capacities). Farmers' organizations, pastoral associations, credit groups, and water users' organizations, for example, are all critical for development, but the methods and knowledge to help construct them on a large scale have not been codified and made widely available. Informal organizations existing in traditional societies can sometimes be used as a matrix for building stronger formal organizations.35 High-yielding social organiza- tions are no less important for development than high-yielding crop varieties, and intensified agriculture cannot occur without intensified human organization. Sociological methodologies for building farmers' organizations or revitalizing existing ones are scarce. Sociologists and anthropologists should recognize this as a broad opportunity for institu- tional innovation. There are, of course, complex epistemological and ethical questions about such methodologies as those mentioned in the examples from Mexico, Tunisia, and Sri Lanka. Would they be operationally valid across cultures? Any extrapolation would require testing adjustments and critical learning processes before they are actually offered for application. The ethical legitimacy of preparing such methodologies is questioned by some, as on a more general plan the validity and legitimacy of action-research has also been both contested and defended in recurrent debates within the professional social science-community. Denying on ethical grounds the legitimacy of engagement in policy formulation and program work has undercut rather than empowered the transforming influence of social science. As has been correctly emphasized by many social scientists, given "the enormity of social illfare in modem times, .... it is morally indefensible to adopt a noninterventionist stance when human suffering is all pervasive and when sociological technology has meliorative relevance."36 Although these issues are not argued here in detail, my stand in brief is that action-oriented sociological work is ethically legitimate and that such sociologically informed methodologies for action are epistemologically feasible. The extent of crosscultural regularities in agrarian production patterns and social structures sets both the ground for, and the limits of, this feasibility. Overall, these new applied scienceproducts, including methodological instruments, are insufficient and are being elaborated more by happen- 28 SOCIAL SCIENCE KNOWLEDGE stance than by design. This mirrors the underdevelopment of develop- ment sociology and anthropology. The situation also largely reflects the fact that the academic mainstream of these disciplines gives little con- sideration to explicitly working out the theoretical and epistemological tenets of their applied domains. At the beginning of the 1980s, a broad review of advances in develop- ment anthropology concluded that "anthropologists working in develop- ment have not yet created an academic subdiscipline, 'development anthropology', for their work is not characterized by a coherent or distinctive body of theory, concepts, and methods. Development anthro- pology has, however, become an incipient profession and field of study ... (and) has produced a body of technically informed, substantive findings...."37 More or less the same can be said about development sociology, although the theoretical work is here more advanced. Further- more, there is unacceptably little intellectual exchange and synergy between sociologists and anthropologists working on development. Al- though some significant advances have been made during the 1980s toward accumulating building blocks for a subdiscipline, the beginning of the 1990s finds the valid demand for a distinctive body of theory, concepts, and methods still wanting. The idiosyncratic contribution of an individual anthropologist or soci- ologist to a certain project may be very valuable, but if it is mainly the product of this individual alone rather than the translation of a systematic methodology, it remains a piecemeal and particularistic contribution. Development agencies often have to rely excessively on a sociologist's personal aptitudes and on the accident of his or her flair and inspiration in the field, rather than on the discipline's methodological and conceptual tools. This reflects the infancy of the discipline itself. Although the creativity, intuition, and ad hoc judgment of the social analyst are critical for the project (and help develop the discipline itself), in the long term it is essential to have a systematic body of sociological know-how that is transferable and usable in operational work by sociologists and non- sociologists alike. Unless such methodologies are developed, behavioral sciences connected with development will only advance slowly. Two Models: "Enlightenment" and "Social Engineering" In discussing how social science knowledge can influence society more effectively, a distinction is often made between two models: the SOCIAL SCIENCE KNOWLEDGE 29 enlightenmentmodel and the engineering model.38 Generally, the differ- ences between the two get stressed, while their complementarity is overlooked. Enlightenment counts on dissemination of sociological knowledge through education, which is a useful but obviously insuffi- cient strategy. Enlightenment alone implies a tortuous, uncertain, and slow way to return the benefits of social knowledge to society and influence its progress. Moreover, the enlightenment model postulates the dissemination of findings and conclusions as available in academic social science, but it does not respond to the need of operationalizing social knowledge for action purposes. For this and other reasons, complemen- tarity between the two models is necessary. The social engineering action model is rooted in knowledge of the social fabric and dynamics. It postulates the translation of social science knowledge into new know- how and change tools, and it uses this knowledge purposively to organize new social action and relationships. For a while, social scientists have shied away from using the very concept of social engineering to avoid the unwarranted twin misinterpre- tations of applied social science as manipulation or as condescending paternalism. Lately, however, as the action-orientation in social research matures, the term is returning with renewed intellectual strength and validated usefulness. In fact, as Hirschman noted, it was on the crest of radical thinking and revolutionary action that the very idea of social engineering and of the perfectibility of the social order first arose- namely, at a time when "it was . . . novel to think that human happiness can be engineered by changing the social order."'9 Since that time, social knowledge has grown more precise, expanded, and deepened to levels at which it is increasingly usable as a guide to plan and as a means to democratize the planning process itself by facilitating broader participa- tion in it of the development actors themselves. This meaning of social engineering is ethically and professionally acceptable. What should be avoided and opposed is not social engineering itself, but its abuse for reprehensible goals. As Rossi and Whyte wrote in a balanced definition, social engineering consists of attempts to use the body of sociological knowledge in the design of policies or institutions to accomplish some purpose. Social engineering can be accomplished for a mission-ori- ented agency or for some group opposed to the existing organizational structure, or it may be undertaken separately from either.... When conducted close to the policy-making centers, it is often termed social policy analysis.... When practiced by groups in opposition to current regimes, social engineering becomes social criticism.40 30 SOCIAL SCIENCE KNOWLEDGE It is worth noting that nowadays social engineering is by far not a monopoly of social scientists. In fact, social scientists do a very limited amount of the "social engineering" that inevitably occurs in every plan and social policy decision in contemporary society. When technical specialists, economists, or managers decide on a development program without so much as consulting a social scientist, they do ad hoc social engineering of their own. When subsequently they implement that plan, they again do social engineering. Unfortunately, however, most often they aren't even aware that at least part of what they do is social engineering, the same way Moliere's Monsieur Jourdain wasn't aware that he was making prose everyday, by simply speaking. However, just as Monsieur Jourdain made rather bad, ungrammatical, and halting prose, present day planners and technicians quite often do poor social engineer- ing, unassisted by the professional competence derived from sociological and anthropological knowledge. Planning agencies or policy bodies should be wary of relying on mechanical engineers to do social engineering. They should ensure the same level of professional competence for the social components of projects as they provide for the technical components. The applied social scientist who responsibly takes on the challenge of social engineering, for instance by planning for social development, provides an important service: he or she replaces the amateurish, do-it-yourself brand of social engineering of the nonsocial scientist with the state-of-the-art tools of understanding offered by a field of professional expertise. Certainly, social engineering does not decide upon or establish the goals of devel- opment. But with a clear understanding of what "putting people first" signifies, it can be employed to chart the relationship between means and goals in programs. For instance, in poverty-oriented rural development programs, it can bring about better strategies to support poverty allevia- tion. Another area concerns local institutions, where informal leadership identified through ethnographic analysis can provide efficient structures for reaching villages. Although social engineering is only one way in which behavioral sciences can influence social action, it is the one that compels social scientists to descend from the realm of generalities to produce operation- ally usable know-how. It also requires sociologists to think through carefully the consequences of their recommendations rather than assume condescendingly that they know what is best for the people. In this vein, the sociologist should conceive of his or her role as not just a producer SOCIAL SCIENCE KNOWLEDGE 31 of expert solutions, but rather as a facilitator whose task is to free and bring to bear the huge innovation potential of the "plannees" themselves. The social scientist is the only kind of expert who is professionally trained to "listen to the people." Social knowledge thus developed becomes a "hearing system" able to amplify the listening for managers and policy- makers, too. Hard dilemmas and controversial tradeoffs confront the development researchers doing social engineering. Sometimes they are required to compromise in their quest for data, to leap over unknown parameters and, yet, to provide their best judgment and advice with only imperfect information. But it is simply not possible to always know everything before doing anything. Dilemmas caused by imperfect knowledge are best addressed when recognized squarely and realistically, without ex- aggerated claims about the "scientism" of all that is done under the auspices of development anthropology or sociology. There must be concern for building in learning mechanisms and flexible adjustment procedures. The strictures of the planning process itself will never easily allow the ideal setting for generating and using social knowledge. Experienced sociologists and anthropologists have struggled against such strictures; they have emphasized the need for applied work to both live with and overcome imperfect knowledge. Reflecting on his own work as sociologist-planner and simultaneously "certified" scholar, Her- bert J. Gans described well the practitioner's dilemmas and the need for creative answers: Most of the questions which must be answered before planning can take place on a rational basis have not yet been sufficiently studied; yet the planners cannot wait for further research. Sociologists who participate in guided mobility programs must be able to come to conclusions on the basis of past research, a modicum of impressionistic observations, and a large amount of freewheeling hypothesizing-that is, guessing. They must gamble further by being willing to build the products of this highly unscientific approach into experimental programs. There is no doubt that this type of sociological endeavor will lay the practitioner open to criticism from colleagues in the discipline as being unscientific or controversial, but it will be countered by appreciation and the surrender of an ancient stereotype about the unwillingness of sociologists to come to conclusions on the part of the planner. More- over, the sociologist must revamp the concepts that he uses so that they 32 SOCIAL SCIENCE KNOWLEDGE can answer questions posed by the plan and in such a way that they will lead to ideas and techniques for action programs.41 Breaking new ground by working on the cultural variables of major development programs, applied social researchers have the rare chance of making "social inventions," to use Whyte's felicitous concept; this is what they do, for instance, when they chart "new sets of procedures for shaping human interactions and activities and the relations of humans to the natural and social environment"42 or when they develop new "incen- tive structures"43 apt to help improve the involvement of social actors in purposive development activities. Without discounting the caution dictated by limited knowledge, soci- ologists should become more operationally prescriptive. Sociologists and anthropologists often have many hundred "don'ts" to only five "do's." By focusing on program design and execution as entrance points, social scientists will force themselves to be more pragmatic, more operationally useful, and more versatile in development work. They will also become aware of variables and relationships that otherwise would have escaped their attention. In sum, my overall argument is that applied social scientists have to learn to generate new products that are usable by development practi- tioners. These new products should not be regarded as replacements for the traditional products of research (such as taxonomies, explanatory hypotheses, concepts, and theories), but as supplementing them with methodologies for social action. Such new social science products and approaches-whether they are called social technologies, social engi- neering, sociotechniques or, in Firth's term, human engineering"- would respond to the needs of induced development, enrich the traditional spectrum of social science products, and make social knowl- edge more effective. The scope for such creative contributions is virtually limitless and they are needed urgently. Institutionalizing Development Social Science To generate new intellectual products, the noneconomic social disci- plines have to work hard in their own gardens. Expansion of their research agenda and more concern with their own institutionalization are essential. SOCIAL SCIENCE KNOWLEDGE 33 Promising new research areas and issues are emerging, both within sociology and anthropology and at their frontiers with other sciences: agricultural sociology; common property resource management; collective action; faniing systems research; crop sociology; social forestry; the devel- opment role of the state and of the nongovermnental organizations; and others. Every one of these research areas is relevant for induced develop- ment and holds substantial potential for expanding applied social research. The metaphor of entrance points in the process of planning develop- ment has obvious implications for the institutionalization ofdevelopment social science and the actual settings within which development sociol- ogy and anthropology are practiced and taught. The process of institu- tionalization of development sociology and anthropology can be conceived, paraphrasing Robert Merton's analysis of the sociology of science, as a process of intensified "interplay between the cognitive and the professional identities"45 of the development social sciences. Solidifying our "cognitive identity" requires not just applied work; to a decisive extent, this cognitive identity will depend upon the general progress of basic theory and research in sociology and anthropology. Indeed, it is on the shoulders of basic research only that applied research- ers can have the secure pedestal of a distinct body of knowledge, supplying them the lenses to see the wider horizons. In turn, solidifying our "professional identity" requires other proc- esses; in this respect, Merton has called attention to the definition of institutionalization suggested by Edward Shils that meticulously cap- tures a set of organizational and intellectual dimensions: By institutionalization of an intellectual activity I mean the relatively dense interaction of persons who perform that activity. The interaction has a structure: the more intense the interaction, the more its structure makes place for authority which makes decisions regarding assess- ment, admission, promotion, allocation. The high degree of institution- alization of an intellectual activity entails its teaching and administered organization. The organization regulates access through a scrutiny of qualification, provides for organized assessment of performance, and allocates facilities, opportunities, and rewards for performance-for example, study, teaching, investigation, publications, appointment, and so forth. It also entails the organized support of the activity from outside the particular institution and the reception or use of the results of the activity beyond the boundaries of the institution.46 34 SOCIAL SCIENCE KNOWLEDGE In the case of development sociology-anthropology, at least three processes appear critical: first, the position of the social analyst should be formally institutionalized within the organizational settings of techni- cal, administrative, and development agencies; second, substantive changes must be made in the training of sociologists and anthropologists oriented to development work; and third, the equivalent of a sociological renaissance is needed in university curricula for training technical spe- cialists and economists for development work. As long as professional social researchers remain outside technical and administrative agencies, knocking on physical doors to gain intellectual entry, the actual use of sociological knowledge in planned development will be hampered by more obstacles than if sociologists were among the insiders. Their inclusion is necessary to reduce organizational ethnocen- trism on both sides. Some agencies have begun to institutionalize socio- logical skills, yet these cases are still few and far between. National (governmental) and international organizations are lagging behind de- spite their professed creed or given mandate. There are, for instance, many livestock departments in agricultural ministries all over the world, which are properly staffed with veterinarians to deal with cattle but lack any sociological staff trained to understand the social organization of pastoral populations. Staff sociologists could certainly enhance the ca- pacity of these agencies to work with cattle owners and thus improve animal husbandry. It is of exceptional importance that indigenous sociologists and anthro- pologists from developing countries participate intensively in applied development activities. Unfortunately, however, this is still far from happening. In India, for instance, and this is true for virtually all devel- oping countries, "most organizations, including those that recognize the value of anthropological contributions, are functioning without qualified anthropologists on a regular basis.... Obviously this situation must change if government-directed development is to benefit from anthropo- logical knowledge."47 I have no naive illusions that the inclusion of sociologists or anthro- pologists in technical settings will solve all social problems, but in their absence many programs remain socially underdesigned and register a high rate of economic, technical, and sociopolitical failure. Cooperation across disciplinary fences is difficult enough; across additional bureau- cratic walls it becomes virtually impossible. The issue is not just one of SOCIAL SCIENCE KNOWLEDGE 35 philosophical recognition but also one of resource allocation. The social scientist may have to play second violin to the technical experts, which is perfectly acceptable; in other circumstances he may fulfill the role of project manager;48 but only the well-orchestrated joint efforts of techni- cal and social experts can produce harmonized development work. The institutionalization of the social professions will generate various pat- terned models of interaction with other disciplines and enhance the quality of development planning. The academic training of sociologists and anthropologists should be profoundly restructured if producing professionals with an action-ori- ented outlook is to be addressed responsibly. Enough has been written on this issue to make repetition unnecessary, yet it is unfortunate that the social science academic establishment reacts so slowly to this impera- tive.49 True enough, all the textbooks for training such action-oriented sociologists and anthropologists are not yet on the shelves.50 But there is little time to wait. Moreover, if the opportunities opened up by the pressure of practical demands are used, then both empirical and analyt- ical materials for textbook synthesis would accumulate faster. Last but not least, in my own experience at the World Bank and in different countries, an enduring obstacle to the influx of sociological knowledge into development work has been that many technical experts lack understanding of what social science and social engineering could bring to their own efforts. The magnitude of this intellectual obstacle on a global scale is underestimated. The gap persists. In fact, it is being recreated with every class graduating from technical and economic institutes, because of the manner in which technical experts are "grown" in the groves of academe. Biologists and economists, agronomists and veterinarians, urban planners or foresters, and industrial or irrigation engineers who tomorrow will have a strong say in the design and execution of development programs, are often being trained today as though people did not matter for the solution of technical issues. Thus, they remain ignorant of the sociostructural and cultural dimensions of technical/productionprocesses because of outdated training philosophies and practices. The experts produced by this training are being deprived of a crucial lens-the social one-for looking at and understanding their own tech- nical field. They are not being prepared to cooperate later with the social experts, don't know what to ask from them, and remain unaware of what 36 SOCIAL SCIENCE KNOWLEDGE they are entitled, as technical specialists, to receive from the social specialist. Correcting this situation is not a task for only a year or two, but rather will require a generation at least. If any renaissance is in store for social sciences in their development role, it will not take place unless social science knowledge (not just introductory principles, but the sociology of the specific subarea of technical activity) is diffused among technical specialists as well. Teach- ing social sciences to students in fields other than sociology and anthro- pology is at least as important and consequential as teaching future sociologists. To sum up, putting people first is not simply a fashionable slogan but is a formidable work program for social sciences. It is also a heuristic device demanding always that we identify, in every seemingly "techni- cal," "financial," or "administrative" intervention, the sociological angle and the variables pertinent to the social organization affected or targeted by the intervention. Sociologists have to face the nuts and bolts of development activities, to roll up their sleeves and deal with the mundane, pragmatic questions of translating plans into realities in a sociologically sound manner. They need to link data generation, action-oriented re- search, social analysis, design for social action, and evaluation into a continuum, and thus stretch sociology's contributions far beyond simple pronouncements. The planning models for rural development are far from perfect, and although sociologists should learn to work within existing frameworks they must at the same time change them with their input. Financially induced change programs need sociological knowledge and must incor- porate the sociocultural variables. Financial resources are not necessarily the key ingredient in all development programs. Sometimes they are the least important. The range of entrance points for sociological knowl- edge and skills should be expanded to all segments of development planning, from policymaking to execution and evaluation, and from theorizing to social engineering. It is essential to design purposively for social action. The conventional range of operationally usable products generated by social scientists is still narrow and insufficient; forward-looking action methodologies should enrich the domain. The support for participation will be more effective if passionate advocacy is accompanied by social methodology. The newly emerging research orientations are more inter- disciplinary than the old ones; they deserve the support and commitment SOCIAL SCIENCE KNOWLEDGE 37 of development-oriented social scientists. Training philosophies must change as a crucial step to avoid producing new cohorts of socially incompetent technical experts or technically illiterate sociologists. Such changes and the new orientation toward increasing the action relevance of the social sciences will result in a better response to the fundamental calling of social sciences: not only to analyze and explain, but also to assist in transfonning society and improving people's lives. Notes 1. For the present argument, I use the terns sociology and social anthropology interchange- ably. I believe that the broad substantive overlap between what sociologists and social anthropologists actually do within the framework of development programs justifies this use; the differences between the overall perspectives of the two disciplines, although real, are not treated here. References to social sciences in this paper should be generally read as references to nonecononic social sciences-particularly anthropology and sociology-but do not imply that there are no differences in approach between disciplines. 2. Other such intemnal factors, which cannot be elaborated here, but deserve wider discus- sion, include the state of the discipline's theory and body of knowledge, the quality of its practitioners' applied craftsmanship, the patterns of their professional fomnal organization, and obviously their views and value judgments on whether or how applied social research should be conducted. For the status of applied anthropology in the United States, see a broad and very instructive historical overview on disciplinary growth and intemal problems in: William L. Partridge and Elizabeth M. Eddy, "The Development of Applied Anthropology in America," in E. M. Eddy and W. L. Partridge, eds., Applied Anthropology in America, 2d ed. (New York: Columbia University Press, 1987). 3. Walter Goldschmidt, "Anthropology as a Policy Science," in W. Goldschmidt, ed., Public Policy: A Dialogue (a special publication of the American Anthropological Associa- tion, no. 21, 1986). 