DIRECTIONS IN DEVELOPMENT The Econonmics 0of Involuntary: Resettlement: Questins 'and Challenges EDITED BY MICHAEL M:. CERNEA _ t-~!-~19081 7-2IM ciiMarch 1999 2 I - if~~~~~~ ::~~~~~~~~ ..... . 4 . ,.A Other World Bank Publications on Resettlement African Involuntary Population Resettlement in a Global Context by Michael M. Cernea, Environment Department, Social Assessment Series 43 (1997) Development Projects and Impoverishment Risks edited by Hari Mohan Mathur and David Marsden (1998) Hydropower Dams and Social Impacts: A Sociological Perspective by Michael M. Cernea, Environment Department, Social Assessment Series 44 (1997) Invol,untary Resettlement: An Annotated Reference Bibliographyfor Development Research by Scott E. Guggenheim, Environment Working Paper 64 (1994) Involuntary Resettlement in Africa edited by Cynthia C. Cook, World Bank Technical Paper 227 (1994) Involuntary Resettlement in Development Projects: Policy Guidelines in World Bank-Financed Projects by Michael M. Cernea, World Bank Technical Paper 80 (1996) Managing Projects That Involve Resettlement: Case Studies from Rajasthan, India by Hari Mohan Mathur, Economic Development Institute (1997) Reconstructing Livelihoods, Theory and Practice: Resettlers' and Refugees' Experiences edited by Michael M. Cernea and Christopher McDowell (1999) Resettlement and Development: The Bankwide Review of Projects Involving Involuntary Resettlement, 1986-1993 Environment Department (1994) Settlement and Development in the River Blindness Control Zone: Case Study, Burkina Faso by Della E. McMillan, Jean-Baptiste Nana, and Kimseyinga Savadogo, World Bank Working Paper 200 (1993) DIRECTIONS IN DEVELOPMENT The Economics of Involuntary Resettlement Questions and Challenges Edited by Michael M. Cernea The World Bank Washington, D.C. (C 1999 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433 All rights reserved Manufactured in the United States of America First printing March 1999 The material in this publication is copyrighted. The World Bank encourages dissemination of its work and will normally grant permission promptly. Permission to photocopy items for internal or personal use, for the internal or personal use of specific clients, or for educational classroom use is granted by the World Bank, provided that the appropriate fee is paid directly to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, U.S.A., telephone 978-750-8400, fax 978-750-4470. Please contact the Copy- right Clearance Center before photocopying items. For permission to reprint individual articles or chapters, please fax your re- quest with complete information to the Republication Department, Copyright Clearance Center, fax 978-750-4470. All other queries on rights and licenses should be addressed to the Office of the Publisher, World Bank at the address above or faxed to 202-522-2422. Cover: background, a rising reservoir from the El Penol Dam in Colombia floods a village, and only the church is still visible (Juan Mesa Villa); fore- ground, squatters construct housing in Lusaka, Zambia (World Bank photo). Library of Congress Cataloging-in-Publication Data The economics of involuntary resettlement:questions and challenges / edited by Michael M. Cernea. p. cm. - (Directions in development) Includes bibliographical references. ISBN 0-8213-3798-X 1. Economic development projects-Developing countries. 2. Developing countries-Economic policy. 3. Forced migration- Economic aspects-Developing countries. 4. World Bank- Developing countries. I. Cernea, Michael M. II. Series: Directions in development (Washington, D.C.) HC59.72.E44E285 1998 333.3'1'091724-dc2l 98-34774 CIP Contents Acknowledgements ix Abbreviations and Acronyms xi INTRODUCTION. Mutual Reinforcement: Linking Economic and Social Knowledge about Resettlement Michael M. Cernea 1 1= Why Economic Analysis Is Essential to Resettlement: A Sociologist's View Michael M. Cernea 5 Editor's Note 5 Insularity versus Partnership 6 Social Knowledge about Resettlement: Recent Progress 7 Economic Knowledge: The Challenges Ahead 8 Brief Review of the Economic Literature 10 Perverse Effects of Insufficient Economic Analysis 13 The Anatomy of Impoverishment: Risks, Safeguards, and Recovery 15 Risk Analysis and Behavioral Risks 15 Impoverishment Risks to People 16 Limitations of the Cost-Benefit Method 19 Underfinancing of Recovery and Development 22 The Promise of an Economics of Recovery 24 Priority Economic Research Areas 24 Additional Lines of Argument 25 Social Contributions to Economic Analysis 26 APPENDIX. Brief Overview of the State of the Art in Social Research on Involuntary Resettlement 28 Intensive, "Thick" Knowledge 30 More Extensive Knowledge 30 iii iv THE ECONOMICS OF INVOLUNTARY RESETTLEMENT From Description to Prescription 32 Policy Formulation 33 Measurements and Quantification 34 New Theory and Models 35 Resistance to Displacement 35 Acknowledgements 36 References 38 2. Methodological Issues in the Economic Analysis for Involuntary Resettlement Operations David W. Pearce 50 Editor's Note 50 Conclusions from Past Resettlement Experiences 52 Does Resettlement Differ from Other Project Externalities? 54 Issues in the Economics of Resettlement 55 Avoiding Involuntary Resettlement 55 Property Rights 57 The No-Worse-Off Criterion 58 Full and Proper Compensation 59 Financing Resettlement 71 The Policy Context 72 Resettlement and Project Rates of Return 73 Guidelines for Gathering and Analyzing Economic Information for Resettlement 73 References 79 3. Comparing the Economic Planning for Voluntary and Involuntary Resettlement John H. Eriksen 83 Editor's Note 83 Programmatic Characteristics of the Projects 86 Bank Financial Participation in Resettlement 87 Relationship between Scale and Success 87 Average Cost of Resettlement per Participant 89 Duration of Resettlement Operations 93 Use of Bank Resources for Resettlement 94 Nature and Quality of Baseline Surveys 96 Quality of Resettlement Plans 97 Differences among Implementing Agencies 99 CONTENTS V Resettlers' Participation in Decisionmaking 106 Accuracy of Cost Projections 107 Inclusion of Transition Costs and Grants 108 Agricultural Characteristics of the Projects 109 Analysis of the Resource Base 109 Appraising Sources of Household Income 110 Participant Selection Criteria 111 Range of Income-Generating Options 112 Farm Models Envisaged 112 Allocation of Land Holdings: Size, Criteria, and Land Tenure 113 Types of Infrastructure Provided 114 Sources of Resettlement Services 115 Degree of Risk Assessment 115 Conclusions and Recommendations 116 APPENDIX. Characteristics of the Agriculturally Based Resettlement Projects Surveyed 121 References 145 4. Economic Dimensions of Urban Resettlement: Experiences from Latin America Maria Clara Mejia 147 Editor's Note 147 Overview of Urban Displacement and Resettlement in Latin America 149 Magnitude of Urban Population Displacement 149 Causes and Rationale for Involuntary Displacement in Urban Areas 150 Specific Conditions in the Urban Context 152 Economic Issues in House Replacement Alternatives 155 Land Titles and Property Compensation 155 Monetary Compensation 156 Socioeconomic Development 156 Subsidized Solutions versus Partial Payment 157 Housing Replacement Alternatives 158 Socioeconomic Profile of the Affected Population 162 Economic Reconstruction and Participation 164 Principal Difficulties in Planning and Implementation 166 Planning 166 Implementation 167 Legal Difficulties 169 Economic Difficulties 170 viii THE ECONOMICS OF INVOLUNTARY RESETTLEMENT 3.2. Characteristics of the Paired Involuntary and Voluntary Resettlement Projects 90 3.3. Comparison of Effective Length of Projects with Resettlement Operations 93 3.4. World Bank Staff Inputs in Involuntary and Voluntary Resettlement Projects in Staff Weeks and by Skill Mix 94 3.5. A Comparison of the Design and Management of Involuntary and Voluntary Resettlement 118 3A.1. Resource Base 122 3A.2. Household Income 124 3A.3. Type of Resettlement Undertaken 126 3A.4. Range of Resettlement Options 128 3A.5. Type of Farm Model Envisaged 130 3A.6. Size of Land Holding Allocated 132 3A.7. Criteria for Land Allocation 134 3A.8. Conditions of Land Tenure 136 3A.9. Type of Infrastructure Provided 138 3A.10. Source of Resettlement Services 140 3A.11. Type of Risk Assessment 142 3A.12. Restriction Placed on Settlers 144 4.1. Population Displacement in Selected Urban Development Projects 151 4.2. Compensation Alternatives Offered to the Displaced Population 158 4.3. Housing Alternatives 161 4.4. Resettlement Costs per Family 172 4.5. Cost of Resettlement as a Percentage of Total Cost 173 4.6. World Bank Participation in Financing Resettlement 175 4.7. Project Optimization to Minimize Displacement 179 Acknowledgements The ideas that led to the preparation of this volume have been much Tdiscussed by the anthropologists, sociologists, economists, and technical specialists who form the "thematic group" for resettlement issues of the World Bank's Social Development Family. We concluded-and set out to make this case here-that more economic research and analyt- ical tools, in addition to sociological and anthropological knowledge, are necessary for understanding and effectively addressing the complex issues of involuntary population resettlement. The authors of this interdisciplinary book acknowledge therefore with great appreciation the support of the Bank's Social Development Department led by Gloria Davis and of the resettlement thematic group led by Maninder Gill. We much benefited from the personal stimulating intellectual interactions with Dan Aronson, Cynthia Cook, Gloria Davis, Ashraf Ghani, Dan Gibson, Maritta Koch-Weser, Maninder Gill, Scott Guggenheim, Ayse Kudat, William Partridge, Robert Picciotto, Ellen Schaengold, Ismail Serageldin, Andrew Steer, and other colleagues whose ideas are abundantly reflected in this volume. We also express our gratitude to the many development practitioners from various countries who helped us during our fieldwork to better grasp and define the intricacies of displacement and relocation. How- ever, the ultimate source of whatever insights this volume communi- cates are the resettlers themselves. They shared with us their firsthand perceptions of what forced relocation is and cast the crude light of real life on the economic and financial aspects of displacement and recovery. The authors and the editor are also indebted to many colleagues and outside readers who reviewed chapters in manuscript-in particular to Andrew Steer, Scott Guggenheim, and Maninder Gill-and helped en- rich them. Warren A. Van Wicklin III, in the initial preparation phase, and Milan Lin-Rodrigo, in the last phase of readying the manuscript for publication, offered help that is much appreciated. Throughout the entire effort, Gracie Ochieng bore the brunt of processing and reprocess- ing the text, all done with her typical diligence and steadiness. Gaudencio Dizon ensured the professional desktopping of the entire ix x THE ECONOMICS OF INVOLUNTARY RESETTLEMENT volume. Alicia Hetzner initially, and Bonnie Bradford with Virginia Hitchcook over the previous year, ably coordinated the technical pro- cess of publication. The background picture reproduced on the volume's cover was gra- ciously contributed by the Colombian sociologist Juan Mesa Villa, who took it himself 20 years ago, in 1978, when he carried out one of the first resettlement studies in Latin America on the population displacement from the reservoir area of the El Penol Dam. It is customary, and fully adequate in this case, to also recognize that whatever is missing or is not well stated in the argument developed in this volume is solely our own responsibility. Abbreviations and Acronyms ADB Asian Development Bank ADEL Agencia de Desarrollo Local CBA Cost-benefit analysis CDHU Housing and Urban Development Company of the State of Sao Paulo CEHAB State Housing Company of Rio de Janeiro CHESF Companhia Hidro Eletrica do Sao Francisco CIDA Canadian International Development Agency COHAB State Housing Company of Minas Gerais DAC Development Assistance Committee DDP Development Discussion Paper DflD Department for International Development (UK) EPZ Export processing zone FAO Food and Agriculture Organization of the United Nations GTZ German Agency for Technical Cooperation ICLARM International Center for Living Aquatic Resources Management IDA International Development Association IDB Inter-American Development Bank IIED International Institute for Environment and Development lIPS International Institute for Population Studies INCRA Instituto Nacional de Colonizacae e Reforma Agraria ISI Indian Social Institute IWGIA International Work Group for Indigenous Affairs kWh Kilowatt-hours LAC Latin American and the Caribbean Regional Office of the World Bank LEGOP Legal Department-Operations NGO Nongovernmental organization NORAD Norwegian Development Agency NRCR National Research Center on Resettlement (China) NTPC National Thermal Power Corporation (India) OD Operational directive xi Xii THE ECONOMICS OF INVOLUNTARY RESETTLEMENT ODA Overseas Development Administration OECD Organisation for Economic Co-operation and Development OECF Overseas Economic Cooperation Fund (Japan) OED Operations Evaluation Department of the World Bank OMS Operation manual statement PAP Project-affected people PAR Performance audit report PCR Project completion report PLN National Electricity Authority (Indonesia) PPF Project preparation facility PROSAM Programa de Saneamiento R&R Resettlement and rehabilitation RSP Refugee Studies Programme (UK) SA Social assessment SAR Staff appraisal report SEHAB Housing Department of Sao Paulo SfAA Society for Applied Anthropology TOI Times of India UGP Project managing unit UNHCR United Nations High Commissioner for Refugees UNICEF United Nations Children's Fund UNIDO United Nations Industrial Organization WTAC Willingness to accept compensation WTP Willingness to pay INTRODUCTION Mutual Reinforcement: Linking Economic and Social Knowledge about Resettlement Michael M. Cernea very important advance in development work yields lessons Ethat point to new directions, fostering a dialogue between prac- tice and research that encourages creative reexamination of past ap- proaches. This volume is the product of such a dialogue on the complicated issue of involuntary population resettlement. The volume calls for overcoming an insular social perspective on resettlement and for building an "alliance" between economic and sociological research about resettlement. It examines the economic tools for planning resettle- ment and searches for ways to refine them. Historically, the disciplines that have most explored resettlement processes and have informed resettlement policy and operations have been social anthropology and sociology. But for reasons discussed in this book, the anthropological analysis of resettlement has remained somehow isolated and insufficiently complemented by parallel inquiry in the economic disciplines. Although involuntary resettlements are a prerequisite for some infrastructural projects and exact multiple undesirable costs, develop- ment economics has paid little attention to the economic and financial underpinnings of resettlement. Awareness about the serious economic consequences of displacement has kept increasing, yet the economic study of resettlement is unjustifiably still lagging. Evidence about the externalization of project costs to resettlers also has kept accumulating, yet the methodology of financial analyses has somehow failed to 1 2 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT internalize them and to find effective remedies. This lack of integration has deprived many resettlement projects of the economic analytical tools and financial means necessary to succeed. Two important conclusions emerge from both operational work and social research: first, involuntary resettlement programs need to be informed better by economic research and theory; second, in the rou- tines of project design, the economic and financial analyses must be ad- justed to address the complexities of displacement and recovery. At this time, development thinking at large is intensely concerned with broadening the frameworks of knowledge and action, particularly with incorporating social dimensions-social transformation-into the development paradigm. This is the substance of the new "Comprehen- sive Development Framework," which proposes a ". . . more inclusive picture of development. We cannot adopt a system in which the macro- economic and financial is considered apart from the structural, social and human aspects, and vice versa. Integration of each of these subjects is imperative" (Wolfensohn 1999; author's emphasis). The same social con- cerns are also at the core of the proposed "New Development Para- digm," which emerges from "both the failures of past conceptions and the changes in the world that lead to the necessity of a new conception" (Stiglitz 1998). This volume participates in this broad policy and intellectual debate about reorienting the development paradigm toward social inclusion and social development by focusing on one aspect: the need to bridge the gap between economic and social knowledge in addressing an important chal- lenge faced by many current and forthcoming development pro- grams-population resettlement. The volume is devoted to the argument for a more direct and involved role for economics in studying the social and economic dimensions and effects of involuntary population resettlement. The authors of this volume bring the perspectives of four different scientific disciplines-economics, sociology, anthropology, and politi- cal science-yet all converge in making the same basic case: they argue for an organic synergy and mutual reinforcement between economic and social knowledge in resettlement work. Bringing in the tools of eco- nomics to complement the sociological and technical analysis of reset- tlement processes is essential not only for explaining their anatomy better, but also for guiding decisionmaking and investments. Informed by the perspectives of these disciplines, the volume's coau- thors offer theoretical and methodological considerations, rein- forcing empirical evidence and lessons learned from actual projects. The studies explore broad policy issues as well as "project-kitchen" issues, such as the procedures of economic planning and analysis, valuation methods, cost identification, resource allocation, and benefit distribution. MUTUAL REINFORCEMENT 3 The first chapter, written by the volume's editor, examines the current state of the art in both social and economic research on resettle- ment. From this angle, it builds up the argument for more specialized economic inquiry as well as for cross-disciplinary knowledge develop- ment. The chapter also highlights the limitations of cost-benefit analysis and the risks of impoverishment through displacement. The evidence shows that the underfinancing of resettlement is pervasive and that it undercuts project outcomes as well as the improvement of resettlers' lives. The author argues that the challenge to economics in projectizing involuntary resettlement is to shift from the shortsighted economics of merely compensating the displacees to an economics of support for resettlers' full recovery and for growth enhancement. The second chapter, written by David W. Pearce, addresses key methodological issues in the economic and financial analyses required for involuntary resettlement. In examining existing economic methods and current practices, the author draws parallels with environmental economics, discusses the unsatisfactory treatment of externalities, and makes important recommendations for improving the economic and fi- nancial foundation of resettlement operations. The third chapter, written by John H. Eriksen, takes the method- ological analysis further by discussing the practicalities of resettlement planning. It is based on the most detailed comparative analysis carried out to date between voluntary resettlement programs and involuntary resettlement programs-particularly of rural populations. The analysis reveals unjustified disparities between the two categories of resettle- ment in the planning and financing of comparable projects cofinanced by the World Bank and points to recurrent undertreatment of the forced resettlement projects. The author documents the large potential for im- proving the planning and financing of involuntary resettlement by adopting methods already tested in voluntary settlements. The fourth chapter, written by Maria Clara Mejia, is dedicated to the economic aspects of urban resettlement, based on experiences in several Latin American countries. It addresses numerous social, economic, and financial variables such as grants, subsidies, compensation issues, affordability of new houses, credits to resettlers, employment in the ur- ban context, and so forth. It demonstrates that in urban resettle- ment-no less than in rural-difficult economic dilemmas must be weighed and resolved to allow the social reinclusion of those displaced. The fifth chapter, written by Lakshman K. Mahapatra, focuses on poverty reduction in resettlements in India, testing the adequacy of the risks and reconstruction model of resettlement and applying the risk-focused framework to a vast body of empirical findings. The author discusses each of the major risks of impoverishment through 4 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT displacement, showing how these risks convert into real impoverish- ment, then outlines the counteractions necessary for avoiding or miti- gating economic and social risks. The chapter also highlights the contrast between India's vast social research and its limited economic research on involuntary resettlement. The final chapter, written by Warren Van Wicklin III, explores the crucial issue of benefit sharing in projects entailing forced resettlement. The chapter demonstrates that contrary to widespread assumptions, fi- nancing resettlers' recovery can be accomplished not only through up-front preproject budgetary allocations, but also through explicit provisions for channeling some of the project-generated benefits to people adversely affected. The author examines the options for be- nefit sharing one by one and recommends employing them as one key strategy for increasing the financial feasibility of resettlement with development. Together, the volume's seven sections aim to inform about unre- solved methodological and operational economic issues, and to further stimulate research and theory, particularly cross-disciplinary economic and social inquiry. The fundamental policy and goal, even where involuntary resettle- ment becomes unavoidable, must remain inclusive development. This means that resettlers should not be excluded from the benefits of the de- velopments they make possible but should instead be explicitly in- cluded. In short, they should share in the gains, not just the pains, of development. In pursuing a closer "partnership" and advocating mutual rein- forcement among economics, anthropology, and sociology in resettle- ment analysis, this volume is only a beginning: once joint research expands, there will be more to explore than the issues we are raising here. The authors of this volume are firmly convinced that a more re- fined and deeper treatment of the complex economics of resettlement by all concerned is essential for achieving resettlement with development. We hope that many researchers and practitioners will respond to this call for better information and comprehensive knowledge and will carry the effort further. References Stiglitz, Joseph E. 1998. "Towards a New Paradigm for Development." Presented as the 1998 Prebisch Lecture at UNCTAD, Geneva, October 19. Processed. Wolfensohn, James D. 1991. "A Proposal for a Comprehensive Development Framework." Memorandum to the Board, Management, and Staff of the World Bank Group, Washington, D.C., January 21, 1999. Processed. 1 Why Economic Analysis Is Essential to Resettlement: A Sociologist's View Michael M. Cernea Editor's This chapter examines the current state of the art in Note resettlement research, comparing the progress in socio-anthropological knowledge about resettle- ment with economic knowledge. The comparison questions whether the economic knowledge and the analytical methods used for planning and financing resettlement are adequate for achieving the goals of resettlement policy. Historically, research on involuntary resettlement has emerged primarily within the fields of anthropology and sociology. Economic research on displacement and resettlement, the author ar- gues, is virtually missing-but would be indispensable to improv- ing resettlement outcomes. Seven characteristics of recent progress in social research are defined and contrasted with the paucity of economic research on re- settlement. Displacements impose major economic and social risks upon the affected people. Absence of economic research partly ex- plains the obsolete methodology of economic and financial analysis employed for planning involuntary resettlements and thus ac- counts for some of the enormous difficulties and failures of such op- erations. The paper argues that the method of cost-benefit analysis, and the conventional project risk and sensitivity analyses used in projects entailing resettlement, are incapable of answering dis- placement's economic and financial challenges and in practice tol- erate the structural underfinancing of resettlement operations. The author calls for a constructive "alliance" between eco- nomic and sociological knowledge on resettlement and argues that 5 6 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT in-depth economic knowledge is indispensable to achieving two fundamental goals of resettlement policy: red uced displacement and development of resettlers' livelihoods, once they have moved. The chapter concludes with a discussion of the areas in which economic research is called upon to make important contributions: valuation of losses, risk analysis, cost analysis, internalization of costs, poverty mapping, distributional inequities, design of safety nets, and rationale for financial investments in reconstruction. The author calls for a novel "economics of recovery and development", to replace the long outdated "economics of compensation" that still guides the majority of resettlement operations, particularly in pro- jects financed only from domestic sources. Developing such a novel "economics of recovery" demands a much closer intellectual and operational partnership than currently exists among economists, sociologists, and technical specialists. The primary goal of any involuntary resettlement process is to pre- vent impoverishment and to improve the livelihood of resettlers. To do this, governments and technical agencies must understand the eco- nomics of dispossession, impoverishment, and recovery and plan for growth at the relocation site. The key point of this chapter is that dis- placed populations face a specific set of risks. These atypical risks are not addressed in routine project economic analysis. Robust empirical evi- dence has shown that, in most cases, these overlooked and little under- stood risks result in cumulated deprivations and severe impoverishment. Insularity versus Partnership Decades of sociological and socio-anthropological research have gener- ated a substantial body of social knowledge about resettlement. This knowledge has been translated into policy and operational prescrip- tions that have improved many involuntary resettlement operations. However, by its very nature, anthropological research primarily fo- cuses on sociocultural dimensions. It does not regularly use the tools of economic analysis to explore the economic and financial variables of re- settlement. Nor does it explore how those variables are incorporated into project budgets. Research findings and lessons from practice indi- cate that far more focus on the economic and financial components of re- settlement is needed. Socio-anthropological knowledge, however important, is not suffi- cient alone to guide all the economic and financial steps involved in WHY ECONOMIC ANALYSIS Is ESSENTIAL 7 displacement and resettlement. Anthropologists should not claim reset- tlement as their "exclusive" territory and should stay clear of any at- tempt to monopolize the research on resettlement. More economic knowledge about resettlement is imperative for better addressing the fi- nancial and budgetary implications of forced relocation. It is also impor- tant to include the political analysis of resettlement events and the scholarly perspective of human geography on such processes (Cemea 1996b). I had the opportunity in 1994 and 1995 to publicly signal the exag- gerated pursuit of exclusive "professional jurisdiction" over substan- tive research areas in resettlement. My point was critical of the insular tendency to "colonize" resettlement issues as an exclusive domain for anthropology. In substantive terms, I argued, important dimensions of resettlement are in the court of our colleagues, the economists. Misunderstand- ing or sheer ignorance in many agencies and countries about the complex economics of displacement and recovery are simply appalling. Many pitfalls in current practice can be traced to the sorry state of the economic research on resettlement, to the flawed prescriptions for economic and financial analysis, and for planning in this domain. Perhaps anthropologists and sociolo- gists have too exclusively colonized resettlement. We must become more inviting of complementary perspectives (Cernea 1995). In this chapter I discuss in some detail the need for integrating the social and economic knowledge about resettlement. In my view, anthro- pologists and economists alike have yet to fully recognize this need. I also review the progress to date in social knowledge, pointing to the ar- eas that require more economic research and project economic analysis. I argue that mutual reinforcement between the economic, financial, and social analyses of resettlement components in projects is indispensable for avoiding their failure. Social Knowledge about Resettlement: Recent Progress Progress in social research on involuntary resettlement has been noth- ing short of explosive during the past two decades. Anthropologists and sociologists started the study of these processes some six or seven de- cades ago, and they maintained a virtual monopoly in this area of in- quiry. Though progress was slow and sporadic in the first four decades, it has picked up great speed, particularly after development agencies started to use social research findings in formulating their resettlement policies and operational activities. 8 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT The appendix to this chapter offers an overview of the recent effer- vescent developments in social research on resettlement. Drawing from this overview, I have identified seven principal characteristics of these recent developments in social knowledge: 1. Social knowledge about processes of involuntary resettlement is more intensive and substantive, "thicker" than ever before. The understanding of displacement's socioeconomic content has been considerably enriched, refined, and systematized. 2. Research on resettlement has developed extensively as well: it covers new sectors of the economy that involve forcible reloca- tions (forestry, urban infrastructure, thermal plants, and so forth) and has been expanded to include previously unexplored geographic areas. 3. Resettlement research has multiplied its products and services, moving from descriptions to prescriptions, from academic analysis to operational research. 4. Research has scaled up, crossing the threshold from micro to macro, from case ethnographies on localized instances of re- settlement to general-sectoral, national, and international- policy frameworks for resettlement. 5. To overcome the limitations inherent in narratives and "soft" methods, resettlement researchers are crafting more precise measurements and indicators for quantifying the magnitude and consequences of resettlement. 6. New theoretical models, based on the ever-growing bodies of em- pirical data on resettlement, inform the practical planning of in- duced resettlement more effectively and generate hypotheses for further research. 7. Resettlers' growing resistance to, and the political implications of, displacement are topically integrated into resettlement re- search and theory; resettlers' participation in the reconstruction process has started to be explored. These seven characteristics reflect essential advances in the state of resettlement social research (see also bibliographic references in the ap- pendix). Such knowledge gains may become gradually codified in new institutional procedures for managing resettlement, previously nonex- istent, and have already led to improvements in resettlement programs. Economic Knowledge: The Challenges Ahead Contrasting with the progress in social research, economic research on the variables of resettlement is lagging far behind. Integration between WHY ECONOMIC ANALYSIS Is ESSENTIAL 9 socio-anthropological knowledge and economic knowledge about re- settlement has far to go, largely because such economic knowledge has yet to be developed. Despite the better social understanding of resettlement, it is dismay- ing that, in practice, so many resettlements programs still go so wrong in so many places-to the detriment of such large numbers of people. Nearly every inquiry presents fresh evidence of "depriving the under- privileged," to use Pandey's forceful expression. And failure to improve or even restore resettlers' livelihoods is frequent. Fresh evidence of im- poverishment after displacement can be found in a series of field sur- veys recently published by the World Bank's Operations Evaluation Department (OED), covering India (Karnataka state), Togo (Nangbeto Dam), and Indonesia (Kedung Ombo) (OED 1998). Why then-despite all the policies, studies, and public debates-is failure in resettlement still so frequent? As in all complex situations, no single cause explains all outcomes. Here too there are several, but I want to focus on causes related to the state of economic knowledge about resettlement. We may consider this from two perspectives: as theory-how economics conceptualizes reset- tlement-and as methodology-how economic methods are applied for costing forced displacement and for determining the financial provisioning of sustainable relocation. Basic research by professional economists on population displace- ments, to the best of my knowledge, currently is not carried out any- where. Development economics still has to respond in full to the challenge of analyzing not only the economic dimensions of dislocation, but also the economic content of other social costs of development. When one of the few books about the social costs of structural adjustment pro- grams came out (Cornia, Jolly, and Stewart 1987), it sent a virtual shock wave through development economics. Population displacement is just another one of those "social" costs; however, the economic book on dis- placements' costs has yet to be written. The shortage of economic theory is mirrored at the practical level in a shortage of specialized methiods or techniques to be used in the economic and financial analysis of resettlement operations at project level. Ex- isting general methods for such analyses have not been tailored to the specifics of displacement. Current methods were proved in projects the world over to be faulty in assessing both the costs and the benefits of re- settlement, yet corrections to these methods have not been worked out. The routine methodology of cost-benefit analysis obscures rather than highlights displacement losses (as shown further in this chapter and in the following chapters by Pearce and Eriksen). 10 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Brief Review of the Economic Literature In a seminal address about the "New Development Paradigm," Stiglitz (1998) forcefully argues that development economics must go beyond the narrowness of the past "Washington consensus" and defines devel- opment as "the transformation of whole societies." The alternative para- digm he proposes is based on a broader conception of development, with a con- comitantly broader vision of development strategies.... Many previous development strategies have focused on pieces of this transformation, but because they have failed to see the broader context, they have failed, and often miserably. Most of these have focused narrowly on economics (Stiglitz 1998). For this paradigm to be adopted and implemented, social pro- cesses-and the social outcomes of economic development-would have to command a much more intense analytical attention from the economic sciences. This should apply also to the "social costs" of devel- opment in general, not just to one such cost as involuntary resettlement is. More economic knowledge on social variables in development is be- coming indispensable to understanding development scientifically, with its benefits and risks, and to influencing and managing change. Key to these changes-wrote Stiglitz-is the movement to "scientific" ways of thinking, identifying critical variables that affect outcomes, attempting to make inferences based on avail- able data, recognizing what we know and what we don't know (Stiglitz 1998; author's emphasis). Certainly, "what we know and what we still don't know" about the economics of resettlement is only a small fraction of the vast body of eco- nomic knowledge needed to inform induced development projects. Nonetheless, this particular body of knowledge becomes, in fact, deci- sive when governments plan and finance resettlements: the presence or absence of such economic knowledge, its adequacy or inadequacy, can spell success or failure in practice. And evidence shows that such knowl- edge is largely missing. Leading economists and sociologists have explicitly called for bridging the overall disciplinary "gulf" between economic and socio- logical-anthropological studies (Lipton 1992; Ruttan 1988; Smelser and Swedberg 1994). For instance, pointing to the vast domain of commu- nity studies, a well-known Indian economist, Pradan Bardhan, inci- sively noted that WHY ECONOMIC ANALYSIS IS ESSENTIAL 11 the methodological gulf is particularly wide between econo- mists and anthropologists. I have often observed between the two groups an attitude of mutual indifference, or worse. On the rare occasions when they meet they usually talk past each other (Bardhan 1989, p. 7). Supporting Bardhan's argument and comparing the same two disciplines, Michael Lipton suggests, in turn, that there is a crying need for economists' survey techniques and theorizing to be based in anthropologists' study of socially mediated relation- ships.... Part of the way forward... is the grounding of mi- cro-economics in an anthropologically and politically researched understanding of intergroup, inter-role processes, not only in models of expected utility maximization (Lipton 1992). Such insufficient theoretical and methodological "conversation" between economists and anthropologists is directly dysfunctional in certain research subareas, such as development-caused displacements and their impoverishing effects. The neglect of resettlement research by economists was specifically deplored by G. Edward Schuh: Economists have tended to neglect the problems of what an- thropologists and sociologists often refer to as "oustees" or dis- placed people-those who have to resettle in the face of large development projects that wipe out the economic activities from which they have earned their livelihood. This neglect is somewhat puzzling. Perhaps it occurs because the resettlement problem tends to be viewed as a social problem, and therefore one to be dealt with by anthropologists and sociologists. These latter groups of social scientists have brought an important per- spective to such projects... But greater involvement of develop- ment economists should give more of a forward and dynamic perspective. It should also help put front and center the idea that it isn't enough just to re-establish those ousted from their previ- ous situations. Rather, the goal should be to re-establish the oustees in such a fashion that they can experience sustained eco- nomic growth in the future, or sustained increases in per capita income (Schuh 1993, p. 59). David Pearce, a leading British economist credited as one of the cre- ators of environmental economics, notes that resettlement issues have been "something of a Cinderella issue" in the economic appraisal of de- velopment projects (see Pearce, this volume). In turn, John Eriksen has 12 THE EcoNoMIcs OF INVOLUNTARY RESETTLEMENT documented Pearce's severe assessment by demonstrating, project by project, the unsatisfactory quality of economic analysis of involuntary resettlement components (see Eriksen, this volume). The roots of this sit- uation can probably be traced to the classic economic textbooks about development projects (Little-Mirrlees 1974), as well as to recent ones (Fabre 1997), which simply skate over population displacements by pro- jects despite their frequency, as do many other manuals of development economics (Torado 1989). Paradoxically, despite their valuable contributions to the economics of hydropower, economists have traditionally overlooked population displacements as a dysfunctional by-product of dam building. A clear example is the response given by a noted British economist, J. L. Joy, when he was asked "What does an economist want to know about dams?" Joy answered with a lengthy article, listing tens and tens of variables and categories of data that an economist "would want to know" for making economic analyses and planning decisions on dam-building options (Joy 1968). However, displacement by reservoirs, canals, and roads was missing from Joy's list. Except for a passing reference to general social aspects, all indicators were of an economic or technological nature. Whether forced displacement is of a magnitude that demands alterna- tive investment options was an issue he never "wanted to know." Nei- ther did he grant a place in his list to compensation of people's assets, to income restoration, or to economic reestablishment issues. Certainly not all economists think alike. However, sadly, it seems though that J. L. Joy's answer reflected a situation common at the time rather than only a personal opinion. The net result of overlooking for decades the displacement effects of infrastructural investments is that economic conceptualizations of reset- tlement have remained skeletal. Consequently, the methods and indica- tors supposed to translate such understanding in prescriptions for the economic and financial analysis of resettlement at project level are unre- fined. As long as such rudimentary methods continue to be used in eco- nomic analysis and planning, past errors will be recycled. Two examples from India illustrate this situation. As recently as 1995, a book by four economists entitled the Economic Dimensions of the Sardar Sarovar Project (Alagh and others 1995) undertook a reappraisal study of the economics and financing of this now famous project. The book includes interesting new data on the dam, the canal network, power and irrigation benefits, and environmental costs and benefits, in- cluding over 100 pages of tables. Yet resettlement is covered by only a few mentions (pp. 96-98) and some annexes. For a project noted world- wide for its flaws in policy and execution, a project with over 200,000 WHY ECONOMIC ANALYSIS Is ESSENTIAL 13 people slated for displacement and at severe risk of impoverishment, no analysis is made in the book of the costs incurred by the displaced peo- ple or the investments required for their economic recovery. Not even baseline data on resettlers' preproject incomes and economic condition is reported. No projection of sustainable farming models or of alterna- tive strategies is provided to guide the restoration and improvement of the oustees' incomes. Another example is the manual Econoniic Appraisal of Irrigation Projects in India, published by the New Dehli Institute of Economic Growth (Sinha and Bathia 1982). As is well known, many large-scale irrigation projects in India cause loss of land and forced displacement of landed and landless families from reservoirs, canal areas, and adja- cent road systems, and such displacement must be considered during project preparation and appraisal. Yet this volume, in its over 480 pages dedicated to "economic appraisal," found no place to analyze the economic consequences and remedies of displacement. The in- complete "guidance" that this, and similar textbooks, provide to practitioners leads to fallaciously incomplete appraisals of resettle- ment operations. Perverse Effects of Insufficient Economic Analysis This brief review of the economic literature is certainly not exhaustive. One can find now and then some studies on the economic aspects of re- settlement (Paranjpye 1988, Goyal 1996); in China, the interest in the economic dimensions of recovery is stronger (Shi, Xun, and Yu 1996, Wu and Shi 1996, Shi 1996b). Overall, however, the contrast is sharp be- tween the sociological and economic literatures. Examining both literatures, I concluded that the least addressed in research and analysis, as well as in financial planning, are the sources of economic recovery. This conclusion may be a bit puzzling to those who know well, as I do, that anthropologists and sociologists have certainly illuminated not only the cultural but also many of the economic aspects of resettlement. Social and cultural researchers (noneconomists) have substituted ad hoc for economists as well. They have often done a remarkable job generat- ing insights into the economic implications of resettlement. Yet after studying hundreds of projects involving resettlement, I am convinced that the expertise of professional economists, using a solid economic ineth- odology for resettlement projects, is indispensable. Therefore, forging a professional partnership between economists and sociologists in study- ing these issues is the next important step. 14 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT It should not be a surprise that insufficient economic research and knowledge result in inadequate methods for operational economic anal- ysis at project level. Coercive mass expropriation of assets by the state causes losses not identified by the existing methodology for project eco- nomic analysis (Tamondong-Helin 1996). Current property acquisition procedures do not provide compensations sufficient to enable displacees to purchase replacement lands and other assets of compara- ble quality or size, because they either exclude or underpay many af- fected people. In this way, the current methodology externalizes part of the cost, instead of internalizing all costs within project budgets. This is not a minor issue, though I'd call it one of the best-kept "secrets" of reset- tlement misplanning and failure. When resettlement costs are underestimated (whether deliberately or by error), allocations of financial resources for displacees fall short, often by large amounts. Thus displacees and hosts are forced to bear "out of pocket" a share of the cost, reducing the standard of living for all. Exacerbating the problems are budget constraints, often limited or ab- sent political commitment, inflexible allocation procedures, and the fact that procedures are limited to instruments of compensation economics. The fundamental financial issue, however, is investing in resettle- ment with development. Even if all material losses were compensated at their market value, the cash equivalent generally would fall far short of the amount needed to restart the economic and productive activities (new farms, small business, and so on) of resettlers in a new setting at levels that provide equal or improved income. To improve the liveli- hood of resettlers above their previous levels-which is the essence of resettlement with development-additional investments are needed to en- sure recovery and development. The only available economic methodology specially tailored for costing resettlement was published in 1988 by the World Bank. Follow- ing the Bank's first internal resettlement review (Cernea 1986), specific guidelines for the economic and financial analysis of resettlement com- ponents were prepared, discussed widely, and finally issued as an at- tempt to adapt the general logic of economic and financial analysis in Bank-assisted projects to the unique characteristics of dislocation and relocation (see Cernea 1988, annex 2). These innovative guidelines pre- pared by a group of Bank economists working together with sociolo- gists, were-and probably remain-the best economic guidelines on resettlement available in the world literature. But they may also rank among the most ignored. The inadequacy of economic analysis has been highlighted in some World Bank docu- ments, including self-criticism targeting the Bank's practices, as well as criticism of the development economics field at large (World Bank WHY ECONOMIC ANALYSIS Is ESSENTIAL 15 1994). Moreover, these 1988 guidelines have not been adopted in non-Bank-financed projects, which constitute the vast majority-over 95 percent-of interventions causing resettlement worldwide. The neg- ative impacts are unmistakable, particularly in large countries with multiple infrastructure projects, such as India, Pakistan, and Brazil. Several dilemmas in forced resettlement directly challenge conven- tional economic thinking and call for creative research and revised ana- lytical methods. We will discuss some of them-about preventing underfinancing of resettlement, anticipating impoverishment risks, cor- recting cost-benefit analysis, and revising valuation techniques. The Anatomy of Impoverishment: Risks, Safeguards, and Recovery Economic research and operational economic analysis could vastly im- prove resettlement outcomes by focusing on how to overcome the risks of impoverishment inherent in displacementa