4. Anthony HalL "Sociology and Foreign Aid: Rhetoric and Reality," in A. Hall and J. Midgley, eds., Development Policies. Sociological Perspectives (Manchester and New York: Manchester University Press, 1988). 5. Vemon W. Ruttan, one of the few notable American economists who set out explicitly to explore what development economics "can leam from anthropology,' noted that "almost no attention has been devoted by economists to the role of cultural endowments." He went on to observe that "professional opinion in economics has not dealt kindly with the reputations of those development economists who have made serious efforts to incorporate cultural variables into development theory or into the analysis of the development process. Their work has typically been favorably reviewed and then ignored. ... But in spite of the failure of research on the economic implications of cultural endowments to find a secure place in economic development literature or thought, the conviction that "culture matters" remains pervasive in the underworld of development thought and practice. The fact that the scholars and practitioners of development are forced to deal with cultural endowments at an intuitive level rather than in analytical terms should be regarded as a deficiency in professional capacity rather than as evidence that culture does not matter." (See "Cultural Endowments and 38 SOCIAL SCIENCE KNOWLEDGE Economic Development: What Can We Learn from Anthropology?" in Economic Develop- ment and Cultural Change, vol. 36, no. 3, 1988, pp. 250, 255-56). 6. For an analytical effort at "deconstructing" the concept of planned intervention and at reformulating its theoretical underpinnings, see Norman Long and J. D. van der Ploeg, "Demythologizing Planned Intervention: An Actor's Perspective," in Sociologia Ruralis, vol. 29, no. 3/4, 1989 (Van Gorcum, Assen, the Netherlands). 7. Albert 0. Hirschman, Development Projects Observed (Washington, D.C.: Brookings Institution, 1967). 8. Dennis A. Rondinelli, Development Projects as Policy Experiments. An Adaptive Approach to Development Administration (London and New York: Methuen, 1983); see, in particular, Rondinelli's analysis of projects as public policy vehicles in social experimentation. 9. Yves Goussault suggests the following as a possible definition: "The sociology of development is a sociology of intervention. It is basically linked to the social changes provoked by the interventions of states and capital in various social sectors as well as in the overall structures of societies. In this capacity, it is a sociology of strategies...." (Yves Goussault, "Olenestla sociologie du developpement?"Revue TiersMonde, vol. 23, no. 90, 1982, p.242). 10. See Robert Chambers for a provocative discussion of the concept of professional "reversal" needed in development work. "Bureaucratic Reversals and Local Diversity," IDS Bulletin, vol. 19, no. 6 (1988). 11. Mark Granovetter, "Economic Action and Social Structure: The Problem of Embeddedness,"AmericanJournalofSociology, vol. 91, no. 3, 1985. 12. The following comment was made by an anthropologist who reviewed an earlier version of this chapter: "Indeed, money is far from being everythmig. In the large institutional development project I did for the Treasury in Saudi Arabia, I discovered that even given unlimited funds, certain social and organizational changes were often impossible through money alone. In fact, the funds themselves became a problem...." (Theodore E. Downing, letter to the author). 13. Robert K. Merton, "The Role of Applied Social Science in the Fomiation of Policy: A Research Memorandum,"Philosophy ofScience vol. 16, no.3, 1949, p.178 (emphasis added). 14. Michael M. Cemea, "Farmer Organizations and Institution Building for Sustainable Agricultural Development," in Regional Development Dialogue, no. 2, 1988. 15. Vemon W. Ruttan, "Social Science Knowledge and Institutional Change," American Journal ofAgricultural Economics, December 1984. 16. Robert A. Scott and A. R. Shore, Why Sociology Does Not Apply: Sociology in Public Policy (New York and Oxford: Elsevier, 1979), p. 35. 17. Scott and Shore 1979, p. 2. 18. Edward C. Green insightfully captured such self-defeating disciplinary biases, when he wrote: "for their part, anthropologists may view economists, agronomists, engineers, and the like as narrowly focused technicians whose rigid professional mindsets and cultural distance from local populations prevent them from coping with human factors or seeing the larger picture, especially when dealing with cultures quite different from their own." (Themes in the Practice of Development Anthropology, in Edward C. Green, ed., Practicing Develop- mentAnthropology, Boulder and London: Westview Press, 1986, pp. 7-8). 19. Warren C. Baum, The Project Cycle (Washington, D.C.: World Bank, 1982). 20. For a detailed description of the preparation stage for an involuntary resettlement project, see Michael M. Cemea, InvoluntaryResettlement in DevelopmentProjects, Annex 1, World Bank Technical Paper No. 80 (Washington, D.C., 1988). 21. In 1984, explicit guidelines to analyze these four sets of factors during the appraisal of projects were formally introduced in the World Bank's intemal policy and procedural SOCIAL SCIENCE KNOWLEDGE 39 directives. These guidelines are mandatory for Bank staff; they also strongly influence the work of planning staff of borrowing agencies in developing countries. These formal appraisal guidelines mandate a more in-depth analysis of the basic social variables than what was described as the "social aspects analysis" in the well-known manual published by the Bank: Economic Analysis of Agricultural Projects, edited by J. Price Gittinger (Baltimore and London: Johns Hopkins University Press, 1982), see pp. 15-16. 22. Solon T. KimbalL "Anthropology as a Policy Science," in E. M. Eddy and W. L. Partridge, eds., Applied Anthropology in America, 2d ed. (New York: Columbia University Press, 1987, emphasis added). 23. See Thomas Weaver, "Anthropology as a Policy Science, Part I and IL" Human Organization, vol. 44, nos. 2,3 (1985). See also Ralph Griuo and Allan Rew, eds., Social An- thropology and Development Policy (London and New York: Tavistock, 1985). 24. Anthony Hail and James Midgley, eds., Development Policies: Sociological Perspec- tives (Manchester and New York: Manchester University Press, 1988). 25. The statement, "Social Issues Associated with Involuntary Resettlement in Bank- Financed Projects," became part of the World Bank's intemal operational manual in February 1980. 26. Michael M. Cernea, "Anthropology, Policy and Involuntary Resettlement," British Association for Social Anthropology in Policy and Practice Newsletter 4(1989). 27. The basic tenets of this policy are presented in detail in the paper "Involuntary Resettlement in Development Projects: Policy Guidelines in Bank-Financed Projects" by Michael M. Cemea, World Bank, Agriculture Department, Washington, D.C., 1988. See also chapters 5 and 6 of this volume. 28. Harriet Zuckermann, "Uses and Control of Knowledge: Implications for the Social Fabric," in James F. Short, Jr., ed., The Social Fabric: Discussions and lssues (Beverly Hills: Sage Publications, 1986). 29. These guidelines, "Tribal People in Bank-Financed Development Projects," became part of the World Bank's intemal operational manual in 1982. 30. Steven McNabb, "Logical Inconsistencies ...," inHuman Organization, vol. 48, no. 2 (1988). 31. The question skeptically asked by Gelia Castillo-"how participatory is participatory development?"-is warranted and should be asked about every development program. (See How Participatory Is Participatory Development? A Review of the Philippine Experience (Manila: Institute for Development Studies, 1983). 32. See more details in Michael M. Cemea, "The 'Production' of a Social Methodology," in E. M. Eddy and W. L. Partridge, eds., Applied Anthropology in America (New York: Columbia University Press, 1987). 33. Michael M. Horowitz, personal communication, January 1990. Several reports dealing with this activity are available from the iDA: among these are Gordon Appleby, "Criteria and Methodology forthe Delimitation of Water-Short Areas in Central Tunisia," 1987; Mohamed Fakhfakh and others, "Projet Eau Potable en Tunisie Centrale," Cartographie des Ressources en Eau et Population," 1987; Nicholas S. Hopkins, "Les Associations d'Usagers de l'Eau Potable dans la Tunisie Centrale, et la nouvelle Unite d'Autogestion au sein de l'or," 1986. A final report by Muneera Salem-Murdock and Michael M. Horowitz describing how this methodology was produced is forthcoming. 34. Norman Uphoff, "Participatory Evaluation of Farmer Organizations' Capacity for Development Tasks," Agricultural Administration and Extension, vol. 30 (1988). Uphoff suggests that the methodology should be adaptable and generalizable for group credit pro- grams, range or forest management, agricultural extension, and so forth. "All that is needed is 40 SOCLAL SCIENCE KNOWLEDGE some modification in the questions that are formulated, though not even in the way the questions are derived and pretested" (p. 44). 35. Milton J. Esman and Norman T. Uphoff have extensively treated the topic of local organizations and their potential in Local Organizations-Intermediaries in Rural Develop- ment (Ithaca, New York: Comell University Press, 1984). 36. James Midgley, "Sociology and Development Policy," in Anthony Hall and James Midgley, eds., DevelopmentPolicies: Sociological Perspectives (Manchester and New York: Manchester University Press, 1988). 37. Allan Hoben, "Anthropologists and Development," Annual Review of Anthropology (1982), pp. 349-50. 38. Morris Janowitz, Political Conflict: Essays in Political Sociology (Chicago: Quad- rangle Books, 1970), p. 247. 39. Albert 0. Hirschman, "Rival Interpretations of Market Society: Civilization, Destruc- tive or Feeble?" Journal ofEconomicLiterature, vol. 20 (December 1982), p. 1463. 40. Peter H. Rossi and W. F. Whyte, "The Applied Side of Sociology, 'in H. E. Freeman, R. R. Dynes, P. H. Rossi, and W. F. Whyte, eds., Applied Sociology: Roles and Activities of Sociologists in Diverse Settings (San Francisco and London: Jossey-Bass Publishers, 1983), p. 10. 41. Herbert J. Gans, "Urban Poverty and Social Planning," in P. F. Lazarsfeld, W. S. Sewell, and H. L. Wilenski, eds., The Uses of Sociology (New York: Basic Books, 1967), p. 448. 42. William Foote Whyte, "Social Inventions for Solving Human Problems," American SociologicalReview, voL 47 (February 1982). 43. Edward H. Greeley, "Project Development in Kenya," in Edward Green, ed., Practic- ing Development Anthropology (Boulder: Westview Press, 1986). Greeley underscores that the first lesson he learned from his experience as an anthropologist working on usAm projects in Kenya was the importance of focusing "on the incentives of individuals, groups, and institutions." 44. Raymond Firth, "Engagement and Detachment: Reflections on Applying Social An- thropology to Social Affairs ,"Human Organization, vol. 40, no. 3 (1981). 45. Robert K. Merton, The Sociology of Science: An Episodic Memoir (Carbondale and Edwardsville: Southem Illinois University Press, 1979), p. 7 (emphasis added). 46. Edward Shils, cf. Merton (1979). 47. H. Mohan Mathur, Anthropology and Development in Traditional Societies (New Delhi: Vikas Publishing House, 1989). 48. Gerald Murray discussed the role of the anthropologist as manager in a social forestry project in Haiti that was managed over several years by several anthropologists in sequence (see G. Murray, "The Domestication of Wood in Haiti: A Case Study in Applied Evolution," in R. M. Wulffand S. J. Fiske, eds.,AnthropologicalPraxis (Boulder, Colo.: Westview Press, 1987). 49. W. Goldschmidt has been among the few to raise his voice in criticism regarding the responsibility of the "more senior anthropologists" who have cultivated remoteness from applied tasks among their students: "Anthropologists. . . have not prepared themselves for the serious and difficult task of translating their deep understanding (of national cultures) into the workday realities of decision making and the crossfire that goes with such a role. Or perhaps, I should say, we more senior anthropologists have failed them, we have failed to prepare them for such tasks, and thus it is we who are to blame for not making them ready when they are wanted. Rarely in our curricula are there programs that translate theories into actionable SOCIAL SCIENCE KNOWLEDGE 41 policies.... It is necessary for the anthropologists to prepare their students forpublic service" (Goldschmidt, 1986, p. 4). 50. For a notable exception, see William L. Partridge, ed., Training Manual in Develop- mentAnthropology, Publication No. 17 (Washington, D.C.: American Anthropological Asso- ciation and Society for Applied Anthropology, 1984). PART I Irrigation Projects Editor's Note About one-third of the world's food supply is being produced from only one-sixth of its cultivated area-the lands that have been provided with irrigation. Globally, irrigated areas more than tripled between 1950-90. Satisfying the increasing demands generated by both population growth and higher per capita consumption will heavily depend upon improving the productivity of both irrigated and rainfed lands. The water flowing in irrigation systems is as much a social product of human organization as it is a natural commodity. As the authors of the following three chapters jointly argue, sociological issues are embedded in the operation of all irrigation systems, small or large: people must organize socially in order to secure water, transport it, divide it into usable shares, enforce rules for its distribution, pay for it, and dispose of unused portions. Therefore, the in-depth understanding of technical and agro- nomic problems of irrigated agriculture is impossible without under- standing the social organization in which it is embedded. However, financially induced irrigation development programs have often focused only on technical and physical components: dams and canals, control systems, water levels, and drainage. Institutional con- cerns, when present, were limited mostly to strengthening central water agencies. It is being increasingly recognized now that major institutional weaknesses undermine the operation and maintenance of the physical infrastructure, that irrigation schemes perform below expectation, and that some even cause serious adverse environmental effects. These unintended effects often result not from technical causes, but from inattention to the social organization of the water users. 43 44 IRRIGATION PROJECIS In the following chapters, social scientists with broad field experience in irrigation discuss several new models for identifying the social-organ- izational components intrinsic to irrigation development. Although their views do not coincide in all respects, the authors jointly stress the centrality of social organizational structures and propose sociological models to explain the operation of these structures. These models are directly usable, in turn, in designing either new irrigation systems or programs to rehabilitate the physical, organizational, and institutional arrangements of existing irrigation schemes. Water Coward is concerned with the correlation between technical and social changes, especially when projects aim to rehabilitate or improve already existing irrigation systems. All too often, planners ride rough- shod overall locally developed irrigation systems and self-management techniques in the rush to provide improved alternatives. Particularly interesting are Coward's recently discovered examples of indigenous social mechanisms for water allocation that cover the area of several villages. Coward argues convincingly that a more careful incorporation of existing social, technical, and managerial arrangements into irrigation project designs would make for higher success rates, avoid undermining existing grassroots institutions, and lower the maintenance costs of irrigation systems. David Freeman and Max Lowdermilk also concentrate on the human networks which must organize and manage irrigation systems. Their structural-functional model looks at the different organizational levels needed to coordinate the different tasks involved in making irrigation infrastructure meet farm requirements. Freeman and Lowdermilk argue that their task-focused model can be usefully applied at three different levels: the farm, the command (irrigation) area, and the state bureau- cracy. Freeman and Lowdermilk emphasize that the mid-level organiza- tional structures, located between local and national management, are the most neglected in the irrigation design. The contribution of these two frameworks lies in using a sociological perspective to understand irrigation organizations-formal and informal ones. Both chapters underscore the need to rely on fieldwork-based descriptions of the customary rules in order to understand how irrigation systems work. By clarifying the possible ways to involve better the users in system operation, and by clarifying the social principles that explain why the system is operated in a particular way, the social analyst can determine how to make system improvements more effective. The chap- IRRIGATION PROJECIS 45 ters draw upon ethnographic fieldwork: Coward in India and the Philip- pines; Freeman and Lowdermilk in Pakistan. Benjamin Bagadion and Frances Korten, in their chapter, start from similar premises about the utility of discovering the social structure of local irrigation activities. They argue that the organizational fit between local institutions and state agencies should be based on complementarity rather than competition. Their particular concern is the empowerment of water users through establishing their own grassroots associations, and the coordination of local users' organizations with the state agency in charge of administering irrigation. Such coordination is not built over- night. Bagadion and Korten look penetratingly at what steps must be taken at the agency level to allow it first to recognize and then to deal with local capabilities. The chapter documents the experiences of the Philippine National Irrigation Administration. Using a learning process approach, the agency gradually understood how to address social needs at the village level and how to interact with the water users, and then developed the necessary support systems for those field activities. The training of community organizers was reoriented to include social science skills, and social analysts learned to cooperate with engineers and other technical experts, with beneficial results. The case study on the water users associations in the Philippines, begun a decade ago, is brought up to date with recent field data, which show that the measurable benefits of using community organizers in promoting grassroot organizations exceed the costs. The type of analysis advocated in these three chapters is useful for the preparation and implementation stages of irrigation projects. Detailing the contours of the social organization of irrigation systems would pernit field staff and project planners to develop the structures into which new technologies could be imbedded. This, of course, implies a planning strategy considerably different from most current models. Harnessing local organizational innovations, talent, and experience in project design has the additional advantage of avoiding the costs and bureaucracy associated with unnecessary paternalistic control by gov- ernments. Social analysis, if properly inserted in the process for gener- ating and implementing a project, is able to design and help construct socio-organizational structures. The local groupings identified during this process should be recognized, assisted, and strengthened in their own efforts for self development. 2 Planning Technical and Social Change in Irrigated Areas E. Walter Coward, Jr. Much of the state-financed contemporary irrigation development is directed toward improving systems that already exist. This includes the rehabilitation or improvement of (usually) large state-operated irrigation works, as well as technological change in smaller irrigation works operated by local groups. In either of these situations, the effective design and implementation of appropriate development interventions must begin with a thorough understanding of the sociotechnical context. Projects to rehabilitate or improve systems should ideally proceed from a comprehensive understanding of the physical apparatus and the associated social organization of existing irrigation activities. From this understanding ought to emerge a refined analysis of the irrigation prob- lems that need attention, and a strategy for implementation which recog- nizes both the local resources at hand (including knowledge, experience, and institutional capacity) and the preferred alterations in the existing state of affairs. In all cases where irrigation systems have been in operation, there are institutional and social arrangements which organize fundamental tasks such as distributing water among users or maintaining the canals. These tasks may be performed by water users themselves or in conjunction with others such as local government officials or irrigation agency staff; the social arrangements may be formal or informal, highly individualistic 46 TECHNICAL AND SOCLAL CHANGE 47 or collective, chaotic or controlled-and, of course, judged (by outsiders or insiders) as effective or not. Nonetheless, we begin with the assump- tion that where canals (or wells, springs, or whatever) serve more than a single person, patterns of social interaction govern the use of those facilities. Two other initial assumptions are equally important. One is that project implementation itself is also a social process that follows various rules and is conducted by individuals occupying specific roles in various organizations-namely, engineers of the irrigation agency, experts from donor organizations, and so on. The second assumption supposes that the final, post-rehabilitation situation requires a set of social arrangements to complement the new or rebuilt physical apparatus provided by the project. These future social arrangements may be implicit or explicit in the design and the project documents and, of course, may be formal or informal, dominated by the agency staff or not, may require individual or collective action, and so on. Most important, the future arrangements may or may not correspond with the preproject social patterns. Agency planning for irrigation projects is already a lengthy and complex process covering (sometimes superficially) an array of social, technical, environmental, and economic factors. Given this complexity, one should hesitate to recommend including more in the feasibility study. Nonetheless, particularly in the case of rehabilitation, projectpreparation papers and appraisal reports often provide only the most meager infor- mation on the existing social arrangements for irrigation, or on the intended social organizational patterns to be associated with the im- proved facilities. Because the physical apparatus of an irrigation system and the social structure for its use are necessarily intertwined, rehabilitation projects run a high risk of failure if they deliberately plan technical change but either ignore, or merely assume, the necessary social changes. Some- times, failure is expressed in unimproved production levels; frequently, in the lack of farmer commitment to using and maintaining the new works. To correct this fundamental weakness in project design, prepara- tion procedures should require more attention to existing irrigation institutions and organizations-the cultural ideas and social arrange- ments that organize activities such as system maintenance and water distribution. 