fundamental issue still not included in routine economic project analyses. Risk Analysis and Behavioral Risks Risk analysis is one of the sophisticated instruments employed in eco- nomic analysis for designing and financing development interventions. But the primary questions are (a) which risks are considered and (b) whether the distribution of risks between project actors and stake- holders is taken into account. The "selection" of risks (Douglas and Wildawski 1982, Rosa 1998) that are seen to merit policy attention and economic-technical analysis in development projects is influenced by political and cultural factors. Conventional economic risk analysis selec- tively focuses only on the risks to capital investments, but not on various kinds of "post-normal risks" (Rosa 1998) that displacements impose upon affected people. Conventional economic risk analysis evaluates the sources, magni- tude, and effects of risks that may reduce the rate of return to capital in- vestments in development projects. It estimates the switching values of key variables (such as duration of project implementation, cost over- runs, and availability of local cofinancing). It also undertakes a sensitiv- ity analysis for each key risk, testing the sensitivity of a project's net present value to possible changes in the variables. When necessary, sen- sitivity analyses are used to develop alternative investment scenarios, including contingency actions, to minimize risk and ensure best return on investment. But if good development practice calls for capital to be 16 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT insured against political risk, currency exchange risk, implementation risk, and so on, so also should the local populations "be insured against the unintended, counter-development risks and consequences" (Downing 1998) of projects. Among the types of project financial risks routinely analyzed at pro- ject outset, one category is defined in economics as "behavioral risks." In the phrase "behavioral risks," however, the notion of "behavior" does not refer to risks that the project imposed on people's normal behavior by compelling them to uproot. Again, the concept, refers to risks to the project's capital investments that can emerge from changes in people's un- related and hard-to-anticipate behavior (for example, assumed con- sumption of project products, employment preferences, etc.). Moreover, it is also common practice for governments to provide guarantees against various risks incurred by investors in infrastructure projects. The state takes responsibility for such risks in order to protect and encourage the private investors (see a detailed discussion of risk calculation methodology in infrastructural projects in Irwin and associ- ates 1997). Yet when the same private investments create risks to such primary stakeholders as the residents of the project area, by expropriat- ing and displacing them, the state does not provide comparable protec- tion against risks to these affected people. Except compensation, most governments do not use any refined economic and legal methodology to institute risk insurance measures for such primary stakeholders. In conclusion, we note that while economic analysis and sensitivity tests are generally designed to identify, measure, and counteract risks to the project and project investors, they are not conversely designed to measure the risks posed by the project to the other project actors, such as the displaced people. Obviously, this is a unidirectional consideration of the relations between stakeholders and projects. This conflicts with the objective of safeguarding people's interests and protecting them against major risks stemming from projects. Human perception and re- sponse to risks, risk aversion, and experience in dealing with risks are essential parameters. Therefore, risk analysis should be defined and practiced as an inherently interdisciplinary activity, a domain of cooper- ation between economics and sociology. The current methodology of risk analysis at project level must be broadened to recognize risk distri- bution among all project actors and address equitably the direct risks to area people as well. These major risks are defined next. Impoverishment Risks to People As stated at the outset, the key point for our risk discussion is that popu- lations displaced by projects face a set of risks that are atypical for other WHY ECONOMIC ANALYSIS Is ESSENTIAL 17 categories of projects. These specific project risks must be pondered from both perspectives-economic and sociocultural. What are these specific risks to people? Understanding how impoverishment risks occur and how to coun- ter them requires deconstructing the anatomy of impoverishment and defining the key determinants of income reconstruction. After synthe- sizing much empirical evidence, I have concluded that the onset of im- poverishment can be represented through a model of eight interlinked potential risks intrinsic to displacement (Cernea 1990a, 1997, forthcom- ing). When not counteracted, these fundamental risks converge and combine into economic, social, and cultural impoverishment. Thus, the model captures the loss by displaced people of all types of capital-natu- ral, manmade (physical), human, and social-as well as loss of opportu- nities and entitlements held at the prior location. The model is also a tool for risk management, suggesting strategies for countering the risks. The eight risks occur with varying intensity in different contents. Concisely, they are defined below: LANDLESSNESS. Expropriation of land removes the main founda- tion upon which people's productive systems, commercial activities, and livelihoods are constructed. This is the principal form of decapitalization and pauperization for most rural and many urban displacees, who lose this way both natural and manmade capital. JOBLESSNESS. Loss of employment and wages occurs more in ur- ban areas, but it also affects rural people, depriving landless laborers, service workers, artisans, and small business owners of sources of in- come. Unemployment or underemployment resulting from resettle- ment tends to linger long after physical relocation. HOMELESSNESS. Loss of housing and shelter is temporary for the majority of displacees, but threatens to become chronic for the most vul- nerable. Considered in a broader cultural sense, homelessness is also placelessness, loss of a group's cultural space and identity (Downing 1994, 1995). MARGINALIZATION. Marginalization occurs when families lose eco- nomic power and spiral downward; it sets in long before physical dis- placement, when new investments in the condemned areas are prohibited. Middle-income farm households become small landhold- ers; small shopkeepers and craftsmen are downsized and slip below poverty thresholds. Economic marginalization is often accompanied by social and psychological marginalization, expressed in a drop in social 18 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT status, oustees' loss of confidence in society and in themselves, and a feeling of injustice and increased vulnerability. FOOD INSECURITY. Forced displacement increases the risk that peo- ple will fall into chronic food insecurity, defined as calorie-protein in- take levels below the minimum necessary for normal growth and work. Sudden drops in food crop availability and income are endemic to phys- ical relocation, and hunger or undernourishment lingers as a long-term effect. INCREASED MORBIDITY AND MORTALITY. The health of oustees tends to deteriorate rapidly from the outbreak of relocation-related parasitic and vectorbome diseases (malaria, schistosomiasis), from malnutrition, and from increased stress and psychological traumas. Unsafe water supply and waste disposal tend to proliferate infectious diseases, and morbidity decreases capacity and incomes. This risk is highest for the weakest population segments-infants, children, and the elderly. Loss OF ACCESS TO COMMON PROPERTY. Loss of access to commonly owned assets (forested lands, bodies of water, grazing lands, and so on)-typically overlooked and uncompensated for in government schemes (with few exceptions, particularly China)-is another major cause of income deterioration for oustees, particularly for the assetless. SOCIAL DISARTICULATION. Community dispersal means disman- tling of structures of social organization and loss of mutual help net- works. Although this loss of social capital is harder to quantify, it impoverishes and disempowers oustees in enduring ways, not recog- nized by project-level economic analysis. Other common risks are loss of access to some basic public services as well as the temporary or perma- nent loss of schooling for children (see Mahapatra in this volume and Mathur 1998). These high-probability risks will certainly become realities if un- heeded. However, like every risk forecast derived from past experience, this predictive model can become a "self-destroying prophecy" (Merton 1957). Impoverishment risks can be avoided or mitigated if they are an- ticipated, correctly identified, and counteracted. If the conventional forms of economic and financial risk analysis are broadened to include project-caused risks to people, as recommended earlier in this paper, the probability of this self-destroying effect is heightened. When done jointly, project economic and social analysis can ascer- tain and quantify the presence of these basic risks during early project WHY ECONOMIC ANALYSIS Is ESSENTIAL 19 preparation, as their shape and intensity vary depending on project de- sign and site circumstances. The combined tools of economists, sociolo- gists, and technical specialists for designing safeguards-including compensatory allocations, benefit sharing, and growth-enhancing in- vestments oriented toward resettlers' development-must become in- tegral to a project's strategy and budget calculations. The impoverishment risks and reconstruction model reveals not only the risks but also the ways to positively reestablish those displaced. Indeed, if we reverse this matrix of risks, the model shows what needs to be done to restore and improve livelihoods: the rehabilitation that must follow displacement. Specifically, for instance, landlessness risks should be eliminated through land reestablishment; homelessness, through sound shelter programs; joblessness, through employment cre- ation; and social disarticulation, through community reconstruction and host-resettler inclusionary strategies. Furthermore, provisions to ensure that those displaced share in the specific benefits generated by the program that caused them to relocate should be included in projects. Practical modalities to do so include resettling reservoir outsees to the newly irrigated lands downstream (rather than around the reservoir), earmarking a share of the revenues from selling energy for resettle- ments or providing access to affordable housing in newly constructed urban areas (more options for benefit sharing are described by van Wicklin in this volume and Cemea 1988,1991). It is obvious from the above discussion of project-caused risks to peo- ple that these risks cannot be tamed solely through compensation for con- demned assets. Yet the strategy of nearly all displacements worldwide is still centered on expropriation payments and not yet on income reestab- lishment. Expropriation laws provide for compensation for condemned assets, but they do not aim, nor do they promise, to reestablish prior levels of living. Therefore, in contrast with prevailing practice, our conclusion here is that the intrinsic socioeconomic risks of resettlement can and must be brought under control only through an encompassing strategy of safe- guarding livelihoods and reestablishing income sources. Limitations of the Cost-Benefit Method The economic methodology employed to justify projects limited to com- pensation approaches is based solely on standard cost-benefit and role of return analyses, which do not address the mitigation of the conse- quences of displacement at the individual level. This is one of the most profound roots of failure of resettlement under such projects. If indeed the core risks are decapitalization and pushing the resettlers out of their prior economically productive positions, then to be adequate, the 20 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT response strategy should be backed up by development investments tailored to the resettlers' capacities and based on growth-oriented eco- nomic analysis and financial provisioning. Cost-benefit analysis (CBA) justifies a project economically when the sum of a project's benefits outweighs the sum of project costs. But CBA overlooks distribution patterns-distribution of both costs and ben- efits. It does not ask who is paying the costs, who specifically is getting the benefits, or who is losing. It only assesses the "total" effect of the pro- ject design to determine how it stacks up relative to other investment alternatives. But harm caused to individual families-the displaced families- cannot be "outweighed" or explained away by benefits to other families, or by the aggregate of project benefits, independent of their allocation. Cost-benefit analysis does not answer the risks accruing to various sub- sets of individuals. Real impoverishment risks are distributed differ- ently than project benefits (Cernea 1997). Aggregate benefits may be real, but they do not automatically offset each individual's costs. Empirical research by anthropologists over decades has docu- mented multiple types of "social costs." It has also revealed the com- monly inadequate valuation of oustees' condemned assets. The concept of "social costs" ("social" is probably a misnomer here), is of- ten misunderstood as referring only to the nonmeasurable costs. In fact, "social" costs define the project costs that are left to accrue to the society outside the project, as opposed to the project's direct internal costs (labor, materials, equipment, and so forth). But social costs are fully project costs in that they would not accrue without the project. Many of them are measurable. These "external" or "societal" costs in- clude (but are not limited to) all that is lost by the forcibly displaced: land, houses, trees, crops, jobs, productive time, cultural assets, com- mon property goods, shrines and places of prayer, burial grounds, and access to social services. Internalizing these "external" costs into the project's overall costs should take the forms of (a) compensations to the displacees at re- placement cost, plus (b) investments for facilitating recovery and ac- celerated takeoff for those uprooted. Unless all categories of losses, as well as the costs of restarting comparable income-generating activities, are budgeted fully into the project, externalization continues to occur. Sac- rifices are imposed on resettlers, and the project transforms the dis- placed into net losers who have to bear many of the "hidden" social costs of projects. It is true that some (not all) costs are not easy to quantify, but invok- ing "intangibles" is often used as an excuse for avoiding any valuation. "Unmeasurable" should not become "forgettable." Ignoring such costs, WHY ECONOMIC ANALYSIS Is ESSENTIAL 21 intentionally or not, "minimizes" the accounted costs of a project, with- out reducing real costs. The cost-benefit method can be easily manipulated or influenced by (a) excluding costs caused by the project, (b) by the way in which costs and benefits are valued when direct market prices cannot be observed or are not conclusive, or (c) by the choice of the discount rate to estimate the present value of a condemned asset. But the point I want to emphasize is that even at its best, withouit distortions, the standard cost-benefit method is incapable of answering the economic and ethical questions involved in forced displacement. Relying on cost-benefit analysis is methodologically questionable not only from a social perspective, but also from a market perspective. Market valuation is based on voluntary exchange between a willing seller and a willing buyer. Resettlement is involuntary: it imposes ad- ministrative acquisition of assets. Nor is resettlement a marginal market valuation decision for those affected: on the contrary, it imposes a total life change. Involuntary resettlement components are included in certain devel- opment projects not by choice but by necessity. They are supposed to be included only after all possible alternatives to avoid and minimize dis- placement have been examined. Since the projects that involve displace- ment are a special subcategory of complex projects, they should undertake a special economic andfinancial analysisfor the displacement, relocation, and socioecononmic reestablishment component, an analysis dis- tinct from the routine cost-benefit analysis. Costs and benefits must be calculated distinctly for each vulnerable group and beneficiary group, not across all populations on the whole "project." The princi- ple of recognizing differential impacts is, in fact, the reason for adopt- ing safeguarding policies (such as resettlement policy). In fact, recent economic thinking (Londero 1996) is returning to the earlier pioneer- ing interest in the social side of the cost-benefit analogy (as expressed in Squire and van der Tak 1975; see also other viewpoints discussed in Rav 1984) for introducing estimates of distributional effects into cost-benefit analysis. Considerable thinking and research has been devoted in recent years in the World Bank to "reviving project appraisal" and improving the economic analysis of projects by shifting from rate of return calcu- lations to a broader examination of the rationale and outcomes of pub- lic financial provisions for projects that result in private benefits (Devarajan, Squire, and Suthiwart-Narueput 1995). Such research demonstrated that projects assisted by the Bank in which costs are borneby thepublicsectorwhilebenefitsare enjoyedby the private sec- tor tend to overestimate netbenefits (ibid., 1995). The undesirable effects 22 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT of such overestimating are aggravated by insufficient attention to is- sues of distributional weight. The efforts to improve the quality of eco- nomic work in the World Bank continue, particularly in areas like public spending programs, infrastructure financing (and entailed risks), microeconomics, public choice, principal-agent issues, safe- guards and safety nets (Baird 1996), with expected gains also for the economic understanding of project-caused displacements. Involuntary resettlement clearly complicates projects, so it should be equally clear that the project economic analysis would have to be modified to meet the requirements of such more demanding project contents. Differential calculation of project costs and benefits for the seg- ment of population displaced is fully feasible. Other refined analytical procedures can be designed as well. Of great concern also should be the fact that, in the absence of rigor- ous analytical demands, insufficient compensation can be easily camou- flaged through nontransparent project budgets. I have seen many project budgets submitted by country agencies to the World Bank that, at first sight, appeared to have a sizable budget for the "R&R" compo- nent. But no breakdown of such budgets is provided. Upon closer exam- ination, however, surprises emerge, showing that the nontransparent budgets create the appearance of high average expenditures per resettler family, by lumping all relocation expenses (infrastructure and human, public assets and family-owned assets) into one consolidated sum. Close examination soon reveals that the bulk of those budgets-up to 70 or 80 percent-represents the financing for condemned public assets (roads, bridges, government buildings, public utilities, military assets, and so on) that are to be relocated or expanded. Only a small percentage of the consolidated R&R budget goes to the affected families. Using such nontransparent budgets to avoid public inspection and debate is a bad practice. The point is that without requirements for an equitable struc- ture of resettlement budgets, oblique accounting will continue to "for- get" the "intangibles." Finally, according to the World Bank's resettlement policy, resettle- ment components in development projects are not expected to pay for themselves. This is one more reason why they should undergo a sepa- rate analysis. This analysis would factor in the costs of recovery and the investments needed for resettlers to reach specified income levels. Underfinancing of Recovery and Development Whether or not the methodology of economic and financial analyses at project level is adequate is not just an academic matter. The operational consequences upon financing levels for resettlement are enormous. WHY ECONOMIC ANALYSIS Is ESSENTIAL 23 Lack of a rigorous and generally accepted methodology leaves room for large degrees of imprecision and arbitrariness in allocations. The result is the underfinancing of resettlement operations in project after project. Such underfinancing directly deprives projects and resettlers of the material means needed for postdisplacement recovery and development. In exploring the causes of success or failure in a subset of 30 projects included in the Bank's resettlement review (World Bank 1994), Guggenheim found an unexpectedly high degree of variance in per-family project allocations of financial resources for resettlement. The variations were not due to intrinsic site differences, but rather to haphazard project approaches. The effects were striking. Analysis showed that resettlement projects with high financial allocations were free of major difficulties, while virtually all of the projects with a low al- location rate experienced serious implementation difficulties. The same wide variance in project expenditure per displaced family was revealed recently, once again, by OED's research of eight Bank-assisted projects in India, China, Thailand, Togo, and Brazil (World Bank 1998). It, too, showed a correlation between resettlement outcomes and level of financing. Yet, despite the general good quality of this OED study, it failed to point to the methodological weaknesses in the economic and financial analyses of resettlement in the evaluated projects. The underfinancing of involuntary resettlement also appears alarmingly arbitrary in John Eriksen's study comparing voluntary set- tlement projects with involuntary (see Eriksen in this volume). By com- paring pairs of projects in five countries, he demonstrates that in almost all cases the economic and financial analysis was performed better, with an orientation towards growth, in voluintary resettlement projects. Para- doxically, voluntary resettlers were more abundantly supplied, finan- cially and technically, than the people compelled to resettle against their will. This comparative analysis demonstrates that, independent of the complexity of forced displacement, failures originate largely in subjec- tive bias and errors in financing strategy. The study concluded that fix- ing the problem requires changes in the planning paradigm and in technical and financial support for involuntary resettlement. Increasing financing for growth-oriented resettlement would bene- fit not only resettlers, but overall project outcomes as well. World Bank research found that 70 hydropower dams co-financed by the Bank regis- tered average cost overruns of 30 percent (inflation adjusted). But cost overruns are often at least partially due to up-front underestimates of the real cost of resettlement. Furthermore, underfinancing of resettle- ment frequently leads to delays in the start of power generation and irri- gation, which inflict losses to the project that are larger altogether than 24 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT the amount needed up-front to adequately finance the resettlement and prevent delays. Our research found that a one-year delay in revenue will reduce the difference between benefits and costs by almost one-third; a two-year delay, by more than half (World Bank 1994). Therefore, it pays to finance resettlement correctly from the outset. Relying on those research findings, World Bank management adopted the recommendations of the 1993-94 Resettlement Task Force to increase the Bank's share in co-financing resettlement operations, particularly for growth-enhancing investments directed to resettlers. "To prevent impoverishment, all future Bank-assisted projects will in- ternalize the full cost of resettlement and of the investments required for income restoration within total project costs" (World Bank 1994). The Bank firmly recommends to its borrowing governments to increase re- settlement financing from domestic resources as well. The Promise of an Economics of Recovery To sum up, the overall argument of this chapter is that resettlement com- ponents in projects would gain enormously from economic research on resettlement and new economic analytical tools. Numerous topics in the economics of resettlement area hold great promise: the task is to gain momentum in the necessary professional economic research. Priority Economic Research Areas A relevant precedent convincingly suggests the great contribution that economics can make. Until recently environmental economics was little researched. Environmental losses and mitigation costs were externalized out of project budgets and overlooked in planning. But when the civic and political recognition of environmental concerns grew to become a powerful force, mainstream economics could no longer afford to ignore environmental losses. The result: both economic theory and environmen- tal practice gained substantially. The economics of the environment has leapt ahead during the past two to three decades, helping to better quan- tify losses, prescribe economic remedies, and ground environmental pro- jects and policies in solid economic and financial analysis. Clearly, similar steps are much needed to build up the economics of resettlement into an economics of recoverv. Such economics has fertile ground from which to grow: its basic pre- mises lie in welfare economics, environmental economics, partly in pub- lic goods economics, and more generally in development economics and political economy. WHY ECONOMIC ANALYSIS Is ESSENTIAL 25 Specific areas where the economic concepts and methods for ad- dressing involuntary resettlement need better articulation are: * Undervaluation of losses; the market's role versus administrative expropriations . Risks of decapitalization to the displacees and risk-insurance options * Impoverishment of resettlers, relative or absolute, and the design of effective safety nets . Patterns of externalization and internalization of displacements and recovery costs (including transaction costs to displacees) * Criteria for financing resettlement components in projects a Distributional inequities in projects entailing relocation * Reestablishment of the productivity of displacees through growth- enhancing financial investments, additional to compensation. This substantive agenda offers economists a vast territory that holds generous promise for research and for consistent translation of knowledge into analytical tools for projects. Such research would also help reorient the current economics of resettlement from a compensation-centered econom- ics confined to repaying losses toward an economics of recovery and growth enhancement. In such a reorientation the overall research priority is to study the reestablishment segment of the displacement-resettlement con- tinuum, complementing the analysis of impoverishment with that of the fi- nancial support needed from governments. The dramas of sudden displacement have understandably attracted more observers than the slow and arduous efforts of rebuilding liveli- hoods. However, if economic research is to illuminate people's way out of displacement dramas, studying coping strategies and the rebuilding of production systems could help accelerate the recovery of future resettlers-the key to successful resettlement. Longitudinal and com- parative studies stand to discover what works and what fails to help dis- mantled societies and local cultures reassemble and function within new environments. Additional Lines of Argument Without going into great detail, there are several other arguments that support the call for deeper economic knowledge about resettlement. First, macroeconomic policy analyses for domains relevant to resettle- ment could introduce pricing limits on the demand for commodities whose production causes displacement. Pricing policies should support 26 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT tariffs that reflect all external costs, social and environmental. This is bound to make demand management more effective and may signifi- cantly diminish the alleged "need" for displacement. Second, full internalization of cost and correct financial analysis and risk consideration will reduce the number of projects causing displace- ment by weeding out those that prove the investment unprofitable. That weeding process also opens the way for alternative less disruptive and more profitable investments. Third, the overall economics of projects will improve by avoiding cost overruns caused by delays and forgone utility. Further improve- ments could arise from budgeting for targeted investments that pro- duce a new stream of long-term benefits from the resettlement with development components. Fourth, by further researching the economic and social dynamics of reestablishment, other levers and resources (of a nonfinancial nature) that reduce costs and accelerate recovery will come to light. The process of income restoration is far from being just a money matter: it is a com- plex social readjustment process that largely depends on cultural fac- tors, too, including resettlers' forms of social organization, degree of cohesion, local leadership, mobilization of resettlers' energy, entrepre- neurship, participation, and motivation. These arguments strengthen the overall case for better economic analysis of resettlement. Strong political commitment from governments and development sponsors (including the private sector) is also critical because resource allocation is ultimately a political matter. As government agencies use the weight of the state and the force of the law to impose expropria- tion and displacement, it is incumbent upon governments also to enable those displaced to get back on theirfeet and share in the benefitsfor which they are displaced. Social Contributions to Economic Analysis The ball, however, is not passed only into the economists' court: it also remains in the court of anthropologists, sociologists, and political scien- tists, who must do much more to promote cross-disciplinary coopera- tion with economists. The historic priority of the noneconomic social sciences in this area remains an asset that should not be turned into the liability of disciplinary isolation or sectarianism. In the past, to obtain recognition for resettlement issues in develop- ment, anthropologists have had to fight intellectually, and sometimes institutionally, against "externalizations" of various kinds, including their own. In the World Bank, for example, professional "jurisdiction" WHY ECONOMIC ANALYSIS Is ESSENTIAL 27 based on competence has been sought and successfully obtained by an- thropologists, out of concern that without due understanding of sociocultural issues the nature of forced population displacement would not be grasped. But the involvement of economists is already long overdue. Economic research should be invited and integrated, so that resettlement difficulties can be better understood and resolved. Inside the World Bank we have taken this road, with partial suc- cess to date. However, on a world scale, beyond the Bank, the task of bringing in economic research and analysis has a long way to go, but is imperative. In addition to cross-disciplinary cooperation, anthropologists and sociologists must look for new insights in their own fields, addressing in more detail the multiple economic variables of displacement from the perspective of their own specialties. Poverty and impoverishment are the heart of the matter in displacement, and one of the tools for analyz- ing them is the impoverishment risks and reconstruction framework, outlined earlier in this chapter (see, for more detail, Cernea 1997). This conceptual model is increasingly being used for both in-depth socioeco- nomic analysis and improved planning (Thangaraj 1996; Ota 1996; Spiegel 1997; Pandey and others 1998; Mathur and Marsden 1998; Mohapatra, Ota, and Mohanti 1998; Mahapatra 1999). Modeling impoverishment risks has led to further explorations about economic counteractions to risks, such as searching for a system of risk insurance, outlining new kinds of "insurance products" and safety net policy provisions (Downing 1998), or linking the risks and re- construction model with the "basic needs" framework in conceptualiz- ing recovery (Basu 1996; see also on broader safety net experiences, Subbarao and others 1996). Anthropological fieldwork should also generate long series of em- pirical data-using the displaced household as unit of analysis and household budgets as the measuring tool-recording incomes and ex- penditures, before and after displacement. Such data are surprisingly scarce. Gender-based economic compar- isons are another angle for innovative research, as some good studies (Kbenig and Diarra 1999, Feeney 1995) have demonstrated. Sociological research in Spain has found that when the would-be resettlers them- selves "shifted" their arguments about the risks caused by an impend- ing dam from the "cultural" arguments to the economic and technical risks, their case became much more convincing to the government au- thorities-what Mairal and Bergna (1996) called a successful "argumental detour." Longitudinal studies on resettled populations, like those advocated by Scudder and Colson, can be powerful in mea- suring economic variables over time, particularly along the various 28 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT stages of the Scudder-Colson model (Scudder and Colson 1982; Scudder 1985,1993). And, of course, we shouldn't forget anthropology's special- ized branch-economic anthropology-which is well positioned to in- tegrate the cultural and the economic in resettlement inquiry. The sober analysis of global trends tells us that involuntary resettle- ment will not disappear. It is actually likely to increase because of urban growth, demographic trends, infrastructural investments, irrigation needs, and policy-induced (as distinct from project-induced) popula- tion displacements. But if a certain degree of displacement cannot be avoided, then the standing challenge is to improve its standards and the overall body of knowledge-both economic and social-apt to guide it. APPENDIX Brief Overview of the State of the Art in Social Research on Involuntary Resettlement Historically, the study of involuntary population resettlements as a con- sequence of purposive development programs began about six or seven decades ago, but only during the 1960s did it start to expand in earnest. Anthropologists and sociologists took the lead in this research from the outset, initiating resettlement research long before the issue became subject to public policy and analysis by other disciplines. Empirical sociological studies on involuntary resettlement have been carried out in both developed and developing countries. In devel- oped countries, one of the first seminal contributions was Herbert Gans' research in the 1960s in the United States on urban involuntary reloca- tion in the Boston area (Gans 1968). Laying the early foundations of re- search and knowledge about involuntary resettlement in developing countries are anthropologists Roy-Burman (1961) in India, Colson and Scudder's field studies in the 1960s on the Gwembe Tonga (see Colson 1971; Scudder 1962, 1973), Brokensha and Scudder (1968), Butcher (1971), sociologist Juan Fernando Mesa in Colombia (1969), political sci- entist Chambers (1970) in Africa, and others. Yet officials-including governments and international agencies, policymakers, and economic planners alike-have overlooked this scholarly knowledge and its implications for general development policymaking, as well as for the practice of project design. Only two de- cades ago, around 1979, was this general official indifference finally bro- ken by the World Bank. At that time the Bank took the initiative to adopt an operational policy on resettlement (see beginning of chapter 1 and Cernea 1986, 1988). This was the first time that a major international WHY ECONOMIC ANALYSIS Is ESSENTIAL 29 development agency turned to social research on resettlement and to so- cial scientists for crafting a policy approach to the thorny, layered prob- lems of involuntary displacement. Based on in-house proposals, Bank management issued explicit guidelines on how staff should treat invol- untary resettlement under Bank-financed projects and on the Bank's ex- pectations from borrowing governments (World Bank 1980; Cernea 1986, 1993-94). This major policy breakthrough also institutionally and intellectu- ally affirmed the contribution of social knowledge to policymaking. From then on, operational social research on resettlement has been strongly promoted for Bank-assisted projects, at headquarters, and in many developing countries. Some of the world's best resettlement social specialists were invited to contribute (as regular staff or as consultants), and since then, the World Bank has become the world's most important hub of social research on resettlement. The social scientists associated with it have produced a vast number of policy and applied studies, theo- retical publications, field surveys, and impact evaluation analyses, while incorporating their knowledge into development processes through "hands on" work. Relying on the expanding contributions of its social sci- entists, resettlement institutionally became recognized as a "topic which touches upon the central aspects of the Bank's work" (Qureshi 1989). Interestingly enough, legal minds inside the Bank responded much faster to these social issues than its economic theoreticians. Issues of le- gal philosophy and the Bank's position were analyzed for the first time and were conceptually defined (Shihata 1991, 1993) in terms that were novel to the existing legal literature worldwide. Accordingly, legal pro- cedures regarding resettlement were patterned for the "legal agree- ments" concluded between the Bank and governments for loans and credits (Escudero 1989). More recently, the Bank's Legal Department issued explicit and de- tailed "Guidelines for Lawyers" (LEGOP 1996) to address legal and pol- icy issues in projects involving resettlement, guidelines derived from the Bank's resettlement policy document. Notably, the attention of many governments was forcefully drawn to the resettlement issues through policy and legal dialogues initiated by the Bank. Operational needs often stimulate new intellectual inquiry. Social science research on involuntary resettlement is a case in point. Its growth has been explosive-in dissertations and books, major studies in journals, national and international conferences, policy documents and evaluation reports, university syllabi, training courses, and briefs for parliamentary hearings or courts of law. Over the past 15 years this growth has practically established a topical resettlement subfield in development anthropology and sociology, causing the intellectual 30 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT resettlement-development nexus to swell. The first annotated biblio- graphy of the global resettlement literature, published by Scott Guggenheim (1994), impressively documents this growth. The public perception of involuntary displacement as a problem of global dimension owes much to this surge in research and publications. In turn, the growing resistance of many affected populations against displacement and the militancy of nongovernmental organizations have endeavored to recast forced displacement as a key political and economic issue in development. What then defines the major advances over the past 15 to 20 years of social research on involuntary resettlement? I characterize this gain in research and knowledge as follows. Intensive, "Thick" Knowledge Today, social knowledge about displacement is shrinking the gaps, go- ing deeper than ever before. To use a common concept in anthropology, it is "thick knowledge." Cognitive gains have been made through refin- ing the understanding of how displacement occurs, in content and in consequences, though our knowledge of the meandering processes of recovery (or lack thereof) that follow remains less advanced. This remarkable deepening of knowledge is primarily the product of an unprecedented hands-on involvement of scores of social scientists during the past 15 years in actual development programs that entail dis- placement. Groups of sociologists and anthropologists left behind their roles as "observers" of such processes and became "participant actors," using their expertise to prevent or minimize displacement. They also help to correct its adverse consequences on the ground by improving the plan- ning, supervision, ongoing analysis, and expost evaluation of operations. Although this type of operational action research is still little heralded, from my vantage point in overseeing the implementation of the World Bank resettlement policy for over 20 years I can stand witness to the good analytical quality, detailed action-oriented prescriptions, and sharpened operational expertise achieved by scores of resettlement so- cial scientists (Bank staff and consultants). They are making a tangible contribution to protecting the welfare of people caught in involuntary displacement processes. More Extensive Knowledge Basic social knowledge about population resettlement has also grown extensively by expanding geographic coverage, which increases oppor- tunities for comparative analysis among countries, and by including WHY ECONOMIC ANALYSIS Is ESSENTIAL 31 previously unstudied sectors of the economy that also trigger involun- tary resettlement. During the 1960s and 1970s, most of the knowledge about resettle- ment resulted from research on urban displacements (done mainly by sociologists in industrial countries) and research on dam-caused dis- placements (done mainly by anthropologists in developing countries). Over the past 15 years, however, considerable empirical research has been devoted to displacements occurring in other sectors, affecting dif- ferent population groups. Forced relocations unrelated to development projects have also been more systematically studied. Among the sectors of the economy encompassed in recent social re- search are forestry (Tumbull 1987; Fernandes, Das, and Rao 1989); min- ing, especially open-pit mining and the construction of thermal power plants (Ray 1994a, 1994b; Hall 1991; and the large literature on Singrauli); biosphere reserves and parks established for environmental rea- sons (West and Brechin 1991, Brandon and Wells 1992); coniversionis in land use from pasture to agriculture (Lane 1996); transportation corridors (roads, highways, power transmission lines, canals); urban and environ- mental infrastructure in developing countries (Cernea 1989; Jellinek 1991; Davidson and others 1993; Reddi 1992, 1994); politically mandated mass relocation (Clay 1988, Pankhurst 1992, Dieci and Viezolli 1992, de Wet 1993); or resettlement caused by structural adjustment reforms. Research on resettlement has also expanded geographically, to in- clude all continents. Much of this expansion has been to countries where no such inquiry has been carried out before. And in countries with a tra- dition in such research, the expansion has been to new topics. In Africa, for example, experiences with high dams on the Volta, Zambezi, and Nile rivers have been the basis of classic anthropological writings on displacement. But a second generation of studies has produced empiri- cal information from new countries that illuminates new facets of reset- tlement (see the 1994 volume on resettlement in Africa by Cook 1994, Cook and Ivarsdotter 1994, Adams 1991, Mburugu 1994, McMillan 1995, Oyedipe 1987, Scudder 1993, Salem-Murdock 1989, Hansen 1990, Roder 1994, Cemea 1997, and many others). In Latin America, the pioneering studies of Alfonso Villa Rojas and Fernando Camaro Barbachano on the Mazatec resettlement have been followed by significant research on dam-related and urban resettlement (Partridge 1993; Partridge, Brown, and Nugent 1982; Bartolome and Barabas 1973, 1990; Wali 1989; de Santos and de Andrade 1990; Bartolome 1992; Serra 1993; Hall 1987; Guggenheim 1994; Mejia 1996; see also Mejia in this volume; Ribeiro 1994; Posey 1996). In Europe, recent writings on resettlement address more the politi- cal and moral issues involved in development aid given by European 32 THE ECONOMfCS OF INVOLUNTARY RESETTLEMENT donors to developing countries (Usher 1997, Lassailly-facob 1998, Conac 1995) than the displacement process taking place in European countries (except for Turkey, see Kudat, Bayram, and Hajai 1996, 1997). In fact, involuntary resettlements continue to take place in some largely industrialized, densely populated countries (to build railways for the new high-speed train, in France; for airports, in Japan and in France). But research on such processes, to my knowledge, is still too limited. In Asia, the resettlement literature has registered the largest and un- doubtedly the fastest growth in India. India's social scientists have de- voted much research to the impact of displacement on tribal groups and scheduled castes (Fernandes, Das, and Rao 1989; Mahapatra 1994; Pandey 1998a). The growing literature on dam-related displacements, particularly on Narmada Sardar Sarovar dam, enriched the debate with new topics-from the role of the state to resettlers' resistance (Thukral 1992; Fernandes and Thukral 1989; Joshi 1987, 1991; Singh 1992; Baboo 1992; Mathur 1994; Baviskar 1995; Fisher 1995; Gill 1995; Dreze, Samson, and Singh 1997; Mathur and Marsden 1998; Mohapatra, Ota, and Mohanty 1998; see also Mahapatra, in this volume, for a discussion of resettlement research in India). Also a series of statistical syntheses, state by state, was started with a volume on resettlement in Orissa State (Fernandes and Asif 1997). In- creasing specialization in resettlement research is illustrated by the landmark volume on legal frameworks governing resettlement practice published by Femandes and Paranjpye (1997) and other legal studies (Ramanathan 1995, 1996), by books and studies on displacement in the mining sector (Ray 1994a, Pandey 1998b, Areeparampil 1996), and by studies in other industries (Reddi 1994). China's creation of a National Research Center on Resettlement (NRCR) has multiplied exponentially the studies of Chinese experi- ences in resettlement (NRCR 1995, 1996; Huang 1984; Wangxiang 1993; Shi 1996a, 1996b), proving that country's continuous interest in measur- ing the economic impacts of displacements and their remedies. Ex- panding resettlement research carried out in China by outside scholars complements the work of Chinese researchers (Barth and Williams 1994; Meikle and Zhu 1999). Altogether, the broadened empirical base creates favorable circum- stances for new theorizing and generalization in resettlement. From Description to Prescription Resettlement research has made considerable strides in advancing from descriptions to prescriptions and from academic study to operational re- search and applied work-a profound change in the state of the art. WHY ECONOMIC ANALYSIS Is ESSENTIAL 33 Early writings on resettlement did not venture far into the territory of recommendations. One exception was the early "handbook" written by Butcher (1971) for the Food and Agricultural Organization (FAO), which, despite its high quality, was in practice totally ignored (Butcher 1979) by FAO and by other development agencies. Over the past 15 years, however, the anthropologists and sociologists from the World Bank, at work on hundreds and hundreds of projects, have studied not only how displacement affects resettlers but also how countless agen- cies around the world plan, finance, manage, and execute resettlement processes (Butcher 1990; Cernea 1986, 1997; Cook 1994; Guggenheim 1989, 1994; Mejia 1996; Partridge 1989; Rew and Driver 1986; World Bank 1994; and other writings, most of which are not yet published). This latter research has led to new knowledge about the administrative, institutional, and legal mechanics of resettlement operations, including the roots of failure and of success. This firmer prescriptive posture taken by social scientists shows their growing confidence and the maturing of resettlement as a field of study. The group of resettlement specialists at the World Bank has syn- thesized much of these experiences in a first electronic "guidebook" on resettlement work (Gill, Gibson, Field, and Schaengold 1998), which makes available a rich inventory of approaches and methods. In turn, the Asian Development Bank (ADB) has sponsored the elaboration of several training manuals for practitioners on project resettlement issues (ADB 1995,1998a, 1998b), which help to disseminate policy ideas and to improve procedures. Practitioners of applied anthropology and sociol- ogy are making some of their finest contributions ever, triggering land- mark changes in how resettlement processes take place. Policy Formulation Social research on involuntary resettlement has come of age by crossing the difficult threshold from micro to macro, moving from writing piece- meal ethnographies on localized resettlement to crafting general policy frameworks. Case-focused ethnographies describe past processes. Policies guide forthcoming activities. Of course, case studies remain a staple of today's research, but there is much to gain from articulating credible policy frameworks. The technical and conceptual work of professional social scientists has been decisive, not only in advancing the World Bank's re- settlement policy, but also the policy of all donor agencies representing the 25 countries in the Organisation for Economic Cooperation and De- velopment (OECD 1992), the resettlement policy of the Asian Develop- ment Bank (ADB 1995), and other agencies. 