48 IRRIGATION PROJECTS Conceptualizing the Sociology of hrigation A sociological perspective on irrigation commences with two funda- mental concepts (which are also part of project preparation terminology): institutions and social organization. "Institution" is a concept varyingly used in both sociology and in everyday language, but here I employ it to refer to ideal behavior and role expectations and as a generic concept for the variety of rules that help pattern social behavior: norms, folkways, customs, conventions, etiquette, and law. Economists sometimes use a similar notion when they define an institution as a behavioral rule. In this sense, the rule of continuous irrigation, the custom of performing a ritual ceremony at the headwork of a local irrigation system, and the law requiring payment of a water fee are all examples of irrigation institutions. In addition to these institutions, there are, in any human group, actual patterns of social interaction which are referred to as the social organi- zation. These patterns of behavior are sometimes formal, purposive, and enduring enough to warrant the use of a group name: the Royal Irrigation Department, Subak Tarnblang, the San Lorenzo Farmers' Irrigation Cooperative Association, Inc., or the Muda Area Development Author- ity. Of course, social organization is also composed of patterns and groups less formal, purposive, or enduring: an evening meeting between irrigation authorities and an assembly of water users, a partnership between two farmers which allows one to move water across the fields of another, or a temporary band of farmers who share a common lateral working to clean a canal. An understanding of the basic relationship between institutions and social organization requires recognition of the frequent inconsistency between what people believe should occur (the institutional element) and what actually occurs (the organizational element). The basic "lack of close correspondence between the 'ideal' and the 'actual' in many and pervasive contexts of social behavior" is one important force for change in either the institutional or social organization arrangements.! A major reason for this inconsistency is that changes in the social or physical environment make it difficult or impossible to act in certain established ways or make it easy or possible to act in certain new ways. Change in either the institutional or the social organizational element creates de- mand for change in the other. TECHNICAL AND SOCIAL CHANGE 49 Institutions and social organization are fused through the basic concept of role which in turn is composed of two elements: role expectations (the institutional dimension) and role periormance (the social organizational dimension). In part, a role can be thought of as the cluster of expectations associated with a given function. In addition, a role is associated with actual patterns of action. Roles help one to predict the actions and reactions of others and thus they enable social patterns and social organization to emerge. For example, in an irrigation system, a cluster of institutions associated with the function of water distribution may be found in the role of the ditchtender and another cluster found in the role of the water user. The presence of these two roles allows social organi- zation to take shape in an irrigation system in the form of patterned relationships between the ditchtenders and water users. This perspective of institutions and organizations allows us to consider patterns of social organization in a particular irrigation system or between systems even though no formal irrigation associations may exist. The general concepts of rules, roles, and groups are useful in under- standing human behavior in many contexts and for a variety of purposes. To be useful in understanding irrigation, however, these concepts need to be matched with critical tasks or actions that occur in irrigation systems. Although many tasks must be organized to operate an irrigation system, five are of fundamental importance.2 Three of these tasks are somewhat specific: (1) wateracquisition-the task of obtaining water for the irrigation system through regular or extraordinary means; (2) water allocation-the task of dividing and distributing the system's supply to its users; and (3) system mainte- nance-the task of repairing, cleaning, and otherwise reconditioning the physical apparatus of the system. The remaining two tasks are more general and may, in fact, be directly related to one or more of the above tasks: (4) resource mobilization-the task of activating and accumulating labor, materials, funds, and other resources needed to implement such tasks as system maintenance; and (5) conflict management-the task of containing and adjudicating (though not necessarily resolving) disputes and disagreements arising from operations such as water allocation. The above list of tasks can be used to create an analytical scheme for understanding the institutional and organizational aspects of one or more existing irrigation schemes before rehabilitation, and to examine the 50 IRRIGATION PROJECIS institutional and organizational assumptions embodied in the post- rehabilitation physical works. The matching of the basic concepts of rules, roles, and groups with the five fundamental tasks is illustrated in table 2-1. Social analysts participating inthe preparationphase of arehabilitation project could use this analytical scheme to detail the existing social organization of irrigation systems and to clarify the organizational and institutional changes likely to be implied in various technological alter- ations introduced by the project. Studying Irrigation Institutions and Organizations Two main factors explain, in my view, why the planning for rehabili- tation projects usually includes little analysis of social organization. The irrigation agency itself-the typical source of rehabilitation project pro- posals-frequently makes assumptions that blunt the demand for organ- izational analysis. If the works to be rehabilitated are publicly managed, the problems to be solved are typically identified by the agency as poor or absent infrastructure and inadequate social organization. If the reha- bilitation is directed toward small, locally managed systems, the organi- zations currently operating the systems are often judged by the agency to be no longer successful or to be unsuitable for operating "modem" irrigation works. In either case, there is little or no demand to explore the current organizational arrangements. However, even when applied social scientists have been engaged in project preparation work, their research may be difficult for planners to use. This can occur even when the project designers have called for sociological input-which usually is a request for a general analysis of "socioeconomic" factors. When this is the request, what is it that sociologists supply? Socioeconomic analysis, nearly always, is done by conducting a socioeconomic survey. Surveys are efficient ways of collecting and aggregating information about the characteristics of individual farmers or of individual farm units-information that is necessary to project planning as well as to some later project evaluation. But surveys are ineffective in revealing important dimensions of the structure of collec- tive action related to irrigation, such as arrangements for water distribu- tion or processes for mobilizing resources for maintenance. Thus, as with Table 2-1. Irrigation System Tasks, by Institutional and Organizational Element Institutional and Task organizational Water Water System Resource Conflict element acquisition allocation maintenance mobilization management Key rules Rules for acquiring Rules for allocating Rules for what Rules for Rules for avoiding extra water supplies water between repairs need to be mobilizing labor, or resolving for the systems subunits of system, done, where, and materials, money, disputes between farms, and so on by whom or other resources systems, zones needed to perform of a system, or system tasks and individuals for responding to shortfalls in resources Important roles Roles for planning Roles for Roles for Roles for Roles for mediating and implementing establishing and identifying implementing and disputes, making water acquisition implementing maintenance jobs monitoring the judgments, and activities water allocation and supervising resource enforcing sanctions policies repairs mobilization process Significant social Groups that seek Groups that Groups that Groups that collect Groups that groups additional water influence water provide routine or specific resources participate in supplies allocation policies emergency repairs settling disputes and implement to system and in enforcing water distribution sanctions 52 IRRIGATION PROJECrS many other endeavors in sociology, a full sociological description of any particular irrigation situation requires multiple and complementary methods of acquiring the data. In this case, socioeconomic surveys must be supple- mented with the in-depth regional field studies discussed below. For a number of reasons, the socioeconomic survey remains the dom- inant tool for data collection in project design efforts, despite some inadequacies. First, the researchers implementing such studies are often economists by training and are interested in collecting data on some so- cial characteristics (such as age, family size, or farm size), but not in doing sociological or organizational analysis of the irrigation relation- ships. Second, the survey is usually perceived to fit better within the time frame for the feasibility phase of project design. It is viewed as a quick research style-even though, in actuality, the length of time required for such tedious tasks as sampling or preparation and analysis of large data sets can often extend the period significantly. In contrast, field studies are often viewed as being too long, requiring perhaps twelve or more months of actual field observations. While many comprehensive ethno- graphic field studies do require extended periods, it is also possible for field studies focused on more specific themes, such as irrigation activi- ties, to accumulate important data in periods comparable to those of other data collection efforts typically included in feasibility studies. A third important reason relates directly to the participation of sociol- ogists themselves. Because of the dominant theoretical and methodolog- ical approaches in the discipline of sociology, many sociologists approach project design with a heavy emphasis on the perceptions and characteristics of individual farmers which, when combined with famil- iar reliance on sample survey methods, reinforces the likelihood that the socioeconomic survey will be used exclusively. Analogously, anthropol- ogists tend to lean toward the use of field case study techniques (though they might also incline toward some "comprehensive" village study in the area). Thus, substantial forces continue to press for the classic socioeco- nomic survey as an integral part of the feasibility study. The point, however, is not to substitute some other single method, but to comple- ment the socioeconomic survey3 with other research approaches such as the regional field study to which we now turn our attention. Some recent irrigation studies have begun using a methodology- herein called the regional field study-that overcomes certain limitations TECHNICAL AND SOCIAL CHANGE 53 found in both the socioeconomic survey and in field studies that focus only on a limited number of irrigation systems. There are two dimensions to this methodology, though a given regional study may not pursue both. The first dimension involves studying a large number of individual systems in a defined region (which may be cultural, ecological, admin- istrative, or some combination of all three) to understand the distribution of common social and technical features as well as to identify any important subcategories of system types. The second dimension identi- fies and understands any regional mechanisms that articulate or mediate the activities of otherwise separate systems. Regional field studies are designed to give in-depth attention to col- lective patterns, group actions, and institutional arrangements for irriga- tion that may already operate in the proposed project area. This information focuses on the social organization of irrigation and is gath- ered through discussions with key informants, observation of group activities (such as meetings and work parties), and the review of existing records and documents. A later section of this chapter focuses on two recent examples of regional field study in settings with a high density of local irrigation systems-of which many, but not all, have small commands of 100 hectares or less. But beforehand, it is useful to illustrate briefly the practical problems that can arise when the state acts without sufficient institutional and organizational analysis in regions that possess rich local irrigation experience. Ignoring Local Irrigation Institutions and Organization To illustrate the propensity to ignore what might exist at the local level, I will examine two irrigation projects. The first example is the Palsiguan River Multi-Purpose Project in the Ilocos Region of the northern Philip- pines. In its initial phase, this project involved foreign assistance in the rehabilitation of 172 local irrigation systems, known in the region as zanjeras, covering approximately 10,000 hectares. Technical and eco- nomic feasibility studies were conducted.4 However, no sociological studies addressed the organization and operation of local systems, nor did the feasibility report incorporate any findings on those associations from the studies published by several social scientists. Instead, the report contained only a very brief comment indicating that some traditional 54 IRRIGATION PROJECf S irrigation existed in the project area but that the systems were ineffective (and, by inference, unimportant). In the absence of detailed organizational information and due to the misrepresentation of the elaborate network of local irrigation institutions, the project designers assumed they were creating irrigation de novo, rather than providing assistance to ongoing local systems. The de novo orientation led to the design of a 1,000 hectare pilot area in which major canals were radically realigned. This rearranged the internal segments and boundaries of the local systems and, though it was poorly recognized, the social organization required significant modification to fit these new canal layouts. As Visaya reported, "The design of the pilot project was based on maximum engineering efficiency, without considering the existing communal irrigation systems and the irrigation organizations in the area. [A] majority, if not all [of the] designed canals are new ones, crisscrossing the canals of the existing zanjera systems. The proposed rotational areas, consequently, disregarded existing area boundaries of the irrigation associations."5 The unavoidable question is, How could this have happened? How could such a large misunderstanding occur? I assume that projects are designed in this manner for several related reasons-all having to do with the nature of the planning agencies. First, there is bureaucratic arro- gance-the assumption of superior engineering skills and technology that can certainly produce a better system of irrigation (how different the process might be if the designers approached the task with more humil- ity-"perhaps we will not be able to improve things very much"). Second, there is the problem of the planners' definitions-irrigation is viewed primarily, if not exclusively, as a technical apparatus and the design of an irrigation system is seen as a technical-engineering process. If farmer organization is considered, it is viewed as secondary to, and needing to adjust to, the infrastructure. Third, there is agency centraliza- tion-the designers are either part of, or closely associated with, a centralized government bureaucracy which intends to design and operate its irrigation system. Any concern with local and regional irrigation experiences and ar- rangements is clearly subordinated to this combination of factors. Given the scarce resources, limited time, and large agenda associated with most feasibility studies, it is not surprising that the sociology of irrigation in the rehabilitation project area is overlooked or addressed superficially. TECHNICAL AND SOCIAL CHANGE 55 One finds a similar approach in the Village Irrigation Rehabilitation Project in Sri Lanka, financed by the World Bank. This project attempted to improve the physical facilities and management arrangements in approximately 1,200 village irrigation systems, most of which, but not all, are tank systems.6 As with the Philippine case, it appears that the planning for this project was not based on an assessment of the existing local institutional and organizational arrangements for irrigation tasks. At least such an assess- ment is not presented in the Bank's appraisal report. In fact, the appraisal report, in its brief discussion of village irrigation and its context in Sri Lanka, uses rather vague information about village tanks and their operation! It states that many schemes have deteriorated or have been abandoned; that head sluices are seldom operated systematically; and that the decline of traditional management partly explains the deteriora- tion and underutilization of village irrigation systems. If the above statements are accurate, they are significant symptoms of something gone wrong. One might expect that rehabilitation of the tanks would be directed at the causes of these symptoms. But, the "causal factor" on which the project seems to rest is that the tank community has lost the power to act effectively in irrigation matters. This, of course, may be an accurate assumption-though the basis for it is left unclear, as are the reasons for the community impotence. In particular, it is not clear to what extent inaction or erratic action by the government has contributed to this decline in local capacity. This last point is the more important when one considers the funda- mental assumption on which the project rests and the major objective that derives from it. The basic assumption is: "Since it is difficult for individual farmers to change this system, increased guidance by govern- ment will probably be necessary to make the best use of available resources.'4 On this powerful assumption, the project logically adopts as its main objective "strengthening the major government institutions involved in village irrigation." The significance of this assumption is best illustrated by recasting the proposition (obviously paraphrasing the original): "Since it is difficult for individual farmers to change this system, and since outside govern- ment assistance alone is unlikely to be effective, it will be necessary to strengthen both government capacities and the capacity of the local group to make these changes." It seems likely that this altered assumption 56 IRRIGAllON PROJECIS would lead to an additional project objective explicitly concerned with strengthening capacity at the village level to handle irrigation affairs. As with the Philippine case, there was available for Sri Lanka some sociological field research on village tanks that might have been con- sulted for the feasibility study. These range from the classic study of Pul Eliya by Sir Edmund Leach to a more contemporary study of a single tank by two Sri Lankan researchers.9 The latter conclude: Although the Kelegama tank and its operation were solely for the benefit of the users of the tank, these beneficiaries were made to feel that the ultimate control and custodianship of the tank lay with a somewhat nebulous body outside, namely, that is the State, whose powers and responsibilities were exercised by the Agrarian Services Department through its officials at the regional and village level. They report that the community has failed to evolve new institutional arrangements that are consonant with the contemporary state presence, and there is thus a certain degree of community malaise with regard to tank maintenance and operation. Somewhat at variance with the assump- tions of the village irrigation project, these authors suggest that state involvement, at least of a particular kind, may increase, rather than reduce community disengagement. These examples illustrate an obvious but fundamental point-the factors investigated in the preproject studies represent the designer's judgments and understandings of what is important for the irrigation systems. Conventional thinking about irrigation subordinates social or- ganization to other factors, such as the technical-engineering, which are perceived to be more critical. The result is that local irrigation organiza- tion, even when it exists, may be ignored, as in the Philippine case, or may be assumed defunct, as in the Sri Lankan case. In sum, often there is simply no demand for sociological analysis of local irrigation institutions and organizations because designers are oriented toward other matters: the engineering works, the financial arrangements, the irrigation agency's resources, and so on. This narrow approach is consistent with the technocratic image that most planners hold of irrigation development. In both of the cases discussed above, the events of early project implementation heightened awareness of, and concern for, local institu- tions and social organization. In the Philippine case, the local irrigation groups offered dramatic resistance when the irrigation agency began TECHICAL AND SOCIAL CHANGE 57 implementing its program in the pilot area. When these troubles arose, several outsiders, including some social scientists familiar with the remarkable capacities of the zanjeras, began sharing with the agency the results of past field studies. These standard academic studies were not done for the purpose of project planning but, nonetheless, contained highly relevant information. These data, as well as the monitoring studies subsequently conducted as part of the project, were then incorporated into project implementation. The best example of this reorientation is seen in the project manager's statement of the reformulated project goal: "To rehabilitate existing zanjera communal irrigation systems and con- struct facilities for new areas, through the maximum participation of the farmers in planning and implementation of the project, and improve the farmer's capability to effectively operate and maintain their irrigation systems."l0 In the case of Sri Lanka, since the initiation of the project, the Department of Agrarian Services and the Agrarian Research and Train- ing Institute have conducted detailed field studies in a few selected tanks, both to increase their knowledge of existing local organizational capac- ities and to understand more fully the government-community interac- tions produced by the village tank program. Regional Field Studies of Local Irrigation: Two Examples Many regions in Asia and elsewhere (for example, Ilocos in the Philippines, Tamil Nadu in South India, or the Peruvian Andes) are characterized by a high density or frequency of local irrigation systems. Their organizational shapes, while displaying variations, possess a de- gree of commonality resembling one another to the extent that we can recognize a Balinese subak, an Omani aflaj, or a Himachali kuhl. In part, of course, these resemblances may derive from similarities in the phys- ical habitat, the cultural elements of the place, the historical processes that have occurred, or the political economy of the region. In short, the local systems are to be seen in light of the regional circumstances, past and present, in which they operate. Thus far, regional analyses of local irrigation systems have been infrequent, though several scholars now have such work under way" I and a few studies have recently been completed (two of these are discussed in the following sections). In this section, I propose to explore this new 58 IRRIGATION PROJECIrS line of research on local irrigation, illustrating its application in two cases, and considering its policy relevance for designing effective state assis- tance to such systems. Baskets of Stones: Local Irrigation in a District in Northern Sumatra Research was recently completed examining government assistance to local irrigation systems in a region in northern Sumatra; Pidie District in the Province of Aceh.12 Pidie was selected for this work because of the importance of irrigated agriculture, the prevalence of local irrigation networks and the variety of government irrigation assistance programs available in this region. The research which was conducted in three sequential phases-an inventory, followed by several case studies, and then by a survey of forty-nine local systems-provides a full description of regional commonalities and significant variations in local irrigation. Most irrigation works in Pidie use surface water as their supply and nearly all of the systems utilize simple diversion structures to acquire this water. The structures that divert water are called seuneulop, or more familiarly, neulop, regardless of their size or location-the headwork on the river course or along the main canal at the point where a branch canal begins. The term seuneulop (or neulop) also is used to refer to the entire irrigation system. For constructing diversions, the most commonly used materials are bamboo, stones from the riverbed, wide timbers, or some combination of these. On some of the larger rivers in the narrow coastal plain area of Pidie, the neulop is of a particular type called beuriyeueng- a structure made by first weaving large cone-shaped baskets of bamboo which are then filled with stones from the riverbed. These baskets of stones are a kind of indigenous crate structure. The canals (called leung) of these local systems are of unlined, earthen construction. Typically, in addition to the main canal, there are various branch or secondary canals. At each of the numerous branches of this canal network, one generally finds a proportioning weir (a neulop, as discussed above). The combination of neulops and canal network pro- vides the physical apparatus that permits relatively complex management and distribution of water within a system when such management is required. Rotational water distribution during times of shortage is common. TECHNICAL AND SOCIAL CHANGE 59 Most local systems in Pidie have small commands-the forty-nine systems we surveyed had from 17 to 426 hectares and 70 percent of them irrigated less than 150 hectares in the wet season. However, in Pidie, there is also a special set of local systems whose overall command is very large, covering thousands of hectares and tens of villages. Leung Treung Campli is one such large network. It conveys water through a main canal approximately 15 to 20 kilometers in length. The canal passes through four subdistricts and irrigates about 2,000 hectares in over 100 hamlets. Many of these large systems have an origin myth that features the prominent role of some special religious leader (ulama) of an earlier century who is regarded as having been central in inspiring and mobiliz- ing people to construct the large system. An important institutional component in the management of local irrigation systems in Pidie is the keujreun arrangement. The people of Pidie conceptualize the keujreun as operating simultaneously at several levels of civil administration and at several levels in the irrigation network. At the hamlet (gampong) level, the institution is represented by a man referred to as the keujreun blang who works closely with the hamlet head and the water users of that area. At the next higher level of civil administration, the village cluster (mukim), the institution is repre- sented by someone called keujreun muda (there is some variation in the terminology), with responsibilities for irrigation activities throughout the mukim. At the next level, the keujreun syik (again there is some variation inthe terminology) operates at the subdistrict level with its administrative head (the camat). In this way, the keujreun, while having responsibilities specific to irrigation, operate as an element of the civil administration. Also, the keujreun arrangement is sufficiently flexible to cover both administrative and hydraulic boundaries. For example, at the village level, several situations may exist. There may be facilities of an irrigation network that serve multiple villages. If so, typically, there will be a keujreun blang for each village served. If that network covers a group of villages, all of which fall within a single mukim, then the activities of the several keujreun blang will be coordinated by a mukim-level keujreun. At the hamlet level, the keujreun blang usually is selected by the farmers of that area with the approval of the village head. At the higher levels, the keujreun are appointed by the administrative heads of the units from among the existing keujreun at the lower levels. While working closely with the various heads of the civil administra- tive units, the keujreun are also accountable to the water users, since they 60 IRRIGATION PROJECTS receive payments for their activities from them and not from the govern- ment. These payments are called beuheuek bruek umong. The literal meaning of this phrase is the "half of a coconut shell (bruek) share (beuheuek) of unmilled rice to be paid by each ricefield plot (umong)." In Neulop Amud, one of our case study locations, farmers made an annual payment to the keujreun blang of twelve kilogramns of unmilled rice for each quarter hectare of rice fields that they irrigated. In Seuneulop Beuracan, a network that has received little state assis- tance, we found the duties of the keujreun blang to include a wide range of activities-coordinating canal cleaning and system repair, distributing water within the village lands when rotation was in use, involvement in the various irrigation and rice production ceremonies (khandurn, and settling minor disputes among the irrigators. This range of reported duties is consistent with those detailed in an ordinance issued in 1973 by the head (bupati) of Pidie district. This ordinance, which deals with the organization of rice production activities in the district, includes an extended discussion of the keujreun institution. In it, the stated respon- sibilities of the keujreun blang are (1) supervising repair of the facilities, including the management of the farmers' collective work groups, (2) managing the distribution of water, (3) settling disagreements among farmers regarding water distribution, and (4) supervising and controlling timing of the planting of the rice nurseries and the subsequent transplant- ing to the fields. Not all irrigation systems in Pidie have the keujreun arrangement. Our survey indicated that 65 percent of the systems utilized the keujreun; the rest did not. There are two situations in which the keujreun is absent. One is in very small systems involving only a few farmers who informally coordinate their irrigation operation and upkeep activities. The other situation involves numerous water users, but the local government au- thorities coordinate operations and maintenance without the aid of a keujreun. The traditional view of neulop networks in Pidie is that they are public infrastructure, public utilities as it were, whose maintenance and opera- tion lies jointly with the civil administrators, irrigation caretakers (the keujreun), religious leaders, and the related users, rather than as facilities owned by