34 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Social science has played a similar role in formulating national or sectoral resettlement policies in Brazil, Central African Republic, Co- lombia, Guyana, Indonesia, Jamaica, Uganda, Vietnam, and other coun- tries. Most significant is the current growing debate in India over the adoption of a national resettlement policy and of related state policies (see Fernandes and Paranjpye 1997). But we should not overstate the achievements. While social science has helped to articulate resettlement policies in several countries, policy adoption has yet to occur routinely everywhere. Political and bureau- cratic obstacles continue to impede the incorporation of social science research into policy and law. In scores of developing countries, resettle- ment policies and laws do not exist and are far from being adopted-or even contemplated (Cernea 1996). Measurements and Quantification Much of the past "softness" in social science fieldwork has been replaced by statistics building. More precise measurements of displacement and reloca- tion processes and their empirically observable impacts have succeeded in estimating the magnitudes of national and global development-induced displacements, producing a comprehensive image of the process. Through collecting, verifying, and combining a vast body of data, our group at the World Bank has generated the most significant world- wide estimate. Though conservative, it is nothing less than stunning: we found that about 10 million people annually enter the cycle of forced dis- placement and relocation just in the sectors of dam construction and ur- ban and transportation development (World Bank 1994). That means that in the past decade about 100 million people have been displaced. This magnitude is comparable to the refugee crisis, long recognized as a major international problem. Development-caused displace- ments-which appear piecemeal and to total far less than the number of war and natural disaster refugees (estimated by UNHCR at 15 to 20 mil- lion worldwide)-turn out to involve more people than all the world's refugee flows. Of course, refugees and development displacees are not "numbers" that compete with each other: they are global parallel dra- mas, sometimes intertwined. But the quantifications are part of the "so- cial construction" of forced resettlement as a worldwide issue, and the comparison increases public attention to resettlement issues. We must also note that the estimate of 100 million people displaced over 10 years does not include displacements from forests and reserve parks, mining and thermal power plants, and other comparable situa- tions. Reliable aggregate data for all sectors worldwide are needed to capture more fully the magnitude of the process. WHY ECONOMIC ANALYSIS Is ESSENTIAL 35 New Theory and Models The Scudder-Colson framework has informed and inspired the field- work of numerous resettlement researchers since its publication in the early 1980s (Scudder and Colson 1982, Scudder 1985). This diachronic model identifies four stages of a successful resettlement operation-re- cruitment, transition, potential development, and incorporation-with each stage tracking resettlers' behavioral response patterns to the dis- ruptions and stress caused by relocation. When resettlements are not successful, the stage framework does not apply. In the 1990s, theoretical work on resettlement evolved further, and the impoverishment risks and livelihood reconstruction model was for- mulated (Cernea 1990a, 1991,1995,1996,1997, and 1998b). This model is synchronic, yet it also captures the process of moving from displace- ment to relocation and recovery. It conceptualizes the economic and social content of both displacement and reestablishment processes. In this model, the behavior of displaced populations is explained as a response to a range of potential impoverishment risks (or realities)-economic, cultural, and social-and it incorporates the risks of decapitalization and community unraveling. This theoretical framework can be used as a tool for predicting, diagnosis and planning, problem solving, and research on resettlement. Most important, the model suggests action strategies and focuses on the heart of the matter in resettlement: preventing pau- perization and reconstructing livelihoods. This conceptual model has been widely discussed and adopted as a guide for preparing resettlement and recovery components in projects. It is currently at work in numerous field studies and in strategies to counter impoverishment in India, Vietnam, the Philippines, and China (see Agnihotri 1996; Basu 1996; Downing 1996; Joseph 1998; Pandey 1998; Mathur and Marsden 1998; Mohapatra, Ota, and Mohanty 1998; Ota 1996; Thangaraj 1996; Spiegel 1997; Nayak 1999; see also Mahapatra, in this volume). Some researchers have also explored its application to the study of refugee populations (Kibreab 1999). Resistance to Displacement Resistance to development has heightened in many countries, and op- position led by nongovernmental organizations (NGOs) has become more politically oriented during the past decade. But it was not until this decade that anthropological research "discovered" this resistance as a big theme and started to focus its analytical lens upon its structure, growth, composition, and impacts (Oliver-Smith 1991b, 1996; Good 1992; Gray 1996; Kothari 1995; Posey 1996; Dhagamwar, Thukral, and 36 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Singh 1995). Such opposition has been encouraged by an expanding and powerful international critique of dam building, solidly documented in studies generated by environmental NGOs (McCully 1996, Udall 1995). This expanding segment of the resettlement literature is increasingly in- fluential and has contributed directly to the establishment of the World Commission of Dams in 1998, among other outcomes. The several characteristics of knowledge progress I have defined here summarize the current state of the art in social research on resettle- ment. The anthropological and sociological communities can be proud to have collectively expanded in depth and breadth, "intensive" and "extensive" knowledge expansion; to have shifted from academic anal- ysis alone to operational research on a vast scale, from description to prediction and prescription, from deploring ill effects to also charting ways to overcome them; to have moved up from case micro- ethnographies to macro policy formulation; from narrating discrete re- settlement instances to measuring and quantifying aggregates and worldwide trends; to have renewed its theory with poverty-center frameworks and analysis; and to have expanded its topics and perspec- tive by including the study of resettlers' resistance to displacement, so- cial opposition, and NGO-led movements. The "on the ground" positive effects are embodied in many tangible improvements accom- plished in resettlement programs and policies worldwide. Acknowledgements From my many economist colleagues, I drew considerable wisdom and realism about the economic anatomy of development. From my own ex- perience, I realized that the highest benefits are obtained when sociolo- gists and economists reinforce each other in linking the economic and social analyses of development interventions. Enriched with these lessons, I want to thank specifically those econ- omists who, in preparing studies for this volume or in discussing reset- tlement issues throughout the years, as well as in other circumstances of my own development work, have contributed their advice, guidance, suggestions, or criticisms. Many of their comments have been taken di- rectly into account in articulating the argument of this chapter. My debt is substantial to Irma Adelman, Leif Christoffersen, Herman Daly, John Page, Michel Petit, Robert Picciotto, Joanne Salop, G. Edward Schuh, Andrew Steer, Paul Streeten, Herman van der Tak, Vijay Vyas, Montague Yudelman, and, in fact, to a much larger number of economist col- leagues than I can name here. WHY ECONOMIC ANALYSIS Is ESSENTIAL 37 Suggestions and feedback from many resettlement specialists, particularly my late friend David Butcher, Ashraf Ghani, Scott Guggenheim, William Partridge, Ted Scudder, and, not in little mea- sure, Ruth Cernea, have strengthened my argument for the need to de- velop not only social but also economic knowledge about resettlement. My thanks go to all of them. Part of this paper was included in the Honorary Elizabeth Colson Lecture that I was invited to present at the University of Oxford, RSP, in May 1998. I thank also those whose questions and comments on that lecture helped to sharpen my argument. Notes 1. This point was also developed and documented by Robert Picciotto with OED-generated evidence in a lecture given in June 1998 on the findings of OED's 1998 study on resettlement experiences. 2. This economic book was published three years after the famous Morse re- port (Morse, Berger, Brody, and Gamble 1992) was made public. That independ- ent report severely criticized the socioeconomics of the Narmada Sardar Sarovar project, particularly on resettlement issues. Yet the book by Alagh and others and their economic reappraisal didn't give any response to the critique of Sardar Sarovar's economics by the Morse report. 3. Most governments use the might of the law and their institutional instru- ments to forcibly displace people, but they do not sufficiently use their institu- tional capacities to facilitate "land for land" alternatives. They rely instead on cash compensation. 4. See Guidelines for the Economic and Financial analysis of Project Compo- nents Addressing Involuntary Resettlement. Annex 2 in Cernea, Michael, Involun- tary Resettlement in Development Projects: Policy Guidelines in World Bank--Financed Projects, World Bank Technical Paper No. 80,1988, Washington, D.C. 5. The only significant country exception is China, where internal guidelines require planners to do a meticulous inventory of all displacement costs and an equally meticulous projection of the economic components of livelihood recovery. 6. The authors of that study recommended, among other measures, the "rein- statement of a Central Projects Unit (in the World Bank) to both provide assis- tance to project analysts and exert a degree of quality control" in the interest of the Bank's "development impact and the quality of its portfolio" (Devarajan, Squire, and Suthiwart-Narueput 1995, p. 27). 7. We have moved since to a more encompassing concept in the World Bank, that of "resettlement specialist," who is not always and necessarily an anthropol- ogist; in some cases she or he is an economist or technical expert. 38 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT 8. We also know that sociocultural variables have been frequently overlooked in economic analyses in many other areas as well. Since the early 1980s an in- creasing number of anthropologists and sociologists have argued that tradi- tional economic analysis misrepresents reality by failing to acknowledge essential social factors (Swedberg 1993, Granovetter 1988, Smelser and Swedberg 1994). Such concerns have propelled exciting developments in the field of economic sociology, illuminating essential economic variables and pro- cesses through social analysis. 9. The Bank's resettlement policy describes the advisory role of the Legal De- partment with respect to policy and legal issues arising from the involuntary re- settlement of people. It requires that the Bank's legal experts review the domestic laws and regulations of the borrowing country and the legal aspects of project resettlement plans. Further, the participation of the Legal Department is required throughout the project cycle. Along with these policy provisions, the Legal Department has defined the lawyers' responsibilities and the kind of ad- vice and inputs they must provide, as their professional obligation, during the identification, appraisal, negotiation, and implementation of projects involving resettlement operations. The guidelines for lawyers refer to such important as- pects as defining the project concept, land acquisition, valuation and compensa- tion for assets, the rights of the project-affected people, eligibility criteria; provision of rehabilitation measures; monitoring and impact evaluation ar- rangements, and so forth. The guidelines also define what must be included, as a minimum, in the legal agreements between the Bank and the loan- or credit- receiving country, so as to establish the obligations of the borrower and the im- plementing agency as a legal contract between the Bank and the borrowing country (see LEGOP 1996, Sherif Omar Hasan "OD 4.30-Involuntary Resettle- ment: Guidelines for Lawyers," prepared by Carlos Escudero). 10. The bibliographic references provided for these domains, and for the bur- geoning research in various countries, are intended as examples rather than as a comprehensive listing of all relevant publications. Certainly, an updated inter- national bibliography, to following up on Guggenheim (1994), would be a useful tool for further comparative research and analysis. 11. Comparable examples can be provided from the work of other practicing social scientists in various national and international agencies or research cen- ters, such as the Indian Social Institute, New Delhi; the Institute of Development Anthropology, Binghamton, New York; the Refugee Studies Programme of the University of Oxford, England; the National Research Center on Resettlement in China; and others. References The word "processed" describes informally reproduced works that may not be commonly available through libraries. WHY ECONOMIC ANALYSIS Is ESSENTIAL 39 Adams, W. M. 1991. Wasting the Rain: Rivers, People, and Planning in Africa. Lon- don: Earthscan Publications. ADB (Asian Development Bank). 1995. Indonesia: Operational Guidelinesfor Reset- tlement Management in Road Projects: A Handbook. Manila: ADB and Govern- ment of Indonesia, Directorate General of Highways. .1998a. Handbook on Resettlement: A Guide to Good Practice. Manila: ADB, Of- fice of Environment and Social Development, processed. Prepared by M. Zaman. _ . 1998b. Manitalfor Land Acquisition and Resettlement Management in Water Resoutrces Development Projects. (draft) Manila: ADB and Government of Indo- nesia, Directorate General of Water Resources Development. Prepared by M. Zaman and Sheladia Associates, Inc. Agnihotri, Anita. 1996. "The Orissa Resettlement and Rehabilitation of Pro- ject-Affected Persons Policy, 1994: An Analysis of its Robustness with Refer- ence to the Impoverishment Risk Model." In A. B. Ota and A. Agnihotri, eds., Involutntary Displacement in Dam Projects. New Delhi: Prachi Prakashan. Alagh, Y. K., R. D. Desai, G. S. Guha, and S. P. Kashyap. 1995. Economic Dimen- sions of the Sardar Sarovar Project. Delhi: Har-Anand Publications. Areeparampil, Mathew. 1996. "Displacement due to Mining in Kharkhand." Economic and Political Weekly, 31:24. Baboo, Balgovind. 1992. Technology and Social Transformation: The Case of the Hirakuid Multi-Purpose Dam Project in Orissa. Delhi: Concept Publishing. Baird, Mark. 1996. "Economists and Economic Work in the World Bank," DEC Notes, Policy directions, nr. 24. World Bank, Washington, D.C. Processed. Bardhan, Pranab, ed. 1989. Conversations between Economists and Anthropologists: Methodological Issuies in Measutring Economic Change in Rulral India. New Delhi: Oxford University Press. Barth, Fredrik, and T. R. Williams. 1994. "Initial Resettlement Planning and Ac- tivity (1992-1994) in a Large Scale Hydropower Process: The Ertan Dam in Southwest China." World Bank, Washington, D.C. Processed. Bartolome, Miguel, and Alicia Barabas. 1973. "Hydraulic Development and Ethnocide: The Mazatec and Chinantec People of Oaxaca, Mexico," Copenhagen: IWGIA. .1 1990. La presa Cerro de Oro y el ingeniero El Gran Dios, Mexico DF: Instituto Nacional Indigenista. Basu, Malika. 1996. "Basic Needs Approach in Displacement Situations." Main- stream, 24, July-August. Baviskar, Anita. 1995. In the Belly of the Beast. New Delhi: Oxford University Press. Brandon, Katrina, and Michael Wells. 1992. People and Parks: Linking Protected Area Management with Local Commutnities. Washington, D.C.: World Bank. Brokensha, David, and Thayer Scudder. 1968. "Resettlement." In N. Rubin and W. M. Warren, eds. Dams in Africa: An Interdisciplinary Study of Man-Made Lakes in Africa. London: Frank Cass. 40 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Butcher, David. 1971. An Operational Manual for Resettlement: A Systematic Ap- proach to the Resettlement Problem Created by Man-Made Lakes, with Special Rele- vance for West Africa. Rome: FAO. . 1979. Personal communication. . 1990. "Review of the Treatment of Environmental Aspects of Bank En- ergy Projects." PRE Working Paper. Industry and Energy Department, World Bank, Washington, D.C. Processed. Cernea, Michael M. 1986. "Involuntary Resettlement in Bank-Assisted Projects: A Review of the Application of Bank Policies and Procedures in FY97-85 Pro- jects." World Bank, Agriculture and Rural Development Department, Wash- ington, D.C. Processed. . 1988. Involutntary Resettlement in Development Projects: Policy Guidelines in W'Vorld Bank-Financed Projects. World Bank Technical Paper 80. Washington, D.C.: World Bank. .1989. "Metropolitan Development and Compulsory Population Relocation: Policy Issues and Project Experiences." Regional Development Dialogue, 10(4). . 1990a. Poverty Risks from Popuilation Displacement in Water Resource Pro- jects. DDP 355. Cambridge, Mass.: Harvard University, HIID. . 1991. "Involuntary Resettlement: Social Research, Policy, and Plan- ning." In Michael M. Cernea, ed., Putting People First: Sociological Variables in Development. 2nd ed. New York: Oxford University Press. . 1993. "Social Science Research and the Crafting of Policy on Population Resettlement." Knowledge and Policy: The International Journal of Knowledge Transfer and Utilization. Fall/winter 1993/94, 6:3-4, 176-200. . 1995. "Understanding and Preventing Impoverishment: The State of Knowledge in Resettlement." Journal of Refiugee Studies. December. . 1996a. "Public Policy Responses to Development-Induced Population Displacement." Economic and Political Weekly. (India) 31 (June 15): 1515-23. .1996b. "Bridging the Research Divide: Studying Refugees and Develop- ment Oustees." In Tim Allen, ed., In Search of Cool Ground: Displacement and Homecoming in Northeast Africa. London: James Currey. . 1997. "The Risks and Reconstruction Model for Resettling Displaced Populations." World Development 25 (October): 1569-87. .1998a. "La sociologie des deplacements forces: un modele theorique." In V. Laissailly-Jacobs, ed., Communauites deracinees dans les pays du Slid. Paris: Autrepart, ORSTOM. .1998b. "Economics and the Private Sector: Open Issues in Resettlement Re- search." The 1998 Colson Lecture, University of Oxford, RSP, May 13. Processed. Forthcoming. "A Theoretical Model for Population Displacement and Resettlement." In Michael M. Cernea and Chris McDowell, eds., Recon- structing Livelihoods: Theory and Practice-Resettlers' and Refuigees' Experiences. Washington, D.C.: World Bank. WHY ECONOMIC ANALYSIS Is ESSENTIAL 41 Chambers, Robert, ed. 1970. The Volta Resettlement Experience. London: Pall Mall Press. Clay, Jason W. 1988. "Villagization in Ethiopia." In J. Clay, S. Steingraber, and P. Niggli, eds., The Spoils of Famine. Cambridge, Mass.: Cultural Survival. Colson, Elizabeth. 1971. The Social Consequences of Resettlement: The Impact of the Kariba Resettlement 071 the Guwembe Tonga. Manchester, U.K.: Manchester Uni- versity Press. Cook, Cynthia C., ed., 1994. Involuntary Resettlement in Africa. Washington, D.C.: World Bank. Cook, Cynthia C., and Kristine Ivarsdotter. 1994. "Development and Displace- ment: Resettlement Review for the Africa Region." World Bank, Africa Tech- nical Department, Washington, D.C. Processed. Conac, Franqoise, ed. 1995. Barrages internationatixetcooperation, Paris: Karthala. Cornia, Giovanni Andrea, Richard Jolly, and Frances Stewart, eds. 1987. Adjlust- ment with a Hutman Face. New York: Oxford University Press. Davidson, Forbes, and others. 1993. Relocation and Resettlement Manuial: A Gutide to Managing and Planning Relocation. Rotterdam: IHS. de Santos, Leinad, and Lucia M. M. de Andrade, eds. 1990. Hydroelectric Dams o07 Brazil's Xingu RiverandIndigenouisPeople. London: Cambridge UniversityPress. Devarajan, S., L. Squire, and S. Suthiwart. 1995. "Reviving Project Appraisal at the World Bank." World Bank, Policy Research Department, Washington, D.C. Processed. de Wet, C. 1993. "A Spatial Analysis of Involuntary Community Relocation: A South African Case Study." In M. Cernea and S. Guggenheim, eds., Antlhropo- logical Approaches to Resettlement: Policy, Practice, and Theory. Boulder, Colo.: Westview Press. Dhagamwar, Vasuda, E. Ganguly Thukral, and M. Singh.1995. "The Sardar Sarovar Project: A Study in Sustainable Development." In W. Fisher, ed., To- ward Sustainable Development: Struggling over India's Narmada River. Armonk, N.Y.: M. E. Sharpe. Diara, T., D. Koenig, Y. F. Kone, and M. F. Maiga. 1995. "Reinstallation et devel- opment dans la zone du barrage de Manantali." Bamako. Processed. Dieci, Paolo, and Claudia Viezolli, eds. 1992. Resettlement and Rutral Development in Ethiopia. Milano: Franco Angeli. Douglas, Mary, and Aaron Wildawski. 1982. Risk and Culture: The Selections of Technological and Environmental Dangers. Berkeley, CA: University of Califor- nia Press. Downing, Theodore E. 1994. "Social Geometrics: A Theory of Social Displace- ment in Resettlement." Paper presented at the International Congress of the Americanists. Stockholm/Uppsala, Sweden. Processed. . 1996. "Mitigating Social Impoverishment When People Are Involun- tarily Displaced." In C. McDowell, ed., Understanding Impoverishment. Oxford and Providence: Berghahn. 42 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT _____. 1998. "Protecting the Widow's Mite: Designing Local Environmental Risk Insurance." Paper presented at the Annual SRAA Meetings, San Jose, Puerto Rico. Processed. Dreze, J., M. Samson, and S. Singh, eds. 1997. The Dam and the Nation: Displace- ment and Resettlement in the Narmada Valley. Delhi: Oxford University Press. Escudero, Carlos R. 1988. "Involuntary Resettlement in Bank-Assisted Projects: An Introduction to Legal Issues." World Bank, Legal Department, Washing- ton, D.C. Processed. Fabre, Pierre. 1997. Financial and Economic Analysis of Development Projects: Man- tal. Luxembourg: European Commission, Methods and Instruments for Pro- ject Cycle Management. Feeney, Patricia. 1995. Displacement and the Rights of Women. Oxford: Oxfam. Femandes, Walter. Forthcoming 1999. "From Marginalization to Sharing the Project Benefits." In Michael M. Cernea and Chris McDowell, eds., Recon- structing Livelihoods. Washington, D.C.: World Bank. Fernandes, Walter, and Mohammed Asif. 1997. Development-Induced Displace- ment and Rehabilitation in Orissa 1951 to 1995: A Database on its Extent and Na- ture. New Delhi: ISI. Femandes, Walter, J. C. Das, and Sam Rao. 1989. "Displacement and Rehabilita- tion: An Estimate of Extent and Prospects." In W. Femandes and E. Ganguly Thukral, eds., Development, Displacement and Rehabilitation. New Delhi: ISI. Fernandes, Walter, and Vijay Paranjpye, eds. 1997. Rehabilitation Policy and Law in India: A Right to Livelihood. Delhi: ISI and Pune: Econet. Femandes, Walter, and Enakshi Ganguly Thukral, eds. 1989. Development, Dis- placement and Rehabilitation. New Delhi: ISI. Fisher, William F., ed. 1995. Towards Sutstainable Development? Struggling Over In- dia's Narmada River. Armonk, N.Y.: M. E. Sharpe. Gans, Herbert J. 1968. People and Plans: Essays on Urban Problems and Solutions. New York: Basic Books. Gibson, Daniel R. 1993. "The Politics of Involuntary Resettlement: World Bank- Supported Projects in Asia." Ph.D. dissertation, Department of Political Sci- ence, Duke University. Processed. Gill, Maninder S. 1995. "Resettlement and Rehabilitation in Maharashtra for the Sardar Sarovar Project." In W. F. Fisher, ed. Towards Sustainable Development? Struggling Over India's Narmada River. Armonk, N.Y.: M. E. Sharpe. Gill, M., D. Gibson, A. Fields, and E. Schaengold. 1998. "Resettlement Guide- book." World Bank, Washington, D.C. Processed. Goyal, Sangeeta. 1996. "Economic Perspectives on Resettlement and Rehabilita- tion." Economic and Political Weekly, June 15. Gray, Andrew. 1996. "Indigenous Resistance to Involuntary Relocation." In Christopher McDowell, ed., Understanding Impoverishment: The Consequences of Development-Induced Displacement. Providence: Berghahn. WHY ECONOMiC ANALYSIS Is ESSENTIAL 43 Guggenheim, Scott E. 1989. "Development and the Dynamics of Displacement." In Al Fernandes, ed., Workshop on Rehabilitation of Persons Displaced by Develop- ment Projects. Bangalore: Myrada. .1 994. "Involuntary Resettlement: An Annotated Reference Bibliography for Development Research." Environment Working Paper 64, World Bank, Washington, D.C. Processed. Hall, Anthony. 1987. "Agrarian Crisis in Brazilian Amazonia: The Grande Carajas Programme." Journal of Development Stuidies. 23(4, July): 522-552. .1991. "O Programa Grande Caragas genese e evolucao." In Jean Hebette, ed., 0 Cerco Esta be Fechand. Vozes FASE NAEA: Petropolis-Rio-Belem. - 1992. "From Victims to Victors: NGOs and the Politics of Employment at Itaparica." In David Hulme and Mike Edwards, eds., Making A Difference. London: Earthscan. Hansen, Art. 1990. "Long-term Consequences of Two African Refugee Settle- ment Strategies." Paper presented at the meetings of the Society for Applied Anthropology. York, U.K. Processed. Huang, Yao Fei. 1984. "On the Reform of Resettlement Policies in China." Chi- nese Journal of Water Conservation 6. Irwin, Timothy and associates. 1997. Dealing with Putblic Risk in Private Infrastruc- tiore, Washington, D.C.: World Bank. Jellinek, Lea. 1991. The Wheel of Fortune. TheHistoryofa Poor Community in Jakarta. London: Allen and Unwin. Joseph, J. 1998. "Evolving a Retrofit Economic Rehabilitation Policy Model Using Impoverishment Risks Analysis-Experience of Maharashtra Composite Irri- gation Project III." In H. M. Mathur and D. Marsden, eds., Development Projects and Impoverishment Risks in Resettlement. New Delhi: Sage Publications. Joshi, Vidyut. 1987. Stubmerging Villages: Problems and Prospects. New Delhi: Ajanta Publications. _ - 1991. Rehabilitation-A Promise to Keep: The Case of the Sardar Sarovar Pro- ject. Ahmedabad: Tax Publication. Joy, J. L. 1968. "What an Economist Wants to Know about Dams?" In Rubin Neville and William M. Warren, eds. Dams in Africa. An Interdisciplinary Study of Man-Made Lakes In Africa. New York: Augustus Kelley Publishers. Kibreab, Gaim. Forthcoming 1999. "Common Property Resources and Involun- tary Resettlement." In Michael M. Cernea and Chris McDowell, eds., Recon- structing Livelihoods. Washington, D.C.: World Bank.. Koenig, Dolores, and Tieman Diara. Forthcoming 1999. "The Effects of Resettle- ment on Assets and Common Property Resources." In Michael M. Cernea and Chris McDowell, eds., Reconstructing Livelihoods. Washington, D.C.: World Bank. Kothari, Smithu. 1995. "Damning the Narmada and the Politics of Develop- ment." In William F. Fisher, ed., Towards Suistainable Development? Struggling Over India's Narmada River. Armonk, N.Y.: M. E. Sharpe. 44 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Kudat, Ayse, Mumtaz Bayram, and T. Hazar, eds. 1996. Rutral Resettlement in The- ory and Practice (in Turkish). Ankara: Anatolian Development Foundation. .1997. Urban Resettlement (in Turkish). Ankara: Turkish Rural and Urban Development Foundation. Lane, Charles. 1996. Pasttres Lost. Nairobi, Kenya: Initiative Publishers. Laissailly-Jacobs, V., ed. 1998 Commuinaut&es deracinees dans les pays du Sotd. Paris: Autrepart, OSTROM. LEGOP (Legal Department - Operations). 1996. OD 4.30- Involuintary Resettle- ment: Guidelinesfor Lawyers. Prepared by Carlos Escudero. Lipton, Michael. 1992. "Economics and Anthropology: Grounding Models in Relationships." World Development 20 (10): 1541-46. Little, I. M. D., and J. A. Mirrlees. 1974. Project Appraisal and Planning for De- veloping Countries. New York: Basic Books. Londero, Elio. 1996. Benefits and Beneficiaries: An Introduiction to Estimating Distri- buttional Effects in Cost-Benefit Analysis. IDB. Washington, D.C.: Johns Hopkins University Press. Mahapatra, L. K. 1994. Tribal Development in India. Myth and Reality. New Delhi: Vikas Publishing House. Mairal, Gaspar and J. A. Bergua. 1996. "From Economicism to Culturalism: The Social Impact of Dam Projects in the River Esera." Paper presented at the Bien- nial Conference of the European Association of Social Anthropologists, Barce- lona. Processed. Mathur, Hari M., in collaboration with M. Cemea, eds. 1994. Development, Displace- ment and Resettlement: Focuts on Asian Experiences. Delhi: Vikas Publishing House. . 1998. "Loss of Access to Basic Public Services." In H. M. Mathur and David Marsden, eds., Development Projects and Impoverishment Risks. New Delhi: Oxford University Press. Mathur, Hari Mohan, and David Marsden, eds. 1998. Development Projects and Impoverishment Risks. New Delhi: Oxford University Press. Mburugu, E. K. 1993. "Dislocation of Settled Communities in the Development Process: The Case of Kiambere Hydroelectric Project." In Cynthia C. Cook, ed., Involuntary Resettlement in Africa. World Bank Technical Paper 227, Wash- ington, D.C.: World Bank. McCully, Patrick. 1996. Silenced Rivers. London: Zed Books. McDowell, Christopher, ed. 1997. Understanding Impoverishment: The Consequences of Development-Induced Displacement. Providence: Berghahn. McMillan, Della E., 1995. Sahel Visions: Planned Settlement and River Blindness Control in Burkina Faso. Tucson: University of Arizona Press. Meikle, Sheilah, and Youxuan Zhu. Forthcoming 1999. "Employment for Displacees in the Socialist Market Economy of China." In Michael M. Cernea and Chris McDowell, eds., Reconstructing Livelihoods. Washington, D.C.: World Bank. WHY ECONOMIC ANALYSIS Is ESSENTIAL 45 Mejia, Maria Clara. Forthcoming 1999. "Involuntary Resettlement and Eco- nomic Recovery: The Case of the Brickmakers Displaced by the Yacyreta Hy- droelectric Project Argentina-Paraguay." In Michael M. Cernea and Chris McDowell, eds., Reconstruicturing Livelihzoods. Washington, D.C.: World Bank. Merton, Robert K. 1957. Social Theory and Social Structure. New York: Free Press. Mesa, Juan Fernando. 1969. "Emergencia Social de El Penol." Medellin. Pro- cessed. Mohapatra, P., A. B. Ota, and R. N. Mohanty, eds. 1998. Development-Induced Dis- placement and Rehabilitation. New Delhi/Bhubaneswar: Prachi Prakashan. Morse, Bradford, T. Berger, H. Brody, and T. Gamble. 1992. Sardar Sarovar: Report of the Independent Review. Ottawa: Resource Futures International, Inc. Nayak, Ranjit. Forthcoming. "Risks Associated with Landlessness: Exploration towards Socially Friendly Resettlement." In Michael M. Cernea and Chris McDowell, eds., Reconstructing Livelihoods. Washington, D.C.: World Bank. NRCR (National Research Center for Resettlement). 1995. Resettlement and Reha- bilitation. Nanjing, China: Hohai University Press. . 1996. Papers on Resettlement and Development. Nanjing, China. OECD. 1992. Gutidelinesfor Aid Agencies on Involuntary Displacement and Resettle- ment in Development Projects. Paris: DAC. Oliver-Smith, Anthony. 1991a. "Successes and Failures in Post-Disaster Reset- tlement." Disaster, 15(1, March): 12-23. - 1991b. "Involuntary Resettlement, Resistance and Political Empower- ment." Journal of Refugee Studies 4 (2): 132-149. _ . 1996. "Fighting for a Place: The Policy Implications of Resistence to Development-Induced Resettlement." In Chris McDowell, ed., Understanding Impoverishment. Providence: Berghahn. Ota, A. B. 1996. "Countering the Impoverishment Risk: The Case of the Rengali Dam Project." In A. B. Ota and A. Agnihotri, eds., Involuntitary Displacement in Dam Projects. New Delhi: Prachi Prakashan. Oyedipe, F. P. A. 1987. "The Relative Success of the Kainji Resettlement Scheme as Compared with that of Volta." Researchfor Development (Nigeria) 3:41-54. Pandey, Balaji. 1998. Displaced Development: Impact of Open-Cast Mining on1 Womeni. Delhi: Friederich, Ebert, and Stiftung. Pandey, Balaji, and others. 1998. Depriving the Underprivileged for Development. Bhubaneswar, Orissa: Institute for Socio-Economic Development. Pankhurst, Alula. 1992. Resettlement and Famine in Ethiopia: The Villagers' Experi- ence. Manchester: Manchester University Press. Paranjpye, Vijay. 1988. Evaluating The Tehri Dam: An Extended Cost Benefit Ap- praisal. New Delhi: Indian National Trust for Art and Cultural Heritage. .1990. High Dams on the Narmnada: A Holistic Analysis of the River Valley Pro- jects. New Delhi: Indian National Trust for Art and Cultural Heritage. 46 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Parasuranam, S., and N. Purendra Prasad. 1997. "Differential Distribution of So- cial Cost." Economic and Political Weekly (August): 2218-22. Partridge, William. 1989. "Population Displacement Due to Development Pro- jects." Journal of Refugee Stuidies 2 (3). _.____ 1993. "Successful Involuntary Resettlement: Lessons from the Costa Ri- can Arenal Hydroelectric Project." In Michael M. Cernea and Scott E. Guggenheim, eds., Anthropological Approaches to Resettlement: Policy, Practice, Theory. Boulder, Colo.: Westview Press. Partridge, William L., A. B. Brown, and J. B. Nugent. 1982. "The Papaloapan Dam and Resettlement Project: Human Ecology and Health Impacts." In Art Hansen and Anthony Oliver-Smith, eds., Involutntary Migration and Resettle- ment, Boulder, Colo.: Westview Press. Paul, James C. N. 1988. "International Development Agencies, Human Rights and Human Development Projects." Denver loutrnal of International Law and Policy 17 (1): 67. Posey, Darrell A. 1996. "The Kayap6 Indian Protests against Amazonian Dams: Successes, Alliances, and Unending Battles." In C. McDowell, ed., Understand- ing Impoverishment. Providence: Berghahn. Qureshi, Moeen. 1989. "Resettlement, Development, and the World Bank." In The Bank's World (May): 12-13. Ramanathan, Usha. 1995. "Displacement and Rehabilitation: Towards a Na- tional Policy." Lokayan Bulletin, 11(5): 41-56. .1996. "Displacement and the Law." Economicand Political Weekly 31 (24): 1486-92. Ray, Anandrup. 1984. Cost-Benefit Analysis: Issues and Methodologies. Baltimore: Johns Hopkins University Press. Ray, Rita. 1994a. Underground Drama: The Social Ecology of Two Chromite Mines in Orissa. Delhi: Ajanta Publications. _____. 1994b. "Demographic and Socio-Economic Study of the Proposed Naraj Thermal Power Project." Utkal University, Bhubaneswar, India. Pro- cessed. -_ 1995. "Evaluation of Rehabilitation and Resettlement of Talcher Super Thermal Power Project." A study conducted for NTPC, New Delhi. Processed. Reddi, I. U. B. 1992. Displacement and Rehabilitation. New Delhi: Mittal Publications. . 1994. Induistrial Development and Problems of the Uprooted. New Delhi: Rawat Publications. Rew, Alan W., and P. A. Driver. 1986. "Evaluation of the Victoria Dam Project in Sri Lanka." Vol. III. Initial Evaluation of the Social and Environmental Im- pacts of the Victoria Dam Project, Annexes J and K. Processed. Riberio, Gustavo Lins. 1994. Transnational Capitalism and Hydropolitics in Argen- tina: The Yacyretd High Dam. Gainesville: University Press of Florida. WHY ECONOMIC ANALYSIS Is ESSENTIAL 47 Roder, W. 1994. Human Adjustment to Kainji Reservoir in Nigeria: An Assessment of the Economics and Environmental Consequences of a Major Man-Made Lake in Af- rica. Lanham: University Press of America. Rosa, Eugene A. 1998. "Metatheoretical Foundations for Post-Normal Risk." Journal of Risk Research 1 (1). Roy-Burman, B. K. 1961. Social Processes in the Industrialization of Rourkhela. New Delhi: Census of India. Ruttan, Vernon W. 1988. "Cultural Endowments and Economic Development: What Can We Learn from Anthropology?" Economic Development and Cultural Change 36 (3, April): S248-S271. Salem-Murdock, Muneera. 1989. Arabs and Nubians in New Halfa. Salt Lake City: University of Utah Press. Schuh, G. Edward. 1993. "Involuntary Resettlement, Human Capital, and Eco- nomic Development" In Michael M. Cernea. and Scott Guggenheim, eds. An- thropological Approaches to Resettlement: Policy, Practice, Theory. Boulder, Colo.: Westview Press. Scott, R. Parry. 1992. "Dams, Forced Resettlement, and the Transformation of the Peasant Economy in the Sao Francisco River Valley, Brazil." Paper presented at the XVII International Congress of the Latin American Studies Association (LASA), Los Angeles. Processed. Scudder, Thayer. 1962. The Ecology of the Gwembe Tonga. Manchester, U.K.: Man- chester University Press. _ - 1973. "The Human Ecology of Big Projects: River Basin Development and Resettlement." In B. Siegel, ed., Annual Review of Anthropology. Palo Alto: Annual Review. - 1985. "A Sociological Framework for the Analysis of New Land Settle- ments." In Michael M. Cernea, ed. Putting People First: Sociological Variables in Rural Development. New York: Oxford University Press. - 1993. "Development-Induced Relocation and Refugee Studies: 37 Years of Change and Continuity among Zambia's Gwembe Tonga." Journal of Refu- gee Stutdies 6 (2): 123-52. - 1996. "Development-Induced Impoverishment, Resistance, and River- Basin Development." In Chris McDowell, ed. Understanding Impoverishment: The Consequences of Development-Induced Displacement. Providence: Berghahn. Scudder, Thayer, and Elizabeth Colson. 1982. "From Welfare to Development: A Conceptual Framework for Analysis of Dislocated People." In Art Hansen and Anthony Oliver-Smith, eds., Involuntary Migration and Resettlement. Boul- der, Colo.: Westview Press. Serra, Maria Teresa Fernandes. 1993. "Resettlement Planning in the Brazilian Power Sector: Recent Changes in Approach." In Michael M. Cernea and Scott E. Guggenheim, eds., Anthropological Approaches to Resettlement: Policy, Prac- tice, Theory. Boulder, Coto.: Westview Press. 48 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Shi, Guoqing. 1996a. "Comprehensive Evaluation on Resettlers' Production and Living Level in Gaojiajing of Dongfeng Hydroelectric Stations." In NRCR, Papers on Resettlement and Development. Nanjing. . 1996b. "The Comprehensive Evaluation Method and Its Application of Production and Living Standard for Rural Resettlers in Reservoir Area." In NRCR, Papers on Resettlement and Development. Nanjing. Shi, Guoqing, and Hu Weison. 1994. Comprehensive Evaluation and Monitoring of Displaced Persons: Standards of Living and Production. Nanjing: NRCR. Shi, Guoqing, Huoping Xun, and Wenxue Yu. 1996. "Research on Economic Evaluation of Reservoir Protection." In NRCR, Papers on Resettlement and De- velopment. Nanjing. Shihata, Ibrahim F. I., 1988. "The World Bank and Human Rights: An Analysis of the Legal Issues and the Record of Achievements." Denver Jouirnal offlnterna- tional Law & Politics 17 (1). . 1991. "Involuntary Resettlement in World Bank-Financed Projects." In Ibrahim Shihata, ed., The World Bank in a Changing World. The Netherlands: Martinus Njhoff Publishers. . 1993. "Legal Aspects of Involuntary Population Resettlement." In Michael M Cernea and Scott E. Guggenheim, eds., Anthropological Approaches to Resettlement: Policy, Practice, Theory. Boulder, Colo.: Westview Press. Singh, Mridula, and Associates. 1992. Displacement by Sardar Sarovarand Theri: A Comparative Study of Two Dams. Delhi: MARG. Sinha, Basawan, and Ramesh Bhatia. 1982. Economic Appraisal of Irrigation Project in India. New Delhi: Agricole Publishing Academy. Smelser, Neil J., and Richard Swedberg, eds. 1994. The Handbook ofEconomic Soci- ology. New Jersey: Princeton University Press. Spiegel, H. 1997. Personal communication. Squire, Lyn, and Herman van der Tak. 1975. Economic Analysis of Projects. Balti- more, MD: Johns Hopkins University Press. Stiglitz, Joseph E. 1998. "Towards a New Paradigm for Development: Strat- egies, Policies, and Processes." Presented as the Prebisch Lecture at UNCTAD, Geneva, October 19, 1998. Processed. Subbarao, Kalanidhi, and others. 1996. Safety Net Programs and Poverty Reduc- tions: Lessonsfrom Cross-Country Experience. Washington, D.C.: World Bank. Swedberg, Richard, ed. 1993. Explorations in Economic Sociology. New York: Rus- sell Sage Foundation. Tamondong-Helin, Susan D. 1996. "State Power as a Medium of Impover- ishment: The Case of Pantabangan Dam Resettlement in the Philippines." In Chris McDowell, ed., Understanding Impoverishment: The Consequences of Development-Induced Displacement. Providence: Berghahn. Thangaraj, Sam. 1996. "Impoverishment Risk Analysis: A Methodological Tool for Participatory Resettlement Planning." In Chris McDowell, ed., Under- WHY ECONOMIC ANALYSIS Is ESSENTIAL 49 standing Displacement: The Conisequiences of Development-Induced Displacement. Providence: Berghahn. . 1998. "Impoverishment Risks in Involuntary Resettlement: An Over- view." In Hari Mohan Mathur and David Marsden, eds., Development Projects andlmpoverishmentRisks in Resettlement. New Delhi: Oxford UniversityPress. Thukral, Enakshi Ganguly, ed. 1992. Big Dams, Displaced People: Rivers of Sorrow, Rivers of Cliange. New Delhi: Sage Publications. Torado, Michael P. 1989. Economic Development in the Third World. Fourth Edi- tion. New York: Longman. Turnbull, Colin M. 1987. The Mouintain People. Touchstone Books. Udall, Lori. 1995. "The International Narmada Campaign: A Case of Sustained Advocacy." In William F. Fisher, ed., Towards Sustainable Development? Strug- gling over India's Narmada River. Armonk, N.Y.: M. E. Sharpe. Usher, Ann Danaiya. 1997. Dams as Aid: A PoliticalAnatomy ofNordic Developmenit Thinking. London: Routledge. Wali, Alaka. 1989. Kilowatts and Crisis: Hydroelectric Powerand Social Dislocation in Eastern Panama. Boulder, Colo.: Westview Press. West, Patrick, and Stephen Brechin, eds. 1991. Resident Peoples and National Parks. Tucson: University of Arizona Press. Wangxiang, Chen. 1993. "Resettlement Associated with Hydro Projects in China." International Water Power & Dam. 45 (2):19-20. World Bank. 1980. Operational Manual Statement 2.33. "Social Issues Associ- ated with Involuntary Resettlement in Bank-Financed Projects." Washington, D.C. Processed. .1990. Operational Directive 4.30: "Involuntary Resettlement." Washing- ton, D.C. Processed. _= -. 1994. "Resettlement and Development: The Bankwide Review of Pro- jects Involving Involuntary Resettlement 1986-1993." (2nd ed. 1996.) Envi- ronment Department, Washington, D.C. Processed. - 1998. "Recent Experience with Involuntary Resettlement. Overview." Operations Evaluation Department, Washington, D.C. Wu, Zougfe, and Shi Guoqing. 1996. "Study on the Planning Theory of Produc- tion Development in Reservoir Resettlement." In NRCR, Papers on Resettle- ment and Development. Nanjing. 2 Methodological Issues in the Economic Analysis for Involuntary Resettlement Operations David W. Pearce Editor's Environmental economics has evolved in recent de- Note cades into a powerful field that has helped to quan- tify the losses caused by environmental degradation and to ground environmental protection policies in rigorous analysis. In this chapter David Pearce, a founder of envi- ronmental economics, discusses the methodological issues in- volved in applying basic economic analysis to involuntary population resettlements-an area that has yet to be solidly re- searched. Resettlement issues, he notes, have been "something of a Cinderella issue" in the economic literature; even the leading man- uals of development economics and project work have "virtually ignored the subject." Pearce's study is revealing in its critique of the lack of economic reflection on resettlement and in its positive meth- odological suggestions for improving that situation. The author draws on recent experiences in using environmental valuation tech- niques to identify and reduce resettlement costs. The author finds that compulsory resettlement does not differ essentially from other project externalities; thus losses should be in- temalized in project budgets, not externalized to those displaced. The first step: accurate valuation of assets and incomes lost, both public and private. Pearce makes several recommendations for im- proving the routine economic analysis of resettlement operations, in- cluding standard full cost minimization in project design, where full 50 METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 51 cost refers to the sum of economnic, social, and environmental costs; trade-off analysis, where resettlement minimization is at the ex- pense of project benefits; and recognition that a $1 loss for project "losers" (displaced people) has a higher social value than $1 gains to project "beneficiaries." In support of the anthropological findings about typically in- adequate compensations, Pearce also unveils the cognitive and methodological fallacies that account for the persistent undercom- pensation of dislocated people in many developing countries. His analysis concludes with a discussion of the income curve of dis- placed families, where he demonstrates that, contrary to wide- spread but erroneous assumptions, asset replacement through compensation will not prevent resettlers from becoming "worse off." To make the "catching up" process possible, new assets must be created through investment. The involuntary resettlement of large numbers of people because of major water resource development projects has been described as "the single most serious counterdevelopmental social consequence of water resource development" (Cemea 1990). What is true of water re- sources projects is also true of many other types of development projects in the urban, energy, and transportation sectors. Self-evidently, if the means of compensation such "losers" receive from major projects is inadequate, significant loss of well-being must ensue simply because of the numbers affected. Because baseline surveys have been inadequate in the past, the asset and income base of the popu- lations affected is not known with certainty. By and large, however, those affected are the poor. Thus, an additional concern about the effects of displacement is based on equity grounds. The well-being lost by those who are displaced by a project, such as a dam, is an example of an externality, or more strictly, of an external cost. Project evaluators must account for this externality and add it to the pro- ject's internal costs (the costs of materials, labor, and so on), otherwise investment and policy choices will be economically inefficient. World Bank guidelines go beyond this accounting requirement and call for the externality to be internalized by requiring that the costs of resettlement be calculated as part of the project's cost and that affected people be compensated. This distinction between accounting for externalities and actual compensation for externalities is important. 