some private group of coproprietors. Initial decisions regard- ing the annual refurbishing and use of these facilities are made in a forum organized at the subdistrict administrative level by the camat. In this meeting, discussions of the coming agricultural season are held and dates TECHNICAL AND SOCIAL CHANGE 61 are set for key activities, such as repairing and cleaning the neulop networks, initiating land preparation activities, and the various ritual activities to accompany these actions. Thus, the neulop network is highly intertwined with matters of civic and spiritual order as well as agronomic practice. Its state of readiness and smooth operation are both necessary for, and a reflection of, order in the realm of governance and culture. The effectiveness of the neulop network is dependent on the synchronized individual behavior of multi- ple users as well as their timely collective behavior-a pulsating rhythm of aggregating and disaggregating in which people sometimes come together to act jointly, while at other times act singly but in a coordinated fashion. One sees then that the order needed for effective wet rice production and neulop network operation is socially constructed by the Acehnese of Pidie. The Acehnese use an aphorism that irrigated agriculture requires men to act like brothers. However, unlike other regions of Southeast Asia in which the required social solidarity is supported through the formation of a specialized irrigation group (for example the subak of Bali or the zanjera of Ilocos Norte), in Aceh order is shaped through the actions of the civil administrative heads and the specialized irrigation institution called keujreun-accountable to both the civil administration and the water users. While the keujreun do receive contributions from the water users, they are not the staff of some membership-based organization formed by the villagers. Rather, they are intermediate between government and farm- ers. The keujreun are arms of the local government officials and they derive part of their authority to direct farmers from their association with these local authorities but, however, local government has no budget to pay the keujreun. Thus, they depend on the farmers to provide support for their services. In this way, the keujreun also are accountable to the farmers. In short, irrigation order can be seen as one element of general public order. State assistance to local irrigation in Pidie has been widespread. Nearly six out of every ten local systems in our survey had received some form of state aid. With a few important exceptions, this state assistance has served to complement rather than demobilize local resources. Our case studies revealed the frequency with which resources supplied by the state are combined with local resources for the purpose of improving existing structures and canal networks. For example, in Neu- 62 IRRIGATION PROJECTlS lop Taka Adan, where water from a swampy area is used for irrigation, in the budget year 1980-81, the camat combined the village subsidy funds from ten villages to purchase materials for improving the dike used to divert water from the swamp. Labor that farmers provided for the construction work supplemented the state assistance. Overall, our research findings produce a positive picture of local irrigation in the Pidie region. These local systems function in the diverse ecological settings of the district-the hill areas, the fertile coastal zone, and even the low-lying swampy areas. In all of these locations, facilities are not only maintained, but also improved with locally mobilized resources. Local government officials, frequently assisted by local irri- gation caretakers (keujreun), exercise leadership in mobilizing these resources and in the operation of the systems. These irrigation activities are punctuated with a special rhythm and given cultural meaning and order through the myriad ritual ceremonies that occur at the subdistrict, neulop network, and, sometimes, individual field levels. While farmer involvement in the operation and maintenance of these irrigation works is considerable, it is not accurate to label them farmer-managed systems. More to the mark would be to identify them as "locally managed," since it is the local authorities, often with assistance from the keujreun, who perform these functions. This regional picture of irrigation in Pidie suggests important consid- erations for formulating policies in support of local irrigation develop- ment. For example, the widespread, but not universal, presence of a customary institution for local irrigation management suggests little need for the introduction by the irrigation agency or foreign donors of a standard water user organization model from another region. The earlier effort of the district government to strengthen the keujreun arrangements is more appropriate. There are also opportunities for regional policies to build on the widespread customary practices of system management through water rotation. No doubt, improvements could be made in selected systems, but programs for improvement should recognize the local experience that exists and the principles in which these practices are rooted. Programs built on this approach have a much greater possibility of being sustained than those that assume little relevant local experience. Plans for assisting these local systems also should recognize the history of, and continuing capacity for, local resource mobilization that exists- TECHNICAL AND SOCLAL CHANGE 63 local labor, funds, materials, and so on are regularly organized and applied to system development. Future state assistance should be de- signed to complement and not to replace or erode these valuable custom- ary activities. The Book of Irrigation Customs: Irrigation Rights in Palampur The Kangra valley of Himachal Pradesh (India) is one of the most intensely irrigated regions of this mountainous state. Most of the existing irrigation facilities are so-called private kuhls, small gravity flow net- works that are locally operated and maintained by the water users themselves. Recently, research was conducted in this region to explore the importance of water rights which were recorded during the British colonial period to the functioning of the contemporary systems.'3 As with the Pidie case discussed above, the results of the Kangra research also confirm that the understanding of kuhl irrigation stands to gain from taking a regional perspective. The Historical Record The functioning of these local irrigation networks was noted by the British as early as the middle of the nineteenth century, when they initiated procedures for assessing taxes in their newly acquired territory. As part of these procedures, they also recorded, and hence established within their legal framework, irrigation rights. The instructions provided to the staff implementing the tax assessment included directions to record rights to wells and other sources of irrigation. Thus, detailed records of customary rights and rules for operating these hydraulic works were produced. Our research in one subdistrict of the Kangra region, Palamnpur, focused on the analysis of these historical records combined with field- work in a number of operating kuhls. In the revenue office of Palampur, one can find an important historical register called the Riwaj-i-Abpashi (the Book of Irrigation Customs). The title page includes the following statement: General Rules of Irrigation, Tehsil Palampur, Kangra District, 1918. It is written in Urdu and based on the findings of the 1915-16 Settlement Report. The register contains the following sections: 64 IRRIGATION PROJECTS * A table of contents * An introductory section discussing general rules for kuhl irrigation in Tehsil Palampur * A glossary of irrigation terms * A map of the Tehsil showing the location of rivers and streams * An index of 298 kuhls existing in the Tehsil and arranged by river course * A section containing details regarding each of the indexed kuhls * A listing of smaller kuhls (with no additional details) * A concluding section. The General Irrigation Rules Six general rules are discussed in this opening section of the register. The first rule deals with the construction of diversion structures on the rivers or streams. The farmer-built structures of today probably appear very similar to those constructed at the time these rules were recorded. They are temporary structures, built with the use of stones found in the river bed. In some cases, the stones are supplemented with the use of mud, grass, and other natural materials used to reduce leakage. Usually they are damaged one or more times during the wet season, when river flows increase significantly, and thus they need repair either within the season or before the start of the next cropping season. The first rule simply states that if the diversion needs to be repaired, the work is to be done by the "members" of the kuhl. If the diver- sion needs to be rebuilt, it can be constructed anywhere between the nearest upstream diversion and the next diversion below. This latterpoint clearly deals with the rights of the various kuhls along the stream and limits the rights of any particular kuhl to diverting at some point between the diversions located immediately above and below the one being rebuilt. In short, the order of placement along the river course cannot be altered. The second general rule deals with the matter of constructing a new diversion. It states that this is not to be done without obtaining permission from the (British) government as well as the local raja. In general, permrission will not be given to construct a new diversion upstream from an existing one. However, if it is determined that there is surplus water in a river, permission could be granted to build a new diversion upstream with the provision that, in times of future scarcity, the earlier diversion TECHNICAL AND SOCIAL CHANGE 65 would have priority to meet its needs. Likewise, if a new kuhl is built between two existing diversions, in times of scarcity the two original diversions would have priority in meeting their irrigation needs. This second rule also discusses the characteristics of materials that can be used in constructing the new diversion. First, the materials must be approved by the members of the existing kuhls. Second, the materials for the new diversion cannot be different from those used in the diversion immediately above (thus protecting the rights of the lower kuhl). Finally, this rule notes that the landowner has three compensation alternatives if the channel for the new system passes through his land. He can receive a payment from the builders of the new kuhl, or he can make use of water from the kuhl to irrigate his land, or he can opt to become part of the group constructing the kuhl, which will later receive payment for pro- viding the irrigation. The third rule deals more specifically with the issue of distributing water among the severalkuhlsthatmaybedrawingwaterfrom acommon stream. To illustrate the sharing principle that applies in this case, the example of Baner Khad (river) is given in the record. In 1914, due to insufficient rain, water flow was low in this river. After the first sixteen kuhls along this river had diverted their water, there was none remaining for the lower kuhls. The Deputy Commissioner instituted the rule that the first sixteen kuhls had to allow at least one nallah of water for each of the lower kuhls. It also was noted that this principle had been used in earlier times. In 1868, farmers on Baner Khad had agreed to the same water-sharing arrangement. The fourth rule concerns the use of seasonal creeks (also called nullahs) for irrigation purposes. The general rule allows villages where the creek originates to use as much of the seasonal flow as they require. T'he unused water can be claimed by the next lower village on the creek. However, water from the creek can only be used on the riparian lands of that stream. The water from the seasonal creek cannot be diverted to some nearby stream to augment the supply in it. A fifth general rule relates to the organization of repair and mainte- nance activities for a kuhl. The general principle is that the "last" village is responsible for maintenance. But last is defined so that, in a set of four villages spread from the head to the tail of the canal, the last village will work alone from the tail upstream to the boundary of village three. At that point, village three will join village four to perform maintenance activities upstream to the boundary of village two, at which point it joins 66 IRRIGATION PROJECIrS the maintenance work, and so on. Thus, "last" is redefined and expanded as the maintenance work progresses from the tail to the head. Each water user must supply labor in the form of a male at least twelve years of age. Those who do not supply labor will be charged the cost of hiring a replacement. If investments are to be made in systems improvements, funds will be collected proportional to the size of the irrigated holding. The last general rule concems water distribution procedures within a kuhl and states generally that these irrigation rules vary from system to system and are described in the section of the register containing details on the customary rules of the individual systems (section 6 in the Book of Irigation Customs). The irrigation register also contains a glossary of irrigation terms. A total of twenty-four terms are presented and defined. These terms deal with essential matters such as measurements of water, irrigation struc- tures or technologies, agronomic terms, and the names of different irrigation cycles and their purposes. This set of irrigation vocabulary suggests the existence of a districtwide "irrigation culture," recognizable by its common teminology and the common material items, techniques, and concepts. The shared vocabulary contributes to the discourse among the users of a single kuhl as well as between kuhl groups. The main canal of the Bharul Kuhl system irrigates about 145 hectares belonging to five hamlet clusters. Irrigation water is particularly scarce during the summer planting season. To cope with the water constraint, the people of Bharul Kuhl use customary rules of system operations, simple irrigation technology, and suitable agronomic practices. An intrahamlet distributional system assigns water to the five hamlets based on their location along the main canal. Each hamlet must, there- fore, select agronomic technologies according to the available supply of water and labor. After the summer rice crop sowing is finished, each cluster receives an assured water supply appropriate for the crops that it planted. Also consistent with the historical record, Bharul kuhl is served by a local irrigation foreman, the kohli. His responsibilities include two major activities: (a) mobilizing labor and materials and supervising routine and emergency repairs to the diversion structure, the main canal, and the water control structures along it, and (b) supervising the distribution of water along the main canal, including supervision of the guarding activ- ities frequently required when cultivators in the lower reaches of the network receive their water supply. TECHNICAL AND SOCIAL CHANGE 67 Repairing and maintaining the physical facilities of Bharul kuhl is not an onerous task. The diversion structure is made of stones found in the bed of the Awah River. The main canal is unlined and courses through undulating countryside. When the canal reaches a gully, or seasonal creek, the canal water is channeled into the natural stream and then rediverted somewhere downstream back into the continuation of the canal, thus avoiding the need for complicated structures at troublesome, creek-crossing locations. The river water is relatively free of silt and thus little siltation occurs in the canal. Most bothersome is weed growth along the canal, which at times can be heavy. In addition to providing labor for repairing and maintaining these main system facilities, farmers must also provide labor and other resources for repair of the branch facilities within each of their hamlet clusters. In addition, farmers in the lower clusters have to provide labor for the guarding activities. Over the last several decades, the state has been providing assistance to kuhls such as Bharul in the Kangra region-in the last decade at an increasing pace. While this assistance has sometimes been useful, dis- cussion here of an unsuccessful case illustrates the shortcoming of a policy that is uninformed by regional realities. Birari is an old kuhl that diverts water from the one of the largest rivers in the Kangra region, the Neugal. At the hamlet boundary where the main canal enters the village, stands a small Hindu temple dedicated to the goddess Birari for whom the kuhl also is named. Historically, the water supply in this kuhl has been good since the Neugal is a perennial river receiving snowmelt from the nearby Dhala Dhar range and Birari kuhl is in an upstream location. Some years ago the state irrigation agency approached the people of Birari kuhl with a plan to extend the main canal for approximately thirteen kilometers, to serve a set of hamlets which now irrigate from several seasonal creeks in their area. The Birari people report being told by the agency at the time of these discussions that, as part of this project, the water flow into Birari kuhl would be augmented to meet the additional needs of the new area. Extending the canal for this purpose was not an entirely new idea. The Birari inhabitants indicate that a number of years previously, some people from the proposed extension area had approached Birari about extending the canal. Birari proposed that the extension people assume full responsibility for maintaining the entire main canal, including the portion then maintained by the Birari users. In exchange for this arrange- ment, the Birari inhabitants would have rights to two, or two and one-half, 68 IRRIGATION PROJECIS nallahs of water (a local measure of water volume described in the Book of Irrigation Customs glossary). No agreement was reached between the two groups at that time. The original canal now has been extended by the irrigation agency to reach the new area. But, only infrequently does water actually reach this tail area. While the kuhl now is nominally under the control of the irrigation agency, the Birari people, located at the head portion, effec- tively operate it and water is supplied to the tail area only when there is surplus relative to the needs of the original rights-holders. The disagreement between the agency and the Birari people is now in court. The dispute turns on the matter of "ownership" of the kuhl. The Birari people argue that their ancestors built the kuhl and they tradition- ally operated and maintained it-thus it is a private kuhl belonging to them. The agency argues that, based on earlier state assistance to the kuhl, it became a government kuhl and, consequently, they are in a position to modify and improve it as necessary. There is a clear lesson for local irrigation development policies in the Palampur region when this case is placed in regional perspective. As a result of the colonial and subsequent settlement procedures and the continuing activities of the Revenue Department, the people of Palampur are aware of their legal rights to water and their ownership of the means of irrigation in the form of the diversion structures and the canals of their kuhl systems. Under certain conditions, they are prepared to take strong actions to prevent the invalidation of these rights. Thus, state actions in existing irrigation networks are not likely to be successful if they attempt to erode rather than enhance existing water rights. Regional Field Studies and Project Planning The two field studies just summarized, carried out on a regional basis rather than on only one or two irrigation systems, illustrate the range of data on patterns of social organization that can be assembled in areas dtick with local irrigation systems. In this section, I want to step back from the location-specific data presented in those cases and ask the question: What of this infornation is relevant for project planning? First, in bodt cases we see there are regional institutional arrangements in place that deserve attention by the project planners. In Pidie, the existence of the district ordinance which legitimates the traditional TECHNICAL AND SOCIAL CHANGE 69 irrigation supervisor role (the keujreun) is an essential institutional resource for local irrigation in this region. It should neither be ignored nor unnecessarily modified, though it is certainly plausible that some changes might be required, in the name of rehabilitation or irrigation modernization. In Palampur, the water rights defined in the records of customary irrigation (the Book of Irrigation Customs) are equally fun- damental underpinnings for everyday iriigation matters in the systems of this region and should be respected as far as possible in any state interventions. At a minimum, ignoring these rights greatly complicates the outcome of state assistance. Second, a point that derives from the regional approach is the ability to see both common patterns and impor- tant differences across systems. For example, in Pidie, we see a common element in that, for all cases, the various heads of the civil administrative units are involved in key activities related to irrigation management. But, while the keujreun institution is widespread, it is not ubiquitous-our regional survey indicated that 35 percent of the systems were without this irrigation supervisor. Good planning in Pidie will build on both the com- mon elements and the differences found in local systems in the region. Third, the regional sociological study, if targeted on a region that coincides with the territory of a meaningful unit of the bureaucracy, can give planners and program implementors confidence that they under- stand the situation of local irrigation in the region where they are working-rather than be left uncertain about the extent to which condi- tions found in one or a few systems studied apply throughout the area. The virtue of such sociological studies is that they document for the planner and designer not just the existing physical works of the systems but the associated patterns of behavior and social relations. Through in-depth field studies, the research can display the rich texture of the sociology of irrigation systems, revealing the deep-seated rules for carrying out basic irrigation tasks, the expectations that various irrigation actors have of one another, and the functional fit between the engineer's canals and turnouts and the sociologist's institutions and organizations. Regional field studies can concentrate specifically on these institutional and organizational arrangements, and the qualitative picture they provide can alert the planner to the existing social organization of irrigation, which is a resource just as the existing physical works are seen as an important component of the given situation. This may sound to some as though field studies are meant to paralyze the planner-to inhibit any, perhaps all, changes for fear that they will 70 IRRIGATION PROJECTS adversely affect the status quo. This is not the intent! With or without state-assisted rehabilitation work, many changes are constantly occur- ring in irrigation systems and adjustments are continually being made by and to the institutional and organizational arrangements. The point is that changes need to be made with sensitivity and awareness of their potential implications-with an understanding that irrigation systems are usually composed of a close congruence between the physical works and the institutional and organizational arrangements. Understanding the nature of that correspondence in the particular project provides a realistic base for planning new changes. Even with this amount of knowledge it is difficult to anticipate the various implications and outcomes of proposed changes. This is why it is imperative to involve the people of the project area in the very process of planning and design (also see chapters 3, 12, and 13). Some Suggestions for Action When planning and designing rehabilitation projects for regions with existing local irrigation works, it is critical to understand the established social arrangements for implementing fundamental irrigation tasks. Ithas been suggested that information on these existing institutional and social patterns will be difficult to obtain through the usual method of the socioeconomic survey, but can be obtained by trained analysts using a focused regional field study approach. Such field studies should collect information required to complete the cells identified in table 2-1, both by studying a large number of individual systems in the region and by identifying any regional mechanisms that relate to these cells. The result would complement the statistical portrait produced by a socioeconomic survey with the structural principles and social organizational arrange- ments illuminated by the sociological analysis of the regional field study. To implement regional field studies of this kind, the following recom- mendations are put forward. First, the terms of reference for the socio- economic portion of a rehabilitation project preparation study should specifically identify the need to examine the existing social organization related to fundamental irrigation tasks. Second, project feasibility studies and appraisal reports should include a subsection on the existing local irrigation organization in the project area. This subsection should report on the preproject organization of TECHNICAL AND SOCIAL CHANGE 71 irrigation tasks, contrasting them with the social arrangements the plan- ners intend to have upon completion of the project. There should also be a clear statement of the steps necessary to achieve identified changes of institution and social organization. Third, national and international agencies concerned with the design of irrigation rehabilitation projects should consider holding periodic workshops for socioeconomic analysts and others working on studies for irrigation project proposals. The proposal specialists could then discuss their experiences with applying an organizational approach and identify better procedures for data collection and report presentation. A detailed analysis of local activities by region will inform local project proposals and will shape thinking about irrigation organization. The accumulation of information on how much local groups are doing and, in some cases, how they are doing it, could influence policies regarding the appropriate mix of state and local responsibilities and rights in irrigation development. Knowing more about what local irrigation groups are doing could influence our thinking about what they ought to do. Notes 1. Wilbert E. Moore, Social Change (Englewood Cliffs, NJ.: Prentice-Hall, 1965). 2. In previous writing I have emphasized only three fundamental tasks. Subsequent fieldwork has demonstrated the need to expand this list. The task of resource mobilization was first suggested to me by Romana de los Reyes ("Sociocultural Patterns and Irrigation Organization: The Management of a Philippine Community Irrigation System," Ph.D. disser- tation, University of Califormia at Berkeley, 1982). A detailed exploration of this system function has been presented by Uraivan Tan-Kim-Yong ("Resource Mobilization in Tradi- tional Irrigation Systems of Northem Thailand: A Comparison between the Lowland and the Upland Irigation Communities," Ph.D. dissertation, Cornell University, 1983). The task of water acquisition is discussed by Mark Svendsen ("Water Management Strategies and Practices at the Tertiary Level: Three Philippine Irrigation Systems," Ph.D. dissertation, Cornell University, 1983); John Duewel ("Promoting Participatory Approaches to Cultivating Water User Associations: Two Case Studies from Central Java," paper presented at the Work- shop on Investment Decisions to Further Develop and Make Use of Southeast Asia's Irrigation Resources, Bangkok, 1981); and Ruth Meinzen-Dick ("Local Management of Tank Inrigation in South India: Organizations and Operations," M.S. thesis, Cornell University, 1983). 