52 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Conclusions from Past Resettlement Experiences In 1993-94 the World Bank reviewed its portfolio of projects entailing in- voluntary resettlement and the performance of governments and the Bank itself in those projects (World Bank 1994). Earlier reviews (Butcher 1987; Cemea 1986, 1988, 1990) had already led to substantial modifica- tions of the Bank's initial (1980) policy and procedures, which are codi- fied in the Bank's June 1990 Operational Directive (OD) 4.30 on involuntary resettlement. Nonetheless, many still view the Bank's oper- ational treatment of resettlement in the context of development projects as unsatisfactory. This study, which was initially started as part of the Bank's (1994) review, focuses on the relationship between economic methodology and resettlement, specifically, the Bank's guidelines on the economic and financial analysis of resettlement projects as set out in Cernea (1988, annex 2). Based on the large volume of material examined, my main conclu- sions are: aThat the methodological arguments that underlie some of the re- luctance to elevate the problems of resettlement in project ap- praisal are without foundation and that routine project design based on social cost minimization might help reduce the scale of the resettlement problem. . That full and proper compensation, consistent with the Bank's OD 4.30 and the concept of resettlement as a development oppor- tunity, can be thought of in terms of some recent literature on the economics of sustainable development. This reinforces the require- ment that involuntary resettlement be treated as a development ac- tivity in itself, rather than as a relief or salvage operation. * That additional means of determining full and proper compensa- tion are available to governments, over and above the Bank's em- phasis on valuing private and public assets and income. These methods draw on recent experience with environmental valuation techniques in developing countries; government agencies dealing with resettlement and Bank staff should experiment with them to ensure that underestimation of resettlement costs is minimized. Such underestimation arises often because (a) country planners do not follow the existing guidelines properly (World Bank 1994; see also chapters by Cernea and Eriksen in this volume), and (b) the ex- isting guidelines for economic analysis fail to capture the full social costs of dislocation. These full social costs include the loss of nonpriced environmental and cultural assets, the loss of social METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 53 cohesion, the loss of market access, and psychological damage from dislocation. * That governmental procedures and Bank guidelines need to give more emphasis to the role of nonmarket assets in the "before devel- opment" incomes of displaced people and to the process of asset growth in preproject communities. . That the focus on resettlers should not overshadow the needs of host communities. Some evidence indicates that inadequate host community assistance can result in unsatisfactory resettlement. . That unwarranted displacement of families may also result from the domestic policy context in which projects are formulated. While projects are blamed for the human costs associated with de- velopment, responsibility also lies with the policies in the country in question. More generally, environmental economics and the welfare econom- ics that underlies most Bank development practice at the microeco- nomic level emphatically reinforce the importance sociologists and anthropologists attach to the design of socially acceptable resettlement programs. Many inadequacies of resettlement operations arise from failure to implement the existing guidelines formulated by the Bank. Quite often, government agencies that agree with the Bank's resettlement policy guidelines at project outset fall back on their old routines during imple- mentation and do not carry out the policy agreements made at project inception. In other words, as several reviews indicate (Cernea 1990; Eriksen in this volume; Gutman 1993; World Bank 1993b, 1994), resettle- ment inadequacies derive from bad practice, not from bad policy. In particular, many projects tend to systematically underestimate the costs of resettlement (World Bank 1993b, 1994). We know that cost underestimation tends to be a feature of public projects generally. Reset- tlement projects are unlikely to be an exception. However, better and clearer guidelines could address some of the sources of these underesti- mates. More important, better guidelines might raise the probability of resettlement projects being successful. Ultimately, bad practice is a mat- ter of political will in the countries in question and of the determination of funding agencies, such as the Bank, to enforce covenants and agree- ments. Where that domestic political will is lacking, it is important that it is not given any support from weakness in the guidelines. In terms of the priority given to it, resettlement has been something of a Cinderella issue as concerns project appraisal. Remarkably, until re- cently, the economic literature virtually ignored the subject, the main 54 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT exception being the discussion in Cernea (1988, particularly annex 2). Thus, the main economic manuals used for project appraisal in develop- ing countries-for example, Curry and Weiss (1993), Gittinger (1982), Squire and van der Tak (1975), UNIDO (1972), and Ward and Deren (1991)-do not address issues of compensation for displacement. Although the World Bank's policy guidelines focus, explicitly and primarily, on the social issues involved in displacement and relocation, they also directly address some basic economic issues of income recov- ery and contain many economic implications. The following are rele- vant to this study and will be discussed in more detail later: . Projects should avoid or minimize involuntary resettlement (OD 4.30, paragraph 3a). * Project designers should regard both customary and formal prop- erty rights as criteria for eligibility for compensation (OD 4.30, paragraphs 3e and 17). • Resettled people should be better off, or at least no worse off, after resettlement (OD 4.30, paragraph 4), and project designers should focus on resettlement as a development opportunity. * For the above point to hold, full and proper assessment of compen- sation must be carried out through the valuation of private and public assets and income (OD 4.30, paragraph 3b). Does Resettlement Differ from Other Project Externalities? The Bank's policy guidelines require that actual compensation be paid for displacement and lost assets. However, cost-benefit analysis theory, whose principles underlie Bank project and policy appraisal methodol- ogy, does not prescribe the payment of actual compensation to "losers.' In theoretical terms, it is only necessary for the project's benefits to be able to compensate for the losses and still have some gains left over. No actual transfers from gainers to losers are required. This is the ac- counting procedure referred to at the beginning of this chapter. If the Bank accepts this hypothetical compensation principle, why should re- settlement costs be special? Or to put it another way, if OD 4.30 man- dates actual compensation for resettlement, why not mandate compensation for other externalities, such as environmental impacts?' This question has several answers as follows: . Resettlement involves the loss of property rights to land and re- sources, although these rights may range on a broad spectrum, from secure and well-articulated rights to customary rights not en- METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 55 shrined in any formal law or registration. In contrast, environmen- tal impacts are not usually the subject of such rights. * Involuntary displacements may differ from other impacts because of the large numbers of people affected. * Displacement often affects the very poor, providing a strong equity argument for special attention. As Bank policy already mandates actual compensation, these con- siderations may seem academic, but they do raise the issue of why other impacts, notably environmental impacts, are not subject to actual com- pensation. One answer is that perhaps they should be where they are significant, for example, where they affect large numbers of people or the group in question is poor. This could be the case for downstream ef- fects of hydroelectric dams. In industrial countries, domestic compensation practices vary widely. Many countries operate compensation policies based on prop- erty values and removal costs only. For environmental impacts, their usual practice is not to compensate at all, except perhaps through noise insulation grants for houses near airports, but generally not for roads. An interesting departure from this practice is Germany's Nature Protec- tion Law, which requires that environmental impacts be offset through the creation of equivalent environmental assets elsewhere. Thus, ar- guing that all significant externalities be compensated is not absurd, although one does have to acknowledge the complexities of implement- ing such a policy. Issues in the Economics of Resettlement Avoiding Involuntary Resettlement Economics lends support to the idea that projects should first be screened to ensure that resettlement is minimized. Thus, Bank projects should attempt to avoid involuntary resettlements where possible. In avoiding involuntary resettlement, two contexts need to be distin- guished. The first is where resettlement is minimized for a given project benefit. For example, if a project to produce 100 units of electricity has two possible designs, one of which involves moving 100 people while the other involves moving 120 people, then the screening process should select the project that involves displacing fewer people. This usually amounts to ensuring that the concept of an efficient project de- sign is extended to incorporate resettlement as a cost. Typically, efficient design is taken to mean minimizing the private costs of construction and 56 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT operation for a given level of project benefit; that is, minimizing the costs of capital, raw materials, labor, and energy. However, this fundamental feature of good project appraisal must be extended to become the princi- ple of minimizing all costs, economic, social, and environmental, of pro- ducing benefits from a development project. Typically, cost-benefit guidelines emphasize maximizing the net benefits from a project, or at least ensuring that its benefits are greater than its costs.2 Implicit in these requirements is the idea that costs are no higher than they need be, but in practice, project designers often over- look this requirement. Project design often emerges without detailed scrutiny of the cost dimension, and even when such scrutiny is carried out, it tends to relate to private costs, not to the wider concept of costs, that is, economic plus social plus environmental costs. Thus, one of the major social costs, resettlement, can easily be overlooked when project designers are trying to minimize costs. Failure to minimize these costs means that some resettlement occurs needlessly.3 The second context differs from the first in that a tradeoff is involved between minimizing resettlement costs and the benefits of the project. In the first context, no tradeoff is involved: wasteful expenditures are being avoided. Tradeoffs can occur when, for example, in a hydroelectric pro- ject the height of the dam is lowered to reduce the area of inundation, and hence the number of people displaced. While lowering the dam's height need not sacrifice electricity output, it may do so. The situation then is that project designers must compare displacement costs, the human costs associated with resettlement, with project benefits. If the affected parties are equally "deserving," for example, electricity consumers are as poor as displaced families, then project designers can carry out a di- rect comparison of costs. If full compensation for displaced families is less than the value of the electricity, then maintaining the dam's height and moving the families will be beneficial. If the compensation costs are higher than the added value of the electricity, then the designers should lower the dam. Some projects have gone through this kind of analysis, but inspection of many projects involving human displacement does not reveal evidence of project designers routinely calculating these tradeoffs. Although this form of tradeoff analysis assumes that all parties are equally deserving, one can make a strong case for applying greater weight to the loss of well-being of the displaced than to the gain of the electricity consumers. In the first place, electricity beneficiaries may of- ten be better off to begin with than those displaced. Second, and perhaps more important, a fundamental asymmetry exists between the value of a loss of well-being (displaced people) and the value of a gain in well-being (electricity consumers). This basic asymmetry is dealt with in METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 57 more detail later. For now, note simply that a $1 loss may be far more im- portant in social terms than a $1 gain, even where the gains and losses accrue to people with the same incomes or wealth.4 This discussion suggests several recommendations for routine eco- nomic analysis of resettlement operations in developing (as well as de- veloped) country projects. Thus, any set of guidelines on resettlement should emphasize the following: * The fundamental need to engage in standard full cost minimiza- tion in project design, where full cost refers to the sum of economic, social, and environmental costs . The need to engage in tradeoff analysis where resettlement minimization is at the expense of project benefits . The need to recognize that a $1 loss for project losers (displaced people in this case) has a higher social value than the $1 gain to pro- ject beneficiaries. None of these recommendations is inconsistent with the economics of project appraisal as prescribed by the Bank. They are, in fact, logical outcomes of that economics. Property Rights The emphasis in OD 4.30 on treating people with customary property rights in the same way, as far as is feasible, as people with full legal rights is important. It is also consistent with the Bank's policy guidelines on indigenous people (World Bank 1991). Clearly, however, when is- sues of actual compensation for displaced people arise, those with less well defined tenure and resource rights are often treated less well than those with secure rights (Cernea 1990). This asymmetry of treatment is often conveniently justified by the absence of legal registration docu- ments and so on, but has more to do with minimizing the monetary, rather than the human, cost of resettlement. There are two strong reasons for treating customary rights in the same way as legal rights. First, project analysis is concerned with the ag- gregate well-being of people. Gains and losses in well-being are no dif- ferent if customary rights are involved than if full legal rights are involved. Second, by conferring full recognition on holders of custom- ary rights, the resettlement project is likely to generate better protection of land and natural resources. While social analysts have shown that the traditional distinction between common property and private owner- ship in terms of sustainable land and resource use is false (Bromley 1991, Bromley and Cernea 1989, Stevenson 1991), a better definition of land 58 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT and resource rights generally acts as a means of ensuring sustainable agricultural practices.5 Thus, recognizing customary rights is not just a matter of fairness (in itself a powerful enough argument) but also an issue of economic efficiency and sustainability. The recommendation that follows from this is that in developing countries guidelines for resettlement could find strong economic justifications for the need to recognize custom- ary rights as being on a par with legal rights. Such treatment is consis- tent with the concepts of economic efficiency and sustainable development. Of course, determining what the structure of rights actually is in a particular local context is immensely complex. There is always the risk that outsiders will seek to exploit the availability of compensation fund- ing by claiming customary rights where none exist. This is a hazard of any situation in which compensation for harm is paid and is an example of a much wider phenomenon that economists call rent seeking. It is not peculiar to resettlement. The No-Worse-Off Criterion The exact meaning of being no worse off as a result of resettlement re- quires some clarification. Two considerations are involved: the cor- rect application of the welfare economics underlying cost-benefit analysis and the implications of the philosophy of sustainable devel- opment. It is tempting to compare incomes and assets before and after the project and use that as a test of whether or not resettlers (and hosts) are better or worse off. This is inadequate for two reasons. The first is that we need to know how much better off resettlers would have been had the development not occurred. An estimate of their incomes in the project's absence is at best only a minimum estimate, because it is necessary to project what the situation would have been without the project. This is not necessarily the same as the situation before the project. The second reason is that the methods for valuing assets and income do not allow for the displaced people's own valuations of as- sets and access to employment. Moreover, practice in domestic pro- jects in many countries seldom relies on genuine consultation of affected groups. The World Bank's policy requirement that those resettled, at the very least, not become any worse off and its focus on resettlement as a development package receives strong support from the recent literature on sustainable development. The essence of sustainable development is (a) that future generations should be no worse off than those currently METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 59 living, and (b) that the currently least advantaged in society should be given special attention and, at the very least, should be no worse off be- cause of development. That is, both intragenerational and intergenera- tional equity are of paramount importance (Pearce, Barbier, and Markandya 1990; World Commission on Environment and Develop- ment 1987). The implications of sustainable development for cost-benefit analysis are still being worked out in the literature (Pearce, Markandya, and Barbier 1990; von Amsberg 1993), but conventional interpretations of cost-benefit analysis are inconsistent with these pre- cepts of sustainable development. Essentially, these equity consider- ations require procedures for actually compensating future generations and the poor now. The idea that resettled people should be no worse off as a result of being displaced is thus consistent with the basic philosophy of sustain- able development. This consistency is reinforced by considering the other feature of OD 4.30, namely, the emphasis on transforming resettle- ment into a development opportunity. The literature on sustainable de- velopment shows that one of the conditions for intergenerational equity is that stocks of capital assets be no less in the future than they are now, where asset stocks include the everyday concept of capital (machines, roads, and so on) as manmade capital, the stock of skills and knowledge (human capital), and the stock of environmental assets (land, trees, re- sources, and so on) (Pearce 1993; Solow 1986, 1993). By extension, then, the intragenerational requirement has as its condition that the asset base of the poor should not decline as a result of development. The basic logic of the constant and increasing capital base argument is that capital pro- vides the capability to develop. Thus the Bank's policy emphasis on restoring the economic base of the resettled (OD 4.30, paragraph 4) is wholly consistent with the idea of ensuring that the same, or an expanded, capital base is con- ferred on the disadvantaged. That is, treating resettlement as a devel- opment opportunity is an example of what is meant in practice by sustainable development. Full and Proper Compensation The economic costs of resettlement comprise three elements: (a) com- pensation to those displaced and associated rehabilitation investment costs, (b) compensation to the host community, and (c) administrative costs. We will focus on the first two elements. Several potentially large sources of undercompensation exist even if governments and Bank staff pursue the existing guidelines. These are: 60 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT * Undercompensation because of the time lag between determining compensation and the time of resettlement * Failure to account for nonmarket income and costs, which in turn comprise nonpriced environmental services, cultural assets, the psychological costs of dislocation, the value of community social cohesion, and the value of market access * Lost consumer surplus from existing assets. Project designers can overcome at least some of these sources of undercompensation by using valuation techniques borrowed from en- vironmental economics. UNDERCOMPENSATION DUE TO DELAYS. OD 4.30 and the existing eco- nomic guidelines (Cernea 1988, annex 2) approach the scale of desirable compensation for resettled people as follows: cost of replacement for losses plus costs of translocation plus transitional costs plus develop- ment or rehabilitation costs. There are three sources of bias when delays are significant. The first arises if the living standards of the displaced community were rising at the time compensation was determined. Delays will therefore mean that the community would have been even better off at the time of the actual relocation. The second bias arises if, as is often the case, the project itself raises land values and the displaced population has to buy into the new neighboring land rather than being allocated land directly. The third arises from a failure to account for inflation. THE COUNTERFACTUAL PROBLEM. Figure 2.1 illustrates the cost con- cepts involved in resettlement. The horizontal axis shows time and the vertical axis shows monetary units (which is not to imply that cash is al- ways the appropriate means of compensation). The curve NR (for no re- settlement) shows the path of asset accumulation without the project. This development curve, assumed for the moment to be rising, is inter- rupted by the project that will result in displacement. People's assets fall immediately because of the displacement and remain at a new low level for a further period as economic activity is built up at the new site (the transition period). If the resettlement package works, then a new devel- opment path is established as curve R. To meet the no-worse-off crite- rion, development path Rhas to involve faster development for a period than would otherwise have been the case to catch up with where those resettled would have been at time t+x. It is the full replacement of capital assets that ensures the takeoff on path R, and it is the rehabilitation component of the package that METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 61 Figure 2.1. $ Assets NR Rl 0 t t+n t+x Time ensures that path R is steeper than path NR. Essentially, if the com- munity was experiencing rising living standards before the project because of asset accumulation, replacing existing assets is not enough: the catching-up process will involve some additional asset creation. If path R continues to be steeper than NR after t+x, then the relocated com- munity is better off as a result of the resettlement project, which is con- sistent with the Bank's aims. By contrast, consider what would happen if the project made provi- sion only for replacing assets. After the transition period (t to t+n), devel- opment would recommence along path R', but the resettled community would be worse off than it otherwise would have been, even though it is developing at the same rate as if the project had never occurred. Strictly speaking, compensation for income loss caused by disloca- tion is the shaded area of figure 2.1; that is, the compensation arising from the fact that dislocation interrupted the community's develop- ment. This income loss compensation needs to be distinguished from as- set loss compensation, which involves replacing capital assets as they existed at time t or setting up new and different assets in the host com- munity. As several sources note (for example, World Bank 1993b), com- pensation for asset loss must be at replacement cost, not the worth of the asset. It is asset compensation that enables the achievement of path R. Asset replacement is an adequate criterion for compensation only if project designers can assume that the displaced community was not 62 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT experiencing growth and was not likely to experience growth, as shown in figure 2.2. Here assets are constant at time t and are expected to be con- stant in the future. Hence replacing assets will ensure a continued con- stant flow of income. The other situation in which replacement of assets is sufficient is if the adjustment period is extremely brief; that is, if the gap between t and t+n is short. However, delays are often substantial. Income compensation must be managed carefully, because the problem of dependency has been widely noted in communities that come to rely on the cash and food handouts in the new area (Butcher 1990; Cernea 1988). One way to avoid this problem is to place the obliga- tion for reestablishing a new income-earning capacity primarily on the developer through the rehabilitation part of the resettlement scheme. CHANGING LAND VALUES. The second source of bias arises because compensation at market values may be inadequate, because the devel- opment itself raises land values between the time compensation is agreed on and the time of dislocation (Butcher 1990). If compensation is in cash and is based on the situation before the price rise, then the dislocated population bears a direct cost. This must be accounted for and compensation must be adjusted, perhaps through contingency funds. If compensation for property loss includes a direct allocation of the new land, either the developer or the government bears the cost. If Figure 2.2. $ Assets NR 4 1 I 0 t t+ n Time METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 63 the developer seeks to reduce the commitment to provide new land of the same quality as those resettled had before the project, then some of the cost burden is transferred onto the resettlers. This can be avoided by carefully specifying both the quantity and quality of new land to be allo- cated to resettlers. Where this is done, the price rise should not affect resettlers. In some contexts they may actually gain, because the land price increases may reflect real land productivity gains from the project, for example from irrigation. INFLATION. While the need to calculate all resettlement costs in real terms should be self-evident, in some situations no attempt has been made to estimate inflation rates. In China, for example, past projects have not taken inflation into account, with the result that delays in mak- ing payments reduced their real value to displaced populations (World Bank 1993a). The theoretical ideal for payments has two main problems. The first problem is that the socioeconomic baseline survey at time t would need to assess not only assets and income at that point, but would have to secure some idea of how those assets and incomes would be likely to change over time. Historical evidence will almost certainly be absent unless the community happens to have been the subject of a previous survey, so there will be little evidence on which to base projections. A follow-up sur- vey of four Bank projects by the Bank's Operations Evaluation Depart- ment demonstrated the importance of changes in asset formation (World Bank 1993b). In the Ghana Kpong hydroelectric scheme needs had clearly been underestimated because of changes in assets and population be- tween the time of the survey and the time of the resettlement. The second problem is that it is not easy to judge what the entire ad- justment period is likely to be, so the size of the income compensation will also be unknown. No systematic evidence on adjustment periods is available. It is therefore unwise to assume that adjustment will be rapid. At best we can say that if there is any reason to assume that people's liv- ing standards were improving prior to the project and if the adjustment period is lengthy, then compensation at replacement cost for asset loss is a minimum requirement. UNDERCOMPENSATION BECAUSE OF LIMITED CONCEPTS OF INCOME AND ASSETS. A proper theoretical classification of payments associated with displaced people would be as follows: * Asset replacement costs * Income compensation for the entire adjustment period 64 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT • Additional investment costs, over and above asset replacement, designed to secure an accelerated development path to compensate for the forgone growth of past income * Any costs involved in actually moving the affected population, in- cluding administrative and transaction costs. A comparison of the asset and income list above with the pro forma cost tables in Cemea (1988, annex 2) reveals that the latter omit the in- come compensation element except in a reference to "subsistence pack- ages" (Cernea 1988, table 2). The additional asset formation or rehabilitation/development component is not clear in the pro forma ta- bles either. The advantage of making it clear would be to draw attention to the need to ensure the inclusion of the extra rehabilitation financing. The importance of this component is that it appears to be significant for the success of resettlement schemes; that is, schemes where people's liv- ing standards are better after the project than before. Schemes that em- ployed only asset compensation have been markedly less successful (World Bank 1993a, 1994,1998). Revised pro forma tables are suggested later. The terms income and assets need to be interpreted more broadly than has been the case to date. Income, for example, may be in cash or in kind, and it may be psychological in nature, for example, cultural and religious benefits. Similarly, assets may take the form of physical assets such as plows, tools, and buildings; livestock; access to the market place; and even nonphysical social goods, such as a community's social net- works for mutual help (Cemea 1990). Note that assets and incomes are closely related. For the farmer, land values will tend to be the capitalized value of the income to be earned from that land, assuming that any economic distortions (e.g., subsidies) have been netted out. The farmer's income will exceed the annuitized value of the land by amounts determined by off-farm in- comes and nonmarket income from fuelwood, plants, wildmeat, and so on. The landless will not have land assets, but what they lose is access to employment, and hence income. Their incomes are the annual value of themselves as a human capital asset. Once this is recognized, it is possi- ble to extend the compensation concept to both the landless and small businesses displaced by projects. For small businesses, assets will con- sist of nontransferable assets, usually the buildings, and transferable as- sets, such as tools and market access. In practice, then, it is important to determine the nature of all the as- sets and income sources forgone because of dislocation. An example of nonmonetary income is the range of forest products that people relocated METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 65 from tropical forest areas are likely to lose. A survey of the communities affected by the Pelagus /Bakun hydroelectric projects in Sarawak showed that important sources of nonmarket income were hunting and fishing, on which some groups depended substantially for food, and which would be lost on relocation (German Agency for Technical Cooperation and SAMA Consortium 1982). On actual relocation in Sarawak, those re- located also cited the loss of nonmarket benefits as a main source of "re- gret" (Furtado and associates 1991). They collected other forest products, such as rattan, dammar, garu, and jelutong, for cash sales, which would tend to be captured in any competent socioeconomic survey. The con- cepts of assets and income need to be kept broad to embrace nonmarket activities, the benefits of which may be lost on relocation. Communities may also lose the pattern of social organization that prevailed before the project. Resettlement projects often treat the dis- placed community not as a social unit, but as an aggregation of individ- uals and families, paying little attention to the social structure of the community. On relocation, traditional communal rules and regulations may break down as traditional leaders lose their power over the alloca- tion of rights to hunt and fish, to land, and so on-rights that effectively become the province of the government agency handling the resettle- ment. This is a social cost of resettlement for which compensation is ex- tremely difficult. Careful project design and, where possible, integration of particular ethnic and social groups with similar host groups can minimize the problem. UNDERCOMPENSATION BECAUSE OF NEGLECT OF THE CONSUMER SURPLUS. Welfare economics has shown that market prices do not cap- ture the full economic benefits of a good or asset to the consumer or owner. The difference between the "willingness to pay" for the benefit in question and the market price is known as the consumer surplus. Fig- ure 2.3 illustrates this concept. When the market price is Pm, the individ- ual's willingness to pay is shown by the demand curve. The shaded area is the consumer surplus. If an asset is lost, proper compensation must be based not on the market price, but on the market price plus the con- sumer surplus. If the asset is replaced, the loss is the consumer surplus. Finally, if the asset has no market value because it is not traded (as with cultural or religious assets, for example), the surplus is again the rele- vant measure (the price is zero, but the surplus is positive). The size of this surplus in the resettlement context is unknown be- cause we have no real studies in developing countries on which to base an estimate. Some capital investments involving dislocation have been the subject of willingness to accept compensation (WTAC) surveys in 66 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Figure 2.3 Price While the market price is Pm, the individual's willingness to pay is shown by the demand curve. The shaded area is consumer surplus. Pm Qm \uantity the industrial world. Thus, London's Third Airport study of 1969-71 asked householders potentially affected by the development for their WTAC to be moved (Dasgupta and Pearce 1972, Schofield 1987). Unfor- tunately, the questionnaire used was flawed in that it did not indicate how far affected households would have to move. Moreover, it was an early example of the use of contingent valuation. Questionnaires have become far more sophisticated since that time. OVERCOMING THE BIAS TO UNDERCOMPENSATION. It is evident that asset compensation could well involve undercompensation. The same may be true of the income approach. The issue is whether there are any ways to overcome this uncertainty about what the full and proper level of compensation should be. One approach is to supplement any asset or income inventory with a survey of willingness to pay (WTP) and WTAC. WTP and WTAC surveys are common in environmental economics, where economists use them primarily to elicit economic valuations of nonmarketed environmental services, for example, the benefits of pol- lution reduction. They are just as important for marketed services or quasi-marketed services, for example, water supply. This applicability of WTA and WTAC surveys, or contingent valuation, to the spectrum of marketed and nonmarketed benefits (and costs) is important in the re- settlement context.6 Another failing of many socioeconomic baseline METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 67 surveys, as already noted, is the omission of nonmarket sources of in- come, including cultural values and social cohesion. These are as much assets as anything else, and need to be accounted for. In the same way, there are psychological costs with dislocation. Once these broader concepts of income and assets are recognized, it will be evident that compensation based on some official assessment of employment and property will understate the true value of the assets in the preproject situation. If contingent valuation can be made reliable, it has the virtue that it captures the dislocated population's own valuation of its assets and income construed in the broad manner described ear- lier. That is, it can capture the nonmonetary concepts of income and as- sets, including psychological costs, environmental services, market access, and cultural property. Contingent valuation works by asking respondents to state their WTP to secure a given benefit and WTAC for giving up some benefit, usually (and preferably) in face-to-face interviews. A number of poten- tial sources of bias exist, including hypothetical bias if the respondent is not facing the actual choice in question. However, modern question- naire design has overcome many of the biases; investigators can test for the biases that remain to see if under- or overestimates are likely to arise. The hypothetical bias is likely to be minimal in the resettlement content because the threat of dislocation is a real one. (Hypothetical bias is not wholly absent, as several surveys have shown that people do not believe the event will happen until the day for relocation actually arrives [World Bank 1993b]). As resettlement involves a loss, the relevant measure is the WTAC. In virtually all contingent valuation studies the means of compensation is money. That is not always appropriate for the main losses-land and housing-in the resettlement context, for reasons Cemea (1988) and oth- ers have pointed out. However, nothing in the contingent valuation ap- proach says that money has to be the means of compensation. Nevertheless, cash compensation is appropriate for some losses that cannot be recreated instantly, for example, trees, and may work well for communities above the poverty line and where institutions are avail- able for investing the money in new homes and so on (World Bank 1993b). One possibility would be to survey communities about services that might be offered under a rehabilitation package and see what level of these might compensate for forgone preproject services. Such a ques- tionnaire approach could be confined to nonmarket assets and services, or it could be extended to cover all assets. One obvious problem with WTAC questions is the potential for people to give inflated responses in the expectation that the responses will influence the actual level of 68 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT compensation. Therefore, some thought needs to be given to mecha- nisms that will minimize such biases. The contingent valuation literature has another important feature. In studies in industrial countries a fair degree of evidence indicates that WTAC measures exceed WTP measures, often significantly (Gregory 1986, also box 2.1). While the reasons for this disparity are disputed Box 2.1. Contingent Valuation and Loss Aversion Contingent valuation involves carefully structured questionnaires that ask respondents to indicate their WTP for a benefit or WTAC for a cost. Contin- gent valuation has been used extensively in environmental contexts, and ap- plications have increased significantly in developing countries, mainly, but not exclusively, in connection with water supply and sanitation investments. WTP questions are more usual, but where anticipated losses are involved, WTAC questions are more appropriate. Economic theory suggests that WTP and WTAC should not differ much (they are the compensating variation and equivalent variation of welfare economics), but contingent valuation studies reveal a significant difference between them in practice, a result that may be caused by loss aversion. Some results from case studies in industrial econo- mies are shown below for hypothetical contexts, that is, "what if" contexts where the answers do not involve any real transaction, and for real exchange contexts, where the survey uses actual money or goods. Author Context WTAC/WTP ratio Hypothetical surveys Hammack and Brown 1974 Marshes 4.2 Sinclair 1978 Fishing 2.9 Banford and others 1979 Fishing pier, postal service 2.7 to 2.8 Bishop and Heberlein 1970 Hunting 4.8 Rowe 1980 Visibility 2.6 Brookshire 1980 Hunting 2.6 Heberlein and Bishop 1985 Hunting 16.5 Real exchanges Knetsch and Sinden 1985 Lottery 4.0 Heberlein and Bishop 1985 Hunting 6.9 Coursey 1987 Tasting 1.4 to 2.6 Brookshire and Coursey 1987 Trees 2.1 to 5.6 Source: Kahneman, Knetsch, and Thaler 1990. METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 69 (Kahneman, Knetsch, and Thaler 1990), many believe that it reflects a general phenomenon of loss aversion, whereby individuals attach greater value to losses than to equivalent gains.7 How far this asymme- try of valuation applies in the developing world is difficult to gauge, be- cause no studies appear to have obtained WTP and WTAC measures. However, if it does apply, then compensation measures may well be un- derstated by reference to the market prices of assets. The WTAC approach might also assist with the problem of valuing public assets. Clearly, relocation involves the loss of access to and shares in communal assets. The Bank's approach is to replace those assets in the host community, replicating (and improving) the asset structure in the preproject situation as far as possible. This is a sound approach, but it could be supplemented by contingent valuation assessments to provide some idea of the valuations of those assets by individuals.8 A further ap- proach to valuing public asset losses is suggested later through the use of rules of thumb for compensating the host community. The WTAC approach may prove unworkable in some contexts. In the resettlement case respondents are effectively asked to state the value of their assets, income, and other sources of well-being in the context of a development that they may well not understand or may not believe will actually take place. As Butcher (in Furtado and associ- ates 1991) notes: The tendency to disbelieve anything will happen, together with their inability to conceive of the man-made lake, creates a very difficult situation in which affected people have to make rational choices on what to do. For many, their world is limited to their village and the land and forest around it; and anything beyond this has no significance for them other than its strangeness. However, the observed ability of such groups to organize them- selves when the threat of development is realized suggests that they may well be able to respond rationally to interviews involving WTAC questions. Moreover, the contingent valuation approach has an addi- tional virtue: its compatibility with participation exercises. Because WTAC questions require that respondents understand the exact sce- nario for relocation, contingent valuation canbe added to baseline social surveys fairly easily. At the same time, the questionnaires can encour- age a sense of involvement in the project on the part of the affected pop- ulation (box 2.2). While there is little experience with WTAC/WTP valuations in de- veloping countries, the evidence that is available suggests that such val- uations can work and that they produce sensible results (box 2.3). 70 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Box 2.2. Negotiating Compensation in China Resettlement procedures in China have some of the characteristics of the kind of bargaining context to be found in contingent valuation procedures, even though contingent valuation is not used. Initial announcements about projects are followed by careful information campaigns targeting the spe- cific populations. Counties negotiate with towns, villages, and households about compensation. Compensation offers are often refused, and negotia- tions subsequently focus on securing acceptance of the offer made or on modifying the offer. The extent of public discussion may vary from one project to another and the mechanisms of such negotiation differ from rural to urban resettlement programs. Source: World Bank 1993a. While the suggested approach obviously has its problems, contin- gent valuation surveys could be experimented with to see how answers match up with the replacement costs of assets. Theoretically, we would expect the WTAC to be greater than the value of replaced assets, reflect- ing the nonmarket element and the respondents' expectations about the change in the quantity (or value) of assets over time. Box 2.3. Contingent Valuation in Madagascar One of the few examples of the use of WTAC approaches in developing countries concerns the potential loss of access to a forest area of Madagascar because of a planned national park. The establishment of the park involves reduced access for nearby villagers, who rely on the forest for various envi- ronmental services. The project planners assessed their losses in two ways: by estimating the market value of the services lost (the opportunity cost ap- proach) and by asking villagers for their WTAC to forgo the uses of the for- est (contingent valuation). The contingent valuation should capture the nonmarket benefits of the forest. The results are set out below. Although shown in dollars, the compensa- tion was measured in terms of rice. Market valiue offorest products (US$ per household) WTAC $91 $108 While the results are similar, it is significant that the contingent valua- tion measure is nearly 20 percent higher than the market price indicator. Source: Kramer and others 1993. METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 71 PAYMENTS TO THE HOST COMMUNITY. As Cernea (1988) notes, the host community also bears costs in assimilating the displaced commu- nity: "empty green field" relocation sites are rare. However, the 1988 pro forma for economic and financial costing of resettlement projects does not overtly recognize the concept of compensating the host community (Cernea 1988, annex 2). While assessing the losses of the relocated group is difficult, it is eas- ier than assessing appropriate compensation for the host community. This is because the relocated group has identifiable losses, whereas while the host community has some known costs, it also faces antici- pated but uncertain costs. These will be especially difficult to assess if the host community has little experience of assimilating settlers. The World Bank's Operations Evaluation Department's review of four pro- jects (World Bank 1993b) also shows that failure to compensate hosts adequately (or at all), as in the Kpong hydroelectric project, can quickly render resettlement unsustainable. Experience with the con- struction of the Narayanpur dam in India also suggests that failure to anticipate environmental demands on the host area can cause signifi- cant social conflicts between host populations and resettlers (World Bank 1989). The most difficult cost to estimate in this context is the effect of the expanded population on the host community's natural resources and environmental services. However, some fairly elementary rules of im- pact assessment can be applied. If fuelwood is an issue, estimating the extra demand should be possible by looking at demand in the host and in the dislocated community. Ideally, a resource inventory of the host area would indicate the potential for meeting the expanded demand. Otherwise, explicit provision will have to be made to establish woodlots for the specific purpose of providing additional supplies. Much the same holds true for water. Compensation might consist of funds for ad- ditional boreholes or even for communal taps. By far the most difficult problem will be estimating the effect of an expanded population on the receiving capacities for human and other waste. Financing Resettlement Resettlement packages need to be financed, and the costs of resettlement must be debited to the cost of the particular development project. Where the project yields benefits, for example, in the form of electricity or irri- gation, beneficiaries should pay for the resettlement costs through a charge on the output, for instance, through higher electricity prices or ir- rigation charges. Because the resettlement costs occur early in the pro- ject cycle and the benefits come later, charges will yield revenues that 72 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT can be used to pay back loans made specifically for the resettlement component of a project. The Policy Context Resettlement occurs mainly in the context of dams for irrigation, hydropower, and potable water. These investments are justified in terms of the net benefits they generate for society at large. Conventional measures of those net benefits are the economic rate of return and the net present value. These measures of social profitability are determined, in turn, by the quantities and prices of the project's outputs and inputs. As far as the Bank is concerned, shadow pricing of the costs and benefits of, say, a hydropower dam should ensure that the project is correctly evalu- ated. In terms of markets' actual workings, developing countries subsi- dize energy, including electricity, fairly systematically (Kosmo 1987, 1989). Electricity prices are, on average, about 30 percent of the costs of supply. If consumers paid the long-run marginal cost of supply, they would consume around 20 percent less electricity (World Bank 1992). Thus, if marginal cost pricing were employed, people would demand less electricity and fewer communities would be dislocated. The problem of resettlement is, therefore, partly a problem of dis- torted energy and water pricing policies in the developing world. Those who are threatened with social dislocation should not pay for these eco- nomic distortions by being forced to bear the burden of economic ineffi- ciency. Project interventions cannot solve these problems. Only macroeconomic and sectoral policy reform can correct the distortions. Provided that those projects involving the Bank are properly evalu- ated, that is, shadow priced, one could argue that the Bank is not contrib- uting to this element of unnecessary displacement. However, the shadow pricing requirement is somewhat stricter than marginal cost pricing. Marginal social costs should form the basis of the correct pricing rule (Goodland, Juras, and Pachauri 1991). This, in tum, means that elec- tricity and water prices should reflect the environmental and social costs of production and distribution, as well as the private costs. This full cost- ing of the outputs of projects is still not practiced on a sufficiently rigor- ous basis (Dixon, Talbot, and Le Moigne 1989). In addition to proper pricing, a careful consideration of the alternatives always needs to be made. As far as multipurpose dams are concemed, defining the alterna- tives is not easy because few alternatives are usually available to dams for irrigation or flood control. However, alternative sources of electric- ity are available, and planners should investigate these thoroughly (World Bank 1992). METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 73 From this discussion, it follows that at least some amount of dis- placement is due to economic distortions in the markets for the outputs from the projects in question, mainly water and electricity. The Bank can minimize its involvement in this unnecessary displacement by careful application of shadow pricing rules, but these must embrace projects' environmental and social consequences. Resettlement and Project Rates of Return One issue that frequently arises is whether resettlement costs make pro- jects uneconomic. Gutman (1993) shows that if resettlement costs amounted to more than 10 percent of total project costs, 50 percent over- runs in resettlement costs could have jeopardized up to 30 percent of a sample of projects reviewed. Where resettlement costs are less than 10 percent of project costs, overruns are not likely to result in unacceptable rates of return. The possibility that high resettlement costs could cause unacceptable rates of return appears to be a cause of some concern. (Simi- lar concerns have been expressed about environmental costs.) However, if the resettlement costs are correctly calculated using the "no worse off" criterion, then projects with unacceptable rates of return should not go ahead anyway. Invoking a rate-of-return criterion for accepting and re- jecting projects, and then denying its relevance when a proper accounting of costs produces a reject decision, would be contradictory. Circumstances that give rise to such concerns are probably those in which the project is thought to be a national priority, perhaps because no alternative is available to meet projected energy or water demand. Apart from the fact that one alternative is always available, which is not to proceed with any development, the option of raising the price of the project's output is also available. While this may seem like a departure from correct shadow pricing, what such a procedure is actually doing is acknowledging the national need for the project and placing a premium on prices to reflect that need. What is critically important is that arguments about national need and unacceptable rates of return should never be used to play down the importance of properly costing the resettlement component, or for that matter, any other component. Guidelines for Gathering and Analyzing Economic Information for Resettlement This section converts the conclusions of the previous sections into a set of principles and guidelines for carrying out the economic analysis of 74 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT resettlement components in projects and for generating the economic information needed for such analysis. The guidelines also take into ac- count the current Bank guidelines, as well as the arguments developed in two other studies in this volume (Cernea and Eriksen) and the find- ings of the Bank-wide resettlement review (World Bank 1994). Box 2.4 lists the cost components of resettlement according to their attributions into a pro forma that may facilitate and standardize the project econo- mist's work. The suggested principles and guidelines are as follows: Box 2.4. A Revised Pro Forma for Assessing Resettlement Costs The costs listed below are those that should be debited to any project that en- tails population displacement. All cost items should be allocated to the year in which they are likely to occur to permit standardized discounted cash flow analysis. For financial appraisal, costs are estimated inclusive of taxes, subsidies, and so on. For economic appraisal, shadow pricing of significant items will be required using available conversion factors. Cost categories vary according to the nature of project and the affected party. Cost category Comment Administrative and transaction costs Baseline surveys of affected populations Ensure that costs are estimated and host communities, census, land in both affected and host areas, registration, participation exercises, and for the full projected reset- staffing, offices, technical assistance. tlement period. Compensation and resettlement costs: marketable assets Lost land: new land where possible. Allow for productivity differ- ences of new land, if any. Allow for potential higher price of new land-see guide- lines. Lost houising: replacement housing. Crops: new land compensates but not Obligation on the developer to for the adjustment period when food ensure that the transition and and cash will be required. adjustment period is minimized. Lost employment: new employment. METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 75 FULL COST ANALYSIS. All projects that result in displacement of peo- ple require full cost analysis, where full cost means resource costs plus environmental costs plus social costs. The failure to engage in full cost analysis will result in unwarranted displacement effects. COST MINIMIZATION. For any given level of project benefit (such as electricity or irrigation water), resettlement costs should be minimized. This can be done through careful analysis of alternative designs and op- timal location of the project. Cost category Comment Small butsiness: new building. See also "market access" below. Public assets Infrastructure: expanded assets in the Includes any provision of host community. community assets to induce acceptance of immigrants. Environmental services Grazing lands, common resource rights: Depends on capacity in host expanded assets in host community. community. Water: cost of new boreholes. Cash Depends on capacity in host payments appropriate if water has to be community. purchased. Fuelwood, fodder, poles; cost of tree- Depends on capacity in host planting schemes in host community. community. Cash payments appropriate if wood and fodder have to be bought, such as in transition period. Compensation and resettlement costs: nonmarket assets Cultural assets. Costs of recreation. Contingent valuation if recreation of assets difficult. Social cohesion Use contingent valuation to Psychological stability determine values for cash or Environmental services in-kind compensation (for ex- ample, wild meat). Rehabilitation costs Investment costs of rehabilitation projects. Costs of extension, new works, retraining, credit, marketing. 76 THE EcoNoMIcs OF INVOLUNTARY RESETTLEMENT TRADEOFFS. Where a tradeoff between the amount or severity of dis- placement and the net benefits of the project can be identified, that tradeoff should be evaluated. A simple indicator might be computed, such as the number of families displaced per $1,000 of annual net bene- fits, so as to show the displacement cost in terms of the overall project benefits achieved. WEIGHTING THE COSTS. Giving a higher weight to the cost of dis- placement, that is, in excess of the monetary value of assets and income lost, is warranted on two grounds: (a) equity, if the displaced population has less income or wealth than the likely beneficiaries, and (b) the proba- bility that loss aversion exists, whereby people place a higher value on a $1 loss than on a $1 gain. ATTENTION TO HOST COMMUNITIES. The concern that resettlers are no worse off after the project than before needs to be extended more explic- itly to host communities where relevant. In determining whether resettlers and host communities are no worse off, attention must be paid to what would have happened in the absence of the project. Current guidelines assume a comparison between levels of well-being before and after the project. However, the relevant comparison is between the level of well-being that would have occurred without the project (allow- ing for the growth of incomes and assets) and the standard of living with the project. Additionally, account must be taken of the feedback effect of the de- velopment on land values: in other words, compensation agreed on be- fore the development is likely to be inadequate if land values rise by the time of the development. Land-for-land compensation avoids resettlers paying the costs of these price increases, provided the quantity and quality of the land they will receive is carefully specified. EXTENT OF INVESTMENT. Some relatively simple rules of thumb might be used to compute the relevant level of compensation for the loss of public assets where a host community is involved. As a general rule, investment in expanded and new assets should take place up to the point where assets per capita or per family are the same or better than in the best off of the two communities. Where resettlement involves a green field site (newly developed land), assets per capita should be no less than they would have been in the displaced community in the ab- sence of the project. No WORSE OFF. The "at least no worse off" criterion has strong justi- fication in the philosophy of sustainable development, which requires METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 77 that special attention be paid to the most disadvantaged members of a society and that development packages be implemented to ensure that displaced communities are no worse off in the future. The most impor- tant features of a rehabilitation package for sustainable development are likely to be secure land tenure and defined resource rights, exten- sion, start-up funds and access to affordable credit. NONMARKET SOURCES OF INCOME. The concept of income used when assessing incomes lost because of relocation must be extended to in- clude nonmarket sources of income. Market values ignore nonmarket and intangible assets. Even when the emphasis is on replacement costs and not just on the market value of lost assets, the risk of understating true losses of well-being remains. VALUATION BY RESETTLERS. An attempt should be made to assess the resettlers' willingness to accept compensation for the move by identify- ing their valuations of, at least, cultural assets and nonmarket benefits. Willingness to accept studies are also relevant to market assets because of the likely presence of consumer surplus, that is, valuations of assets over and above the market price of those assets. CASH COMPENSATION. The form of compensation is likely to vary from context to context. In general, however, cash compensation should be confined to those assets that cannot be easily replicated in the resettle- ment area, for example, trees; where resettlers have the experience to in- vest money in new assets and have access to reasonably functioning markets; or where employment opportunities are available. Such con- texts are more likely to arise in the higher-income countries. ATTENTION TO NATURAL RESOURCES. Efforts must be made to assess the additional demands that resettlers will make on natural resources in the resettled area, especially fuelwood, water, and grazing lands. If the host community has spare natural resource capacity, the resettled area may be able to accommodate new demands, but planners must ensure that they assess the regenerative capacity of resources, such as forests and groundwater. THE POLICY CONTEXT. Renewed attention needs to be paid to the policy context of projects entailing resettlement. Avoidance of reset- tlement can be maximized by ensuring that projects are subject to full social, environmental, and economic analysis; that they avoid exter- nalization; and that they incorporate valuations of all impacts as far as possible. 78 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT WHO PAYS FOR RESETrTLEMENT. Financing for resettlement should be regarded as part of the capital cost of a development project, with spe- cial attention paid to inflation and the means of financing resettlement. Because of inflation, compensation may be inadequate if it is not mea- sured in real terms. A second consideration is how resettlement is to be financed. The "beneficiary pays" principle suggests that the beneficiaries of the devel- opment project should pay the full costs of resettlement, not the people upon which displacement is imposed. This will happen automatically if the developer bears the full costs of resettlement rather than the govern- ment, because the costs are then internalized into the private costs of the development. Where the developer does not bear the full costs, the de- veloping entity can be made to do so through a "development charge," which is a charge levied on the developer to repay a loan granted by the government. Notes 1. Note that those who bear the costs of projects, the "developers," are already more than compensated by the expected flow of benefits from the project. Hence the issue is one of asking about compensation for externalities, costs that are not normally internalized. 2. The benefits and costs are all in terms of discounted values, which means adopting a discount rate. 3. A more technical point is that minimizing the sum of social, economic, and environmental costs does not necessarily mean that each component cost is min- imized. Minimizing one component cost may raise costs in another component, for example, but we ignore this complication here. 4. This is not necessarily a recommendation for the social weighting of gains and losses, although the Bank did experiment with such weights in project ap- praisal for some time (see Squire and van der Tak 1975). 5. Traditionally, development practitioners thought that communal owner- ship resulted in the "tragedy of the commons" (Hardin 1968), but in reality, it is open access, defined as no ownership at all, that gives rise to excessive resource degradation. Communal management is often an excellent protector of re- sources because the community as a whole is perceived to bear the cost of degra- dation, not the individual alone (Baumol and Oates 1988; Bromley and Cemea 1989; Pearce and Turner 1990). 6. The literature on contingent valuation is huge. A definitive bibliography compiled by Carson and associates (1995) has some 2,100 entries. The main theo- retical texts are Cunmmings, Brookshire, and Schulze (1986) and Mitchell and Car- son (1989). Applications of contingent valuation in the developing world have primarily been in the context of water supply and sanitation (see, for example, METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 79 Altaf and others 1992; Boadu 1992; Briscoe and others 1990; North and Griffin 1993; Singh and others 1993; Whittington and others 1987, 1990; Whittington, 1992). For surveys of contingent valuation in developing countries see Pearce and Whitting- ton (1994) and for a technical manual see Pearce, Whittington, and Georgiou (1993). 7. That is, the valuation function is kinked at the point that reflects the individ- ual's existing situation (the reference point) rather than smoothly curved, as is as- sumed in neoclassical microeconomics. Loss aversion extends beyond environmental and dislocation effects and may even explain some stock exchange behavior. The reason that loss aversion is contrary to standard economic theory is that the theory of consumer behavior suggests that the compensating variation measure of consumersurplus (essentially WTP) should be approximately equal to the equivalent variation measure of consumer surplus (WTAC) (see Willig 1976). 8. Moreover, replacing assets may contain a further element of bias in that the surplus value attached to the asset in the preproject context may not be the same as in the relocation area: it could be higher or lower. Contingent valuation would secure some measure of this surplus. References The word "processed" describes informally reproduced works that may not be commonly available through libraries. Altaf, M., H. Jamal, and D. Whittington. 1992. "Willingness to Pay for Water in Rural Punjab." Pakistan UNDP-World Bank Water and Sanitation Program Series No. 4. World Bank, Washington, D.C. Processed. Baumol, W., and W. Oates. 1988. The Theory of Environmental Policy. 2nd edition. London and New York: Cambridge University Press. Boadu, F. 1992. "Contingent Valuation for Household Water in Rural Ghana." Journal of Agricultural Economics 43 (3): 458-65. Briscoe, J., P. F. de Castro, C. Griffin, J. North, and 0. Olsen. 1990. "Towards Eq- uitable and Sustainable Rural Water Supplies: A Contingent Valuation Study in Brazil." The World Bank Economic Review 4 (2): 115-34. Bromley, D. W. 1991. Environment and Economy: Property Rights and Public Policy. Oxford, U.K.: Blackwell. Bromley, D. W., and M. M. Cernea. 1989. The Management of Common Property Natural Resources: Some Conceptual and Operational Fallacies. Discussion Paper 57. Washington, D.C.: World Bank. Butcher, D. 1987. "Report on Status of Resettlement in Bank Financed Projects and Corrective Actions Taken." World Bank, Washington, D.C. Processed. - 1990. "Review of the Treatment of Environmental Aspects of Bank En- ergy Projects." PRE Working Paper, World Bank, Industry and Energy De- partment, Washington, D.C. Processed. 80 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Carson, R., J. Wright, N. Carson, A. Alberini, N. Flores. 1995. Bibliography of Con- tingent Valuation Studies and Papers. La Jolla, CA: Natural Resource Damage Assessment, Inc. Cernea, Michael M. 1986. "Involuntary Resettlement in Bank-Assisted Projects: A Review of the Application of Bank Policies and Procedures in FY79-85 Projects." World Bank, Agriculture and Rural Development Department, Washington, D.C. Processed. _ . 1988. Involuntary Resettlement in Development Projects: Policy Guidelines in World Bank-Financed Projects. Technical Paper 80. Washington, D.C.: World Bank. . 1990. Poverty Risks from Population Displacement in Water Resources Devel- opment. Development Discussion Paper 355. Cambridge, Massachusetts: Har- vard Institute for International Development. Cummings, R., D. Brookshire, and W. Schulze. 1986. Valuing Environmental Goods: An Assessment of the Contingent Valuation Method. Totowa, New Jersey: Rowman and Allanheld. Curry, S., and J. Weiss. 1993. Project Analysis in Developing Countries. Basingstoke, U.K.: Macmillan. Dasgupta, A. K., and D. W. Pearce. 1972. Cost-Benefit Analysis: Theory and Practice. Basingstoke, U.K.: Macmillan. Dixon, John A., L. Talbot, and G. Le Moigne. 1989. Dams and the Environment: Considerations in World Bank Projects. Washington, D.C.: World Bank. Furtado,J. dos R., A. Markandya,J. Mitchell, and D. W. Pearce. 1991. LargeHydropower Dams in the Tropics: Development Assistance and Integrated Management. London: Centre for Integrated Management for Overseas Development Administration. German Agency for Technical Cooperation and SAMA Consortium. 1982. Pelagus/Bakun Hydro-Electric Projects: Socio-Economic Aspects and Resettlement. Kuching, Malaysia: Sarawak Electricity Supply Corporation. Gittinger, J. P. 1982. Economic Analysis of Agricultural Projects. Baltimore, Mary- land: The Johns Hopkins University Press. Goodland, Robert, A. Juras, and R. Pachauri. 1991. "Can Hydro-Reservoirs in Tropi- cal Moist Forest Be Made Environmentally Acceptable?" Policy and Research Di- vision Working Paper 1991-23, World Bank, Washington, D.C. Processed. Gregory, R. 1986. "Interpreting Measures of Economic Loss: Evidence from Con- tingent Valuation and Experimental Studies." Journal of Environmental Eco- nomics and Management 13: 325-37. Gutman, P. 1993. "Involuntary Resettlement in the Hydro-Power Projects: A Re- view of Bank Performance during 1978-1992." World Bank, Industry and En- ergy Department, Washington, D.C. Processed. Hardin, G. 1968. "The Tragedy of the Commons." Science 162: 1243-48. Kahneman, D., J. Knetsch, and R. Thaler. 1990. "Experimental Tests of the En- dowment Effect and the Coase Theorem." Jouernal of Political Economy 98 (6): 1325-47. METHODOLOGICAL ISSUES IN ECONOMIC ANALYSIS 81 Kosmo, M. 1987. Money to BBurn? The High Cost of Energy Subsidies. Washington, D.C.: World Resources Institute. _ . 1989. "Commercial Energy Subsidies in Developing Countries." Energy Policy 17 (une): 244-53. Kramer, R., M. Munasinghe, N. Sharma, E. Mercer, and P. Shyamsundar. 1993. Valuing a Protected Tropical Forest: A Case Study in Madagascar. Durham, North Carolina: Duke University, Center for Resource and Environmental Policy. Mitchell, R. C., and R. T. Carson. 1989. Using Siurveys to Value Public Goods: The Contingent Valuation Method. Washington, D.C.: Resources for the Future. North, J. M., and C. C. Griffin. 1993. "Water Source as a Housing Characteristic: Hedonic Property Valuation and Willingness to Pay for Water." Water Re- soiurces Research 29 (7): 1923-30. Pearce, D. W. 1999. "Sustainable Development." In D. W. Pearce, ed. Ecological Economics: Essays on the Theory and Practice of Environmental Economics. Lon- don: Edward Elgar. Pearce, D. W., and R. K. Turner. 1990. Economics of Natuiral Resources and the Envi- ronment. Hemel Hempstead, U.K.: Harvester Wheatsheaf; Baltimore, Mary- land: The Johns Hopkins University Press. Pearce, D. W., and D. Whittington. 1994. Economic Valuation of Environmental Costs and Benefits in Developing Countries: A Compendium. Nairobi, Kenya: United Nations Environment Programme. Pearce, D. W., E. Barbier, and A. Markandya. 1990. Sustainable Development: Eco- nomics and Environment in the Third World. London: Earthscan. Pearce, D. W., A. Markandya, and E. Barbier. 1990. "Environmental Sustain- ability and Cost Benefit Analysis." Environment and Planning 22: 1259-66. Pearce, D. W., D. Whittington, and S. Georgiou. 1993. A Technical Manualfor the Environmental Appraisal of Projects and Policies. Paris: Organisation for Eco- nomic Co-operation and Development. Schofield, N. 1987. Cost-Benefit Analysis in Urban and Regional Planning. London: Unwin Hyam. Singh, B., R. Ramasubban, R. Bhatia, J. Briscoe, C. Griffin, and C. Kim. 1993. "Ru- ral Water Supply in Kerala, India." Water Resoutrces Research 29 (7): 1931-42. Solow, R. 1986. "On the Intergenerational Allocation of Natural Resources." Scandinavian Journal of Economics 88 (1): 141-49. - 1993. "An Almost Practical Step toward Sustainability." Lecture given to Resources for the Future, Washington, D.C. Processed. Squire, Lyn, and H. van der Tak. 1975. Economic Analysis of Projects. Baltimore, Maryland: The Johns Hopkins University Press. Stevenson, G. 1991. Common Property Economics: A General Theory and Land Use Applications. Cambridge, U.K.: Cambridge University Press. UNIDO (United Nations Industrial Organisation). 1972. Guidelines for Project Evaluiation. Vienna. 82 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT von Amsberg, J. 1993. "Project Evaluation and the Depletion of Natural Capital: An Application of the Sustainability Principle." Environment Department Working Paper 56. World Bank, Washington, D.C. Processed. Ward, William A. and Barry J. Deren, with Emmanuel H. D'Silva. 1991. The Eco- nomics of Project Analysis: A Practitioner's Guide. Washington, D.C.: World Bank. Whittington, D. 1991. "Willingness to Pay for Improved Sanitation in Kumasi, Ghana." World Bank, Infrastructure and Urban Development Department, Washington, D.C. Processed. . 1992. Willingness to Pay for Water in Ruiral Punjab, Pakistan, UNDP-World Bank Water and Sanitation Program, Report 4. New York: UNDP. Whittington, D., M. Mujwahuzi, G. McMahon, and K. Choe. 1989. Willingness to Pay for Water in Newala District, Tanzania: Strategies for Cost Recovery. Water and Sanitation for Health Project, Field Report 246. Washington, D.C.: U.S. Agency for International Development. Whittington, D., J. Briscoe, X. Mu, and W. Barron. 1990. "Estimating the Willing- ness to Pay for Water Services in Developing Countries: A Case Study of the Use of Contingent Valuation Survey in Southern Haiti." Economic Development and Culttural Change 38 (2): 293-311. Willig, R. 1976. "Consumers' Surplus without Apology." American Economic Re- view 66 (4): 589-97. World Bank. 1989. "Staff Appraisal Report: India-Upper Krishna (Phase II) Irri- gation Project." Report No. 7406-IN. Washington, D.C. Processed. .1990. Operational Directive 4.30: "Involuntary Resettlement." Washing- ton, D.C. Processed. .1991. Operational Directive4.20: "Indigenous People." Washington, D.C. Processed. _ . 1992. World Development Report 1992: Development and the Environment. New York: Oxford University Press. _ . 1993a. "China: Involuntary Resettlement." Report 11641-CHA. China and Mongolia Department. Washington, D.C. Processed. _.1993b. "Early Experiencewith InvoluntarySettlement: Overview." Oper- ations Evaluation Department. Washington, D.C. Processed. _ . 1994. "Resettlement and Development: The Bankwide Review of Projects InvolvingInvoluntaryResettlement 1986-1993." Washington, D.C. Processed. . 1998. "Recent Experience with Involuntary Resettlement: OED Over- view." Washington, D.C. Processed. World Commission on Environment and Development. 1987. Outr Common Fu- tutre. Oxford, U.K.: Oxford University Press. 3 Comparing the Economic Planning for Voluntary and Involuntary Resettlement John H. Eriksen Editor's This chapter provides the most detailed comparative Note studv of voluntary population settlement programs and involtntary resettlement programs undertaken in the literature. Using a careful secondary analysis of data, the author reveals wide disparities in the ways in which sev- eral governments have treated the two types of resettlement over essentially the same time period. In almost all cases, these differ- ences have been to the detriment of those forcibly displaced. Coun- try economic planners appear to systematically undertreat involuntary relocation. The few positive exceptions signaled by the author not only reinforce the findings, but also convincingly dem- onstrate that involuntary resettlement does not necessarily have to become a development setback. These exceptions prove that the process can protect, even benefit, those relocated. The study uses a simple yet ingenious methodology. Five sets of development projects were identified in five countries according to criteria that make comparisons among the projects legitimate. Each pair consists of a project supporting voluntary resettlement and a project in the same country containing an involuntary reset- tlement component. For some countries, more than a pair of projects could be identified, so a total of 18 projects were studied. Appraisal reports and completion reports provided both ex ante and ex post. All of the projects were cofinanced by the World Bank and offer some of the largest sets of data available for development pro- jects-much more than the data typically available for projects fi- 83 84 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT nanced from domestic sources alone. Several variables-from type of resettlement to household income, from type of farm model en- visaged to type of risk assessment-are examined, as are some of the assertions expressed in the literature about differences between voluntary and involuntary resettlement projects in design, strategy, economic analysis, and financial allocations. The findings of this comparative analysis persuasively demon- strate that, independent of the complexity of forced displacement, failures originate partly from subjective errors of strategy and from inadequate or absent economic and financial analysis of resettle- ment components at project appraisal stage. The study argues that changes of paradigm and method are imperative, because marginal fixes will not improve the outcomes of involuntary resettlements. The facts highlighted in this chapter and its recommendations in- vite the full attention of everyone involved in planning resettle- ments, from decisionmakers to practicing planners, economists and social researchers. T his chapter compares projects supporting voluntary resettlement Twith those supporting involuntary resettlement. The purpose of this comparative analysis is to identify the differences or similari- ties between voluntary and involuntary resettlement operations. This analysis uses five sets (or pairs) of projects in five countries, with each set containing both types of projects (thus each set con- tains two or more projects). Several criteria were used to select ap- propriate project pairs for comparison: both projects had to take place in the same country, during approximately the same period. And both had to have resettlement components that focused on re- storing families' incomes by installing identical or very similar sets of agricultural enterprises. Project pairs also were selected to reflect a broad geographical spectrum, with cases drawn from Asia, Africa, and Latin America. The first section compares the planning approach and characteris- tics of resettlement activities in Brazil, China, India, Indonesia, and Ghana. The second section further analyzes the different farm "models" used in planning project resettlement operations. The analysis of the in- dividual projects also reviews the economic assumptions made about the baseline income of resettlers and planners' expectations as to their future after resettlement, including farm production, off-farm income, and so on. The final section lists recommendations for improving the de- sign and implementation of resettlement operations. COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 85 The specific resettlement activities compared in this study were components of the following World Bank projects: . In Brazil, the Sobradinho involuntary resettlement under the Paulo Afonso IV Hydroelectric Project and the voluntary reset- tlement projects under the Northwest Regional Development (Polonoreste) Program, which includes the Agricultural Develop- ment and Environmental Protection Project, the Mato Grosso Ru- ral Development Project, and the New Settlements Project * In China, the involuntary resettlement under the First and Second Shuikou hydroelectric projects and the voluntary resettlement un- der the Red Soils Area Development Project * In India, the involuntary resettlement under the Gujarat Irrigation projects I and II and the voluntary resettlement under the Rajasthan Canal Command Area Development and the Rajasthan Command Area and Settlement projects . In Indonesia, the involuntary resettlement under the Thirteenth Power (Cirata Dam) Project and the voluntary resettlement under the Transmigration I, II, III, and IV projects • In Ghana, the involuntary resettlement under the Kpong Hydro- electric Project and the agricultural development under the Volta Region Agricultural Development Project. A number of issues arose in connection with the analysis. First, data and time constraints imposed some limitations on this comparative analysis, the comprehensiveness of the available data sets, and the avail- ability of comparable pairs of projects. The selection criteria-country, time period, and sector focus-limited the number of involuntary/vol- untary resettlement pairings appropriate for comparative analysis. Sec- ond, because the study required complete and detailed data sets on project planning and implementation processes over time, four of the five resettlement pairings selected involved projects initiated in the 1970s or early 1980s. Thus the sample did not capture recent improve- ments in the way the Bank handles involuntary resettlement projects. Third, the involuntary resettlements were all components of major in- frastructure development projects, dams, irrigation, or both. These pro- jects, with their large-scale construction activities, demanded massive displacements. As such, they represent only one type of involuntary re- settlement. The study utilized World Bank staff appraisal reports (SARs), project completion reports (PCRs), and project performance au- dit reports (PARs), and in some cases, additional documents from pro- ject files. 86 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Programmatic Characteristics of the Projects The pairs of projects reviewed presented striking contrasts in their approaches to resettlement. In four pairs-those in Brazil, India, In- donesia, and Ghana-project designers and implementers clearly treated the involuntary resettlement component as subordinate to construction processes and schedules (for a discussion of this "engineering bias" in connection with reservoir planning see Cernea 1991) and as an economic externality (see the study by Cemea in this volume)- with poorly identified costs and no defined "benefits." Planning objec- tives appear to have been to internalize as few of the costs of resettle- ment as possible and to ignore the real costs or possible benefits of the resettlement component in internal rate-of-return calculations. SARs and PCRs often discussed involuntary resettlement activities in the same section as the analysis of the project's environmental impact. This way, the mitigation of resettlement problems had to be put in place during the life of the project, not ahead of time, and would not impinge upon or disrupt the central process of infrastructure development. By contrast, in the case of voluntary resettlement, project designers and implementers treated such activities, whether spontaneous or gov- ernment-assisted, as either the central focus of their activities or as an in- tegral part of a long-term, regional process to develop agriculture. In both Brazil and Indonesia, for example, planners and the settlers them- selves saw the transmigration of a large number of families from densely settled to "frontier" areas as vital steps in both regional devel- opment and nation building. In India, the settlement of landless families on newly developed irrigation schemes in desert areas of Rajasthan had a high priority in terms of both regional economic development and set- tlement of border areas. In Ghana, while there was no resettlement per se, area development for smallholder farmers had a high priority within the government. In voluntary resettlement projects, the costs of providing potable water supplies, social infrastructure, agricultural extension services, and road construction between farms and markets were internalized as line items within project budgets. Thus these activities were designed and implemented as building blocks of a larger process of social and economic development. The First and Second Shuikou projects and the Red Soils Project in China present a different approach to planning and implementing in- voluntary and voluntary resettlement activities in that resettlement ac- tivities were handled as central to the project and as an opportunity for long-term agricultural development in a frontier area. COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 87 A review of the paired projects revealed that, in each, the primary determinants of the approach to resettlement were the government's policies and attitudes and the composition of the teams charged with project design and implementation. Evidence suggests that interven- tions by Bank staff during supervisory missions may have mitigated some of the worst impacts of poor resettlement planning after its effects became evident: for example, in the Sobradinho component of the Paulo Afonso Project in Brazil. However, prior to the World Bank's adoption of its resettlement policy, Bank input into project design seems not to have influenced borrowers' handling of resettlement issues. More significant is that several individual voluntary resettlement projects were implemented sequentially under the Northwest Regional Development (Polonoreste) Program in Brazil and the transmigration programs in Indonesia. Thus the greater success of the voluntary reset- tlement programs could be a result of the cumulative effects of the expe- rience gained during individual projects. Bank Financial Participation in Resettlement Table 3.1 shows that the degree of financial participation by the Bank in the projects under review varied considerably; however, a pattern is evi- dent. Bank participation as a percentage of total project costs was gener- ally lower in the involuntary resettlement projects than in the voluntary schemes. And whereas the Bank's financial participation percentage- wise tended to increase (sometimes significantly) between appraisal and loan closure relative to total project costs in voluntary resettlement (often because the total investment in the project ended up less than an- ticipated), in involuntary resettlement costs the Bank's participation re- mained at the same level, or even declined, over the life of the project. An interesting observation is that where Bank financial participa- tion was lowest as a percentage of total project costs, project perfor- mance with respect to involuntary resettlement was judged to be the poorest. For instance, in the Paulo Afonso IV Project, where Bank partic- ipation was only 5.7 percent, performance was so poor and the impacts so negative that the lessons derived provided a major argument for in- stituting the first Bank policy on social issues in involuntary resettle- ment operations (Cernea 1993). Relationship between Scale and Success The number of people resettled varied considerably from country to country and among the paired projects within each country (see table 3.2). The largest numbers of participants were resettled voluntarily 88 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Table 3.1. World Bank Financial Participation in Involuntary and Voluntary Resettlement Total project World Bank fundinu finding Percentage of Bank Country/project (millionsofUS$) (millionsof oUS) participation Brazil Paulo Afonso IV 692.6a 81.0 11.7 Hydroelectric 1,413.9b 81.0 5.7 Agricultural Devel- opment and Envi- 199.3a 67.0 33.6 ronmental 132.2b 71.5 54.1 Protection Mato Grosso Rural 76.8a 26.4 34.6 Development 43 8b 22.7 51.8 New Settlements 182.0a 65.2 35.8 Indonesia Thirteenth Power 625.7a 278.3 44.5 609.2b 269.6 44.3 Transmigration 1 56.8a 30.0 52.8 Transmigration II 242 Oa 144.1 59.6 Transmigration III 185.8a 101.0 54.4 Transmigration IV 121.0a 63.5 52.5 638 gb 38.6 60.4 India Gujarat Irrigation 170.Sa 85.0 49.9 353.8e 85.0 24.0 Gujarat Irrigation II 350 Oa 175.0 50.0 248.7b Rajasthan Canal 174.0a 83.0 47.7 Command Area 183.6b 83.0 45.2 Development Rajasthan Coin- 110.6a 55.0 49.7 mand Area and n.a. n.a. n.a. Settlement China Shuikou 1,087.8a 140.0 12.9 Hydroelectric n.a. n.a. n.a. Second Shuikou 321.Oa 100.0 31.2 Hydroelectric n.a. n.a. n.a. COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 89 Table 3.1. (continued) Total project World Bank funding itnding Percentage of Bank Cotuntry/project (millions of US$) (millions of US$) participation China (continued) Red Soils Area 122.2a 40.0 32.7 Development 110.3b 44.7 40.5 Ghana Kpong Hydro- 236.5a 39.0 16.5 electric 250.6b 39.0 15.6 Volta Region 48.7a 29.5 60.6 Agricultural 43sb 28.8 65.8 Development n.a. Not applicable. Notes: Involuntary resettlement projects in the table are as follows: in Brazil, the Paulo Afonso IV Hydroelectric Project; in Indonesia, the Thirteenth Power Project; in India, the two Gujarat Irrigation Projects; in China, the two Shuikou Hydroelectric Projects; and in Ghana, the Kpong Hydroelectric Project. a. Projected figures from staff appraisal reports. b. Actual figures from project completion reports. in India and Indonesia: some 432,000 people in the two Rajasthan irriga- tion projects in India and 133,850 people under the four transmigration projects in Indonesia. Voluntary resettlement in agricultural development projects in Brazil, India, Indonesia, and Ghana were more successful than the in- voluntary resettlements, according to Bank evaluations, suggesting that the large scale of the resettlement activity does not significantly contrib- ute to failure of the component. Average Cost of Resettlement per Participant Table 3.2 presents a breakdown of the costs of resettlement activities by project, family, and individual. The data vary considerably in terms of quality and comprehensiveness, a fact that creates a number of prob- lems when attempting to analyze the costs of resettlement. The first problem is that in many projects, particularly those in the mid-1970s and early 1980s, the procedures for the financial analysis of projects with re- settlement components did not follow a standard format for enumerat- ing resettlement costs. The involuntary resettlement projects in Brazil, India, and Indonesia initially reported only the costs of land acquisition Table 3.2. Characteristics of the Paired Involuntary and Voluntary Resettlement Projects Resettlement Number of Total resettle- costs per Resettlement Ratio of resettle- families Number of per- ment costs family costs per capita ment costs Country/project resettled sons resettled ($ million) ($) ($) to GDP Brazil Paulo Afonso IV 9,700a 70,000 120.2 12,392 1,717 1.95 Hydroelectric 7,200b 51,969 236.2 32,806 4,545 5.17 Agricultural Development and 18,200a 87,000 199.3 10,951 2,291 1.23 Environmental Protection 11,330b 70,000 132.2 11,668 1,889 10.93 Mato Grosso Rural Development 10,OOOa 47,800 76.4 7,640 1,598 0.78 t 6,497b 39,632 43.8 6,742 1,105 0.54 New Settlements 15,000a 72,000 182.0 12,133 2,528 1.23 Indonesia Thirteenth Power 10,121a 40,000 92.8 9,167 2,319 3.80 11,052 43,655 92.8 8,395 2,125 3.48 Transmigration I 8,369a 33,476 52.8 6,787 1,697 6.29 7,232b 28,928 57.3 7,923 1,981 7.34 Transmigration II 30,000a 120,000 242.0 8,067 2,017 5.04 19,622b 78,488 140.1 7,140 1,785 4.46 Transmigration III 2,OOOa 8,000 185.8 92,900c 23,225c 38.07 2,000b 8,000 131.0 65,500c 16,375c 26.84 Transmigration IV 5,400a 27,000 121.0 22,407 4,482 7.35 4,466b 18,434 63.9 14,317 3,469 5.68 India Gujarat Irrigation and 18,630a 111,780 1.8 95d 16d Gujarat Irrigation II 15'700 94,200 344.1 21,917e 3,653e 19.23 20,050a 120,300 248.7 12,404' 2,067' 10.88 Rajasthan Canal Command 33,000a 198,000 174.0 5,273 879 6.28 Area Development Rajasthan Command 39,000a 234,000 110.6 2,836 473 2.49 Area and Settlement China Shuikou and Second 14,522a 67,239 207.0 14,254 3,079 10.26 Shuikou Hydroelectric Red Soils Area Development 19 OOOa 61,000 122.0 6,421 2,000 6.67 1 3,61 9b 33,646 110.3 8,099 3,278 10.93 Ghana Kpong Hydroelectric I,140a 5,700 11.5 10,439 2,018 n.a. 1,0890 7,000 5.4 4,957 771 n.a. Volta Region Agricultural 60,000a,g 336,000 48.7 812h 145h n.a. Development 90,000b 504,000 43.8 487h 87h n.a. n.a. Not applicable. a. Figures from SAR. b. Figures from PCR. c. Figure includes site selection and other planning costs for an additional 200,000 families-only 2,000 families actually resettled under the project. d. Land acquisition costs only. e. Total cost per beneficiary family and per capita calculated for Gujarat Irrigation Project. f. Total cost per beneficiary family and per capita calculated for Gujarat Irrigation 1J Project. g. Total beneficiary families, no resettlement in this project. h. Total cost per family and per capita. 92 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT and direct compensation. For the Kpong Hydroelectric Project in Ghana, and particularly for the Shuikou I and II projects in China, the fi- nancial analyses for involuntary resettlement were much more meticu- lous at appraisal and appear to have enumerated virtually all the costs of reestablishing the income levels of the displaced populations.' This may explain why the per capita resettlement costs are so similar for the China projects, but so divergent for the paired projects in the other countries. A second difficulty was the large differences between the antici- pated and actual cost figures reported for resettlement activities in all the countries. For example, the appraisal report for the Paulo Afonso IV Project in Brazil projected costs of US$12,392 per family and US$1,717 per capita, whereas the PCR indicated that actual costs had amounted to US$32,806 per family and US$4,545 per capita.2 In most of the other pro- jects, the actual costs of resettlement were lower than the initial esti- mates. Moreover, one sees PCRs simultaneously reporting that (a) on the one hand, the Bank's evaluators correctly concluded that project planners significantly underestimated the actual size of the populations to be displaced, and (b) on the other hand, because of defective project performance in implementing involuntary resettlement, several pro- jects actually provided compensation and services to significantly fewer displaced people than estimated. The PCRs generally attributed this to poor project implementation, revisions in project objectives during im- plementation, and exchange rate fluctuations. A third difficulty was that the number of people included in a par- ticular project tended to change significantly during implementation. In the case of the Northwest Regional Development (Polonoreste) Pro- gram in Brazil, where flows of spontaneous migrants were high, one can understand why it was difficult to obtain an accurate estimate of the size and composition of voluntary resettlement populations ex ante. In the involuntary resettlement projects, however, where the affected areas were defined in advance, the poor specification of both the size and composition of the groups to be displaced appears to have been primar- ily a function of inadequate initial field surveys by project planners (World Bank 1994; 1996, p. 131). Therefore an analysis of the economics of resettlement was difficult, given the constant shifts in funding levels and the number of people to be resettled throughout the implementation of most of the projects sur- veyed. We can conclude (a) that the available statistical data, given the unrefined methods used in financial planning, do not permit us to dem- onstrate a relationship between the amount of money spent on resettle- ment per family or per capita and the project's success at restoring resettlers' incomes, even though individual observations and reports suggest such a relationship; and (b) that when a development strategy COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 93 common to both involuntary and voluntary resettlement activities is ap- plied systematically from the outset, as in China, it is likely to lead to similar per capita costs for projects (excluding the costs of compensating resettlers for their land). Duration of Resettlement Operations Table 3.3 shows the length of Bank involvement in the surveyed resettle- ment projects. Currently, a full comparison of the pairs of projects is pos- sible only for Brazil, Indonesia, and Ghana. In Brazil and Indonesia, no appreciable differences in the duration of implementation are apparent between involuntary and voluntary resettlement activities, except that the voluntary resettlement projects had slightly shorter implementation periods than the infrastructural projects entailing involuntary resettle- ment within the same country. In Ghana, by contrast, the Kpong Hydro- electric Project lasted 69 months, whereas the Volta Region Agricultural Development Project took 111 months. What is probably more signifi- Table 3.3. Comparison of Effective Length of Projects with Resettlement Operations Effective length Loan dates of Bank involve- (Board approval ment in project Project to loan closing) (months) Involutntary resettlement Paulo Afonso IV Hydroelectric 06/04/1974 to 06/30/1983 97 Thirteenth Power 06/02/1983 to 05/08/1991 95 Gujarat Irrigation 05/25/1978 to 06/30/1984 73 Gujarat Irrigation II 04/29/1980 to 04/30/1989 96 Kpong Hydroelectric 03/01/1977 to 12/31/1982 69 Voluntary resettlement Agricultural Development and Environmental Protection 12/01/1981 to 09/06/1990 93 Mato Gross Rural Development 03/25/1982 to 12/31/1988 80 Transmigration 1 07/21/1976 to 04/13/1983 81 Transmnigration II 05/29/1979 to 12/31/1986 91 Transmigration IIi 03/22/1983 to 01/10/1989 69 Transmigration IV 05/24/1983 to 08/08/1991 98 Rajasthan Canal Command Area Development 07/16/1974 to 06/30/1983 107 Red Soils Area Development 09/09/1986 to 06/30/1992 70 Volta Region Agricultural Development 03/20/1980 to 06/30/1989 111 Note: Project dates taken from project completion reports. 94 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT cant is that several individual voluntary resettlement projects were im- plemented sequentially under the Brazilian Northwest Regional Development and the Indonesian Transmigration Programs. Therefore, it is possible that the greater success of the voluntary resettlement pro- grams is a function of the cumulative effects of several individual pro- jects, which learned lessons from prior phases. Use of Bank Resourcesfor Resettlement Table 3.4 presents an estimate of staff weeks and professional skills de- voted to most of the projects reviewed. The levels of Bank staff and con- Table 3.4. World Bank Staff Inputs in Involuntary and Voluntary Resettlement Projects in Staff Weeks and by Skill Mix Inputs into Preappraisal/ Country/project appraisal Negotiations Supervision Brazil Paulo Afonso IV 73.3 7.4 102.3 Hydroelectric E and Eng n.a. n.a. Agricultural 128.8 8.0 153.8 Development and E, A, M, F, Ls, A, E, F, and E, En, An, Eng, Environmental Ed, S, Ts, An Ts, A, M, Dm, Protection Eng, and An H and Ec Mato Grosso Rural 97.4 6.9 139.3 Development E, C, A, H, R, En, n.a. E, C, A, En, H, S, FandI F,I,R,LO,Ls,Ts, An and Cs Indonesia Thirteenth Power 55.9 13.7 104.5 Eng and F n/a Eng, E, F and O Transmigration I n.a. n.a. 43.4 n.a. n.a. A, E, C, Ps, F and H Transmigration II n.a. n.a. n/a n.a. n.a. A, E, An, Ls, Ae and Eng Transmigration III 110.9 16.0 109.5 n.a. n.a. A, C, Ms, Ls, E, Ae and Rs Transmigration IV 49.8 6.7 88.0 E, Eng, A, Ts An, A, E, and E, Eng, A, Ts, and An Rp S and M COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 95 sultants allocated to involuntary and voluntary resettlement projects had to be estimated because there is no uniform system for reporting those levels. The available data indicate, however, that the voluntary re- settlement and agricultural development projects generally benefited from a much broader mix of specialist skills during both preparation and appraisal activities than did the projects involving involuntary re- settlement. Agricultural and livestock specialists, agronomists, soil scientists, foresters, and commodity processing experts appear to have devoted a substantial amount of time to the design and implementation of most voluntary resettlement projects. This is perhaps because the projects were conceived as agricultural or integrated rural development Table 3.4. (continued) Inputs into Preappraisall Country/project appraisal Negotiations Sutpervision India Gujarat Irrigation 172.9 2.1 205.9 n.a. n.a. E, Eng, F and Wm Rajasthan Canal 97.7 7.0 133.6 Command Area A, E, Eng, R n.a. A, E, Eng and R Development and Fr China Red Soils Area De- 112.6 3.7 60.3 velopment E, A, Rc, Ls, Fs, n.a. E, A, Rc, Ls, Fs and Ap, M and S Ap Ghana Kpong n.a. n.a. 16.5 Hydroelectric n.a. n.a. C, F and An Volta Region 74.0 8.0 99.5 Agricultural A, E, Ls, Ex, C n.a. A, E, Ex, M, C Development and F and Fss n.a. Not applicable. Notes: Skill codes are: A = agronomist/agriculturalist; Ae agricultural economist; An anthropologist; Ap = agroprocessing specialist C = civil engineer; Cs = community develop- mentspecialist; Dm = demographer; E = economist; Ec = ecologist; Ed = education specialist; En = environmentalist; Eng = engineer; Ex = agricultural extension specialist; F = financial analyst; Fr = forestry specialist; Fs = fruit crop specialist; Fss = fisheries specialist; H = rural health specialist; I = institutional specialist; Lo = legal specialist; Ls = livestock specialist; M = management specialist; Ms = cartographer/mapping specialist; 0 = unspecified other skill specialist; Ps = procurement specialist; R = road engineer; Rc = rural credit specialist; Rp = rubber processing specialist; Rs = remote sensing specialist; S = soils specialist; Ts = tree crop specialist; and Wm = water management specialist. 