3. The critique of the survey approach, and particularly of its use as the single tool for gathering social information needed for project feasibility studies, is developed by Robert Chambers in chapter 14 of this volume. 4. See Visaya, 'The PRaw and the Zanjeras," paper read at the Workshop on Organization as a Strategic Resource in Irrigation Development, Asian Institute of Management, Manila, November 1982. A full report on the organization and operations of a particular zanjera is 72 IRRIGATION PROJECTS contained in E. Walter Coward, Jr.'s "Principles of Social Organization in an Indigenous Irrigation System," Hunan Organization, vol. 38 (1979), pp. 28-36. 5. "The PRew and the Zanjeras," p. 6. The details of these changes have been monitored by a team of researchers from Central Luzon State University. See Honorato L. Angeles and others, Process Documentation on the Development of the Palsiguan River Multi-Purpose Project (Munoz, Nueva Ecija: Central Luzon State University, 1983). 6. For a description of the project see Jaliya Medagama, "Some Observations on Farmer Involvement in the Village Irrigation Rehabilitation Project in Sri Lanka," paper read at the Workshop on Organization as a Strategic Resource in Irrigation Development, Asian Institute of Management, Manila, November 1982. Recent field studies of village irdigation rehabilita- tion projects include Shyamala Abeyratne and Jayantha Perera, "Changes in Irrigation Man- agement in Small Communities: A Tank and Anicut System in Moneregala District, Sri Lanka," Sri Lanka Journal of Agrarian Studies, vol. 5 (1986), pp. 20-27; and, for a more comprehensive examination within a historical context, Shyamala Abeyratne, "Rice, Rehabil- itation and Rural Change: Social Organization and State Intervention in Small-Scale Irrigation Systems in Sri Lanka," Ph.D. dissertation, Comel University, 1990. 7. World Bank appraisal materials for the Village Irrigation Rehabilitation Project in Sri Lanka, 1981. 8. World Bank appraisal materials. Note that there is a bit of a straw man i this statement Few would expect that "individual farmers" could change things. But, as noted below, this does not necessarily imply a heavy government role. There is another alternative-strong local organization. 9. E. R. Leach, Pul Eliya: A Village in Ceylon (Cambridge: Cambridge University Press, 1971); and WickremaGunasekera and SunimarFemando, "WaterManagement in the Village of Kelegama in Sri Lanka" (Colombo: Marga Institute Final Report, 1981), p. 36. 10. Visaya, 'The PRmn and the Zanjeras," pp. 13-14. 11. Those with interest inthistopic shouldbe aware ofwork inprocess by Nirmal Sengupta of the Madras Institute of Development Studies and -John Ambler of the Comell Irrigation Studies Group. 12. This research was carried out in 1986-87, and was funded by the Water Management Synthesis Project under usAm Contract DAN-4127-C-00-2086-00 to the Consortium for International Development and subcontracted to Comell University. Fieldwork in Indonesia was supervised byDr. AminAziz of the CenterforAgribusiness Development in collaboration with Syaih Kuala University in Aceh Province. For a more detailed discussion of findings, see E. Walter Coward, Jr., "Basket of Stones: Government Assistance and Development of Local Irrigation in a District of Northem Sumatra," Water Management Synthesis Project, WMS Report, no. 80, March 1988. 13. This research was supported by Cornell University and assisted through an award from the Indo-U.S. SubcomTnission on Education and Culture. In India I was affiliated with the Hiimachal Pradesh Agricultural University and provided professional assistance and warm hospitality by ProfessorT. V. Moorti, chairman of the Department of Agricultural Economics. My reliable field assistant was Mr. Rattanchand Bhardwaj. 3 Developing Irrigators' Organizations: A Learning Process Approach Benjamin U. Bagadion and Frances F. Korten Examination of development projects worldwide indicates that social and institutional issues are often not adequately addressed. Although technical and economic considerations receive a great deal of attention, the people-related issues are frequently subordinated or ignored, with detrimental consequences for sustained development.1 In recognition of these failings, many implementing agencies, as well as donor large and small agencies, are giving greater attention to social issues and adding individuals trained in sociology, cultural anthropology, and other noneconomic social sciences to the project design and imple- mentation teams. In many situations, these individuals have spotlighted key needs and helped shape project concepts that better fit the realities of the lives of the intended beneficiaries.2 Quite often, however, a wide gap remains between the needs that are spotlighted and the actual implementation of programs that address those needs. Sometimes this is because adequately addressing the social issues would raise political questions too large for the project to tackle.3 In other cases, the implementing agency is not oriented to deal with those needs. Addressing social issues often involves building new capabilities among the people at the community level. But many government agen- cies assigned to implement large projects have norms, procedures, poli- cies, and attitudes which provide little support for building such capabilities. Thus, socially sensitive individuals who are trying to address 73 74 IRRIGATION PROJECrS social needs find they must work through implementing agencies that are oriented to very different issues. As discussed in the final section of this chapter, the project develop- ment process commonly used by the large donors is likely to be of little help in solving this problem. While project design can be improved by considering social and institutional issues, implementation remains a problem unless the implementing agency can also address these issues. Furthermore, the organizational subsystems that the implementing agency needs for supporting such methods are rarely in place. The agency first has to learn how to address the social needs effectively at the village level, then develop support systems appropriate to the required field- level actions. When new capacities need to be developed, the need is not for a comprehensive plan, but rather for a process. The process must allow for trial and error on a small scale; continuous examination of the village- level work to identify problems, issues, and successful approaches; and adjustments in agency policy, procedures, and organizational structures to accommodate responses to the field-level needs. The process itself must be shaped to develop both the individuals and the organizational systems needed for eventually implementing the new approaches on a broader scale. In short, what is needed is a learning process.4 The conceptualization of a learning process for use in operational programs is an important contribution which the social sciences and professionals with a strong "people orientation" can bring to development policy and practice. In irrigation development these concepts are clearly displayed. There is a growing consensus that farmer participation in irrigation develop- ment and operations is important to the effective operation and mainte- nance of irrigation systems.5 Irrigation bureaucracies, overburdened by the rapid expansion of irrigated area, cannot serve every individual farmer throughout a populous nation or state. Groups of farmers who use the irrigation water are needed to maintain channels and allocate water locally. Further, the farmers' knowledge of local conditions is needed in the development of the physical system. Although there is increasingly wide agreement on these concepts, very few irrigation projects follow them consistently. One reason is that the irrigation agencies executing the projects seldom have the capacity for working with the local people. Staffed mostly with civil engineers, these agencies are geared to constructing the physical DEVELOPING IRRIGATORS' ORGANIZATIONS 75 system. To combine these construction objectives with the social objec- tives of developing strong irrigators' groups requires a variety of adjust- ments in the field-level approach, the management systems, and the policies of the implementing agency. The exact nature of the changes needed in any specific agency would depend on various factors, including its current norms and structures, the culture of the farmers, the nature of the irrigation use, and the policy frameworks within which the agency operates. But in nearly all instances the changes needed are likely to be substantial. While the social scientists on a project development team can suggest that irrigators' associations should be created and involved in the development and management of the irrigation facilities, the suggestion can be implemented only in a most superficial manner if the agency has not systematically developed the capacity that supports such involvement. To develop the agency capacity, a learning process is needed in which the agency first develops strong irrigators' associations on a small scale, examines the problems and issues in so doing, and adjusts its policies, procedures, and personnel to fit the field-level requirements. Such a learning process must begin long before the implementation of a major project. The development of a participatory approach by the National Irrigation Administration (NIA) in the Philippines illustrates such a learning process. After searching for a number of years for better ways of working with farmers, in 1976 the NIA initiated a process which gradually and system- atically created the capacity to develop strong irrigators' associations.6 In September 1981 when the World Bank appraised a $70 million loan for the NIA communal irrigation program, it was able to state that the loan would "support a national program under which farmers would effec- tively participate in planning and construction of communal irrigation systems and after completion would assume full control of the operation and maintenance of the system."7 These were not just empty words in a project document. Rather, they referred to the results of a learning process which had been under way for five years: many field-level procedures had been developed, hundreds of agency personnel were involved in implementing them, the new approaches were being institu- tionalized on a nationwide basis, and the creative capacity existed for further improvement of the program. The Bank's appraisal team included engineers and economists sensi- tive to social issues and a social anthropologist who lived for several days 76 IRRIGATION PROJECIS in a village where the NIA was using the participatory approach. The team members' careful examination of the existing program allowed them to shape the Bank's loan in a variety of ways supportive of the ongoing participatory program. But the team had only to understand and adapt to an existing process, not create it. The learning process which led to developing the national capacity was not part of a limited process of preparing for a Bank loan and was very different from routine procedures for the preparation of similar projects. Below, the discussion presents the path of the learning process in the NIA, the benefits and costs of the participatory approach developed, and the differences between the learn- ing process and the usual project development process. The Learning Process in the Participatory Inigation Program The National Irrigation Administration is the agency primarily respon- sible for assisting Philippine irrigation development. With more than 30,000 employees dispersed throughout the nation's twelve regions and seventy-two provinces, the agency has a leading role in the country's agricultural development program. Its work involves national systems, which are owned, operated, and maintained by the government and usually serve more than 1,000 hectares, and communal systems, which are owned, operated, and maintained by the farmers and usually serve less than 1,000 hectares. As of 1988, the service area of communals totaled about 600,000 hectares, and those of nationals about 500,000 hectares. While communal systems are often initiated by the farmers, the NIA assists them by constructing improvements such as concrete diver- sions, canals, and canal structures. In its basic policies, the Nu has long recognized the importance of organized irrigators' groups. On its national systems the NLA has had a long-standing policy that irrigators' associations are responsible for operation and maintenance of facilities within the areas served by terti- ary-level canals. Once NLA completes its construction assistance, com- munal systems are completely turned over to the local irrigators' associations. Although its policies assume the existence of strong irrigators' associations, before 1975 the agency did not have a systematic process for assisting the establishment and development of such associ- ations. Like many large irrigation agencies throughout the world, the NIA DEVELOPING IRRIGATORS' ORGANIZATIONS 77 viewed its work primarily as the construction of physical facilities and paid minimal attention to forming irrigators' organizations. Many of these associations were hastily created by calling a meeting and electing officers. Unless a strong association already existed in the area, this approach generally resulted in paper organizations that lacked the via- bility for good water management and maintenance. Unease with this approach was widespread in the agency, and in 1975 anew government policy on communals highlighted its inadequacy.8 The policy stated that farmers were to pay for construction or improvement costs incurred by the government on communal irrigation systems. The irrigators' associations were to provide 10 percent of the value of the construction assistance in the form of labor, materials, cash, and rights of way and to repay the remaining costs over a number of years without interest. To implement such a policy required an irrigators' association strong enough to comply with these conditions. However, the gap between policy and implementation was wide. The NIA had limited institutional and staff capacity for developing strong irrigators' associations. As an interim measure it contracted a different governmental agency, the Farm Systems Development Corporation, to organize the farmers on communal systems, while the NA concentrated on physical construction. This arrangement was based on the assump- tions that organizing the farmers and constructing the physical facilities were separate tasks, appropriately carried out by different agencies; that problems in coordination at the field level would be minimal; and that most of the organizing could be done during or after construction. Subsequent experience showed that these assumptions were wrong. The issues that most concerned the farmers during preconstruction and construction were technical ones, such as the location of the diversion and the canals, the timing of the construction activities, and the choice of laborers for the construction work. Because the socio-organizational tasks were not carried out by the agency doing the planning, design, and construction of the physical irrigation system, the social organizers could not deal with the issues of concern to the farmers-they approached the farners empty-handed. To generate commitment and organizational strength, the irrigators' association needed to be organized long before construction, with the organizing and engineering tasks closely inte- grated. Seeing this, in 1976 the NMA attempted to learn how to work in this new mode. 78 IRRIGATION PROJECTS Development of the Learning Process Although the NIA had an ongoing program of some 250 communal irrigation projects a year, it began experimenting with the new approach on a small scale. With partial assistance from the Ford Foundation, the NIA began a pilot project in one municipality in central Luzon. Dubbed the "participatory approach," the basic concept was to field full-time community organizers to the project area well before construction began. The community organizers would live in the project area and help the farmers build or strengthen their association, using the intensive activi- ties of planning, design, and construction to develop the skills of the association and to gain the farmers' full commitment and involvement in the development of their own irrigation system. Although the initial pilot project was beset with numerous problems, the results appeared promising. The farmers were eager to participate in the development of their own system; they had considerable knowledge of the local area to contribute to planning and design; and through their involvement in the planning and construction activities, they developed the decisionmaking skill and organizational structures of their associa- tion.9 The project also revealed that fully integrating the institutional and engineering aspects of the work was difficult and required a variety of adjustments on the part of the NIA. At the time of this pilot project, the NiA contracted social scientists from the Institute of Philippine Culture to undertake a sociological study of fifty-one existing communal irrigation systems throughout the country. The field study was intended to provide knowledge of farmers' ap- proaches to managing irrigation systems, and the results were later used in numerous training programs for NIA personnel.'0 In addition to examining the existing communal systems, the social scientists examnined NA's pilot project to see if their findings could contribute to NIA'S organizing work. Others from outside the mLA also became interested in the pilot project. Faculty from the Asian Institute of Management were intrigued with the management implications of the use of a participatory approach, and agricultural engineers from the International Rice Research Institute were interested in the potential for improved water management that strong irrigators' associations pre- sented. In 1979, when the NLA decided to expand the use of the participatory approach, it convened these individuals and its own key officials into a DEVELOPING IRRIGATORS' ORGANIZATIONS 79 Communal Irrigation Committee, chaired by an NIA assistant administra- tor. This committee, supported in part with funds and staff from the Ford Foundation, was charged with guiding the further development of NIA'S participatory program. Two new pilot projects were started, which were explicitly conceived as laboratories in which the agency would learn how it would need to change in order to implement the approach more widely.'" The field-level activities of the new projects were intensively docu- mented by a full-time social scientist residing in each of the project areas and supervised by the social scientists who had conducted the study of the communals.12 Monthly process documentation reports were exam- ined by the Communal Irrigation Committee for indications of problems, issues, and successful approaches. These were examined not to solve the problems of the specific pilot projects, but rather to develop agency capacities to deal with the problems on a program-wide basis.13 Figure 3-1 depicts the process followed. New procedures and policies supported by field guidelines, manuals, and training were then developed, tested, refined, and gradually introduced throughout the country. Figure 3-1. Process UtilizedforProgramImprovement Examination of field Determination of experience to identify agency capacities problems required to avoid or _aaDeviate each problem Application of the Development of oneauimied scaaites required capacities Application of the required capacities on a national scale 80 IRRIGATION PROJECTS To build the decentralized capacity needed to implement the new approach nationwide, in 1980 twelve new participatory communal proj- ects were started-one in each region of the country. In 1981, 24 new participatory projects were added-two in each region-and the follow- ing year a total of 108. In 1983 the participatory approach became the standard operating procedure on nearly all the NIA-assisted communal irrigation projects, including projects fully funded by the Philippine government and those funded through a World Bank six-year loan which took effect in June 1982. Later, loans covering communal irrigation systems by the Asian Development Bank and a grant from the U.S. Agency for International Development also supported the projects using the participatory approach. By 1987 participatory methods had been used in 1,335 communal irrigation projects covering 180,000 hectares. The participatory concepts were introduced to national systems in 1981. By involving farmers in the rehabilitation of the irrigation system, irrigators' associations were developed, and subsequently they assumed responsibility for major sections of the irrigation system. Use of partic- ipatory methods on national systems expanded thereafter and by 1987 such methods had been introduced on 37 national irrigation systems covering 34,000 hectares.14 As the new methods were introduced to the agency, in-country training programs and workshops were held to develop attitudes and skills appropriate for the mA employees dispersed throughout the regions and Table 3-1. PhasedExpansionoftheParticipatory Projects Workshops and training Participatory projects courses Number Hectares NIA Year added covered' Number participants' 1976-78 1 400 1 25 1979 2 550 2 75 1980 12 1,800 6 150 1981 24 3,600 36 600 1982 108 16,200 50 1,700 1983 200' 30,000 60b 2,006b Total 347 52,550 155 4,550 a. Approxunate number. b. Estimated as of 1983. DEVELOPING IRRIGATORS' ORGANIZATIONS 81 provinces of the nation. Table 3-1 reveals the growth of both the projects and the training for the communals program from 1976 to 1983. This gradual approach to capacity building meant that when the program was applied on a nationwide basis, the personnel, policies, and procedures were in place, appropriate to the needs of the new approach. At the same time, NIA had developed a creative capacity for addressing new needs and problems as they arose. Agency Change The participatory approach required numerous changes in the NiA'S approach at the field level. Some of the key changes in personnel, policies and procedures, and norms, attitudes, and expectations are discussed in this section.15 Personnel Before the participatory approach was introduced, nearly all of NIA'S professional field-level staff in charge of communal projects were trained exclusively in technical subjects such as construction, design, and sur- vey. The participatory approach, however, required the addition of staff oriented toward people and trained in building the problem-solving capacities of local people. Consequently, mA hired community organiz- ers who could work with the farmers to develop irrigators' associations. The organizers came from avariety of educational backgrounds, butmost commonly from applied social sciences such as community development and social work. In 1976 when NIA initiated its first participatory project, it hired six community organizers. Over the ensuing years, the number gradually expanded, by 1987 reaching 395 organizers deployed to communal systems and 167 deployed to national systems. This was not simply a matter of hiring a new category of personnel. The approach the organiz- ers used had to be developed and refined in ways suited to the needs of the farmers and the NMA'S irrigation development program. The organiz- ers had to learn by doing. The gradual approach to expanding the program and the intensive observations of the early pilot projects allowed this new class of person- nel to accumulate learning. Organizers who had worked out the initial 82 IRRIGATION PROJECrS strategies and solved innumerable field-level problems in the initial pilot projects became the supervisors and trainers of the new roster of orga- nizers hired in 1980. When the program expanded again in 1981, the best of the 1980 recruits were added to the pool of supervisors and trainers of the new group. Training programs initially based on theory and anecdotal material gradually became more structured and program-specific until by 1983 detailed manuals based on extensive field experience had been written to assist organizers throughout the agency.'6 The gradual expansion also allowed the progressive integration of the organizers into the NiA agency structure. At first they were simply part of the NIA-Ford Foundation pilot program. It was not clear where they should be placed organizationally or how a government budget for them should be provided. But as the program grew, the field-level organizers became part of the Provincial Irrigation Office staff, directly under the supervision of the provincial irrigation engineer who served as overall technical and institutional manager of the communal irrigation projects in each province. The organizers were budgeted as part of the regular com- munal irrigation program. At the regional level, one experienced commu- nity organizer was placed in each regional office to help recruit and train new organizers and work with provincial irrigation engineers and field- level organizers on site-specific issues. For the participatory projects involving national irrigation systems, the organizers were considered part of the project staff of the particular system on which they worked. This expansion and integration into the NIA structure was not without problems. Living in the barrio and working closely with the farmers, this new class of personnel often saw projects from a different perspective than that of the regular NIA personnel. Their work habits were also different. Because farmer meetings were often held at night or on weekends, they could not keep the normal working hours of government employees. These differences became sources of tension between engi- neers and organizers at various times. As the program became better understood throughout the country, however, provincial irrigation engi- neers began to view the organizers as part of their staff and as helpful to the goals they wanted to accomplish. Policies and Procedures The NIA made six types of change in its field-level procedures, which are reviewed next. DEVELOPING IRRIGATORS' ORGANIZATIONS 83 SOCIOTECHNICAL PROFILES. Before the development of the participatory approach, NIA'S feasibility studies for communal projects consisted of basic technical data on the water, land, and crops, and rough estimates of the construction costs and the extent of the area to be irrigated. Little or no information was available on the existing social arrangements for irrigation among the people, their organizations, leadership, conflicts, interest in NIA assistance, or landholding patterns. The technical data were often hastily gathered with only a few water availability measurements and minimal topographic investigation. One consequence was that many planned projects were discovered to be unfeasible shortly before con- struction was to begin. To make use of the available construction budget, other projects were then substituted at the last minute. The participatory approach required better analysis of project feasibil- ity. Once a community organizer had worked with the irrigators' associ- ation, it was important to be reasonably sure the project would be carried through to construction. It was also important that the NIA have basic social data so that the initial scheme of development could take into account social as well as technical issues. To meet these needs, a sociotechnical profiling approach was devel- oped. Anthropologists and NIA engineers from the Communal Irrigation Committee worked together with NA organizers to develop comprehen- sive guidelines for collecting field-level data on each communal candi- date for construction assistance. By following the guidelines, a regionally based NIA person could produce a profile in one month. It provided a good initial picture of the area to be assisted and the key social and technical problems likely to arise if assistance were made available. Formats were developed for the write-ups and analyses to help ensure that the data were used for decisionmaking.17 The NIA'S use of these sociotechnical profiles accompanied the expan- sion in the participatory projects. By 1981 each regional office began to hold workshops on profile analysis to assist each province in assigning priorities to the communals seeking program support and in determining the key problems to be addressed in each. The sociotechnical profiles varied substantially in quality. In some cases shortcuts were taken to reduce costs, and data were sometimes inadequate for a full analysis. But in general the profiles provided a much improved data base for site selection and reduced the number of last-minute substitutions of projects. By 1983 the workshops on profile analysis had become a standard part of the NIA communal program. 