96 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT activities. In addition, these projects made greater use of rural develop- ment specialists and social scientists in project preparation and supervi- sion. By contrast, few involuntary resettlement projects had broad partic- ipation by Bank staff beyond those needed for the basic analysis re- quired to plan and supervise the projects' physical infrastructure components. Because the Bank did not use social specialists (sociolo- gists and anthropologists) before 1980, when its own resettlement policy was first adopted, the overall average for the projects included in this study is low. Whereas the engineering and construction aspects of the infrastructural projects entailing involuntary resettlement benefited from the participation of Bank staff and consultants throughout the pro- ject cycle, their resettlement components appear to have been designed primarily by recipient country staff, with little Bank oversight. Informa- tion on the mix of local preparation staff is not available, but social spe- cialists were rarely part of local teams. Nature and Quality of Baseline Surveys Baseline surveys for involuntary resettlement projects should accom- plish at least three basic objectives: * They should specify the size and composition of the populations to be displaced. = They should provide a quantitative analysis of existing levels and sources of household income to aid in formulating precise income restoration targets. • They should provide detailed assessments of the income resto- ration potential of the resources, land, irrigation water, occupa- tional training, cash payments, and so on to be made available to the displaced households under each resettlement option proposed. Although data collection efforts improved modestly in the 1980s, the involuntary resettlement projects surveyed generally did a less than satisfactory job of assembling an accurate database on the populations to be displaced. The almost casual attitude of project planners toward displacees becomes clear when one compares the level of data provided in most appraisal documents for physical structures with the cursory, data-free discussion of the resettlement "program.' In most cases, the planners of involuntary resettlement activities at- tempted to specify the size and composition of populations to be COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 97 displaced by extrapolating from existing census data or other local re- cords and making projections of population growth through the period of infrastructure construction. However, the documents for the Thir- teenth Power Project in Indonesia (appraised after the adoption of the Bank's resettlement policy, see table 3.3) refer to a two-year anthropo- logical, socioeconomic, and ecological baseline study conducted by a lo- cal university before actual resettlement activities began. The Paulo Afonso IV Project in Brazil, specifically the resettlement program linked to the creation of the Sobradinho reservoir, exempli- fies the poor initial specification of the size and composition of the pop- ulations to be resettled. Such superficial survey work directly led to not only a serious underestimation of how many people would be af- fected by the reservoir's construction but to a complete misunder- standing of how those people would respond to the proposed resettlement options. Baseline surveys for the involuntary resettlement projects surveyed were particularly deficient in providing accurate information about the levels and sources of household income in the populations to be dis- placed and in valuing fixed assets that would be lost. The persistence of this deficiency in project planning and appraisal is particularly serious given the Bank's current attempts to build into involuntary resettlement projects a structured process of income restoration (see Cernea 1988). If one cannot specify in precise terms the nature and sources of baseline in- come flows and asset holdings within households to be displaced, one has a very limited basis, if any, upon which to formulate a resettlement options package. One also has no baseline from which to set goals, mea- sure "income restoration," and monitor progress. Finally, with the exception of the Shuikou I and II projects in China, the resource packages provided to households do not appear to have undergone serious multidisciplinary analyses to ensure their potential to restore income levels. Nor have the capacities of different types of households to respond to these packages been assessed. Quality of Resettlement Plans The documents on planning for involuntary resettlement demonstrated four different approaches to the problems caused by infrastructure de- velopment, as follows: * In the Paulo Afonso IV Project in Brazil, the resettlement plan had miscellaneous descriptive information (for example, generalized descriptions of the geology, soils, and vegetative cover in the 98 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Sobradinho reservoir area and qualitative and summary descrip- tions of the farming systems in place), along with a general state- ment of the objectives of the resettlement program. The Bank rejected this document. * In the Gujarat irrigation projects, resettlement planning appears to have been limited to a restatement of the government of India's pol- icy on compensation for displaced people and the operational pro- cedures for obtaining such compensation in the case of involuntary resettlement. . In the Thirteenth Power project in Indonesia, the resettlement plan offered four options for displaced people: resettlement from Java to outlying areas of Sumatra and other islands, par- ticipation in the development of aquaculture in the new reser- voir, occupational training, and simple cash compensation. The outcomes: the displacement and resettlement activity was han- dled under the Transmigration Program already in place, the results of the aquaculture program were satisfactory, the pro- posed occupational training program was subsequently can- celed, and the cash compensation program proceeded without major difficulties. * In Ghana, a team from the Faculty of Architecture of Kumasi Uni- versity of Science and Technology prepared "conceptual designs of the resettlement villages and construction of the township" (ap- praisal report). A team from the University of Lagon was con- tracted to "prepare the agricultural resettlement program" (appraisal report for the Kpong Hydroelectric Project). However, as the actual resettlement plan for this project was not found for this review, the influence of these university teams on the quality of the planning is unknown. None of these four so-called resettlement plans proposed a well-articulated and appropriately phased program, nor did they con- tribute to a structured process for restoring the incomes after displace- ment. The case of the Shuikou I and II projects in China presents an inter- esting contrast to the prior approaches. The local planners and the Bank (which sent its senior sociologist on the appraisal mission) treated the resettlement component of the projects as a genuine development op- portunity-not as an uncomfortable and undesirable project compo- nent. They created an upland agricultural development project and treated project planning and implementation similar to that of the Red Soils Area Development voluntary resettlement activity. The plan COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 99 aimed to reclaim uncultivated land on steep hillsides by using bench ter- racing for cultivating tree crops and for creating new village sites, with- out encroaching further on already cultivated lands. The Fujian Provincial Resettlement Office and local agricultural development agencies were assigned responsibility for executing the plan in recogni- tion that the engineering agency in charge of the actual dam construc- tion "lacked the authority or the capabilities needed for managing the resettlement work." Resettlement planning brought together a broad array of local specialists to work with groups of displaced families with the specific objective of arranging resettlement "so that physical and hu- man resources can continue to be at least as productive and living condi- tions at least as convenient as they were before." Differences among Implementing Agencies PAULO AFONSO IV HYDROELECTRIC POWER PROJECT. In the case of the Paulo Afonso IV Hydroelectric Power Project in Brazil, the Companhia Hidro Eletrica do Sao Francisco (CHESF) was responsible for the pro- ject's design and implementation, including that of the resettlement component. At appraisal Bank staff projected that CHESF would be able to "contract out" the responsibility for resettling about half the families to be displaced by the dam construction to the Instituto Nacional de Colonizacae e Reforma Agraria (INCRA). Under this plan, INCRA was to be responsible for resettling the rural population affected by reser- voir construction and CHESF was to be responsible for resettling the urban population affected. However, after discovering that not all the displacees met INCRA's eligibility requirements for the resettle- ment program and that many eligible individuals did not want to mi- grate to INCRA's distant agricultural resettlement sites, CHESF was forced to assume major responsibilities for resettling both rural and urban populations in the immediate area of the Sobradinho reservoir. This was a complex sociocuttural and technical task for which CHESF, as a power agency, was not staffed or otherwise prepared to un- dertake. The PCR for this project concludes that CHESF and the government of Brazil had complied with the strict conditions of the resettlement cov- enant, but with "migration results far different from what had been ex- pected." Moreover, the evaluators stated that: A major lesson of this project related to the importance of the re- settlement component is that planning for resettlement was as important as planning for the hydroelectric project; and that the 100 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT inclusion of covenants providing for resettlement did not en- sure the effective preparation and implementation of such plans. A companion lesson is that, for resettlement efforts on the scale of that experienced at Sobradinho, explicit consideration should be given to establishing a separate project. The status of a separate project would help to assure appropriate levels of gov- ernment funding and interest. GUJARAT IRRIGATION I AND II. For India's Gujarat irrigation pro- jects, which together included financing for about 15 dams, three agen- cies were responsible: the national Ministry of Water Resources, the Central Water Commission, and the State Irrigation Department. The government of Gujarat was responsible for resettling about 5,700 fami- lies (some 30,000 to 35,000 people) from villages that would be sub- merged by the reservoirs. To facilitate the socioeconomic reestablishment of the displaced families, the government was to ap- point a rehabilitation committee for each dam and irrigation scheme that would comprise representatives of the resettled families, govern- ment officials, and voluntary agencies active in the areas. A director of rehabilitation from the State Irrigation Department was to supervise and coordinate the resettlement. However, matters did not turn out as envisaged, and the analysis car- ried out at project completion pointed to organizational, institutional, and financial causes. Once again, local preproject planning had been deeply flawed, in fact utterly incomplete, with regard to resettlement and rehabilitation. During implementation, political commitment to full rees- tablishment was inadequate. Inside the Bank, these projects had also been largely processed before the new policy (1980) took hold. According to the PCR for the Gujarat Medium Irrigation Project (Phase I): Resettlement planning was not considered an integral part of the overall preparation of subprojects under Phase I. Although the legal framework and administrative arrangements existed, resettlement actions tended to be implemented in an ad hoc manner and were not consistently included in the implementa- tion schedules for project construction works. Moreover, since the government lacked effective institutionalized monitoring procedures, it was unable to determine to what extent the com- pensation packages it offered displaced persons enabled them to set up sustainable income-generating arrangements. In rec- ognizing these deficiencies, the government agreed to prepare and implement detailed resettlement plans for nine subprojects under Phase II of the project. The progress of resettlement activi- COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 101 ties is now monitored regularly ... Furthermore, special evalua- tion studies are being undertaken to analyze the standard of living of displaced persons in five subprojects and to provide the government with feedback on its compensation package. Despite the optimistic expectations for improved government work in the Phase II Project, formulated in the Phase I PCR, the subsequent Phase II PCR had to paint a somewhat discouraging picture: The project fell well short of achieving the project's established goals and objectives in both quantitative and qualitative terms despite two extensions, totaling three years . . . Project imple- mentation continued to experience delays, both avoidable and unavoidable, leading to significant performance deficiencies right from the early stages of the Credit.... Furthermore, it would not be reasonable to blame "start up" delays since sev- eral major components were on-going at appraisal. Except for some specific works and components which moved ahead rea- sonably well, the overall implementation of the project was less than satisfactory.... Numerous revised implementation sched- ules were loosely formulated by the government at regular in- tervals during each semi-annual review; still, the subprojects persistently lagged behind the targets which had been revised only a few months before. THIRTEENTH POWER PROJECT. Overall implementation of the Thir- teenth Power project in Indonesia was the responsibility of the National Electricity Authority (PLN), a public corporation charged with the gener- ation, transmission, and distribution of electricity and with the planning, construction, and operation of electricity supply facilities. On-site project implementation was the responsibility of a project manager responsible to PLN headquarters staff in Jakarta. NEWJEC of Japan was the consult- ing engineer and was responsible for the design and construction super- vision on the project. Finally, a special board of consultants (BOC) was appointed to review the investigations and design for the project. A resettlement coordinating board (RCB) coordinated compensa- tion and resettlement for people with homes and other major interests in the reservoir areas. The implementation of resettlement by the National Electricity Board was to (a) execute an environmental and resettlement plan of action satisfactory to the Bank, (b) monitor resettlement and en- virorumental activities closely, and (c) furnish the Bank with quarterly reports on the progress of environmental and resettlement activities. The PCR reported that: 102 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT The resettlement component was based on the experience with the Saguling hydroelectric project . . and was carefully planned. It was implemented on schedule and in accordance with a plan agreed between the Government of Indonesia and the Bank at the time of project appraisal. All families who lost land or property were given fair compensation in cash, based on procedures agreed with the Bank. The owners of private land were compensated based on the market value of the land, and the users of state forest lands were compensated through pur- chase and transfer of equivalent acreage of forest land in other locations. In addition to compensation for the loss of land and other prop- erty, the Cirata project also necessitated the resettlement, as of August 31, 1989, of 8,469 families from the reservoir area and 2,593 families from above the reservoir area.... About 34 per- cent of the affected families were resettled through the Transmi- gration and Nucleus Estate and Smallholder (NES) programs. Those who elected to remain in the area were assisted through training in aquaculture and industrial skills. Both of these train- ing programs were reserved for members of households who had not received any other form of compensation. The aquaculture program, which was modeled on a similar pro- gram developed for the Saguling Hydroelectric Project, was particularly successful. The project included follow-up surveys that were carried out at one-year intervals for three years (the last one in 1990), follow- ing the impoundment of the reservoir. Among the positive im- pacts of the project, the surveys identified electrification and improved roads. Among the negative impacts, the surveys noted the increased population densitv in the host villages, leading to reductions in average size of landholding, working hours per person, income, and nutrition levels. The authors of the surveys concluded that these impacts should be regarded as transitional, given the timing of the surveys shortly after the re- settlement. KPONG HYDROELECTRIC PROJECT. In Ghana, the Volta River Author- ity was responsible for planning and implementing the resettlement program. This authority had one prior experience with resettlement when the Akosombo Dam was built and nearly 80,000 people had to be moved. In this case, serious problems arose with the agricultural pro- gram and the housing and water supply activities, partly because of COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 103 poor planning by the authority, and partly because of its reliance on other government agencies for inputs that were not always forthcom- ing. The net result was that some of the resettlement towns created were deserted or they gradually deteriorated (for an in-depth socioanthropological description and analysis of this resettlement oper- ation, see Chambers 1970). Thus the preparation of the Kpong Project provided a rare opportunity for a power authority to learn from the planning failures of a recent, larger project on the same river and to avoid repeating the same resettlement failures. (For a discussion of the extent to which the Kpong project learned from the resettlement experi- ence gained in Akosombo, see Adu-Aryee 1993). A recent Bank report (World Bank 1993b) states that despite im- proved planning in the preparation and appraisal phases, "Kpong's re- settlement process was deficient in some elements of execution.... One weakness in plan preparation, however, caused settlers dissatisfaction at Kpong-underassessment of needs." The evaluation team concluded that while surveys of populations to be resettled must be done well ahead of the actual resettlement, constant focused monitoring is needed during implementation so that the resettlement process can be adjusted to respond to changing circumstances. The Kpong resettlement team faced nearly all the same problems as at Akosombo, despite specific attempts to address them. These prob- lems included working under time pressure, satisfying land needs, avoiding conflicts, receiving inadequate support from qualified agricul- tural staff, and encountering failures in the handing over of responsibili- ties from the Volta River Authority to assigned local support agencies. The evaluation team concluded that the problems might have been cor- rected without delay under more stable macroeconomic conditions. But even though Bank supervisory missions identified resettlement prob- lems and proposed solutions at the time, the country authorities respon- sible for actual implementation did nothing. In addition: Financial management of resettlement caused avoidable prob- lems and conflicts despite established policies and guidelines. Overall resettlement cost accounting was poor. Valuations and land compensation payments were treated as matters to be dealt with after resettlement. And valuations for compensation were not done, which meant that compensation for land taken for re- settlement was not paid, causing opposition to resettlement. Ac- cording to the evaluators: The overriding lesson from Kpong's resettlement program is that more attention must be paid to plan implementation. There were generally sound policies, ac- ceptable plans, a suitable organization, and a seasoned staff 104 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT with very relevant resettlement experience. Yet the program faced avoidable [economic and financial] problems which were not adequately solved. SHUIKOU HYDROELECTRIC I AND II. In the Shuikou Project in China (phases I and II), several agencies shared responsibilities for engineer- ing and construction design, implementation, and supervision. The Fujian Provincial Electric Power Bureau was to be the project's owner, in accordance with its program of economic reforms and to ensure that large hydroelectric projects like Shuikou are constructed as efficiently as possible. The Ministry of Water Resources and Electric Power set up a special organization, the Shuikou Hydroelectric Project Construction Corporation, to organize project construction and manage the main civil works contract. And the East China Hydroelectric Power Investigation and Design Institute was responsible for project design. Implementation of the resettlement plan was the responsibility of the government of the province of Fujian, which in April 1986 set up the Resettlement Office for the Shuikou Project, with separate divisions for handling the resettlement of state-owned enterprises and the resettle- ment of agricultural households. The Resettlement Office was inde- pendent of the Shuikou Hydroelectric Project Construction Corporation but coordinated its activities with it. It was to have one consolidated budget covering the entire costs of resettlement and was to be assisted by the line bureaus of the provincial and county governments. Resettle- ment suboffices were also established under each county government within the province. The Resettlement Office was responsible for regu- lar monitoring and reporting on resettlement progress and on progress made toward restoring the productive base and welfare of the dislo- cated population, and for the midterm review of resettlement impacts and other studies. The implementing agencies for the Shuikou resettlement were oper- ating in a policy framework that guided resettlement planning and im- plementation. This policy framework was formulated in 1982 as the State Council's Statute for Land Requisition for Construction Use by the State. The basic principles of this country policy are so clearly stated and so relevant to all the other involuntary resettlement projects surveyed that they are repeated here in full: The basic principles of this policy aim at enabling the affected population to become economically and socially re-established on a sound productive basis, to restore their communities, to at- tain at least the same standard of living as prior to relocation, and to fully reconstruct and develop the affected means of pro- COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 105 duction. The inhabitants of an inundated village are to be reset- tled in the same county, and within the same township, if possible, on higher ground at the nearest adjacent location. Re- location at new sites will aim at preserving the integrity of the original social units, village communities, neighborhood groups, and extended families. Resettled persons will be as- sisted to resume, wherever possible, their original occupations, or to get training for off-farm employment in new locations. Compensation will be paid for lost assets and dwellings at real- istic levels (recently increased by governmental decision) and public facilities will be reconstructed. These explicit policy principles guided a systematic planning effort for the resettlement carried out by an interdepartmental task force of the government of the province of Fujian as part of project preparation. This resulted in a full-fledged resettlement and reconstruction plan, based primarily on agricultural development and on local industries, that was included as a component of the project. After having substantially helped design the resettlement strategy at appraisal, Bank social specialists carried out full reviews of the Shuikou resettlement process in April and July 1991. In both cases, the assessment of resettlement was positive, but both missions expressed concern about the approach to monitoring and reporting physical prog- ress and about the possibility of individual hardship because of the time lag between reservoir impoundment and new systems (such as fruit trees and fishing) reaching full production. By 1992, however, these con- cerns had been either addressed or overtaken by events, and physical re- settlement activities were completed on schedule. The authorities also took two additional steps to strengthen the resettlement program: (a) they launched an independent evaluation of the resettlement on a sample of 15 to 20 percent of those resettled to provide timely warning of problems in reestablishing incomes and general welfare; and (b) they provided substantial additional financing by establishing two funds to ensure that the local government could continue to help resettlers fully reestablish their standard of living after external project support ended. CONCLUSIONS. The review of these individual project experiences led to the following conclusions: The primary implementing agencies for all the involuntary reset- tlement projects surveyed had mandates oriented almost exclu- sively toward installing major infrastructure for power generation 106 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT or irrigation. This lack of focus on resettlement in the context of de- velopment projects is not unusual (see Pearce in this volume). The implementing agencies initially had little internal staff capac- ity for, or institutional interest in, handling the resettlement prob- lems caused by their construction activities. This problem was aggravated by the lack of a well-articulated national policy on in- voluntary resettlement (Brazil), by a failure to implement the gen- eral policies that had been developed (India); or by the absence of the macroeconomic and administrative environments within which successful resettlement could be carried out, despite a good initial planning effort (Ghana). . Some of the implementing agencies demonstrated willingness to learn from their previously unsatisfactory experiences with reset- tlement to improve their performance in subsequent projects. For instance, in Ghana, the Volta River Authority drew upon its experi- ences with the Akosombo project to improve preparation of the Kpong project; the National Electricity Authority in Indonesia im- proved its performance in Cirata by drawing on prior experiences with the Saguling project; and even the Companhia Hidro Eletrica do Sao Francisco in Brazil, after a disastrous performance with re- settlement in Sobradinho, improved its handling of resettlement activities significantly in later projects (World Bank 1993a, pp. 15-16). * The Chinese experience at Shuikou was the only involuntary reset- tlement surveyed that succeeded in putting together a complete and effective implementation program for displaced families. This program was executed through local agencies and based on a well-articulated national policy for involuntary resettlement. Resettlers' Participation in Decisionmaking The surveyed documentation provides little evidence that the involun- tary resettlement projects in Brazil, India, and Indonesia made signifi- cant efforts to engage those scheduled for resettlement in the decisionmaking that shaped the their options or modified procedures. In retrospect, project planners' erroneous judgments about how fami- lies destined for resettlement would react to the options presented to them demonstrate a pattern of poor communication between planners and resettlers and of planners' underestimation of the importance of participatory approaches. The involvement of representatives of populations to be displaced in the Shuikou projects in China appears to be, in our study, the only fully successful attempt by a government to set up a genuinely COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 107 interactive process between project management and those affected by displacement. The key to this success appears to be that management of the resettlement component was independent of the infrastructure de- velopment process and decentralized down to the county level, where local government officials had better knowledge of and were more sym- pathetic towards the people being displaced than officials at higher lev- els of government. Accuracy of Cost Projections Given the lack of a uniform Bank format for enumerating the projected costs of resettlement at appraisal, it was impossible to compare actual resettlement costs as reported in PCRs with the costs projected at ap- praisal on a line item basis. Nevertheless, the limited ex ante and ex post cost information available for the projects surveyed indicates that cost projections were not particularly accurate. As table 3.2 shows, in the Paulo Afonso IV Hydroelectric Project in Brazil, planners underestimated resettlement costs by about $20,000 per family, an error of 165 percent. Flawed planning for involuntary re- location, combined with poor estimates of inflationary trends, com- pletely invalidated the initial cost projections, whereas costprojections for the voluntary resettlement projects in Brazil, which were imple- mented at roughly the same time, diverged much less from actual costs. In fact, it is most likely erroneous to assume that the two types of resettlement can be correctly planned with the same approach, as their very socioeconomic nature, requirements, and impacts are deeply different. The principal reason for actual costs being less than the original cost estimates for the voluntary resettlement components of the Transmigra- tion III and IV projects was the premature termination of the program. In the Transmigration I and II projects underestimates apparently oc- curred because these projects served fewer families than anticipated, but per capita costs were higher than estimated. In Ghana, officials involved in the Kpong Hydroelectric Project ini- tially estimated that US$10,439 per family would be needed to resettle 1,140 families, but they actually spent the equivalent of US$4,957 per family to resettle 1,089 families. In this instance, however, the largest part of the difference in costs was attributable to planners' inability to predict inflation and the effects of devaluation of the cedi in the context of a deteriorating macroeconomic situation. In the Gujarat Irrigation Projects in India, cost projections turned out to be inaccurate for two reasons. First, they laid out only the land ac- quisition costs for the project's resettlement activity, without the 1 08 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT additio-nal support needed for the resettlers. And second, the Indian planners redesigned and enlarged the project almost immediately after the Bank appraisal had been completed and the initial credits negoti- ated. In the case of the eight voluntary resettlement or agricultural devel- opment projects that could be evaluated, the percentage error in the cost projections ranged from an underestimate of some 17 percent to an over- estimnate of approximately 40 percent. As table 3.2 shows, cost projections for the projects surveyed are ap- proximately evenly split between those appraisals that underestimated costs and those that overestimated them. The reasons for inaccurate cost estimates varied from project to project, but one common reason was poor correspondence between the projected and actual number of peo- ple served by the projects' resettlement activities, In most cases, pro- jects provided resettlement services to fewer displaced people than anticipated. Inclusion of Transition Costs and Grants The projects entailing involuntary resettlement surveyed in Brazil, In- dia, and Indonesia were initiated in the 1970s and early 1980s. In all cases, cost estimates for resettlement were aggregated under one or two categories: land acquisition and preparation costs and the costs of reset- tlement. There were no line item breakdowns to indicate if any funds were allocated to support displaced persons in transition. In addition, the breakdown between loan and grant funding for resettlement activi- ties was not clearly indicated. In the Kpong Hydroelectric Project in Ghana, the appraisal report broke down resettlement costs into categories that included land pur- chase, land clearing, dike construction, housing, schools, water and sewerage, moving expenses, administrative expenses, and contingen- cies. Unfortunately, subsequent evaluation reports failed to enumerate actual resettlement costs using the same categories. In contrast to the involuntary resettlement operations, several of the voluntary resettlements did enumerate grant funding to migrating fam- ilies for temporary transition costs. The transition support was usually provided in kind as access to cereals and other foodstuffs when cropping systems were being established or as subsidies for special credit arrangements to establish crops or purchase equipment. In these instances, transition support was viewed as part of the income regenera- tion process. COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 109 Agricultural Characteristics of the Projects The set of 12 tables (Tables 3A.1 to 3A.12) consolidated into Annex 1 of the present study compares the principal characteristics of the resettle- ment projects surveyed, variable by variable. For the most part, the spe- cific characteristics reviewed were drawn from Kinsey and Binswanger (1993). The objective was to compare the structuring of involuntary and voluntary resettlement programs within each country and each pro- ject's structure against the variables or characteristics Kinsey and Binswanger identify as being important to the success of resettlement programs. Analysis of the Resource Base For projects involving agriculture, in-depth knowledge of the physical environment is essential for decisions on: • The appropriate agricultural and livestock enterprises to be pro- moted e The types and levels of agricultural inputs necessary to establish and sustain those enterprises over time e The time necessary to establish those enterprises and bring them to a mature and stable state * The levels of income generation possible from those enterprises and the risks involved. Unlike urban resettlement efforts, resettlement aimed at restoring the incomes of farm families must deal with climatic and biological fac- tors as an integral part of project design and implementation. This is par- ticularly true when resettlement occurs in frontier areas on marginal soils, as was the case with most of the resettlement programs reviewed. In these situations, project planners must attempt to collect all relevant data on the resource base in the project area, assess the agronomic impli- cations for crop and livestock enterprises, and, where major knowledge gaps exist, take the time and provide the resources needed to fill in the gaps. None of the involuntary resettlement efforts reviewed benefited from comprehensive analyses of the resources available to displaced families for agricultural enterprises. In Brazil and Indonesia, the two power authorities essentially avoided responsibility for any resource analysis related to agriculture by contracting out the agricultural 110 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT resettlement components of the projects to other agencies. Conse- quently, examination of the natural resource bases in the resettlement areas and the implications for agricultural enterprises did not take place. In the Indian and Chinese involuntary resettlement projects, the re- source analyses were fairly general and not site-specific, but at least some attempt was made to link the proposed farming systems to the re- source bases underlying them. In both cases, it was significant that the area development activities were carried over two project phases, ex- tending the total period available to install the resettled populations. In Ghana, planners designed the siting of new resettlement villages around four basic objectives: (a) to retain villagers' rural, riverine life style, while providing the means for self-development; (b) to supply sufficiently good farmland and water in proximity to the resettlement areas; (c) to maintain health, especially with respect to bilharzia; and (d) to respect land tenure rights and traditional, tribal, and clan group- ings. No attempt was made to change the existing traditional farming systems as part of the resettlement program. Consequently, resource conditions were not analyzed in relation to any specific income restora- tion objective. A Bank consultant visited Ghana to discuss public health aspects related to construction of the Kpong head pond, but no agricul- tural consultants were supplied to help assess the agricultural condi- tions facing settlers displaced by the construction. In contrast to the involuntary resettlement activities, all the volun- tary resettlement projects were planned with a central focus on agricul- ture and rural development. The appraisal reports for these projects contain much more detailed and location-specific discussions of re- source conditions and relate them directly to the agricultural enterprises to be promoted and their income-generating potential. In many cases, the ex ante projections of incomes to be generated from crop and live- stock enterprises proved to be optimistic, but the initial efforts by project planners at least provided points of departure for monitoring and ana- lyzing actual income generation during project implementation. Appraising Sources of Household Income For all the projects reviewed, preproject analyses of the income sources and flows of the families to be resettled, to the extent they existed at all, were poorly done and were rarely based on current field survey data. For the involuntary resettlement programs, displaced families were of- ten classified as rural or urban, as if these labels constituted mutually ex- clusive categories. Such imprecise labeling led to assumptions in planning that all members of rural families were available full-time for COMPARING VOLUNTARY AN-D INVOLUNTARY RESETTLEMENT 111 work in agricultural enterprises and that the incomes of rural families were derived entirely from agriculture. Experience gained from agricultural projects around the world casts doubt on such facile planning assumptions. A better approach for planners would have been to assume that most so-called farm families in developing countries follow a diversified income strategy aimed at ensuring family survival. To define realistic strategies for income regen- eration, planners must develop much more refined profiles of the differ- ent subgroups within any population to be relocated, so that they can better tailor the resettlement options to the affected individuals. This re- quires that resources be made available for high-quality field survey work to set baseline parameters for planning diversified income regen- eration strategies as an integral part of any resettlement effort. Simultaneously with establishing baseline parameters, any income regeneration strategy requires that serious efforts be made to analyze and project family income flows from all employment sources available in the designated resettlement area, and not just from the agricultural enterprises being promoted. While the voluntary resettlement projects generally did abetter job of projecting potential income flows from agri- cultural enterprises, the voluntary resettlement projects in China and Indonesia were the only ones to deal, in even general terms, with nonag- ricultural and off-farm activities. Participant Selection Criteria Although information on participant selection criteria for the voluntary resettlement projects is scant, project documents indicate that most of these projects did have the latitude to select applicants for the available resettlement openings. While the Transmigration Program in Indonesia and the Rajasthan irrigation schemes in India used the most selection criteria, essentially all the voluntary resettlement programs exerted some control in shaping the composition of the groups relocated. While the relevance of some of the selection criteria used as indicators of family success in establishing viable agricultural operations is questionable, the criteria used in the voluntary resettlement programs were generally more detailed than those used in the involuntary resettlements, provid- ing a better basis for eliminating unsuitable candidates. In the involuntary resettlement operations, some of those with legit- imate loss claims were unsuitable by background, skill level, or disposi- tion for resettlement options based on agricultural activities. The problems were compounded because the projects presented the dis- placed families with a narrow range of nonagricultural resettlement op- tions, and thereby reduced the potential for channeling candidates less 112 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT suitable for agricultural enterprises into other types of employment. This indicates that planners of involuntary resettlement operations should pay more attention to the range of resettlement options pre- sented to displaced families, the array and intensity of support services provided to nascent agricultural enterprises, and the time requirements for reestablishing family incomes following displacement. Range of Income-Generating Options For participants in the voluntary resettlement programs, the process of evaluating family employment options essentially took place outside the context of the particular resettlement project. Candidates submitted themselves to a project's selection process and thereby demonstrated a positive commitment to the proffered agricultural opportunities. In other words, they self-selected themselves for the opportunities offered, having weighed them against available alternatives. The involuntarily displaced families had no choice. They were des- tined to suffer losses and had to devise alternative economic survival strategies based in most instances on a narrow range of options, usually cash compensation and outmigration; entry into an agricultural resettle- ment scheme; or, occasionally, participation in training programs for nonagricultural employment. Only in the Shuikou projects in China did project planners attempt to tailor the assistance provided to reestablish- ing families in the full range of income-generating activities they had employed before displacement. In all the other involuntary resettlement projects, those displaced were asked to fit in with the narrow range of options provided through the project. While this approach may have been the most convenient one for project planners, it was not necessarily the most appropriate way to restore family income flows expeditiously. Farm Models Envisaged For all projects where at least one farm model was defined, planning fa- vored the development of smallholder farm operations based primarily on family labor. In the Red Soils Area Development Project in China, col- lective farming was offered as an alternative in one county, but the PCR subsequently reported that it had been unsuccessful. In the Transmigra- tion Program in Indonesia, some of those resettled as smallholder farm- ers participated in contract farming agreements for rubber production. In the involuntary resettlement project components, with the excep- tion of the Kpong Hydroelectric Project in Ghana, the farm models im- plemented involved major changes from the farming systems previously practiced by those displaced-from lowland and riverbank COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 113 farming to hillside systems in Brazil and China, from dryland to irri- gated crop production in India, and from highly intensive crop produc- tion on Java to less intensive crop production on other Indonesian islands. At appraisal, planners of the involuntary resettlement projects in Brazil and Indonesia did not present any specific farm models as the bases for resettlement planning, and in China and India they stipulated single generalized farm models. In Ghana, however, planners did not attempt to change the preexisting traditional farming system when moving displaced populations to new resettlement villages. By contrast, with the exception of the Rajasthan irrigation scheme in India, the voluntary resettlement projects all evaluated two or more dif- ferent farm models. In Brazil, for example, two of the voluntary resettle- ment projects promoted five and seven farm models, respectively, compared to none for the involuntary resettlement at Sobradinho. In ad- dition, the farm models promoted were not only far more detailed at appraisal, they were also more intensively monitored during imple- mentation, and in some cases were modified in the follow-on pro- jects. This was particularly evident in the Transmigration Program in Indonesia and the Northwest Regional Development Program in Brazil. Allocation of Land Holdings: Size, Criteria, and Land Tenure In Brazil, Ghana, and India, planners for neither the involuntary nor the voluntary resettlements attempted to tailor the size of landholdings al- located to the labor or capital endowments of recipient families. The se- lection criteria used for the New Settlements Project in Brazil did, however, favor the recruitment of "large, physically able" families for the 25- to 40-hectare farms allocated. In the Transmigration Program in Indonesia, farm sizes were fixed at 3.5 or 5.0 hectares depending on plot location and the farm model be- ing promoted, and selection criteria were designed to recruit nuclear families with no more than five members and with no members older than 60. Only the projects in China embodied a deliberate policy to allocate land commensurate with capacities-and in the Shuikou projects, the previous occupations-of the families being recruited or displaced. The projects also restricted the size of land allotments so as to foster inten- sive input use and crop production. The objective of this latter policy was to facilitate farm families' ability to incorporate crop residues and organic fertilizers into the soil of their small plots and thereby build up the land's productive capacity more rapidly. 