84 IRRIGATION PROJECTS Figure 3-2. Flow Chart of Preconstruction Activities Week WeekS Wick 12 Week 16 Enginiag g T a p table Restcal canl Reaey baed on dversam a man na!y)lt ann naia and an/ anata~can canal: wtaler tnauneac p:mtidistim pnall diactanina diacaia at fnas' canal; wwa mewAmmem ="mat faact ` assemnbly meetng: fina mbly meeting nmakwut and walknahgh Cmanannity ganivus and CannWaeity cg ans Ic ad taff te ia staff di= pfjec k tnl saistc for codliaali:t withes kantm ame plan schedls FRabilit.y wo v _ ~~~~~~~Fan pu e .m-T abt. aeey psits an n the layout and design of theirirrigationsystemrequiredacommunis tf my / T / ~~~~~~~~~~~~~M=az,c did mmac amm. tify byhlws or e tobmuty fie digh to nine mnonts before cnstanructn. During iti egmize with wh th po e e riaio and ad the wassoc- pkicatio LEAD TIME. 'Me initial pilot projects showed that involving the farmers in the layout and design of their irrigation system required a community organizer to be fielded eight to nine months before construction. During this time the organizer integrated with the community, helped develop or strengthen the irrigators' association, and encouraged the association to discuss the plans for the system with the NIA technical staff. Farmers accompanied the survey team and discussed alternative canal routes in meetings and walk-throughs of proposed routes with the NIA technical staff. Shortly before construction the irrigators' association and the NiA Provincial and Regional Irrigation Offices jointly agreed on the final design. Figure 3-2 shows the flow of the technical and institutional (socio- organizational) activities during the preconstruction period. In some cases, problems interfered with the full implementation of this approach. For some projects the organizers were not fielded to the site for the full nine months of lead time; in others, there were difficulties in scheduling survey teams when the irrigators' association was ready to receive them; and in some there were misunderstandings on the respective roles of the DEVELOPING IRRGATORS' ORGANIZATIONS 85 Week 19 wedk 2Z Week 25 Week 28 P_otenx of mlms, eqp ar d I/; cnmw of acs raFd ~~~ ~' ' meeting d cofe ce p cenfreS e Canvsing, Bidding, and Psn t C(a ¢at /Y aMfeXSlce EWty tsi#mo C-mi- .......... // ~ ~ ~ ~ ~ ~ ~ ~ ~ I enn \o W m ..................................................................___ ____ association, technical staff, and organizers. But as experience was gained in the participatory approach and NIA management systems were adjusted to fit the new needs, these problems lessened substantially. The new procedures meant a major shift in the planning schedules of the provincial irrigation staff. Feasibility work had to be done more than a year earlier than previously and required adjustments in budgeting, scheduling, and the time horizons of provincial staff. Many provinces slipped behind schedule, but throughout the country there was steady progress in developing more orderly and better documented feasibility work and institutional and technical preparation for construction. INSTITTIJONAL AND TECHNICAL COORDINATION. Before the introduction of the participatory approach, NIA field staff did not systematically consult the irrigators' association regarding the development of the physical system. Discussions were sometimes held with individual farm- ers or with the officers of the association, but full and regular interaction with the association on project planning and construction issues was rare. Achieving such regular interaction required a variety of coordination 86 IRRIGATION PROJECrS mechanisms. The community organizers encouraged the irrigators' as- sociation to develop task-specific committees to work with the NIA personnel on survey, design, and construction of the project.18 Organizers needed to consult daily with the project-level engineer and biweekly with the provincial irrigation engineer to help link the technical decision making with the issues the irrigators' association was facing.'9 To explain this coordinated approach, a flow chart based on the experience of the early pilot project showed the parallel activities of the technical and institutional staff in the preconstruction period (figure 3-2). Used in all organizer and engineer training, this helped everyone con- ceptualize theirrespective roles. In subsequent years the chart underwent substantial elaboration, though the essential elements remained the same. The success with which this coordination was carried out varied from project to project. Many engineers were pleased to interact more fully with the farmers, to be welcomed in the villages, and to understand what was happening in the project area. Others found the approach a burden and the farmers' suggestions a threat to their technical expertise. The early pilot projects had revealed the need for structured workshops with NIA engineers and community organizers in which the institutional and technical problems were aired and a common project perspective developed. Consequently, when the program expanded to all the regions, the Communal Irrigation Committee sponsored regional review and planning sessions for participatory communal projects. These work- shops, in addition to the discussions in the field, helped build the technical staff's understanding of the institutional issues and the organizers' un- derstanding of the technical issues.20 The workshops were so helpful that by 1982 the NiA regional offices throughout the country had instituted them on a regular quarterly basis. HIR1NG AND CONTRACrING FOR CONSTRUCTION WORK. NIA personnel were accustomed to hiring and contracting with little or no consultation with the irrigators' associations. But the early participatory projects revealed that these decisions were of great concern to the farmers. Association members were often eager to work as laborers or contractors in the project (though interest varied from one area to another depending on alternative employment opportunities). The association also needed full involvement in the construction work to generate the local equity contri- bution stipulated by agency policy. DEVELOPING RRGATORS'ORGANZATIONS 87 One approach to meeting the concerns of the farmers was to negotiate contracts with the irrigators' associations for much of the work. Under this arrangement, the association contributed equity by having a percent- age of the contract payment withheld and allocated to their equity account, which meant lower than normal daily wages for those who worked. One problem, however, was that workers did not receive their money on time. Although commercial contractors received payment only when the contract was completed, they would normally have "front money" to pay laborers their daily wages. The irrigators' associations, however, had no such funds. Experience showed that farmers could work for about two weeks without pay, but then became so desperate for money to feed their families that they left in search of other work. In some cases efforts were made to develop small contracts that could be completed and paid for within two-week cycles. In other cases mA directly admin- istered the construction and paid the farner-laborers on a volumetric basis (that is, per cubic meter of canal earthwork). Sometimes, when outside contractors were used, conditions were imposed to make this acceptable to the association. In some cases, prior to bidding, potential contractors were informed that the association wanted its members to have priority in getting jobs as laborers, and that a percentage of their pay would be deducted as their equity contribution. The association officials also observed the opening of the bids. With these kinds of arrangements the association was usually much more satisfied with the construction work and made higher equity contribu- tions. FINANCIAL RECORDING AND RECONCILATION. Upon completing a com- munal infigation project, NIA presented the irrigators' association with a statement showing the project costs that were chargeable to the associa- tion and the equity contributions. This was used as a basis for calculating the annual amortization payments due. The association was expected to verify this statement and sign it before the new or reconstructed irrigation system was officially turned over to the association. Experience in the early participatory projects, however, revealed that it was difficult for farmers to verify what was often two or more years of project expenses. In nonparticipatory projects this problem caused widespread unwillingness among the associations to sign the final state- ment, with the result that the projects could not be officially turned over. 88 IRRIGATION PROJECrS The Communal Irigation Committee discussed this problem and in response NIA developed a system of periodic financial reconciliations.21 Not only engineers and organizers, but also NIA accountants and billing clerks were trained in the new way of recording expenses. It was suggested that these reconciliations be done monthly but they were often done about every three months. Throughout the country, NIA personnel indicated that on the project expenses, this new approach led to much greater agreement with the irrigators' association than previously. POSTCONSTRUCTION ASSISTANCE. Before the development of the partic- ipatory program, NiA terminated its assistance to an irrigators' association as soon as construction was completed. But, as a general result of the participatory process, irrigators' associations were sufficiently well- developed that it seemed appropriate to cap the NIA construction assis- tance with help on planning the operation and maintenance of the irrigation system and the financial management of the association. During 1979-81 the attention of the Communal Irrigation Committee had concentrated primarily on the preconstruction and construction stages. But by early 1982 attention turned to these postconstruction issues. Since some research had already been done on these issues,22 working groups were constituted to use that research to develop ap- proaches that NiA could implementthroughoutthe country. By mid-1983, approaches specifically tailored to communal systems had been devel- oped to assist both system operations and financial management. NIA personnel throughout the country were trained to use these materials, offering system management training to irrigators' associations once construction was completed. The financial management materials provided simple forms and in- structions with which the irrigators' association could assess and collect irrigation fees, record receipts and expenses, and report financial infor- ination to its general membership.23 These financial systems were viewed by many observers as important to the success of the associations. Many rural community organizations in the Philippines had collapsed because members suspected their officers of mishandling funds. It was hoped that these simple but precise recording systems would help the irrigators' associations avoid this fate. With regard to irrigation system management, NIA'S previous training for farmers had been done mainly on national systems and was designed to explain the water requirements for crops and NIA'S rules for water DEVELOPING IRRIGATORS' ORGANIZATIONS 89 allocation. Under the participatory program NIA'S role switched from lecturer to facilitator. NIA personnel encouraged the farmers to hold a workshop during which the association officers and members would work out their own management plans. NIA would suggest types of problems the association might discuss; provide formats and data with which the association could assess its current operations; give examples of how other communal associations had addressed such problems; and encourage a structured planning process in which the association deter- mined the steps it would take on these issues. The goal was not to give the associations a ready-made plan, but rather to give them a process by which they could make their own system management plans-a process they could use for many years to meet the system's evolving needs.24 Norms, Attitudes, and Expectations Proper implementation of the participatory approach required some fundamental shifts in the norms and attitudes of NIA personnel. One such shift was in the perspective on basic goals. For many NIA personnel the goal of their work was to construct certain physical facilities. In contrast, the goal of the participatory approach was to build simultaneously both a physical system and a local social capability for using and sustaining that system on a long-term basis. A second fundamental shift was in the relation of the NIA personnel to the fanners. Previously, the farmers had played a passive role in system construction, with NIA personnel making all the key decisions. The participatory approach, however, required a partner relationship between the NIA and the irrigators' association. Technical personnel needed to respect the knowledge and traditions of the farmers, while also recogniz- ing the limitations of that knowledge. They needed skill in communicat- ing their technical knowledge to the farmers so that technical decisions could be made jointly. And they needed to use an interactive style, working through farmer committees to accomplish the goals of the project. Technical staff varied widely in the degree to which they developed the needed attitudes, and consequently in the degree to which their projects followed the true spirit of the participatory approach. Many had come from farm families and quickly came to appreciate the value of the new approach. Others, however, continued to work within the NIA'S previous perspectives and found the new procedures frustrating. 90 RGATION PROJECrS Changes in NIA management systems helped create new expectations which supported the needed shifts in norms and attitudes.25 One change highlighted the data on the equity contributed by the irrigators' associa- tion. Previously the 10 percent local equity requirement was viewed as so difficult to fulfill that it was not used as an indicator of project progress. Physical accomplishment and financial disbursement of construction funds were the primary indicators. However, when the fuller involve- ment of the farmers made higher equity contributions possible (as indicated below in table 3-2), equity data were brought "out of the closet" and the generation of equity contributions became a basic responsibility of every provincial irrigation engineer. Another item relevant to the needed shifts in norms and attitudes was the area actually irrigated once construction was completed. NIA had tended to emphasize "area generated" as the key indicator of perfor- mance. This was an estimate of the area that theoretically could be irrigated by the canals that were constructed-whether the water actually ran in those canals or the farmers made use of them was not an intrinsic part of the indicator.26 The introduction of post construction assessment helped focus more attention on actually irrigated area rather than simply on the construction of physical facilities.27 Postconstruction assistance also helped support these shifts in norms and attitudes since it kept NIA personnel in contact with the farmers during the first two cropping seasons, when the farmers were wrestling with the problems of operating and maintaining the system. Another element supporting the shifts was a stronger emphasis on the collection of amortization payments. The NIA devised cost account sys- Table 3-2. Farmers' Equity Contribution to Construction Costs Participatory Nonparticipatory Basisfor measuring contribution systems (n = 21)a systems (n = 22) Farmers' mean equity contribution per hectare of land irrigated in the 1984-85 wet seasonb 2 357 R 54 Farmers' mean equity contribution per 1984-85 wet season system user R 348 12 44 a. Data on the farmers' contribution to system construction costs were not available for three systems in which construction had not been completed. b. r=5.31;p<0.01. DEVELOPING IRRIGATORS' ORGANIZATIONS 91 tems under which provincial irrigation offices were expected to support their operational costs from their equity and amortization collections- and which therefore gave provincial persormel a strong interest in the longer-term viability of the irrigation groups they assisted. Such fundamental shifts do not take place easily or quickly. It was possible in the NIA because of a variety of factors: a general policy framework supportive of local self-reliance; strong and committed lead- ership from NIA'S top management; the gradual expansion of the pro- gram; the addition of a new type of personnel oriented to social issues; appropriate changes in policies and procedures; extensive training and workshops; and the assistance of an able group of individuals from outside the agency, including social anthropologists, management spe- cialists, and agricultural engineers with the time and creativity to assist in the change process. By 1988 the participatory approach was well established within the communal irrigation program and its use had also been applied to substantial areas of the larger, national systems. Over the twelve year development of the participatory approaches, the agency had developed anew organizational culture more compatible with the need to build local social capacity for irrigation. Benefits of the Participatory Approach: Key Findings In 1985 a sufficient number of communal irrigation systems had been assisted using participatory methods that it was possible to quantitatively evaluate the impact of the new approach. A study was conducted by Romana de los Reyes and Sylvia Jopillo, social scientists from the Institute of Philippine Culture of the Ateneo de Manila University, who headed a team of thirteen researchers which examined forty-six commu- nal irrigation systems. They studied the functionality and productivity of the systems, the structure and activities of the irrigators' associations, and the farmers' contributions to the costs of the system. They also examined both the construction and the institutional costs of assisting the systems. Twenty-four of the systems sampled had been assisted using participatory methods while twenty-two had been assisted using non- participatory methods. The systems were randomly sampled from five regions of the country from among systems irrigating over 50 hectares 92 IRRIGATIONPROlECIS of land which had been constructed between 1981 and 1983 and which were fully or partially operational by June 1984. Before examining the variables reflecting the impact of NIA participa- tory vs. nonparticipatory methods, it is useful to compare the basic characteristics ofthe sample systemspriorto the NIA'S intervention. With the exception of two variables, there were no significant differences in the basic characteristics of the two types of systems or the farmers using them. All sites were gravity fed, run-of-the-river irrigation systems, and forty-one of the forty-six had originally been developed by the farmers themselves. Farmers on all of the systems planted primarily rice and tilled about 1 hectare per family. The land tenure status of the farmers was similar, with 25 percent of the farmers being owner-tillers and the remaining 75 percent tenants. One variable showing a significant difference between participatory and nonparticipatory systems prior to the NIA intervention was the size of the system. Nonparticipatory systems averaged 127 hectares while participatory systems averaged only 88 hectares. The other background variable showing a difference was the accessibility of the system. Par- ticipatory systems were somewhat more remote, averaging 5.5 kilome- ters from the nearest town, while nonparticipatory systems averaged 4.35 kilometers. The nature of the construction done with NIA assistance was basically the same for both types of systems. Forty-three out of the forty-six systems existed before the Num intervention, and the assistance involved upgrading the physical facilities to expand the irrigated area and increase the reliability of water delivery. The remaining three were new irrigation systems. Costs per irrigated hectare for the labor, materials, and super- vision of the participatory and nonparticipatory projects were nearly identical, averaging $769 per hectare.28 In 1981 the NIA was still implementing most of its projects in a nonparticipatory way, however, by 1983 participatory methods had become widely used in the agency. Thus, more nonparticipatory systems were assisted in 1981 (86 percent) while more participatory systems were assisted in 1982 (33 percent) and 1983 (58 percent). Key Findings The following summarizes the key findings from the recent de los Reyes and Jopillo study.29 DEVELOPING IRRIGATORS' ORGANIZATIONS 93 FuNCriONALITY AND PRODUCTIVITY. One of the purposes of the partici- patory approach was to encourage farmers' involvement in the layout and design of the new canals and structures so that their knowledge of the local area and their desires could be incorporated into construction plans. Thus, an important issue in evaluating the impact of such partici- pation was the farmers' degree of satisfaction with the resulting canals and structures. The research data revealed that farmers on participatory systems viewed their new facilities as more functional. On the non- participatory systems farmers abandoned or erased 18 percent of the canals constructed under NLA assistance and assessed an additional 20 percent of the new structures as defective. In contrast, on the participatory systems 9 percent of the canals were abandoned or erased and 13 percent of the structures were viewed as defective. In terms of the expansion of irrigated area, as indicated in table 3-3, the NIA assistance produced a positive effect for both participatory and nonparticipatory systems. In the wet season, the proportion of expansion was similar for the two types of system-18 percent for participatory systems and 17 percent for nonparticipatory. But in the dry season there was a substantial difference between the two types of systems, with the participatory systems expanding their irrigated area by 35 percent, while the nonparticipatory systems expanded by 18 percent. Regarding productivity, as measured by rice yields in the irrigated areas, the participatory systems showed substantially greater increases than the nonparticipatory systems. Table 3-4, contrasts the mean rice yields of these two systems. Prior to the NIA assistance, rice yields in irrigated areas on both participatory and nonparticipatory systems were just over 2.5 tons per hectare per season in both the wet and dry seasons. However, after the assistance, only the participatory systems showed significant gains in yields per hectare, a result that appeared to be a function of greater reliability of water delivery in the participatory systems. ORGANIZATION. Data on the associations that managed the irrigation systems revealed that associations developed through the use of partici- patory methods were much more "rooted" in their communities than those of systems where such methods had not been used (see table 3-5). Associations using participatory systems were more likely to have a second organizational tier-the sector-level unit-based in the various geographical parts of the irrigation system. The sectors spread out the 94 IRRIGATION PROJECTS Table 3-3. Mean Areas Irrigated Before and After NIA'S Construction Assistance Participatory Nonparticipatory Time period systems (n = 24) systems (n = 22) Wet season Before the project 88 ha. 127 ha. Crop year 1984-85 104 ha. 149 ha. Expansion 16 ha. 22 ha. Expansion as a percentage of area 18 17 irrigated before the projecta Dry season Before the project 56 ha. 105 ha. Crop year 1984-85 76 ha. 123 ha. Expansion 20 ha. 18 ha. Expansion as a percentage of area 35 18 irrigated before the projectb a. z = 0.08; p < 0.21. b. z=1.38;p<0.08. Table 3-4. Mean Rice Yields on Irrigated Land Before and After NA's Construction Assistance (tons per hectare) Yield on sample Yields on sample farms in partici- farms in nonpartici- Time period patory systemsa patory systemsb Wet season Year prior to the project 2.84 2.59 Crop year 1984-85 3.05 2.65 Difference in yields after the assistance + 0.21 + 0.06 Dry season Year prior to the project 2.56 2.57 After the project 3.11 2.54 Difference in yields after the assistance + 0.55 - 0.03 a. n = 350 farrns for the wet season and 230 for the dry season. b. n = 295 fanrs for the wet season and 178 for the dry season. DEVELOPING IRRIGATORS' ORGANIZATIONS 95 responsibility for irrigation activities and enabled a more decentralized form of management, critical to the geographically dispersed member- ship for carrying out routine tasks. The greater use of sectors on partici- patory systems resulted in a higher ratio of leaders to members, which helped provide stronger links between the full membership and the management of the irrigation system (see table 3-6). The land tenure status of association leaders also reflects the greater "rootedness" of the leaders of the associations on participatory systems. For both participatory and nonparticipatory systems, about three-fourths of the association members were tenants who owned no land of their own. For participatory systems 47 percent of the central-level leaders also owned no land, while on nonparticipatory systems 35 percent of such leaders owned no land. And for those who did own land, the leaders of nonparticipatory systems tended to have larger parcels than those of participatory systems. These differences indicated that inthe associations of participatory systems, the socioeconomic status of the leadership was closer to that of the ordinary members than was the case for nonparticipa- tory systems. The study also examined the activities of the associations in managing the irrigation systems. An important variable was water management, particularly the degree to which efforts were made to distribute water equitably among the members. A commonly used means of assuring such equity in water distribution is to rotate water according to a schedule, allowing each group in turn to make use of all available water for a specified amount of time. The associations on participatory systems were more likely to use this rotation when water became scarce, and were more likely to employ personnel to supervise the water distribution, than were associations on nonparticipatory systems (see table 3-7). Table 3-5. Organizational Structure Participatory Nonparticipatory systems (n = 24) systems (n = 22) Organizational unit Number Percent Number Percent With central board of directors 24 100 18 82 With sector-level units 24 100 13 59 With central boards composed of sector-level representatives 12 50 3 14 96 IRRIGATION PROJECrS Table 3-6. Mean Number of Leaders per Association Participatory Nonparticipatory Type of leader systems (n = 24) systems (n = 22) Central-level officialsa 10.9 8.8 Sector-level officialsb 6.8 3.5 PersonnefC 2.7 1.8 Totald 15.6 12.0 Ratio of leaders to system users in crop year 1984-85e 1:9 1:14 a. t = 1.59; p < 0.06. b. t=2.7;p < 0.02. c. t=1.09;p<0.14. d. t = 1.87; p < 0.03. The total number of leaders is not a simple summation of the three subtypes of leaders because some individuals held positions at both the central and sector levels. e. t = 1.64; p < 0.05. Regarding maintenance, the study found that the associations on the participatory and nonparticipatory systems mobilized their members for about the same number of days of voluntary group work during the 1984-85 wet season-all averaging about twelve person hours per hectare of irrigated land. However, the associations on participatory systems had more hired personnel (2.7 persons per participatory system versus 1.8 per nonparticipatory system) and consequently had more total person power mobilized for maintenance activities. Data on the financial practices of the associations revealed significant differences between the participatory and nonparticipatory systems. Less than one-fifth of the associations on nonparticipatory systems used vouchers for expenditures, conducted an annual audit of their accounts, prepared financial statements, or monitored payments and collectibles on each member's card. For participatory systems the rate of use of such practices ranged from 21 percent to 50 percent which, while significantly greater for nonparticipatory systems, still indicated considerable room for improvement. FARMERS' CONTRIBUTIONS TO CONSTRUCTION COSTS. Philippine govern- ment policy required farmers to contribute to the costs of construction in two ways. Farmers contributed "equity"-labor, materials, and land-at the time of the construction-and they also made annual repayments of DEVELOPING IRRIGATORS' ORGANIZATIONS 97 Table 3-7. Water Distribution Methods Participatory systems Nonparticipatory (n = 24) systems (n = 22) Method Number Percent Number Percent Rotational distribution inthe wet season 6 25 4 18 Rotational distribution in the dry season 14 58 8 36 Special personnel to distribute water 16 67 9 41 the construction costs, which were amortized over a period of years. For their equity payments farmers were asked to contribute 10 percent of the construction costs or P300 per hectare, whichever was less. Farmers on systems using NIA's participatory methods exceeded the minimum, rais- ing an average of P357 per hectare of irrigated land. In contrast, non- participatory systems raised only F54 per hectare (see table 3-2). Regarding the payment of amortizations as of 1985, associations on participatory systems had paid 82 percent of the amount due, while those on nonparticipatory systems had paid 50 percent. However, most of the nonparticipatory systems had been built a year or two earlier than the participatory systems and, consequently, had more payments due, as indicated in table 3-8. It was not yet known how well the payments on each type of system hold up over time. Costs of the Participatory Approach Under its participatory approach, the NIA conducted a number of activities not done in the nonparticipatory projects. These included writing sociotechnical profiles, fielding community organizers, provid- ing financial management and system management training to the irrigators' associations, as well as recruiting, training, and supervising the organizers and profile writers. The costs of these additional activities, shown in table 3-9, averaged a total of 1P498 ($25) per hectare, or 3 percent of the construction costs of the participatory projects. Interest- ingly, when the total development costs (institutional and construction expenses) of the two types of systems were compared, the participatory systems cost a mere P-48 ($2.40) per hectare more than the nonparticipa- 98 IRRIGATION PROJECI5 tory systems because of the slightly higher costs per hectare for construc- tion of the nonparticipatory systems (see tables 3-9 and 3-10). Comparison of Costs and Benefits As indicated above, substantial benefits were reaped from the small investment in institutional activities, including more functional physical structures, greater increases in rice yields, larger increases in dry season irrigated area, and stronger irrigators' associations. But how do these benefits compare to the costs? The benefits, in terms of recovery of the construction costs, were substantial. The farmers' equity contributions at the time of construction were P303 ($15.15) per hectare greater forparticipatory than fornonparticipatory systems, which represented an immediate recovery of 60 percent of the cost of institu- tional development activities. Moreover, based on the experience of the initial few years, loan repayment rates were P87 ($4.35) per hectare per year higher for participatory than for nonparticipatory systems. If that differential rate of repayment is maintained for three years, the remaining cost of the participatory methods would be completely recovered (as- suming a 10 percent discount rate), and if the higher repayment rate were Table 3-8. Construction Loan Repayment Participatory Nonparticipatorq, Payment systems (n = 17)a systems (n = 19) Mean amortization payment dueC 1A 15,088 12 41,667 Mean amortization payment remitted to NIA P 12,429 1 21,005 Mean percentage of amortization due actually paidd 82 50 a. Of the twenty-four sample participatory systems, four were still under construc- tion during the research fieldwork, one had no loan because the association raised a 30 percent equity, and two were not yet due to begin paying their amortization. b. Of the twenty-two sample nonparticipatory systems, two associations had not accepted the final turnover of the system, while one association had no loan because it raised a 30 percent equity. c. Nonparticipatory systems had larger amortization payments because most had been completed in 1981. Participatory systems were more often completed in 1982 and 1983. d. z=2.17;p<0.05. DEVELOPING IRRIGATORS' ORGANIZATIONS 99 Table 3-9. Institutional Design, and Construction Costs perHectare of Wet Season Irrigated Land Participatory Nonparticipatory systems (n = 24) systems (n = 22) U.S. U.S. Costs Pesos dollars Pesos dollars Institutional costs Profile preparations 36 1.80 - Community organizer salaries and field supplies 293 14.65 Community organizer training and supervision 82 4.10 - Financial and system management training for irrigation associations 87 4.35 - Total institutional costs 498 24.90 - Design & construction costsa Chargeable to farmers 12,666 633.30 12,000 600.00 Not chargeable to farmers 2,484 124.20 3,599 179.95 Total design and construction costs 15,150 757.50 15,600 780.00 Total development costs 15,648 782.40 15,600 780.00 Note: Costs have been standardized in 1984 pesos: g20 =$1.00. a. Includes costs of materials, labor, equipment rental, and technical supervision. Table 3-10. System Construction Costs per Hectare Participatory Nonparticipatory Category (n = 21)a (n = 22) Mean per hectare project costb P15,150 P1l5,599 Mean per hectare chargeable costc P12,666 P12,000 Note: Adjusted to 1984 prices using the implicit price index for government con- struction. Based on area actually irTigated 1984-85 wet season. a. Data on system construction costs were not available for three systems in which a turnover of the system had not been made. b. t = 0.08; n.s. c. t = 0.14; n.s. Excludes equity contribution of association. This is payable in fifty years without interest. 100 IRRIGATION PROJECrS maintained longer, the investment in the participatory methods would provide a direct positive financial benefit to the NIA. In investment terms, if the differential repayment between participatory and nonparticipatory systems remained at P87 for a total of seven years, the rate of return over a ten-year period would be 25 percent.30 For the individual NIA staff member, that financial benefit provided an important reason to make use of the new participatory methods. Provin- cial irrigation offices needed to support their routine operating costs (but not project costs) based on income generated through the equity and amortization payments from communal irrigation associations. Thus, the differences in financial results between the participatory and non- participatory methods were important to NiA staff and were one of the reasons for their strong support of the participatory program. Determining the economic returns from using participatory methods is difficult because of the problems of assigning monetary values to the benefits revealed by the field-level impact data. Even to estimate the benefits simply in terms of increased rice production cannot be done based on the results of the de los Reyes and Jopillo study. Their study revealed that participatory systems produced greater increases in rice yields than nonparticipatory systems due to higher yields for previously irrigated areas, and also due to greater expansion of dry season irrigated area. However, because the study did not examine the related increases in costs associated with the production increases, an economic rate of return cannot be determined for the sample communal irrigation systems. Another type of benefit of the new methods was the goodwill generated between the farmers and the NIA personnel. At the completion of con- struction, systems were officially turned over to the irrigators' associa- tions. On participatory systems, the opening ceremonies were much more likely to be festive village events, reflecting the warm relationships that had developed between the NiA and the farmers through the new ap- proach. For the thousands of NIA personnel and the farmers, that benefit may have been one of the most important of all. Developing Local Capacity The participatory approach described above was developed specific- ally to fit the Philippine context. It cannot be extrapolated directly. But the critical characteristics of the program may be relevant to irrigation DEVELOPING IRRIGATORS' ORGANIZATIONS 101 programs in other settings. In fact, many Asian irrigation leaders have already found many of the concepts and methods used by the NIA highly relevant to the needs of their countries. Based on their observations in the Philippines, leaders from Sri Lanka, Indonesia, Thailand, Nepal, and India have initiated policy changes and pilot projects directed toward making greater use of some or all of the basic concepts outlined below. AUTHORITY OF THE ASSOCIATION. The authority given to an irrigators' association can range from cleaning the ditches to having full responsi- bility for operating and maintaining the entire system. While irrigation agency officials in many countries are eager to have irrigators' associa- tions responsible for cleaning the ditches, some are more skeptical of empowering local associations with control over water allocation. It is sometimes thought that water control should be left to an impartial government authority to avoid inequities in water allocation. Particularly on small-scale systems, government contact is likely to be so limited that the concept of government control is only a myth. The myth, however, officially removes the local authority to run the system, leaving a vacuum where individualism could reign. The appropriate degree of authority to be given to the local irrigators' association needs to vary according to each local program. It should be recognized, however, that the less authority the local association has, the weaker it is likely to be. Farmers are not likely to be eager participants in an association that is simply expected to clean the ditches. Chambers suggested a useful guideline for determining the appropriate degree of authority: "In general, government should unambiguously avoid doing that which communities can do for themselves in their own interest, but should intervene when exceptional problems are beyond a community's power to overcome."31 One of the strengths of the Philippine communal irrigation program is its clear policy regarding the authority of the local associations. Each system of less than 1,000 hectares has one association which owns the irrigation system and has responsibility for operating and maintaining the entire system. The association is a legal entity with water rights registered in its name. This provides the farmers with a clear rationale for committing their time and energy to developing their association.32 USE OF EXISTING IRRIGATION GROUPS. In developing an irrigators' asso- ciation, the organizers can either ignore existing groups or explicitly 102 IRRIGATION PROJECTS build on them. In some places this is not an issue as irrigation groups do not exist. But elsewhere, particularly in the humid tropics, there are likely to be many-even though irrigation bureaucracies tend to be blind to them. A number of factors encourage this blindness. Many irrigation agencies prefer to assume they are generating new irrigated hectarage rather than improving existing systems, and therefore they ignore exist- ing irrigation groups. Another factor is the lack of data about these local groups. Even when local socioeconomic data are collected, they usually focus on characteristics of individual farmers such as farm size, crops grown, and yields, but do not reveal the local social organizational arrangements for irrigation.33 Consequently, existing irrigation groups may be ignored and are wasted as an organizational resource. Although these groups do not always operate as effectively as the government would like, they often have withstood the test of time and evolved useful traditions and leadership roles which are not easily replaced.34 The participatory projects of the Philippine communal irrigation pro- gram explicitly tried to build on existing irrigators' groups. It was normally the local irrigation group that made the request for assistance from the government in the first place. The sociotechnical profile then highlighted current irrigation arrangements in the area, providing the NIA technical and organizational team with the basis for making maximum beneficial use of those arrangements. CONTRIBUTION OF ASSOCIATIONS TO CONSTRUCTION COSTS. The Philippine experience reveals the beneficial effects of requiring some immediate contribution to construction costs from the irrigators' association.35 Since the irrigators' association was expected to contribute an immediate 10 percent of construction costs and subsequently repay the remainder, agency personnel needed to concern themselves with developing irrigators' associations. On the farmers' side, the requirement that their association make a 10 percent contribution was an immediate and tangi- ble organizational task that mobilized membership participation. The arrangement also bestowed greater equality on relations between the agency and the local association. The association's power to withhold its contribution provided some leverage with agency personnel not available in the case of a handout. ASSOCIATION INVOLVEMENT IN PLANNING, LAYOUT, AND CONSTRUCTION. A common assumption is that there is no reason to develop an irrigators' DEVELOPING IRRIGATORS' ORGANIZATIONS 103 association until there is water running in the canals. If this assumption is followed, social organizers are not fielded until after the system is built, and the local people have no involvement in developing the system. The detrimental effects of this approach are discussed in chapters 2 and 4. The Philippine experience reviewed here corresponds to that of nu- merous indigenous irrigation systems around the world and reveals the importance of early involvement in developing the capacities of irrigators' associations to manage and maintain their system.36 Development of Inplementing Agency Capacity The dimensions of program design listed above are likely to contribute to developing strong irrigators' associations. But a program design is meaningless unless the implementing agency can carry it out. The Philippine experience reveals some key questions regarding im- plementational capacity. (These questions do not deal with the purely technical issues, though of course engineering capacities are crucial.) Does the agency have the technology for a rapid assessment of the current irrigation arrangements and other operationally relevant social characteristics of the people to be affected by the system? Does it have a way of using that knowledge in developing the irrigation system? Does it have a way of estimating with reasonable accuracy the area to be irrigated? Does the agency have socially oriented personnel who can develop the irrigators' association, working closely with the technical personnel? Is there a training program that develops field workers' capabilities for the specific tasks of developing associations? Is there a clear framework for relating the socio-organizational and the technical work at each step of a project so that the field-level technical and organizational staff and the farmers understand their respective roles? Are there mechanisms (such as workshops) to help these two types of personnel develop common plans and perspectives? Has the agency closely examined the procedures for its field staff to make sure they fit the operational requirements of the strategy for developing strong irrigators' associations? Are there training programs for technical staff to develop the attitudes, skills, and knowledge suited to the strategy? For many national irrigation agencies the answers to most of the above questions would be no, although recent developments in a number of 104 IRRIGATION PROJECrS Asian countries are beginning to change the situation. When there are no such capacities, those concerned with "putting people first," must con- front the need to develop such capacities within national agencies. The issue poses a dilemma for members of project development or appraisal teams because the project cycle followed by many international or bilateral aid agencies is not well-suited to developing such capacities.37 Project Cycle Issues The learning process described here differs from the project cycle commonly used by major international aid institutions.38 The main elements of the learning process are initial small-scale pilot action, use of the pilot projects to build program-wide agency capacity, and gradual expansion of the pilot projects. Small-scale Pilot Action A number of people at the NIA felt that involving farmers in the preconstruction and construction stages of an irrigation system would help build farmer organizations better able to handle the operation and maintenance activities once construction was completed. But the NIA did not know exactly how to implement such involvement. Three small-scale pilot projects were carried out to learn how. Such learning through action sharply contrasts with the preparation for a large loan, which consists of data collection, not action. Feasibility work uses data to determine whether some set of activities will pro- duce sufficient benefit to justify the investment. Then a set of plans and budgets for a five- to six-year implementation period are drawn up on the assumption that experts can figure out in advance what has to be done. But usually in rural development projects-particularly those intended to involve the beneficiaries in some significant way-the steps to be taken have not been tried before, and no one knows exactly what the steps should be or what management procedures the agency needs to support them. This must be learned. While data collection may be helpful to this process, appropriate methods must be developed through action. Just as a person cannot learn to swim by doing a socioeconomic survey of DEVELOPING IRRIGATORS' ORGANIZATIONS 10S swimmers, an agency cannot learn to implement a new approach through feasibility studies alone. Mechanisms for Building Agency Capacity The NiA developed a variety of mechanisms to use the pilot project experience in building program-wide capacity. A Communal Irrigation Committee composed of key NIA officials and individuals from Philip- pine academic institutions and the Ford Foundation continually exam- ined the pilot project experiences. The committee helped identify agency procedures that interfered with successful work, and it encouraged written documentation of successful methods for later use in training. The committee members also helped develop new agency capacities when field-level experience indicated new approaches were needed and sponsored workshops to spread awareness of the new approaches within the agency. In addition, to bolster NIA training for irrigation associations, the committee examined approaches already in use by some of the most successful indigenous communal irrigation systems.39 To ensure that the pilot project experience was fully captured, a researcher documented the daily interactions among the farmers, and between the farmers and NIA personnel. Thus, the issues and problems occurring in the field could be fully understood by a range of people who could contribute to solutions-not just for a specific project, but for eventual broader implementation in many projects. Large loans by foreign donors occasionally have pilot activities asso- ciated with them. In some cases the pilot is done simultaneously with the main project activity, while in others it is attached to a loan for a different purpose, as a means of testing a new idea for a larger, later loan. Usually the pilot does not contain mechanisms that enable the agency to use the pilot activities to improve its own capacity. The staff that oversees the pilot activity is burdened with meeting disbursement sched- ules on the broader loan, which sometimes forces a pace too fast to allow a thoughtful examination of the small-scale pilot project. There is usually no careful documentation of the field-level activities in the pilot area, nor is there a specially formed group of agency and nonagency individuals committed to extracting from the pilot the knowledge for developing broader implementation capacities. When the pilot and the larger project are carried out simultaneously, it becomes particularly difficult to use the pilot project for developing methods and training with a broader appli- 106 IRRIGATION PROJECTS cation. Even if methods and training are developed from the pilot experience, by the time they are available, the personnel on the other parts of the project have usually already pursued their own approaches.40 Gradual Expansion of the Pilot A third element in NIA'S learning process was the expansion of the pilot projects at a rate gradual enough to build broad program capacity and to ensure that staff had the understanding and capability to implement the new approaches. The first step in the expansion process was to implement one pilot project in each administrative region of the country. This gave key personnel throughout the country some acquaintance with the new approaches so that later, when more projects were added, they would be prepared to oversee their implementation. This gradual expansion al- lowed personnel experienced in the new methods to become the direct supervisors and trainers of others who were implementing the new methods for the first time. It also allowed the gradual shift in NIA policies and procedures to fit the needs of the new program. Planning and funding for the expansion were also done gradually. When examination of the field-level activities indicated the program was ready for some degree of expansion, plans were formulated in workshops by the people who would implement that expansion. Ford Foundation grants totaling $1.4 million over twelve years, matched by NIA funds, were provided at one- to two-year intervals, with each grant fitted to the needs apparent at that stage of the program.41 Flexible budget categories allowed deployment of funds to meet the needs as they emerged from the action. In the case of large loans from international agencies, a gradual expansion is also sometimes built into the plans. But to fit the usual project cycle, this expansion and the associated budget items are pro- granimed in advance over a five- to six-year period. However, when the methods to be used are not yet developed, it is difficult to predict how quickly appropriate methods will be created and what budget items will be needed to support them. Furthermore, what is leamed from the initial activities may call for significant changes in the plans. But plans that have already undergone a variety of bureaucratic approvals may be difficult to change. Another problem is that often a large loan concentrates on a particular geographical area within the country, militating against a focus on nationwide capacity building. Exceptions do exist, however, where a DEVELOPING IRRIGATORS' ORGANIZATIONS 107 loan concentrating on a particular geographical area also includes ele- ments applicable on an agencywide basis.42 Although learning in response to field-level experience can occur in major loans,43 various constraints limit the constructive impact of such learning. The basic project cycle currently used by international aid agencies is best suited to large investments where the capacity to imple- ment an appropriate intervention already exists. But when it comes to working with local people to create self-sustaining development, evidence from the field indicates that large government agencies seldom have the capacity. Appropriate action is not simply waiting for an infusion of funds; rather, the agency's capacity for social development must be created. More flexible funding and arrangements for staff support that allow initial small-scale action, accompanied by intensive, creative attention to field-level experience, followed by gradual expansion are needed if the challenge of creating social development capacities is to be met. Notes A key element in the program described in this paper has been the unwavering support of the successive administrators of the Philippine National Irrigation Administration (NIA), Alfredo Junio, Fiorello Estuar, Cesar L. Tech, Federico Alday, and Jose del Rosario. Each in tum provided the organizational climate needed for the program's initiation, growth, and eventual institutionalization within the NIA. Also deserving of mention are the members of the Communal Irrigation Committee, numerous NIA personnel at all organizational levels, and thousands of farmers whose continual flow of creative ideas kept the program vital and evolving. The authors appreciate the thoughtful comments on earlier drafts of this paper provided by Cesar L. Tech, Rafique Qureshi, Avelino Mejia, Romana de los Reyes, and David C. Korten. They are particularly indebted to Romana de los Reyes and Sylvia Jopillo for the data on development costs and the impact of NIA's assistance on a sample of forty-six participatory and nonparticipatory systems. A fuller examination of the NA's program may be found in Frances F. Korten and Robert Y Siy, eds., Transforming a Bureaucracy: The Experience of the Philippine National Irrigation Administration, (West Hartford, Conn.: Kumarian Press, 1988, and Quezon City, Philippines: Ateneo de Manila University Press, 1989). 