114 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Measures that guarantee long-term access to land through owner- ship or lease rights introduce a major element of stability into develop- ing agricultural systems. In the projects reviewed, documents for the involuntary resettlement projects did not provide clear statements of what land tenure systems were to be put in place, whereas the appraisal reports for the voluntary resettlement projects often contained detailed plans for cadastral surveys and land titling processes. Unfortunately, most of the projects suffered from slow land titling processes, often issu- ing land titles to only a small percentage of new settlers during the life of the project. Where the lack of an effective land titling process directly af- fected farmers' access to agricultural credit services as, for example, in Brazil, this deficiency had doubly stressful consequences for families in the midst of a difficult transition process. Types of Infrastructure Provided The level of infrastructure provided varied significantly both among the involuntary resettlement programs and between the voluntary and in- voluntary resettlements. In Brazil under the Paulo Afonso IV Hydro- electric Project, INCRA was to provide basic infrastructure for displaced families at its distant resettlement scheme, but poor knowledge of the families' intentions led to expenditures on houses and other infrastruc- ture that were never used because families refused to move to the scheme. Faced with unanticipated numbers of families remaining in the vicinity of the reservoir, the World Bank eventually persuaded the Bra- zilian power authority to provide a range of facilities-houses, schools, health facilities, and roads-for displaced families at the reservoir. In Indonesia, some of the families subjected to involuntary resettle- ment were absorbed into the Transmigration Program and had access to the wide range of facilities provided under that program. These in- cluded housing, all-weather roads, wells, village health centers, and community centers. However, when displaced families in Indonesia had the option to take the transmigration route for their resettlement, only a fraction agreed to do so, while many preferred to not move far from their area of origin. No infrastructure appears to have been constructed specifically for the families displaced by the installation of irrigation systems in the Gujarat projects in India. By contrast, the Rajasthan voluntary resettle- ment projects provided for a wide range of infrastructure to be installed in direct support of the agricultural development process. This is, once again, a clear instance of inadequate and unjustified differences in plan- ning approaches and resource allocation. In Ghana, the development of new villages for those displaced by the Kpong Hydroelectric Project COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 115 centered mainly on social infrastructure, housing, schools, and water and sewerage systems, and construction of dikes for irrigation and fish- ing. Infrastructure developed under the Volta Region Agricultural De- velopment Project concentrated on providing area centers to supply agricultural inputs, power and water supply systems, agricultural re- search and extension facilities, and health and education facilities. Finally, planning for the Shuikou Hydroelectric and Red Soils Area Development in China was equally comprehensive with respect to basic facilities provided: roads, water supply systems, and market centers. The Red Soils Project, however, provided a considerably broader array of infrastructure in support of agricultural activities, including ware- houses, sprinkler irrigation systems, and agricultural research and ex- tension facilities. Overall, planners of the voluntary resettlement projects did a better job of detailing and constructing infrastructure that was directly rele- vant to the agricultural development processes envisaged, whereas planners of the involuntary resettlement projects tended to emphasize the replacement of social infrastructure (schools, health facilities, com- munity centers) that families and villages lost as a result of displace- ment. Sources of Resettlement Services The documents for the voluntary resettlement programs provide more details about the importance of these services to the success of the over- all agricultural development process. However, for most of the projects reviewed, the real problem with services for settler families was not with their specification during project preparation and at appraisal, but with their actual delivery during project implementation. The PCRs for many of these projects reiterate the same basic deficiencies in project im- plementation: inadequate support for agricultural research to develop appropriate technology packages for the technical problems of frontier areas, inadequate extension coverage, problems with the supply of agri- cultural inputs, and poor access to agricultural credit for farmers. Degree of Risk Assessment Discussions of risk in the appraisal reports for the involuntary resettle- ment projects were in every case wholly devoted to the investment risks to the Bank and other donors. Despite the literature on the risks and fail- ures of involuntary resettlement, there was no indication that project planners had analyzed the comparative risks to the displaced families or that they had tailored any elements of the resettlement plans to mitigate 116 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT against a perceived risk to the displaced families in the income restora- tion process. At the time when the projects under review were ap- praised, the Bank had not yet developed the impoverishment risks framework which explains how forced displacement can inflict, rather than reduce, poverty (Cemea 1990, World Bank 1994). Plans also failed to recognize the risk to resettled families from in- sufficient time allotted to their recovery of livelihood. The period of as- sistance was limited to the time span necessary to complete the project's civil works components, with no account taken of the time needed to es- tablish agricultural enterprises and stabilize the income restoration pro- cess. In most of the involuntary resettlement components, even the advanced farm families were still establishing their new agricultural op- erations when project support officially ended. Many of those further behind simply gave up on the process. In hindsight, given that the pro- jects were directed at establishing displaced families in frontier areas with marginal soils, the income restoration processes should have been planned in two phases: a system establishment phase of at least four to six years and a system stabilization phase of another five years. Planners of the voluntary resettlement projects produced prepara- tion and appraisal reports that discussed the risks to farmers of initiat- ing agricultural enterprises on soils of uncertain productive capacity and the risks to the attainment of project objectives from inadequate provision of key support services for those farmers. The project planners also attempted to counteract specific risks through deliberately tailored components of the agricultural package. Such counter-risk measures in- cluded supporting agricultural research to produce technological inno- vations adapted to the on-farm environment, undertaking processing and marketing studies to identify appropriate crops, constructing roads to facilitate input supply and crop marketing, and so on. Conclusions and Recommendations This comparison of pairs of voluntary and involuntary projects showed that even though involuntary resettlement is a more difficult undertak- ing than voluntary resettlement, it consistently received less attention from project planners during conventional planning. The exceptions were only the China projects. Not only were voluntary resettlements more carefully prepared, but more financial and physical resources were devoted to their implementation, whereas given the nature of in- voluntary resettlement the reverse should be the case. More financial support is needed in involuntary resettlement because in such situa- tions all families must leave, including those less equipped for the risks and hardships of relocation; no selection to recruit those families most COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 117 likely to succeed in the new environment is possible; forcibly removed families lose their former assets, while voluntary resettlers can maintain them, leave them to kin, or sell them into strong markets; and return mi- gration is not an option for involuntary displacees if the resettlement does not work. Thus, involuntary resettlement should receive more at- tention to ensure adequate planning. The planning of the voluntary resettlements incorporated input from a range of experts, including experts on rural development and ag- riculturists; prepared detailed programs to help farmers adapt to new physical and socioeconomic environments; and internalized resettle- ment costs within project budgets. Planners of involuntary resettlement components of projects treated them as an unfortunate side effect of the main (usually infrastructure) project, with poorly defined costs and no defined benefits. Table 3.5 summarizes several traditional differences in the planning and appraisal of the voluntary and involuntary resettle- ment operations analyzed in this study. The success of involuntary resettlement in terms of restoring settlers' incomes seems to largely depend on planning and implementation. In the early projects we analyzed, the implementing agencies lacked the man- date, resources, commitment, and capacity for an undertaking of this na- ture, and those to be resettled did not participate in making the decisions that would affect their lives-perhaps their very survival. China appears to have been an exception. Project planners gave equal attention and priority to voluntary and involuntary resettlement, handling involuntary resettlement as a genuine opportunity for devel- opment rather than as an additional burden caused by the project activ- ity triggering resettlement. In light of the preceding analysis, several recommendations are pre- sented here, divided between those directed at the programmatic as- pects of resettlement and those directed at the agricultural aspects. With respect to the programmatic aspects of resettlement project, we believe that project planners should: * Provide the resources and take the time during project preparation to define in detail the populations likely to be displaced by infra- structure construction. Base the definition of these populations on data collected from on-site surveys, not on secondary, indirect, or obsolete information. * Define an income restoration objective for each displaced popula- tion in precise terms and then design specific project components to accomplish that objective. . Ensure that any project with a land-based resettlement compo- nent be designed, negotiated, and supervised by a multidiscip- 118 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Table 3.5. A Comparison of the Design and Management of Involuntary and Voluntary Resettlement Aspect of project Involuntary resettlement Voluntary resettlement World Bank Concentration on civil Diverse mix of specialists staff input works for main project used to review technical, component and none of social, and economic Bank's rural development soundness of proposed expertise brought to bear. farm models. Managerial On the infrastructure On resettlement as the focus building component of the core of an agricultural de- project. velopment project in which the new farm's via- bility was critical to the project's overall success. Assessment of No detailed, systematic Resource conditions faced income gener- analysis of resources avail- by farmers related to real- ating potential able to displaced families. istic appraisals of income of new sites generating potential of en- terprises being promoted. Farm models Either single, generalized Several farm models pre- farm model offered or no sented; help given to set- model. tlers to develop off-farm sources of income. Risk Risk assessment restricted Risks to farmers initiating to investment risk faced agricultural enterprises in by donors; risks to dis- new, unfamiliar areas are placed and relocated farm- assessed along with risks ers not analyzed. from uncertain delivery of support, and backup plans prepared. Duration Provisions limnited to time Provisions made to sup- needed to complete main port settlers until full farm physical infrastructure, with development. insufficient assistance pro- vided to ensure stabilization of new farm enterprises. linary team, including competent agricultural and social science specialists. * Treat involuntary resettlement activities as important independent and parallel components of infrastructure projects, not as inconve- nient and subordinate appendages to them. * Adopt realistic time horizons to accomplish the defined income restoration objectives. Break down the resettlement period into a COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 119 system establishment phase of about four to six years and a system stabilization phase estimated to be at least five years. * Offer the widest possible range of compensation and resettle- ment options keyed to the specific needs of the displaced popu- lations and always include some cash payment as part of the option package. . Encourage decentralized management of resettlement activities through established local institutions, including nongovern- mental agencies and local citizens' advocacy groups, backed by established government institutions and resources. Avoid using centralized bureaucracies, such as power authorities, irrigation de- partments, or ministries of agriculture, as the implementing agency for resettlement activities, as they are ill-equipped to deal with those tasks. Also avoid ad hoc institutional structures set up to deal with only the short-term exigencies of relocation and not with the longer-term issues of rehabilitation. • Introduce a standard format for enumerating the full costs and po- tential benefits of any involuntary resettlement activity and insist that these costs and benefits be internalized in all ex ante and ex post financial and economic analyses. * Introduce a standard format for enumerating both donor and re- cipient country staff and consultant inputs used in project preparation, appraisal, negotiations, and supervision by staff months and skills represented, so that staff input can be analyzed and compared. With respect to the agricultural aspects of resettlement, our study suggests that project planners should: * Strengthen site-specific analysis of the resources available to reset- tled families and use appropriate agricultural skill specialists in pro- ject planning and implementation so that the resource analysis can be directly linked to defined agricultural development processes. This would provide a realistic basis for designing project elements that directly address the resource constraints facing settler families and provide a more focused agenda for project monitoring. * Fund detailed analysis of the preproject income flows and sources of farm families to be displaced to provide better baseline defini- tions of the resettlement populations and to set realistic points of departure for planning and monitoring the income restoration pro- cesses. * Set participant selection criteria for the agricultural options pro- vided so as to encourage outmigration from agriculture for fami- 120 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT lies with low prospects of succeeding in the income restoration process. • Broaden the range of resettlement options offered to encourage em- ployment in agricultural support enterprises, such as crop process- ing and marketing, transportation, and input supply. * Encourage experimentation with a range of farm models to better accommodate the varying capacities and capital endowments of families to be resettled and the different natural resource situa- tions. * Tailor land allocations for displaced families to both the physical characteristics of the available resource base and the endowments of individual families, namely, the size of the family's labor force for agricultural work, capacity to invest, and so on. • Ensure greater stability in the income restoration process by pro- viding the legal arrangements needed for rapid land titling or other guarantees of long-term land tenure for displaced families. * Concentrate project resources on providing infrastructure directly related to fostering and sustaining the agricultural income restora- tion process. Developing roads from farms to market centers, rural water supplies, crop processing and marketing facilities, and local agricultural research and extension stations are likely to have a greater impact on livelihood restoration when linked with install- ing a broad array of social services. * Encourage better analysis of and funding for agricultural support services directly relevant to the income restoration process. . Resist attempts to restrict displaced families' use of on-farm re- sources. • Introduce a requirement that project preparation and appraisal documents include comprehensive economic and agricultural risk and constraint analyses, considered from the pragmatic perspec- tive of the resettled farming population. COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 121 APPENDIX Characteristics of the Agriculturally Based Resettlement Projects Surveyed The set of tables that follow provides researchers, practitioners and other interested readers with additional information collected during this comparative research. The characteristics described in these tables for the pairs of compared projects refer to: Table 3A.1. Resource Base Table 3A.2. Household Income Table 3A.3. Type of Resettlement Undertaken Table 3A.4. Range of Resettlement Options Table 3A.5. Type of Farm Model Envisaged Table 3A.6. Size of Land Holding Allocated Table 3A.7. Criteria for Land Allocation Table 3A.8. Conditions of Land Tenure Table 3A.9. Type of Infrastructure Provided Table 3A.10. Source of Resettlement Services Table 3A.11. Type of Risk Assessment Table 3A.12. Restriction Placed on Settlers 122 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Table 3A.1. Resource Base Country/project Analysis of the resource base Brazil Paulo Afonso IV Perfunctory analysis of the geologic, climatic, vegeta- tive, and soil conditions in the vicinity of the reservoir. Information not correlated with productive capacity for income regeneration. Agricultural Devel- Good analyses of land capability and present use, land opment and Envi- tenure conditions, social infrastructure, baseline eco- ronmental logical conditions; research conducted on specific crops Protection and their potential. Good data and discussion on socio-economic conditions faced by settlers and the constraints on various farming systems. Mato Grosso Rural Good set of resource base analyses very similar to those Development in the Agricultural Development and Environmental Protection Project. New Settlements The project is the Phase III follow-on activity to the Ag- ricultural Development and Environmental Protection Project and depends to a large extent on information and data gathered under the Northwest Regional De- velopment Program. Indonesia Thirteenth Power Agricultural resettlement was managed by the national Transmigration Program at sites on islands other than Java. No analysis of resource base and its relationship to income regeneration. Transmigration I The four Transmigration Projects were phased over Transmigration II time but carried out essentially the same activities. Transmigration III Good general descriptions of each resettlement area. Transmigration IV Agricultural research conducted crop trials at various project sites to determine appropriate farming tech- niques and income potential. India Gujarat Irrigation Most information on the resource base is general de- scription at the state, not project, level. No specific dis- cussion of resources nor of an income regeneration process for displaced persons. Gujarat Irrigation Resource analysis similar to Gujarat Irrigation. Most re- II source information related to construction aspects of project, not to agriculture or income regeneration for displaced families. COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 123 Table 3A.1. (continued) Country/project Analysis of the resource base India (continued) Rajasthan Canal Good analysis of soil and water resources in the project Command Area area. Analysis directly relates resources to agricultural Development potential. Research results and long-term crop produc- tion figures used to assess income generation potential. Rajasthan Com- This project is a follow-on of the Rajasthan Canal Com- mand Area and mand Area Development Project. Resource analyses Settlement were extended to new areas to be developed. China Shuikou Hydro- Limited discussions of area resources related primarily electric to construction activities, not the resettlement program. Some discussion of capabilities of hillside lands for crop production, fruit tree production, establishment of economic trees, and permanent forests. Second Shuikou The Second Shuikou Project is a follow-on effort. The Hydroelectric appraisal document does not provide any significant additional information on the resource base for resettle- ment. It does report that the resettlement program was proceeding very satisfactorily. Red Soils Area Good comprehensive analysis of the key resource con- Development ditions in the project area-climate, topography, soils and land use-and the techniques needed to bring the area into viable agricultural production. Ghana Kpong Hydro- Essentially no analysis of the resource base as it related electric to crop, livestock, or fishery enterprises. Volta Region Agri- Good analyses of ecology and climate, soils, infrastruc- cultural Develop- ture and social services, socio-economic situation, land ment use, farm structure, and crop production. 124 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Table 3A.2. Household Income Coutntry/project Appraisal of hotusehold income Brazil Paulo Afonso IV Very general analysis of income groups in poorly de- fined rural population. General description of existing production systems. Agricultural Devel- Limited discussion of preproject household incomes; opment and Envi- good analysis of income generation potentials of farm- ronmental ing models being proposed. Protection Mato Grosso Rural Limited discussion of preproject household incomes; Development good analysis of income generation potentials of farm- ing models being proposed. New Settlements No real discussion of preproject settler income sources and flows; calculations of indicative farm family in- come from agricultural enterprises based on generated farm budgets. Indonesia Thirteenth Power Project financed a two-year anthropological, socioeco- nomic, and ecological baseline and impact study (in- cluding censuses and property inventories). No evidence that information from this study was used in the appraisal report. Transmigration I No evidence that settlers income status and/or sources Transmigration II were investigated in detail, except as an element in the Transmigration III candidate selection process. Considerable work was Transmigration IV devoted to determining the viability of the various farm models used and their income generation poten- tials over time. India Gujarat Irrigation General information from state records on general pop- ulation characteristics. Some discussion of per capita income levels with and without the project, but the in- formation appears to come from secondary sources. Gujarat Irrigation General information from state records on general pop- II ulation characteristics. Some discussion of per capita income levels with and without the project. COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 125 Table 3A.2. (continued) Country/project Appraisal of houtsehold income India (continued) Rajasthan Canal Little discussion of preproject income status of settlers, Command Area but very good discussions of financial requirements of Development anticipated cropping and their relationship to credit needs and farmers' ability to pay water charges. Rajasthan Com- Income analysis discusses two groups of settlers under mand Area and two different farm models. It is of similar quality to the Settlement Phase I project. China Shuikou Hydro- Little discussion of the income status of families being electric displaced other than for purposes of detailing compen- sation. No specific calculations are offered on the pro- jected income regeneration process for displaced persons, although there is a detailed budget for total resettlement costs and one calculation of expected costs and revenues for resettlement in one county. Second Shuikou No additional information is provided. Hydroelectric Red Soils Area Little discussion of preproject incomes and sources Development among settlers. Good preliminary analysis of the in- come generation potentials for eight different farm models. Analyses of market demand for certain crops-mandarin oranges and ramie-may have been suspect. Ghana Kpong Hydro- Essentially no analysis of farm family incomes before electric or after resettlement. Volta Region Agri- Analysis of pre and post project farm budgets. Limited cultural Develop- discussion of family incomes focused entirely on agri- ment cultural enterprises. 126 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Table 3A.3. Type of Resettlement Undertaken Country/project Type of resettlement Brazil Paulo Afonso IV Involuntary resettlement with groups classified as ur- ban or rural. Agricultural Devel- Government-assisted voluntary resettlement, some as- opment and Envi- sistance to families who had migrated spontaneously. ronmental Protec- tion Mato Grosso Rural Govemment-assisted voluntary resettlement, some as- Development sistance to families who had migrated spontaneously. New Settlements Govemment-assisted voluntary resettlement for quali- fied new settlers. Indonesia Thirteenth Power Involuntary resettlement. Agricultural component of resettlement involved out-migration from Java to other islands under the Transmigration Program. Transmigration I Three types of voluntary resettlement. Officially spon- Transmigration II sored migrants had their travel paid and were settled Transmigration III on settlement schemes. Spontaneous migrants paid Transmigration IV their own fares, but were settled on a settlement scheme and received some assistance (food, building materials, and so on). Free migrants who paid their own fares and settled wherever they found available land, other than the settlement schemes. India Gujarat Medium Resettlement involved involuntary displacement of Irrigation families from dam construction and reservoir areas. Gujarat Medium Resettlement involved involuntary displacement of Irrigation II families from dam construction and reservoir areas. Rajasthan Canal Project involved voluntary settlement of landless rural Command Area populations, subject to eligibility conditions. Development Rajasthan Com- Project involved voluntary settlement of landless rural mand Area and populations, subject to eligibility conditions. Farm fam- Settlement ilies from other sites within the existing perimeters were also voluntarily resettled in newly developed areas. COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 127 Table 3A.3. (continued) Country/project Type of resettlement China Shuikou Hydro- Project entailed involuntary resettlement of both urban electric and rural families. Rural families were resettled from lowland sites to hillside farms. Second Shuikou Project entailed involuntary resettlement of both urban Hydroelectric and rural families. Rural families were resettled from lowland sites to hillside farms. Red Soils Area Voluntary resettlement in the Red Soils area based Development upon a set of eligibility criteria. Ghana Kpong Hydro- Project entailed involuntary resettlement of rural farm electric families. Volta Region Agri- No resettlement. Project focused on agricultural devel- cultural Develop- opment of pre-existing smallholder farming systems. ment 128 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Table 3A.4. Range of Resettlement Options Country/project Resettlement options Brazil Paulo Afonso IV To move to distant INCRA resettlement scheme; re- ceive cash settlement and move; remain in the vicinity of the reservoir for temporary employment, farming on the uplands, or work in a fisheries scheme. Agricultural Devel- Four farming models offered to qualified settlers based opment and Envi- on coffee, cocoa, livestock, and rubber production. ronmental Fruit tree model was added later. Protection Mato Grosso Rural Seven farm family income models are analyzed based Development on production of bananas, rubber, coffee, livestock pro- duction, and farm sizes (from less than 20 to more than 100 hectares). New Settlements Generalized farm model for farms of less than 50 hect- ares producing a combination of perennial and annual crops, including coffee, rubber, and tropical fruit trees. Indonesia Thirteenth Power No farm models were discussed in the available docu- ments. Presumably, displaced families were eligible to participate in the same farm models that were used in the Transmigration Program. Transmigration I A wide variety of farm models were designed, imple- Transmigration II mented, and evaluated over the total implementation Transmigration III period. Initial model was based on food and perennial Transmigration IV tree crop production on 5-hectare farms. Other models involved development of low-lying swamp areas, up- land, and forest areas for annual and perennial crops and livestock. Models were adjusted over time based on survey data collected on income generation and technical parameters. India Gujarat Irrigation Two generalized farm models outlined in the appraisal report. Both anticipate smallholder family farms grow- ing a variety of crops under irrigation during the kharif and rabi seasons. Farm size varies from 2 hectares in mainland Gujarat to 5 hectares on Saurashtra. Gujarat Irrigation Displaced families were to receive cash compensation II for lost assets and replacement land. Land purchase costs to be paid only after farmer negotiated purchase and presented evidence to govemment agents. COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 129 Table 3A.4. (continued) Country/project Resettlement options India (continued) Rajasthan Canal Smallholder farms of 6.32 hectares allocated to quali- Command Area fied landless families. Farm model is a two-season irri- Development gated system, with production of a range of crops. Rajasthan Com- Smallholder farms of 6.32 hectares allocated to quali- mand Area and fied landless families. Farm model is a two-season irri- Settlement gated system, with production of a range of crops. China Shuikou Hydro- Specific agricultural model was not presented. Dis- electric placed families were compensated for lost assets and resettled on hillside farms. According to land capabil- ity, farm production was to be a combination of annual crops, fruit trees, and forestry products. Second Shuikou Specific agricultural model was not presented. Dis- Hydroelectric placed families were compensated for lost assets and resettled on hillside farms. According to land capabil- ity, farm production was to be a combination of annual crops, fruit trees, and forestry products. Red Soils Area Smallholder individual farms and cooperative farming Development system. Eight farm models were evaluated, with culti- vated areas ranging from 0.6 to 2.1 hectares of cultiva- ble land, plus woodlots. Crops included irrigated rice, tea, oranges, and forest products. Ghana Kpong Hydro- Resettlement options were limited to cash reimburse- electric ment or moving to new resettlement villages. Volta Region Agri- No resettlement activity in this project. cultural Develop- ment 130 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Table 3A.5. Type of Farm Model Envisaged Country/project Type(s) offarm model envisaged Brazil Paulo Afonso IV No detailed farm model was presented in appraisal documents or resettlement plan. Agricultural Devel- Five smallholder farm models were offered to qualified opment and Envi- settlers based on coffee, cocoa, livestock, and rubber ronmental production. Protection Mato Grosso Rural Seven smallholder farm models were analyzed based Development on production of bananas, rubber, coffee, and livestock production. New Settlements One generalized smallholder farm model was dis- cussed in the appraisal report, but three farm models were evaluated in the financial analysis. Two models based on production of annual crops and coffee and one based on annual crops and fruit trees. Indonesia Thirteenth Power No specific farm model was offered in the appraisal document. Since the displaced families entered the Transmigration Program, the farm models described below pertained to this project. Transmigration I A wide variety of farm models were designed, imple- Transmigration II mented, and evaluated. Transmigration III Transmigration IV India Gujarat Irrigation Farm model under the Gujarat schemes is one for Gujarat Irrigation smallholder family holdings of irrigated cropland. II Model is for diversified crop production over the kharif and rabi seasons. Rajasthan Com- Farm model under the Rajasthan schemes was mand Area Devel- smallholder farms of 6.32 hectares of irrigated land. opment Diversified crop rotations during the kharif and rabi Rajasthan Com- seasons. mand Area and Settlement COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 131 Table 3A.5. (continued) Country/project Type(s) offarm model envisaged China Shuikou Hydro- Shuikou resettlement provided for smallholder hillside electric farms producing, according to land capability, a combi- Second Shuikou nation of annual crops, fruit trees, and forestry prod- Hydroelectric ucts. Red Soils Area Two farm models were used in the Red Soils area. One Development was based on individual smallholder farms and the other on cooperative farming with individual house plots. Eight different cropping models. Crops included irrigated rice, tea, oranges, and forest products. Ghana Kpong Hydro- No new farm model was envisaged. Project planners electric sought to maintain pre-existing "traditional" small- holder farming systems at new village sites. Volta Region Agri- There was no resettlement in this project. Interventions cultural Develop- sought to strengthen existing smallholder farming sys- ment tems on a commodity-by-commodity basis. 132 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Table 3A.6. Size of Land Holding Allocated Country/project Size of land holding allocated Brazil Paulo Afonso IV No detailed information is available on the size of land holdings allocated. Agricultural Devel- Information on the size of land holdings granted is opment and Envi- very confusing. Apparently original model was for ronmental standardized 100-hectare farms, but new settlements Protection had no standard sizes. Instead, size was determined on the basis of soil fertility, topography, and availability of farm labor. Mato Grosso Rural Land holdings for farms varied from less than 20 to Development more than 100 hectares. New Settlements Farm sizes variously reported as less than 50 hectares and between 25 and 40 hectares. Indonesia Thirteenth Power If Transmigration farm models applied, land holdings allocated should have been between 3.5 and 5 hectares per family. Transmigration I Depending on the phase of the Transmigration Pro- Transmigration II gram, land holdings allocated per family varied be- Transmigration III tween 3.5 and 5 hectares. Transmigration IV India Gujarat Irrigation Land holdings allocated ranged from 2 to 5 hectares Gujarat Irrigation per family, depending on location. II Rajasthan Com- Farm size invariable at 6.32 hectares of irrigated land. mand Area Devel- opment Rajasthan Com- mand Area and Settlement COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 133 Table 3A.6. (continued) Country/project Size of land holding allocated China Shuikou Hydro- Size of holdings allocated varied according to family electric size, size of previous holdings, and occupation prior to Second Shuikou displacement. Holdings in all cases were equal to or Hydroelectric greater than those lost through displacement. Red Soils Area Size of holdings included cultivated areas ranging from Development 0.6 to 2.1 hectares, plus woodlots. Ghana Kpong Hydro- No information on the size of plots allocated to reset- electric tled families. Under the "traditional" system, plots ap- parently averaged about 2 hectares. Volta Region Agri- No land was allocated under this project. cultural Develop- ment 134 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Table 3A.7. Criteria for Land Allocation Country/project Criteria for land allocation Brazil Paulo Afonso IV Displaced families classified as "rural" and subject to loss of assets in the reservoir area. Agricultural Devel- Settler eligibility criteria were: area of origin, educa- opment and Envi- tional level, and years of previous agricultural experi- ronmental ence. Later experience showed that success was Protection determined more by whether settlers had been farm owners or managers and whether they had had a stable residence in the past. Mato Grosso Rural No comprehensive set of criteria is listed. Emphasis Development was apparently on 'smallholder" farm families, de- fined as families having access to less than 200 hectares of land. New Settlements Settler eligibility criteria were stated as: physically able candidates with large families who have experience in agriculture but are not landowners, and who have not participated in any previous INCRA settlement project. Indonesia Thirteenth Power Eligible for resettlement if the displaced family lost property or use rights in the reservoir area and meets eligibility requirements. Transmigration I Eligibility criteria for the Transmigration Program: In- Transmigration II donesian citizenship; married; aged 20 to 45, with no Transmigration III family member older than 60 years; agricultural experi- Transmigration IV ence and skill; maximum family size of five; nonpregnant wife; good political record; good police record; agreement to follow transmigration regulations; and has never transmigrated before. India Gujarat Irrigation Displaced families are entitled to land allocations at Gujarat Irrigation least equal to their former holdings. If their holdings II were less than 2 hectares, they were to be allocated at least 2 hectares. COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 135 Table 3A.7. (continued) Country/project Criteria for land allocation India (continied) Rajasthan Com- Priorities for land allocation: pre-1955 landless farmers mand Area Devel- (temporary cultivators); post-1955 farmers (temporary opment cultivators); landless farmers belonging to the village Rajasthan Com- in which land is distributed; landless farmers belong- mand Area and ing to the Tehsils in which the land is distributed; land- Settlement less farmers of other Tehsils in the district; and other landless farmers belonging to other districts of Rajasthan. China Shuikou Hydro- Families displaced by Shuikou construction and/or electric reservoir filling and with rural occupations were eligi- Second Shuikou ble to participate in the resettlement and receive new Hydroelectric terraced hillside land. Red Soils Area Farmers were selected by county governments and Development state farms from among applicants who possess requi- site experience and managerial skills, long-term farm residence, and resources to contribute about 30 percent of the on-farm development costs (including land de- velopment, buildings, crop establishrnent, livestock, and farm machinery and equipment) in the form of cash or labor. Ghana Kpong Hydro- No criteria discussed other than family lost land electric and/or use rights due to project construction activities. Volta Region Agri- No land was allocated under this project. cultural Develop- ment 136 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Table 3A.8. Conditions of Land Tenure Country/project Conditions of land tenture Brazil Paulo Afonso IV Most displaced families apparently never received clear titles to the land allocated. Agricultural Devel- INCRA's performance on land titling was slow. A total opment and Envi- of 14,393 titles were actually issued-about 60 percent ronmental of the total projected. Protection Mato Grosso Rural Area was divided into two tenure regularization pro- Development jects in which legitimacy of land claims was investi- gated by INCRA. Squatters found to have legitimate claims were to receive a "license to occupy" their farms. This was an adequate guarantee to obtain long-term credit and was eventually to lead to granting of an outright land title. New Settlements INCRA had acquired all property rights to the areas considered for settlement and corresponding reserve areas. Project funding is provided for mapping, survey- ing, demarcation, and land titling, but no specific ti- tling process is indicated. Indonesia Thirteenth Power Presumably land tenure arrangements were as per the Transmigration Program. Transmigration I Under the Transmigration Act, settlers are entitled to Transmigration II obtain land rights in accordance with prevailing laws. Transmigration III Settlers receive land in a two step process. First, they Transmigration IV are granted land right of use [Hak Pakail, and, after proving their land has been developed and improved, they can be granted full ownership [Hak Milik]. India Gujarat Medium No clear indication of farmer status of land rights. Ap- Irrigation parently they have land use rights and access to a fixed Gujarat Irrigation water rotational schedule. II COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 137 Table 3A.8. (continued) Countrproject Conditions of land tenure India (continued) Rajasthan Com- Exact status of land tenure rights is unclear. Appar- mand Area Devel- ently most farmers receive their holdings as the result opment of a lottery among qualified farmers. Payments are Rajasthan Com- made against long-term tenure guarantees. Over the mand Area and life of the two projects, payments varied from Rs. Settlement 16,777 to Rs. 29,000 payable over 15 years without in- terest. Full title is given at this time. China Shuikou Hydro- No specific indication of the status of land tenure. Ap- electric praisal report discusses allocation of house plots but Second Shuikou not individual ownership of land. Hydroelectric Red Soils Area Use rights for the land were to be granted by the Development county governments or the state farms on a renewable, long-term basis. Ghana Kpong Hydro- roject did not alter traditional land use rights, which electric vary by ethnic group. Volta Region Agri- Project did not alter existing land use patterns, which cultural Develop- vary by ethnic group and location within the project ment area. 138 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Table 3A.9. Type of Infrastructure Provided Country/project Type of infrastructure provided Brazil Paulo Afonso IV Infrastructure provided by INCRA at the distant reset- tlement site is not detailed, except for the report that INCRA built 3,500 houses for anticipated resettlement population, of which 2,000 remain vacant at the end of the project. Specific infrastructure in the vicinity of the reservoir known to include: settler houses, schools, health facilities, and roads. Agricultural Devel- CODARON was to provide access and farm-to-market opment and Envi- roads, rural service centers, schools, health posts, and ronmental warehouse/drying sheds. Social infrastructure in- Protection cluded schools, rural water supply, health and sanita- tion facilities. Mato Grosso Rural Infrastructure very similar to that provided by the pre- Development vious project. New Settlements Infrastructure provided is very similar to the previous project. Indonesia Thirteenth Power No specific infrastructure development for resettled families is discussed in the appraisal report. Transmigration I Fully financed settlers received a house; all-weather Transmigration II main and village roads were built; one hand-dug well Transmigration III was provided for each of five families; and village Transmigration IV health posts and village centers were constructed. India Gujarat Medium There is no reference in the appraisal documents for ei- Irrigation ther of the Gujarat projects of any infra-structure built Gujarat Irrigation specifically for the families displaced by dam construc- II tion and irrigation tanks. The major infrastructure cre- ated is the irrigation network itself. COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 139 Table 3A.9. (continued) Country/project Type of infrastructutre provided India (continued) Rajasthan Com- Other than the irrigation systems, infrastructure in- mand Area Devel- cluded: village water supplies and major and minor opment roads. Afforestation, water channel lining, and land Rajasthan Com- shaping operations were also included in the projects. mand Area and Settlement China Shuikou Hydro- Infrastructure included: roads, water supplies, market electric centers, and replacement of public buildings lost in Second Shuikou the reservoir area. Building materials were provided Hydroelectric to displaced families to reconstruct houses. Red Soils Area Infrastructure included: enlargement of small reser- Development voirs, canal construction, installation of pumping sta- tions and sprinkler irrigation systems, rural electrification, road and bridge construction, housing, schools and clinics, administrative buildings, ware- houses, laboratories and extension stations, and live- stock sheds. Ghana Kpong Hydro- Infrastructure included: settler housing, schools, water electric supply and sewerage systems, irrigation, and fishing dikes. Volta Region Agri- Infrastructure included: village input supply centers, cultural Develop- research/extension facilities, roads, water supply, ment power distribution networks, schools, and health facil- ities. 140 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Table 3A.10. Source of Resettlement Services Country/project Source of resettlement services Brazil Paulo Afonso IV INCRA was to provide support services in the distant resettlement site. Services in the vicinity were assumed from state agricultural extension and health services. Access to these services was apparently spotty or non- existent for most displaced persons. Agricultural Devel- Full range of services, including agricultural research, opment and Envi- extension, credit, supply of improved seeds, input sup- ronmental ply services, crop processing and marketing, and soil Protection survey and mapping were planned Mato Grosso Rural Services provided are very similar to those under the Development previous project. New Settlements Services provided are very similar to those under the previous project. Indonesia Thirteenth Power No discussion of agricultural or social services to be made available to displaced farm families. Transmigration I A broad range of agricultural and social services were Transmigration II to be provided to settlers, including agricultural re- Transmigration III search, extension, input supply, assistance in coopera- Transmigration IV tive organization, and credit. The projects also had complementary health and education services for the settler communities. India Gujarat Irrigation Resettled families would presumably have the same ac- Gujarat Irrigation cess to the agricultural input supply, research, exten- II sion, and credit services as the other farm families in the Gujarat schemes. COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 141 Table 3A10. (continued) Coutntry/project Souirce of resettlement services India (continuted) Rajasthan Com- Settler families would have access to agricultural input mand Area Devel- supply, research, extension, and credit services in the opment Rajasthan schemes. Rajasthan Com- mand Area and Settlement China Shuikou Hydro- Farmers had access to agricultural research and exten- electric sion, credit, tillage, transport, crop processing, and mar- Second Shuikou keting services. Hydroelectric Red Soils Area Farmers had access to agricultural research and exten- Development sion, credit, tillage, transport, crop processing and mar- keting services. Ghana Kpong Hydro- Project documents do not report introduction of any electric new services for the resettled populations. Volta Region Agri- Project strengthened existing agricultural research and cultural Develop- extension services and created village input supply cen- ment ters, nutrition extension, credit, and irrigation develop- ment services. 142 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Table 3A.11. Type of Risk Assessment Couintry/project Type of risk assessment Brazil Paulo Afonso IV Essentially no ex ante risk analysis for the plight of the displaced persons. Any statements of risk are related to the donor's risk in the project. Agricultural Devel- Brief but intelligent qualitative discussion of crop pro- opment and Envi- duction risks related to soils and crop diseases. Also ronmental discusses weaknesses in the extension service and con- Protection sequences for settlers. Mato Grosso Rural Principal production risk discussed is the unknown na- Development ture of the resources in the frontier areas- soils and climate-and the need for effective agricultural re- search to develop effective cultivation and soil conser- vation techniques. New Settlements Risk statements are similar to the previous projects. Some comments on need to have better selection proce- dures for settlers. Indonesia Thirteenth Power All risk statements are directed at the risk to the donor from hydroelectric projects. No statements on risks to displaced persons. Transmigration I Risk statements over the course of the projects build on Transmigration II implementation experiences and discuss mainly the Transmigration III risks associated with poor institutional performance of Transmigration IV some agencies. Discussion of the relative deficiencies of some of the farm models tried in terms of income gen- eration potential. India Gujarat Irrigation There is some discussion of investment risks implicit in Gujarat Irrigation irrigation projects in India, but there is no discussion of II risks to displaced families or farmers. Rajasthan Com- Risk statements discuss relative absence of risks in the mand Area Devel- project. Conclusion is that there are no greater risks in opment Rajasthan than in other irrigation projects in India. No Rajasthan Com- real discussion of risk at the farm level. mand Area and Settlement COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 143 Table 3A.11. (continued) Country/project Type of risk assessment China Shuikou Hydro- No specific risk statements dealing with farm opera- electric tions. Second Shuikou Hydroelectric Red Soils Area No specific risk statements dealing with farm opera- Development tions. Ghana Kpong Hydro- Planners attempted to reduce risks of a poor resettle- electric ment process by studying the mistakes made in the Akosombo project upstream. Volta Region Agri- Risk statements primarily related to need to provide cultural Develop- agricultural support services to farm families. ment 144 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Table 3A.12. Restriction Placed on Settlers Country/project Restriction placed on settlers Brazil Paulo Afonso IV Not defined. Agricultural Develop- Farmers can cultivate up to a maximum of 10 hect- ment and Environ- ares of their land, of which not more than half can mental Protection be in perennial crops. Mato Grosso Rural Not defined. Development New Settlements Not defined. Indonesia Thirteenth Power Not defined. Transmigration I Settlers had to follow instructions on the allocation Transmigration II of land within their farming systems. Of the Transmigration III 5-hectare allocations, one hectare had to be planted Transmigration IV to rubber; 0.2 hectare was to be used for home gar- dens. Before issuance of full ownership title, settler performance was subject to review by government officials. India Gujarat Irrigation No information on any restrictions placed upon re- Gujarat Irrigation II settled families. Rajasthan Command No information on any restrictions placed upon set- Area Development tler families. Rajasthan Command Area and Settlement China Shuikou Hydroelectric No restrictions indicated. Second Shuikou Hydroelectric Red Soils Area No restrictions indicated. Development Ghana Kpong Hydro- Not defined. electric Volta Region Agricul- None. This was not a resettlement project. tural Development COMPARING VOLUNTARY AND INVOLUNTARY RESETTLEMENT 145 Notes 1. The appraisal of the Shuikou project took place in 1986 immediately after the completion of the Bank's first internal review of resettlement operations and benefited much from the conclusions and effects of that review on the Bank's in-house work (see Cernea 1995). 