1. Many reviews of rural development projects have pointed out that beneficial effects have not been sustained in part because of inadequate attention to broadly defmed social issues. See, for example, Michael M. Cemea, "Fanner Organizations and Institution Building for Sustain- able Development," Regional Development Dialogue, vol. 8, no. 2, Summer 1987 (United Nations Centre for Regional Development, Nagoya, Japan); David C. Korten and Rudi Klaus, eds., People CenteredDevelopment: Contributions toward TheoryandPlanningFrameworkss (West Hartford, Conn.: Kumarian Press, 1984); David C. Korten and Felipe Alfonso, Bureau- cracy and the Poor (West Hartford, Conn.: Kumarian Press, 1983); Elliott R. Morss, John K. Hatch, Donald R. Mickelwait, and Charles F. Sweet, Strategies for Small Farmer Develop- 108 IRRIGATION PROJECTS ment,vol. 1 (Boulder, Colo.: Westview Press, 1976); Bruce F. Johnston and William C. Clark, Redesigning Rural Development: A Strategic Perspective (Baltimore, Md.: Johns Hopkins University Press, 1982); Uma Lele, The Design of Rural Development: Lessons from Africa (Baltimore, Md.: Johns Hopkins University Press, 1975); andRam K. Vepa, "Implementation: The Problem of Achieving Results," inPlanningDevelopmentProjects, Dennis A. Rondinelli, ed. (Stroudsburg, Penn.: Dowden, Hutchinson, and Ross, 1977), pp. 169-90. 2. Walter Coward (see chapter 2) cites the example of the Palsiguan River Multi-Purpose Irrigation Project in the Philippines where social science perspectives eventually helped reorient agency interventions to fit better with the existing social arrangements for irrigation in the project area. This example highlights the importance of project preparation that includes individuals who focus attention on social issues in advance of the elaborate technical engi- neering plans which ignore the local people. A more-detailed report on the Palsiguan case can be found in Robert Y. Siy, "Averting the Bureaucratization of a Community-Managed Resource-The Case of the Zanjeras," in David C. Korten, Community Management: Asian Experiences and Perspectives (West Hartford, Conn.: Kumarian Press, 1986), pp. 260-73. 3. For a useful discussion of some political dimensions of project formulation, see Judith Tendler,RuralProjects through Urban Eyes:An Interpretation ofthe WorldBank's New-Style Rural Development Projects, World Bank Staff Working Paper no. 532 (Washington, D.C., 1982). 4. David C. Korten originally articulated the key elements of a leaming process approach in"Community Organization and Rural Development: A Leaming Process Approach," Public Administration Review, vol.40 (December 1980), pp. 480-511. 5. The need for farmer participation in the development and management of irrigation systems is stressed here in chapters 2 and 4. Other recentpublications of irrigation conferences, donor agencies, irrigation agencies, and irrigation researchers also emphasize the importance of fanner participation. Fora recent comprehensive review of worldwide experience regarding farmer participation in irrigation, see Norman Uphoff, Improving International Irrigation Management withFarmerParticipation: Geting theProcessRight(Boulder, Colo.: Westview Press, 1986). 6. A comprehensive examination of the policy and cultural contexts of the NIA's participa- tory program, the processes it used at the village level and their impact, and the methods and management systems which supported the program can be found in Frances F. Korten and Robert Y. Siy, eds., Transforming a Bureaucracy: The Experience of the Philippine National Irrigation Administration (West Hartford, Conn.: Kumarian Press, 1988, and Quezon City, Philippines: Ateneo de Manila University Press, 1989). 7. Appraisal materials for the World Bank-assisted Communal Irrigation Development Project in the Philippines, 1981. 8. For a more detailed discussion of the macro policy context of the mA, see Benjamin U. Bagadion, "The Evolution of the Policy Context: An Historical Overview," in Frances F. Korten and Robert Y. Siy, eds., Transforming aBureaucracy: The Experience of thePhilippine NationallrrigationAdministration (West Hartford, Conn.: Kumarian Press, 1988, and Quezon City, Philippines: Ateneo de Manila University Press, 1989). 9. Carlos D. Isles and Manuel L. Collado, "Farmer Participation in Communal Irrigation Development: Lessons from Laur," Philippine Agricultural Engineering Journal, vol. 10 (1979), pp. 3-4,9. 10. Fora summary of the findings, seeRomanadelos Reyes,Managing Communal Gravity Systems (Quezon City: Institute of Philippine Culture, 1980). A review of lessons from indig- DEVELOPING IRRIGATORS' ORGANIZATIONS 109 enous irrigation associations which the NIA incorporated into its participatory program can be found in Robert Y. Siy, "A Tradition of Collective Action: Farmers and Irrigation in the Philippines," in Frances F. Korten and Robert Y. Siy, eds., Transforming a Bureaucracy: The Experience of the Philippine National Irrigation Adninistration (West Hartford, Conn.: Kumar- ian Press, 1988, and Quezon City, Philippines: Ateneo de Manila University Press, 1989). 11. For a description of process documentation and the action-related leanings that emerged from it, see Romana de los Reyes, "Process Documentation: Social Science Research in a Learning Process Approach to Program Development," Philippine Sociological Review, vol.32, pp. 105-20. 12. For a brief summary of the events in one of these two projects, see Jeanne Frances I. Ilo, "Farners, Engineers and Organizers: The Taisan Project," in Frances F. Korten and Robert Y. Siy, eds., Transforming a Bureaucracy: The Experience of the Philippine National Irrigation Administration, (West Hartford, Conn.: Kumarian Press, 1988, and Quezon City, Philippines: Ateneo de Manila University Press, 1989). 13. Benjamin U. Bagadion, "Developing Farmers' Participation in Managing Irrigation Systems under the National Irrigation Administration," paper read at Second Social Forestry Forum of the Bureau of Forest Development Upland Development Program, November 1981, Asian Institute of Tourism, Manila. 14. Reports on participatory work on national irrigation systems include: Jeanne Frances I. Uo and Ma. Elena Chiong-Javier, Organizing Farners for Irrigation Management: The Buhi-LaloExperience (Naga City, Philippines: Ateneo de Naga, 1983); and Sylvia G. Jopillo andRomana delosReyes, Proceedingsfrom the wcmAProfileAnalysisWorkshop on National Systems (Quezon City, Institute of Philippine Culture, 1988). 15. Foramore-detaileddescriptionofchangesmadeduringthe feasibility, preconstruction, construction, and operations and maintenance stages of the projects, see Frances F. Korten, Building National Capacity to Develop Water Users' Associations: Experience from the Philippines, World Bank Staff Working Paper no. 528 (Washington, D.C., 1982). 16. In 1988 the NIA was producing revised manuals on every aspect of the communals' program. These manuals included: 1) Manual on the Participatory Approach in Communal Irrigation Projects (Preconstruction Phase); 2) Manual on the Participatory Approach in Communal Irrigation Projects (Construction Phase); 3) Communal Irrigation System Man- agement, A Manualfor Trainers (Operation and Maintenance Phase); 4) Communal Irriga- tion System Management, A Manualfor Facilitators; 5) Simplified Financial Management System Manual for Irrigators' Associations; 6) A Manual of Procedures for Participatory Irrigation Projects. Revisions of earlier work canried out by Sylvia Jopilio, the Communal Irrigation Committee, and Arturo Margalo, were led by a team composed of Susan Leones, Grace Ignacio, Vicky Pineda, Karen Jacob, and Carmelo Cablayan. 17. For a fuller description of the development, content, and impact of the sociotechnical profiles, see Romana de los Reyes, The Socio-Technical Profiles: A Tool for Rapid Rural Appraisal (Quezon City, Philippines: Institute of Philippine Culture, 1984). 18. Different committees were created at different times depending on the needs of the project. Examples included committees for survey and design, bylaws, water permits, Secu- rifies and Exchange Commission registration, right-of-way, manpower survey, quality and quantity control, canvassing, bidding and procurement, and cost control. 19. An example of an issue requiring tight coordination between technical and institutional workwasthe acquisitionof rights-of-way. The communal irrigators' association was expected to obtain right-of-way waivers for all canals from the individual owners of the land. Once 110 IRRIGATION PROJECtS tentative locations of canals were determined the association would begin work on this. But if a particular right-of-way appeared to be unobtainable, the technical staff needed to be informed and alternative routes discussed. If the association was eventually able to obtain the right-of-way the altematives could be abandoned. But if not, new efforts were needed on the alternative routes. Other activities such as the deployment of labor for construction and the receipt of construction materials required similarly close coordination. 20. Asian Institute of Management faculty played an important role in these initial workshops. Representing neither the institutional nor the technical viewpoint, and well- practiced in leading discussions, they were able to draw out both viewpoints and orient the discussions toward problem solving. For a fuller description of some of these differing viewpoints, see Edilberto de Jesus, "Managing Bureaucratic Reorientation," in John C. Ickis, Edilberto de Jesus, and Rushikesh Maru, eds., Beyond Bureaucracy: Strategic Management of Social Development (West Hartford, Conn.: Kumarian Press, 1986). 21. Arturo N. Margallo, CostReconciliation Manual (Construction Phase) (QuezonCity: NIA, 1981). 22. Two studies were funded to develop ideas on how to help irrigators' associations with water management. In 1979 the AI'rcH Research Foundation was contracted by NiA to conduct water management studies on some of NIA's initial pilot participatory projects. In 1980 the Central Luzon State University received a grant from the Ford Foundation to study the water management of an indigenous communal irrigation system. The directors of these studies (Alan Early and Honorato Angeles) then became members of the System Management Working Group, along with key NIA personnel and other outside contributors. This group developed an approach to be used nationwide by NIA personnel. Two studies were also undertaken to develop ideas on how to help irrigators' associations with financial management. In 1980 the BIos Corporation was contracted by NIA to develop a financial management system and test it in three participatory communal projects. In 1982 the Development Academy of the Philippines was contracted by NIA to study financial manage- ment systems of some existing communal irrigation associations which had well-developed record keeping systems. The lessons from these efforts were applied in the financial manage- ment system in which NIA personnel were trained nationwide in 1983. 23. The details of the system are containedinMargallo,SimplifiedFinancialManagement System Manual. 24. This approach is fully described in two NIA manuals: Communal Irrigation System ManagementManualfor Trainers and Communal lrrigation SystemManagementManualfor Facilitators. These manuals cover the development of plans for a cropping calendar, water distribution, maintenance management, conflict management, duties and responsibilities, and farm-level facilities. 25. David Korten provides an analysis of key internal management changes within the NIA which supported the participatory methods in "From Bureaucratic to Strategic Organization" in Frances F. Korten and Robert Y. Siy, eds., Transforming a Bureaucracy: The Experience ofthe PhilippineNationalIrrigationAdministration (West Hartford, Conn.: Kumarian Press, 1988, and Quezon City, Philippines: Ateneo de Manila University Press, 1989). 26. FelipeB. Alfonso, "Assisting Farmer Controlled Development of Communal Irrigation Systems," in David C. Korten and Felipe B. Alfonso, eds.,Bureaucracy and the Poor: Closing the Gap (West Hartford, Conn.: Kumarian Press, 1983). 27. Paddy mapping was helpful in assessing this area-for both the NIA and the irrigators' association. For a description of this technique, see Alan C. Early and Benjamin U. Bagadion, "Custom Fit Design of Farm Ditches: A Participatory Approach to Making Irrigation Systems DEVELOPING IRRIGATORS' ORGANIZATIONS 111 Responsive to the Needs of the Farmers" (Los Banos: International Rice Research Institute, 1982; processed). 28. Costs are reported in standardized 1984 dollars, with a conversion rate of 20 pesos per U.S. dollar. 29. For a comprehensive report of this study, see Romana de los Reyes and Sylvia G. Jopillo,AnEvaluationofthePhilippineParticipatoryCommunallrrigationProgram (Quezon City, Philippines: Institute of Philippine Culture, 1986). A shorterversion of that report is also available in Romana de los Reyes and Sylvia Jopillo, "The Impact of Participation: An Evaluation of NIA'S Communal Irrigation Program," in Frances F. Korten and Robert Y. Siy, eds., Transforming a Bureaucracy: The Experience of the Philippine National Irrigation Administration (West Hartford, Conn.: Kumarian Press, 1988, and Quezon City, Philippines: Ateneo de Manila University Press, 1989). 30. The financial rate of return is calculated assuming that the costs ofthe NiA's institutional activities were incurred as follows: 33 percent in year one; 55 percent in year two; and 12 percent in year three. The farmers' higherequity contribution on participatory systems of I303 is assumed to occur in year two. Repayment on the loan is assumed to begin in year four, with participatory systems paying on the average P87 per hectare per year more than nonparticipa- tory systems and maintaining that incremental rate of repayment through year ten. 31. Robert Chambers, "Men and Water The Organization and Operation of Irrigation,"in B. H.Farmer, ed., GreenRevolution? Technologyand Change inRice-GrowingAreasofTamil Nadu and Sri Lanka (Boulder, Colo.: Westview Press, 1977), pp. 250-63. 32. In recognition of the importance of meaningful rights and authorities for water user organizations, in 1988 the Indonesian government began a program, with assistance from the World Bank, the Asian Development Bank, and the Ford Foundation, to transfer to water-user organizations the full management responsibility of the more than 2,000 government-run irrigation systems of under 500 hectares each. For a more comprehensive discussion of the importance of the rights and authorities of water user organizations see Frances F. Korten, "The Policy Framework for Community Management," in David C. Korten, ed., Community Management: Asian Experience and Perspectives (West Hartford, Conn.: Kumarian Press, 1986). 33. This point is elaborated in chapter 2. 34. This point was emphasized at a conference of Asian irrigation agency personnel, social scientists, and donors; see Walter Coward, Bruce Koppel, and Robert Siy, "Organization as a Strategic Resource in Irrigation Development: A Conference Report" (Honolulu: East-West Center Resource Systems Institute; and Manila: Asian Institute of Management, 1983). 35. A study of thirty-six small-farmer development projects worldwide found that the two components "most important in promoting overall success (were): small farmer involvement in decision making in the implementation phase of a development project; and small fanner resource commitment (labor and cash) to a development project." See Morss and others, Strategies for Small Farmer Development, p. 203. 36. Data on the spontaneous development of indigenous irrigation systems in the Philip- pines indicate that they follow a pattern. Local farners band together to put logs and stones in a river and dig channels to bring water to their fields; the same individuals who initiate the system serve as the leaders for operation and maintenance; during the construction of the system, the farmers informally agree on rules and obligations for operating and maintaining it; later, farmers farther from the water source join in, committing themselves to the same rules and obligations for planning and constructing their own part of the system. Thus the social and physical systems develop simultaneously-often over a period of many years. See Romana 112 IRRIGATION PRO3ECTS de los Reyes and others, Communal Gravity Systers: Four Case Studies (Quezon City: Institute of Philippine Culture, 1980); de los Reyes, Forty-seven Commnunal Gravity Systemns: Organization Profiles (QuezonCity: Institute of Philippine Culture, 1980). 37. In responding to this dilemma some programs have commissioned a different agency, with presumably greater capacities for dealing with people, to do the social organizing. However, such a solution divorces the organizational and technical activities and makes it difficult to involve the farmers in the key questions of planning, design, construction, and operation. 38. In the description of the NuA work, the term "project" was used to refer to a specific irrigation project. In the foUlowing section the term is used in the sense commonly used by the large international aid agencies to refer to a financial package to support a set of activities which may involve many different villages, provinces, or regions. 39. A more complete description of the structure, fimctions, and internal dynamics of the Communal Irrigation Committee can be found in Frances F. Korten, "The Working Group as a Catalyst for Organizational Change," in Frances F. Korten and Robert Y. Siy, eds., Transforming a Bureaucracy: The Experience of the Philippine National Irrigation Adninis- tration (West Hartford, Conn.: Kumarian Press, 1988, and Quezon City, Philippines: Ateneo de Manila University Press, 1989). 40. An interesting example of the use of a pilot activity within a large-scale loan is the Gal Oya irrigation project in Sri Lanka. Built into a large loan from the U.S. Agency for International Development (usAID) for system rehabilitation was experimental work in a pilot area where fanner involvement in rehabilitation, operation, and maintenance was encouraged. Although the pilot was originaUy to be concurrent with the larger project, delays in the larger project gave the pilot a significant head start. In this example, personnel from the Agrarian Research and Training Institute and Cornell University provided the special attention needed for learning from the pilot project, though it also meant that the Irrigation Department was not the central implementor of the pilot activity. For more details on this project, see Norman Uphoff, "Experience with People's Participation in Water Management: Gal Oya, Sri Lanka," in Jean-Claude Garcia-Zamor, Participation in Development Planning and Management: Cases from Africa and Asia (Boulder, Colo.: Westview Press, 1985). 41. The costs of implementing the projects were covered by other sources. Ford Foundation funds, in combination with NuA's own funds, were used for research, training, and workshops to support the learning process elements of the program. 42. For a useful discussion of social experiments and learning that took place during implementation of the large World Bank-funded PIDER project in Mexico, see Michael M. Cernea, A Social Methodology for Community Participation in Local Investments: The Experience of Mexico, World Bank Staff Working Paper no. 598 (Washington, D.C., August 1983), in particular section 2 on "The Production of a Methodology for Participation." 43. For example, the World Bank's Communal Irrigation Development Project in the Philippines included components for developing all of NIA's provincial irrigation offices and for training NIA personnel and farmers for communals regardless of whether they would be implementing the communal subprojects covered by the loan. The project thus has some generalized capacity-building orientation. 4 Middle-level Farmer Organizations as Links between Farms and Central hIrigation Systems David M. Freeman and Max K. Lowdernilk Engineering is not the fundamental problem underlying irrigation development in the LDCS. Engineering principles are known and can be adapted, but the major problem ... is to discover ways to utilize farmer clients more effectively in operations and maintenance and in development programs which will create rural transformation. Rural transformation essentially requires changes in farmers' behavior, mo- tivations, and expectations which is hardly possible until institutions exist to provide them with increased production possibilities and incentives. -Aaron Wiener, ICIDD Bulletin The communal task of creating and operating organizations has always been at the center of societal development. People in all cultures have recognized that they must make permanent arrangements to secure and manage collectively what they could not obtain individually. Irrigated agriculture has always meant the organized, collective attempt to control water efficiently to fulfill crop consumptive needs; and the progress of irrigation systems has always depended upon the design and quality of their respective irrigation organizations. Irrigated agriculture has been disproportionately productive. At pres- ent, only about 18 percent of the world's cultivated land is irrigated, but it produces roughly 33 percent of this planet's human food supply.! But the fact that many landscapes of the world are now dominated by dams, 113 114 IRRIGATION PROJECTS reservoirs, and canals cannot hide a disquieting fact. Many irrigation projects have not served the needs of farmers and agricultural production. Too many accounts of irrigation projects are notable for their reports of failure to meet projected agricultural production targets, of poor maintenance, of disappointing economic returns to investment, and of farmers who not only make little attempt to exploit their expensive water supplies to the degree which had been planned, but who undermine the functioning of the systems developed ostensibly to serve them. Montague Yudelman, reflecting on World Bank experience, has suggested that Bank irrigation projects have seldom met expectations.2 The story of poor irrigation water management unfolds around inefficient water use, distributional inequities, disappointing cropping intensities and yields, and irrigation bureaucracies apparently unaware of farmers' needs to control water. The thesis is that many recurrent problems in large-scale gravity flow irrigation systems stem from a failure to couple social rules with physical tools in local farmer irrigation organizations. To improve matters, farmer water demands must be linked to main system water supply management by an intermediate organization. This discussion thus focuses on designing irrigation organizations in the middle ground between central bureaucracies and farmers. It will formulate strategic propositions which can contribute to improved design of local irrigation organizations. The model of sociological analysis suggested here has emerged from several years of fieldwork by the authors, primarily in irrigation systems in Pakistan and other Asian countries. Its central argument is that the design of the middle-level interface between farmers and the bureaucracy is a strategic determinant of farmers' water control and, therefore, of their productivity. With regard to the practical aspects of designing and implementing irrigation projects, the claim is a strong one: by ferreting out key variables affecting irrigation water control, social obstacles constraining the social and agricultural potential of irrigation can be avoided or removed. The Dry and the Wet: Sociology and rrigation Development planners increasingly recognize that social organization is as fundamental to adequate project design as are accurate economic forecasting and precise technical design. This recognition is creating a MIDDLE-LEVEL ORGANIZATIONAL LINCKAGES 115 constituency for increasing sociological participation in project prepara- tion and implementation. The irrigation systems referred to in this paper are primarily large, public, gravity systems. They have at least three organizational tiers: a centralized public bureaucracy, a local command area organization, and a farm level organization. Irrigation water management in South Asian organizations can be defined as a process by which large, technical bureaucracies capture and control water through central irrigation works, deliver it to local command areas which divide and control it further, and in turn pass it on to the farmer. The essential purpose, therefore, of construction or rehabilitation, allocation, maintenance, and conflict man- agement at all levels of any irrigation system is to provide ultimate control over water to its users-the farmers. The extent to which this purpose is fulfilled indicates the effectiveness of the organization. Only one individual-the farmer-combines the factors of production in a particular field and he or she either succeeds or fails to bring in a crop. Whatever the attributes of upstream organizations, the farmers must possess adequate control over water to place it in the crop root zones when it is most productive. Their requirements for water control too frequently go unnoted, and their attempts to gain effective control are too typically viewed as subverting the interests of more powerful interests at higher levels in the irrigation hierarchy. Irrigation system components have been visualized in various ways.3 There are, however, certain functions basic to all models. Figure 4-1 expresses two strategic aspects of canal systems. First, main system water supply managers approach water control with a fundamentally different set of interests than do farmers. The difference in interest, knowledge, and perspective necessitates an intermediate organizational level which can reconcile the farmer's water demand with central management's supply. Second, figure 4-1 asserts that, at each level of irrigation organi- zation, physical works must be constructed and periodically rehabilitated in a nonroutine manner, water must be allocated, facilities must be routinely maintained, and conflicts arising over each activity must be managed so that the collective effort will be sustained. Irrigation organi- zations must be designed at several levels in order to perform the essential functions noted on figure 4-1. Because the water supply provided by main system managements must mesh (top figure 4-1) with farmer crop water demand (bottom figure 4-1) in some organized fashion (middle figure 4-1), even if middle-level 116 IRRIGATION PROJECTS irrigation organizations are not officially recognized by the state, pat- terned interaction must occur in an organized manner among fanners, and between farners and main system management. Two examples of organizational levels drawn from recent case studies illustrate this point. In the large, highly centralized irrigation system of the Pakistani Punjab, main system management takes responsibility for administering water supply from rivers through primary, secondary, and tertiary canals to a fixed outlet (mogha) for each local watercourse. As the water flows through each mogha, water responsibility shifts from the main system to the local community of irrigators. Middle-level water- courses are collectively administered by farmers on a rotational scheme Figure 4-1. The Organization of Irrigation Systems / Conflict Resoltaion / l l l l ~~~~~~~~~~Central Organizaion Constrction Allocation crial | inrfa | and and | Maintenan a | rehabilitation drainage Ogizi X 47c