2. This absolute numbers appear relatively large because of an artifact in compilation: the total costs of resettlement includes also compensation for infra- structure (for example, roads, bridges, and other items) that does not accrue to the families themselves, but they are part of the total costs computed for the overall component. References The word "processed" describes informally reproduced works that may not be commonly available through libraries. Adu-Aryee, V. Q. 1993. "Resettlement from Akosombo to Kpong." In M. M. Cernea and S. E. Guggenheim, eds., Anthropological Approaches to Resettlement: Policy, Practice, and Theory. Boulder, Colo.: Westview Press. Cernea, M. M. 1988. Involuntary Resettlement in Development Projects: Policy Gulide- lines in World Bank-Financed Projects. Washington, D.C.: World Bank. . 1990. Poverty Risksfrom Population Displacement in Water Resources Devel- opment: Policy and Operational Issues. DDP 335. Cambridge, Mass.: Harvard University, HIID. _- 1991. "Socio-Economic and Cultural Approaches to Involuntary Popula- tion Resettlement." Gutidelines on Lake Management 2:177-88. - 1993. "Anthropological and Sociological Research for Policy Develop- ment on Population Resettlement," in M. M. Cernea and Scott Guggenheim, eds., Anthropological Approaches to Resettlement: Policy, Practice, and Theory. Boulder, Colo.: Westview. - 1995. "Social Integration and Population Displacement: The Contribution of Social Science." Economic and Political Weekly (India) 31 (June 15): 1515-23. Chambers, Robert, ed. 1970. Volta Resettlement. London: Pall Mall Press. CHESF (Companhia Hidro Eletrica do Sao Francisco). 1975. Projeto Sobradinho: Resettlement Planfor the Poputlation Reached by Sobradinho's Reservoir. Sao Paulo, Brazil. FAO (Food and Agriculture Organization of the United Nations)/World Bank. 1975. "Report of the Brazil Mato Grosso Land Settlement Project: Preparation Report." Report No. 44/75 BRA:17. FAO/World Bank Cooperative Programme. Rome. Processed. 146 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT .1980. "Report of the Brazil Rondonia Rural Development Project Prepara- tion Mission," vols. I and II. Report No. 52/80 BRA:21. FAO/World Bank Co- operative Programme. Rome. Processed. Gulyani, S. 1992. "Rethinking Resettlement: Employment, Negotiation, and Land in Singrauli, India." Master's thesis, Department of Urban Studies and Planning, Massachusetts Institute of Technology, Cambridge, Mass. Pro- cessed. Gutman, P. 1993. "Involuntary Resettlement in the Hydro Power Projects: A Re- view of the Bank Performance during 1978 1992." World Bank, Washington, D.C. Processed. Indonesia, Government of the Republic of. 1982. "Physical Planning for Trans- migration Terms of Reference, Phase II, Phase III A, Phase III B." Jakarta. Pro- cessed. Kinsey, B. H., and H. P. Binswanger. 1993. "Characteristics and Performance of Resettlement Programs: A Review." World Bank, Washington, D.C. Processed. Shuikou Reservoir Resettlement Office. 1991. "Shuikou Hydroelectric Project Part II: Supplementary Information on Resettlement." Fujian Provincial Gov- ernment, Fouzhou, China. Processed. World Bank. 1978. "The World Bank and the Settlement of Agricultural Lands: An Issues Paper." Washington, D.C. Processed. _ . 1985. "The Experience of the World Bank with Government-Sponsored Land Settlement." Report No. 5625. Washington, D.C. Processed. .1987. "World Bank Experience with Rural Development 1965-1986." Re- port No. 6883. Washington, D.C. Processed. .1992. "World Bank Approaches to the Environment in Brazil: A Review of Selected Projects, vol. V, The (POLONORESTE) Program." Report No. 10039. Washington, D.C. Processed. . 1993a. "Early Experience with Involuntary Resettlement: Overview." World Bank, Operations Evaluation Department, Washington, D.C. Processed. _.1993b. "Early Experience with Involuntary Resettlement: Impact Evalua- tion on Ghana Kpong Hydroelectric Project (Loan 1380-GH)." World Bank, Operations Evaluation Department, Washington, D.C. Processed. .1994. "Resettlement and Development: The Bankwide Review of Projects In- volving Involuntary Resettlement, 1986-1993." Environment Department, Washington, D.C. Processed. . n.d. "Economic Alternatives for Displaced Populations: Resettlement and the Shuikou Hydroelectric Development Project in China." Washington, D.C. Processed. The author also referred to numerous World Bank appraisal re- ports, project completion reports, and other internal documents relating to the projects reviewed. 4 Economic Dimensions of Urban Resettlement: Experiences from Latin America Maria Clara Mejia Editor's The economic dimensions of urban involuntary re- Note settlement have been studied even less than the eco- nomics of rural population displacement. To this vast "field-in-waiting," the study by Maria Clara Mejia makes a substantive contribution and recommends a rich agenda of topics for further inquiry. After a general overview on urban resettlement in Latin Ameri- can countries, the study focuses on key economic issues regarding house replacement for resettlers. It discusses affordability of new housing, subsidized rehousing solutions with full or partial pay- ment by the project, and other financing alternatives that emerged in Brazil, Argentina, and Colombia. Because failure to understand the complexity of the resettlement process has caused chronic un- derestimation of its costs in urban project budgets, Maria Clara Mejia "deconstructs" the general concept of urban resettlement costs in some 15 specific cost categories germane to resettlement op- erations. Detailed tables provide information on average resettle- ment costs per family and as a percentage of total project costs. The study also discusses the World Bank's participation in re- settlement financing, which in some projects has been increased during implementation to levels higher than those defined at ap- praisal time. The analysis found that reliance on local financing alone for resettlement has increased the likelihood of delays and problems. The availability of external (nonlocal) financing is help- ful because it can meet costs that otherwise tend to be left "unrecog- 147 148 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT nized"; it also enhances the commitment by governments to deliver local financing and facilitates better implementation. Reestablishing income-generating activities for resettlers in ur- ban settings confronts a host of economic problems that are specific to large human agglomerations and different from rural resettle- ment. Among them are the costs of urban services, distance to jobs and cost of transportation, scarce demand for labor, and limited customer bases in newly constructed locations. The study explores possible responses and concludes with numerous practical recom- mendations for the preparation and implementation phases of fu- ture urban projects entailing resettlement. The involuntary displacement and consequent resettlement of popu- Tlations has been a major concern of the World Bank and other inter- national development agencies during the past two decades. During this period, investigators from a variety of scientific fields have con- ducted research on the impacts of official policies and procedures-or lack of policies-for resettled populations. Most such efforts have fo- cused on "megaprojects" for hydroelectricity and irrigation in rural ar- eas, while the experience with involuntary displacement of urban populations, especially in the slums and poverty belts of large cities, has been less systematically assessed. A disciplined effort to understand the specific features of urban resettlement would help in designing opera- tional policies, strategies, and procedures that are compatible with the socioeconomic and political realities of the urban context. Current knowledge of the economic dimensions of urban resettle- ment is rudimentary. This paper draws on recent experience in Latin America with planning, implementing, and evaluating urban resettle- ment operations, discussing the challenges of dealing with the economic costs, financial arrangements, and impacts of the relocation process on family income, employment, and economic activities generally. While the cases reviewed here illustrate many important lessons regarding the economic dimensions of involuntary resettlement, the point we would like to stress is that such economic dimensions require much more attention from professional economists in all resettlement opera- tions. Several of the operations discussed in this study are still under implementation, and the analysis, though empirical, is necessarily preliminary. The cases presented here are drawn from World Bank-assisted ur- ban projects in Latin America and the Caribbean. In the 1970s and 1980s Bank-financed projects involving involuntary resettlement in the region URBAN RESETTLEMENT IN LATIN AMERICA 149 were mostly located in rural locales, but by the middle of the current de- cade the majority of such resettlement-related projects in Latin America were in urban areas. This chapter seeks to improve understanding of the resettlement process in active projects and those under preparation dur- ing the past five to seven years. We readily recognize that the Bank-assisted projects discussed here account for only a small fraction of the total urban involuntary resettle- ment occurring today in Latin America, since the largest part of forced re- location takes place outside Bank projects under domestically financed projects. But the economic and sociocultural issues are more or less the same, notwithstanding the source of financing. With 70 percent of the re- gion's population residing in urban areas, and the pace of urbanization still accelerating, urban development projects and resulting involuntary displacement are increasing as well (Burki and Edwards 1995). During the past decade, Latin American countries and implement- ing agencies have made considerable progress in the design of resettle- ment plans, adoption of participatory strategies for decisionmaking, and restoration and improvement of living standards for those dis- placed. Despite this progress, however, too frequently the scope and na- ture of involuntary displacement are not clearly identified during project preparation. There is little planning for resettlement, the eco- nomic reconstruction of the livelihoods of affected people is inade- quately addressed, financing is insufficient, and unexpected results occur due to failures that could have been avoided or minimized through a comprehensive approach to the economic dimensions of re- settlement. Studying the economic characteristics of urban resettlement opera- tions and analyzing operational problems will help to significantly in- crease benefits, reduce social costs, and enhance the feasibility of urban development projects in Latin America (Cernea 1992). Accordingly, this chapter will analyze some of these characteristics, describe current pro- cedures in projects under implementation, show the main difficulties in the planning and execution of urban resettlement, and make opera- tional recomrmendations for overcoming those difficulties. Overview of Urban Displacement and Resettlement in Latin America Magnitude of Urban Population Displacement One of the most significant changes in development in the past 10 to 12 years has been the growing number of urban, transportation, and water supply investments. The Bankwide resettlement review found that in 150 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT the Bank's overall project portfolio these sectors accounted for 22 per- cent of all projects, 53 percent of all projects involving resettlement, and 47 percent of people displaced (about 950,000 people) (World Bank 1996). The Bank currently has more than 100 urban projects either under implementation or in preparation in Latin America and the Caribbean. Most are directed toward the construction and improvement of ser- vices, urban infrastructure, environmental sanitation, territorial zoning, and management of urban river basins in large and heavily populated cities. Due to the high population density of urban areas, many of these operations will inevitably involve involuntary displacement. Table 4.1 shows urban projects in Latin America and the Caribbean as of September 1995 for which involuntary resettlement operations are under way and for which resettlement plans have been agreed upon by the Bank and the borrower governments. Preliminary estimates indi- cated that a total of over 30 projects that will entail involuntary resettle- ment of urban populations have been scheduled for approval between 1996 and 1999 (World Bank 1995a). Past experiences have shown that due to lack of detailed design, uncertainties in defining technical com- ponents, and a paucity of analysis of the economic dynamics in urban settlements, resettlement needs have been often established only during the advanced stages of preparation or even during implementation. But while the number of households affected is relatively easy to count, the impact of resettlement on the economic and social systems of the urban population is not readily apparent and requires careful examination during the planning phase. The 10 projects in table 4.1 displace almost 100,000 people and serve as a context for our discussion.' These numbers, of course, are likely to change from year to year because of the growing weight of urban pro- jects in the Bank's portfolio, and because there are anticipated projects for which resettlement needs are not yet fully determined. Causes and Rationale for Involuntary Displacement in Urban Areas Displacement of urban populations is generally the result of (a) disor- derly growth and consequent reordering of urban living space, (b) improvement of highly deteriorated sanitation and environmental conditions, (c) improvement of other urban infrastructure facilities, (d) specific social or poverty alleviation policies, and (e) the effects of natural disasters. This last case typically involves moving at-risk popu- lations to avoid future disasters. In this case, the resettlement is an objec- tive in itself and may be highly beneficial to the displaced population, provided sound resettlement plans are implemented. URBAN RESETTLEMENT IN LATIN AMERICA 151 Table 4.1. Population Displacement in Selected Urban Development Projects Displaced Country Project popuilationa Argentina/ Paraguay Highway VIII 2 Argentina/ Paraguay Yacyreta Hydroelectric II 15,000 Brazil Parana Water Quality and Pollution Control 9,000 Brazil Minas Gerais Water Quality and Pollution Control 12,522 Brazil Sao Paulo Water Quality and Pollution Control 24,144 Brazil Rio Flood Reconstruction and Prevention 22,200 Brazil Ceara Urban Development and Water Resource Management 2,000c Brazil Belo Horizonte and Recife Metropolitan Transportation 4,368 Brazil Minas Gerais Urban Development (Ipatinga) 4,500d Colombia Santa Fe 1 Water Supply and Sewerage Rehabilitation and Private Sector Development 390 Total 94,328 Note: Actual and anticipated displacement in projects under implementation or prep- aration as of September 1995; this table reflects the group of projects analyzed in this pa- per; the Bank's portfolio of urban projects with resettlement components in Latin America and the Caribbean includes several other recent projects in Colombia, Peru, Guatemala, Brazil, and Guyana. a. Some numbers were estimated using an average household size of six residents. b. Residents without legal title. No data are available on legal owners to be displaced by the project. c. The first four dams of this project will displace about 2,000 people in rural areas where water supply reservoirs are being built. The total population to be affected by the 39 dams is not yet determined, but an unknown number of urban families will be dis- placed. d. Population displaced in Ipatinga. The number to be displaced in other municipali- ties of Minas Gerais by the next phase is under assessment. The importance of water and sanitation projects in response to soar- ing urban growth in Latin America is beyond question. Such projects improve the health and living conditions of hundreds of thousands of people in metropolitan areas. For example, in the project for Bacia de Guarapiranga in the Sao Paulo region, more than 72,000 families will benefit from rehabilitation of urban slums. The same project will also protect the water supply of more than 3 million families. But such 152 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT projects often imply considerable social and economic costs as well. The reorganization and improvement of urban areas inevitably involves dis- placement of marginalized populations living on public or private lands, in slums, and in highly degraded areas. But the relation between the objectives of urban development pro- jects and involuntary displacement is not always justifiable. Some of the occurring displacements are unnecessary, and they are caused by poor project design, failure to optimize social costs and benefits, or exclusion of relevant actors from the decisionmaking process. For instance, in most of the projects discussed in this chapter, displacement has been re- duced significantly by changing initially proposed project design by us- ing more appropriate technology and broadening public consultation, without seriously affecting project costs or objectives (see table 4.7). Specific Conditions in the Urban Context Is urban displacement and resettlement different from that in rural areas? And if so, how? Diverse and complex factors define the specific features of urban resettlement-the internal conditions of each country, the nature of some projects, the dynamics of urban growth and demographics, the ur- ban land tenure problem, and the institutional arrangements for urban projects. Below are some of the characteristics specific to resettlement in the urban context. e The pattern of settlement is relatively recent. On average, 65 percent of the urban population affected by Bank-assisted projects in Latin America and the Caribbean has lived in the affected area for five years or less, while only about 15 percent of families have lived in the area for more than 20 years.2 That is to say, these populations have already relocated to their present location in the recent past, some as rural migrants, some from town and cities. This feature is in striking contrast to typically long-established rural settlements. • Because of the scarcity and the high price of urban land, it is increas- ingly difficult to find affordable and suitable land for resettlement in urban areas, and real estate prices are subject to runaway specula- tion. This is a serious obstacle. Building upward rather than outward is increasingly seen as the most adequate resettlement alternative. The expenses for doing so may be justifiable when compared with the high sociocultural costs and the disruption of social and eco- nomic activities caused by resettlement to a more distant area. . Urban communities display a high capacity for adaptation, self-management, and resettlement resulting from previous expe- URBAN RESETTLEMENT IN LATIN AMERICA 153 riences in relocating themselves (Peattie 1970). The ability of the ur- ban poor to self-relocate and to manage compensation packages should be studied more deeply, but what is important is that their experience in this regard is much greater than that of rural populations. The Rio Flood Reconstruction Project is a good ex- ample. In the face of delays and bureaucratic inefficiencies, and fearful of squatters, 63 families that were going to be displaced occupied the allocated resettlement site on their own. The self- resettlement was orderly and calm, and it created favorable condi- tions for future negotiations between the communities and project management. The number of families in affected areas is always changing be- cause of pressures on urban living space, inability of the govern- ment to prevent such movements, project delays, and so on. The constant threat of massive new influxes makes traditional field studies, planning processes, and lengthy negotiations difficult. This is especially problematic when projects are planned but then delayed for many years (such as the Yacyreta Hydroelectric Pro- ject and the Rio Flood Reconstruction Project), during which poor people move into the area hoping to receive a resettlement package. The reliance of slum dwellers on the so-called informal economy has not been sufficiently studied or taken into consideration in re- settlement planning. While informal activities allow greater mobil- ity, they also create dependence on the surrounding environment, which provides informal or temporary job opportunities in the commercial and industrial centers of cities. The unsanitary, dan- gerous, and impoverished living conditions of many slum dwellers should not obscure the fact that a vibrant and essential economic life drew them here in the first place (Lomnitz 1977, Peattie 1970, Perlman 1982). Efforts at "urban modernization" or rehabilitation of deteriorated downtown areas underestimate the centrality of this fact. This makes site selection critical to restoring employment, income, and social networks. The socioeconomic and political landscape of cities is constantly changing, affecting the various social actors (institutions, groups, and individuals) and requiring continual adjustments in the agree- ments established in development projects. Relocation of popula- tions and housing solutions for low-income groups has undeniable political significance (Lomnitz 1977). It does not take extensive re- search to understand that, beyond the main objectives, these pro- jects indirectly (a) legalize and regulate the occupation of urban lands, (b) allocate access to public services, (c) create new economic 154 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT opportunities for commercial and industrial businesses, (d) in- crease land values and influence low-income housing, and (e) pro- duce political capital for electoral purposes. * Resettlement "naturally" tends to become intertwined with public housing programs and is therefore subject to official housing agen- cies' parameters on financing strategies and recovery of invest- ments, real estate appraisal, housing standards, and expropriation procedures. This deflects attention from the heart of resettle- ment-reestablishing the economic basis for improved liveli- hoods. . Institutional and administrative complexity. Urban projects usu- ally include a number of components involving many sectors of the economy and a multitude of agencies and institutions. Land use regulation, building codes, environmental zoning, public services provision, and other issues related to resettlement often entail the participation of many agencies. This complicates decisionmaking, dissemination of information, and planning and implementation of resettlement programs. . Slum dwellers are considered third-class citizens. Many public officials tend to view people living in slums as squatters, thieves, prostitutes, and drug dealers, who are therefore less worthy of meaningful consultation and participation strategies or well- designed resettlement plans. Yet slums are like parallel cities whose marginalization and lack of government assistance have fostered specific survival strategies. Relocating their inhabitants in- voluntarily disrupts an entire lifestyle and the social network es- tablished among neighbors (Perlman 1982). * Perhaps one of the most outstanding characteristics of urban reset- tlement processes is the impact on poor families with no legal land titles, who frequently account for the vast majority of the affected population in highly degraded urban areas. Although this feature may also be found in rural resettlement, it is even more challenging in the urban context. There are some special conditions in the case of slum improvement programs that hinder traditional planning processes and create new so- cial and technical challenges: * Resettlement needs can be determined precisely only when project design is ready for each slum, which in many cases practically merges planning and implementation into a single stage. * Additional resettlement may be necessary during construction due to unforeseen property damage and other impacts caused by small URBAN RESETTLEMENT IN LATIN AMERICA 155 variations in the work plan and other strictly technical require- ments. . Displacement of some families is often necessary to provide services and improve the standard of living of the rest of the commu- nity, which divides the population into affected and benefited groups. * A parallel strategy is required to keep out new squatters during the period from project design to registry and relocation of the existing families. In the case of Sao Paulo, for instance, it is estimated that this process usually takes six to 10 months. Another special case is that of linear projects such as construction or improvement of roads, laying out electrical grids, and installing pipes for water and sewers that affect urban areas. On the one hand, these un- dertakings require isolated, noncontiguous relocations that must be planned on a case by case basis. On the other hand, the impact on most houses and property is only partial, and the families can usually stay in the same location as long as their homes are rebuilt. This is quite differ- ent from cases in which the loss of property or housing is total and where the project affects an entire community. Displacement of populations in high-risk areas also has special characteristics. Despite the social and economic costs, resettlement of families frequently threatened by landslides, floods, or risky environ- mental conditions is highly desirable as long as housing provisions are acceptable and the economic recovery program ensures the improve- ment of living conditions. It is a case that requires special procedures and policy considerations. Economic Issues in House Replacement Alternatives During the past four years several countries in Latin America and the Caribbean have designed and implemented a range of policies to ad- dress problems related to urban resettlement. These policies follow the general lines of the World Bank's resettlement policy but also reflect the specifics of each country and each project.3 More important, they are the result of social, economic, and political agreements among the agencies and institutions participating in project implementation. Some of the sa- lient issues addressed are described below. Land Titles and Property Compensation The Bank's policy is to provide compensation regardless of whether the person has legal title. Some countries in Latin America stillbase resettle- 156 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT ment policy on whether the displaced people have legal title to land, while others base their policy on social vulnerability regardless of the le- gal status of the land on which those to be displaced reside.4 Compensa- tion for legal landowners is generally considered a buyer-seller trans- action between the state and private individuals. Houseowners without title, on the other hand, are considered squatters, or "illegal inhabitants," of public or private property who should bear more of the financial burden of resettlement through partial payment of housing solutions, legal prohibitions on property exchange, or lower-quality housing solutions. Commonly, they receive compensa- tion only for personal property but not for the lot itself. However, it is recommended that borrowers recognize de facto property rights as part of the compensation package. Usually replacement of houses and land is the best strategy to ensure recovery of the previous standard of living. Monetary Compensation The Bank's view is that delivering only monetary compensation, while acceptable in select cases, often jeopardizes living standards, especially for low-income groups. Such compensation often ends up in the hands of moneylenders and intermediaries, or is spent on old debts, consumer goods, or other expenses that do not ensure the reestablishment of homes, communities, and social networks. There have been few ex post evaluations of groups that received cash compensation, so the true di- mensions of this problem in Latin America and the Caribbean have not been measured, but current work in Argentina and Colombia is ad- dressed precisely to this lacuna. For those without legal title, monetary compensation often does not allow for full replacement, since it does not include payment for land. Socioeconomic Development Since involuntary displacement can destroy previous means of liveli- hood, all resettlement operations must be development programs. There is need for alternative employment strategies for the displaced, including preferential hiring in the project itself and in the public sector when appropriate. However, most current urban resettlement pro- grams are based only on housing replacement. Economic and social as- sistance programs are mentioned only tangentially, and financing, human resources, and participation of social development agencies are inadequate. Usually it is the social workers directly involved who pro- mote the community and solicit support from state agencies. URBAN RESETTLEMENT IN LATIN AMERICA 157 By contrast, in the El Salvador Earthquake Reconstruction Project, the Italian government collaborated in a resettlement plan that included integrated programs of socioeconomic development. The programs were generated employment (construction at each phase provided em- ployment for the families displaced in the previous phase), created a de- velopment agency to be managed by the relocated families, and generated family income through creation of microenterprises. The case of Yacyreta is another example, with routine socioeconomic assistance programs before, during, and after actual resettlement (Mejia 1996). Subsidized Solutions versus Partial Payment Policies also vary in the degree of subsidy or financing required of the families themselves. In some cases the entire cost of housing and reset- tlement is covered by the project (Minas Gerais, Santa Fe 1, Colombia Power Market, Yacyreta), while in others the displaced families must pay a share of the housing costs (Sao Paulo, Parana, Rio de Janeiro). Without trying to settle the debate over whether to require those dis- placed to cover part of the resettlement cost, the following points should contribute to the discussion. . The Bank's basic policy is that the risks of impoverishment in- volved in forced relocation must be avoided (Cemea 1997). Be- cause the population isforcibly displaced, it should not have to bear any of the resettlement costs and also should have a variety of reset- tlement options. Additional financial burdens must not be placed on low-income families that are already at risk of impoverishment due not only to precarious income but also to labor instability. As an illustration, in projects examined here, between 15 and 25 per- cent of the families affected earn less than the minimum wage. Even those that appear capable of making modest payments might be at risk of losing their homes due to economic crises, inflation, and currency devaluations. Forced imposition of a debt when no other alternative is offered is not acceptable to the Bank. • Faced with declining budgets, needs for investment recovery, and increasing land and construction prices, housing agencies tend to transfer, openly or surreptitiously, part of the resettlement costs to the displaced people. Agencies also fear that not requiring squat- ters to pay part of the resettlement cost will encourage new occupa- tions of public land. * Some civil servants in charge of resettlement think that having to pay part of the housing cost makes the resettled families feel more 158 THE ECONOMAICS OF INVOLUNTARY RESETTLEMENT rooted, increases their sense of ownership, and creates greater so- cial discipline and capacity for self-management. In the projects un- der implementation in Latin America, housing subsidies to families vary from 65 percent for families with income greater than three times the minimum wage to 75 percent for families with up to three times the minimum wage, and as much as 95 percent for fami- lies with less than a single minimum wage. The involuntarily displaced communities themselves have repeat- edly rejected being forced to assume housing costs. In the "Sitio do Livramento" in Rio de Janeiro, 95 percent of the families failed to make their monthly payments due to unstable incomes and the fi- nancing terms imposed. A recent survey by a local housing agency revealed that 70 percent of families that were resettled for more than one year were not making their monthly payments. The Yacyreta project also illustrates reluctance to pay even by families considered well off. Housing Replacement Alternatives Table 4.2 shows the resettlement solutions offered to different groups in urban projects in Latin America. It shows the variety of approaches Table 4.2. Compensation Alternatives Offered to the Displaced Population Country/project Displaced Alternatives Brazil- Legal owner There are no legal owners of land to be dis- Sao Paulo placed in the project. Water Quality and Pollution Landholder 1. Financed house provided by the housing de- Control without legal velopment agency. Alternatives offered de- title pend on family income, in accordance with normal agency regulations. Monthly pay- ments vary from $11 to $226. Compensation for improvements is not paid because the new homes are already highly subsidized. A legal agreement (Termo de Compromiso) is signed upon moving into the house. 2. Letter of credit for up to $15,000. 3. For extremely poor families (less than one minimum wage) exchange for shack worth up to $1,700 in another slum, or construction of home in the same slum, without financing re- striction. URBAN RESETTLEMENT IN LATIN AMERICA 159 Table 4.2 (continued) Country/project Displaced Alternatives Renter Same alternative as for house owners without titles. No indemnification paid to original owners if the units have not been legalized. Brazil-Minas Legal owner Monetary compensation at market replace- Gerais Water ment price. Quality and Landholder 1. Monetary compensation of at least $6,000 for Pollution without legal houses and improvements. Control title 2. House in state housing company develop- ment, with no financial obligation. Resident signs a legal agreement (Posse Provisoria) that prohibits the family from selling the house for ten years. Renter One year rent payment at market prices. Prop- erty owners are given the same option as the house owners without title. Brazil-Parana Legal owner 1. Monetary compensation at market replace- Water Quality ment price. and Pollution 2. Exchange for developed lot with no pay- Control ment by property owner. Landholder 1. Developed lot financed under state housing without legal company rules: 60 to 120 monthly payments title up to 20 percent of minimum salary 2. Modular house financed in conjunction with the developed lot. Renter Basic package of building materials to families recently displaced by floods (Trevisan area). The materials will be financed jointly with the urbanized lot. Brazil-Rio Legal ownera 1. Monetary compensation. Flood Recon- 2. Home in state housing company housing struction and development. Prevention Landholder Unit in the state housing company develop- without legal ment with monthly payments equivalent to 15 title percent of the minimum wage, one-year grace period. Renter 1. Unit delivered by state housing company with monthly payments equivalent to 15 per- cent of the minimum wage, one-year grace pe- riod. 2. Monetary compensation for house and im- provements. Property owner is not compen- sated if units are not legalized. (Table continues on thefollowing page.) 160 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Table 4.2 (continued) Country/project Displaced Alternatives Brazil-Belo Legal owner Monetary compensation at market value or a Horizonte and minimum payment sufficient to buy another Recife Metro- property, whichever is higher. Assistance is politan Trans- provided in finding a replacement property. portation Landholder 1. Compensation at market values. without legal 2. Housing in housing units for poorest fami- title lies. Renter 1. Financial assistance provided to help rent another property. 2. Indemnification for commercial house. Argentina/ Legal owner 1. Indemnification for lot, house, and improve- Paraguay- ments at market replacement price. Yacyreta Hy- 2. Full restitution of the lot, house, and mate- droelectric II rial losses. Exchange of current house for unit built by the project in development with com- munity services and infrastructure. Landholder Monetary compensation for material losses. without legal title Renter Monetary compensation for material losses. Colombia- Legal owner 1. Monetary compensation for 120 percent of Santa Fe 1 replacement value of house and lot. Water Supply 2. Replacement of house at no cost to owner. and Sewerage 3. Monetary compensation for economic Rehabilitation losses in domestic and commercial activities. and Private Sector Devel- 1. Indemnification for 100 percent of house and opment lot replacement value. Landholder 2. Replacement of the property at no cost to without legal owner. title 3. Monetary compensation for effects on eco- nomic activities. Financial assistance equivalent to six months' rent for tenants with less than two years' resi- Renter dence, and twelve months' rent for tenants with more than two years of residence. a. When the parents' marital status is not legalized, the house is registered to the woman. URBAN RESETTLEMENT IN LATIN AMERICA 161 currently used, which depends on the internal policies and the institu- tional and legal frameworks within each country. Table 4.3 presents in more detail the basic menu of strategies avail- able for providing housing. The alternatives actually used depend on families' capacity for self-management of their resources as well as on lo- cal financial and market constraints and possibilities. Each strategy Table 4.3. Housing Alternatives Housing alternatives Description Chess game For families with income below the minimum wage, the project buys a house in the same slum and gives it to the displaced family. The family that sold the house is moved into housing financed by the state housing company. In Sao Paulo, 14 families out of the 80 already resettled have chosen this op- tion. Exchange for partially The family promises to make monthly payments in subsidized housing in accordance with their income level (Rio de Janeiro, developments built by Sao Paulo). housing agencies or un- der contract by private firms Exchange for fully subsi- Those displaced receive a single-family home in a dized house in develop- new housing development built by the project ment built by the project (Minas Gerais, Yacyreta, Sao Paulo). Exchange for single- In Rio de Janeiro, the family promises to make family housing with monthly payments based on its income level. shared social area Apartment buildings Delivery of a financed apartment (Sao Paulo, 80 families). Letter of credit The displaced family receives a document giving them credit to buy an existing house worth up to $15,000 (in Sao Paulo, this tool has yet to be used). Renting with option The renting family moves to a unit that belongs to to buy the local housing agency. The family continues to make monthly rent payments, but part of them goes toward an escrow account to allow for the possibil- ity of purchase in about five years. Financed lot with The renting family moves to a developed urban lot. services The project provides technical assistance to dis- mantle, move, and rebuild the existing home. Lot with services and For families left homeless by flooding, financing is basic building supplies provided to enable the family to buy the building supplies and the lot (Parana). 162 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT involves a different set of pros and cons, and deciding which approach to use is a complicated task that requires careful analysis of the conditions in each project and location, as well as consultation with the population be- ing resettled. Socioeconomic Profile of the Affected Population Socioeconomic characteristics are an important consideration in determin- ing appropriate urban resettlement strategies. The baseline information available for Latin America and the Caribbean urban resettlement projects is drawn from registries and censuses that provide an economic and de- mographic profile on which to base compensation and indemnification. The socioeconomic data presented assist in understanding the risks of impoverishment associated with resettlement, but they are by no means exhaustive. With the aim of contributing to the discussion regarding the appro- priateness or inappropriateness of the strategies and policies adopted in the resettlement process of urban population in Latin America and the Caribbean, socioeconomic characteristics of the displaced population are summarized here. Although each project has census and land regis- try data that at least include the communities to be displaced in the first phases, these data do not clearly delineate the sociocultural character- istics of the displaced population, nor are the socioeconomic data complete. FAMILY INCOME. The Parana Water Quality and Pollution Control Project stands out at the extreme low end of family income-37 percent of the families received less than a single minimum wage, compared with 5 percent in the Rio de Janeiro metropolitan area. In general the great majority of the families to be displaced received less than two times the minimum wage. While the proportion varies, the socioeco- nomic vulnerability of this extremely poor segment is indisputable. Spe- cial social assistance for relocation will be needed. EMPLOYMENT. Average unemployment among the displaced popu- lation is about 15 percent, though in Rio de Janeiro it exceeds 40 percent. Furthermore, about 25 percent of the population on average is underem- ployed or relies solely on incidental work. With 40 percent either unem- ployed or underemployed, this population is highly vulnerable and requires special support during resettlement. It can also be expected that relocation of this group in a new settlement far from its current one will significantly affect job and income opportunity. URBAN RESETTLEMENT IN LATIN AMERICA 163 WOMEN IN URBAN RESETTLEMENT. The gender of heads of house- holds is another important factor, with implications for program design such as financing and titling of real estate, accommodating special socio- economic needs, and building community support infrastructure. Households headed by women who work outside the home have a greater need for kindergarten or day care facilities. For example, in the Rio Flood Reconstruction Project, about 66 percent of households are headed by women, whereas in Parana (Curitiba) only 15 percent are headed by women. Unfortunately, in the other samples there is no breakdown of data by gender. Though gender influences participatory strategies and attainment of economic outcomes, it is insufficiently considered. LENGTH OF RESIDENCE. A high percentage of families have lived in the neighborhoods for five years or less, while about 15 percent have re- sided there for more than 20 years. This influences both the heterogene- ity of the neighborhood and the degree of participation in the community social network. HETEROGENEITY OF THE POPULATION. The heterogeneity of popula- tions in the so-called poverty belts has been clearly confirmed by the technicians responsible for conducting socioeconomic surveys and ne- gotiations with the affected communities. Significant indicators of het- erogeneity include: * Income structure-ranges from less than one minimum wage to more than eight wages per household . Quality and size of houses-from disposable materials to brick; from 18 square meters of floor space to 120 square meters and more; from one floor to as many as four floors Family size-from one person to more than nine * Level of community organization-from rudimentary forms to officially recognized organizations with representation in policymaking bodies at local and state levels. The previous data show the general, widely recognized socioeco- nomic characteristics of slum dwellers: the majority of the population is extremely poor, unemployment and underemployment are marked fea- tures, a significant number of households are headed by women, and there is great internal variation. The point here is that these features, while mentioned in surveys, are poorly reflected in resettlement policies and in resettlement plans. 164 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Economic Reconstruction and Participation Consultation and participation by the affected families is a critical step in designing socially, financially, and technically feasible resettlement programs to restore the economic productivity of resettlers. This is the only way to ensure a fit between proposed economic and social pro- grams and the social and cultural realities in which people live. All the projects reviewed here include consultation with affected communities to guarantee the social feasibility of relocation and rehabilitation pro- grams. The difficulties now are mainly operative: When to consult? Whom to consult? Which actors are most relevant? Through which mechanisms should we consult? Which type of decisions and actions should the communities participate in? How to prevent manipulation of the communities by outside parties? Some of these questions are elabo- rated on below. Is it possible to begin consultation and participation at the earliest stages of the project? There is a myth that participation at the early stages will lead to cost overruns and unnecessary delays. It is also feared that sharing information and consulting with the affected population will create unrealistic expectations and attract new squatters to the pro- ject site. But the experience of current projects in Latin America and the Caribbean have demonstrated that this is not the case. Two exceptions should be mentioned: in the Yacyreta and the Rio Flood reconstruction projects, the massive invasions of people seeking housing were pro- duced by several years of delays in project implementation-a major reason to avoid delay. These and other experiences show that far from hindering the plan- ning process, participation actually reduces tensions, conflicts, and negative social impacts; builds trust and collaboration; avoids delays due to lack of agreement at the community level; and improves the fea- sibility of the project. Participatory mechanisms such as holding a refer- endum on decisions made by community leaders also help prevent manipulation of the community by outside parties. There is no evidence that participatory processes cause delays and cost overruns. Such diffi- culties are more due to planning inadequacy, implementation weak- nesses, or insufficient resources that actually hinder participation. In the Parana Water Quality and Pollution Control Project, for example, pro- ject delays and changes paralyzed the work of a committee composed of a local planning agency and community associations, and heightened criticism and distrust of the project. Consultation mechanisms and strategies vary, but follow the same general pattern: URBAN RESETTLEMENT IN LATIN AMERICA 165 . Preliminary meetings with neighborhood associations or orga- nized interest groups to survey public opinion and gain prelimi- nary consent for the resettlement plan * Agreement on basic policy and plans, especially the resettlement site, housing design, compensation and indemnification, and re- placement of community property • Family-by-family negotiations once the cadastre and census are completed. As the process advances, consultation evolves into more active participation, not only in the decisionmaking process, but also in such areas as construction of housing, house hunting (for self-resettlement), and protection of unoccupied sites. In some cases, after settling into the new site, the communities begin to manage services such as transpor- tation, community centers, schools, hospitals, garbage collection, and road maintenance. In the late 1980s and early 1990s, consultation and participation gen- erally were initiated only after the project had already been approved. Communities were simply informed at the implementation stage that they would be displaced, and under what conditions. Consultation was limited to decisions related to the resettlement itself, and participation was negligible, almost nonexistent. However, in projects approved more recently-such as Colombia Santa Fe 1, Colombia Power Development, and Belo Horizonte and Recife Metropolitan Transportation-and in others under preparation, consultation began at the early stages of project planning. The borrowers agreed to present the resettlement policy and consult directly with the affected community about resettlement as a condition for Bank approval. Likewise, before final approval of Yacyreta II, the resettlement plan underwent an extensive process of consultation, including discussions with community organizations, affected groups, unions, nongovernmental organizations, universities, and others. The economic cost of neglecting consultation and participation can be considerable. Delays in execution of contracts, resulting con- tractor claims and commitment charges due to lack of agreement by the affected population, and protracted negotiations to make unac- ceptable proposals acceptable can elevate project costs to unneces- sary levels. On the other hand, the costs of the consultation process are rarely internalized in the planning and implementation budgets and timetables of the projects under consideration, so that funding consultations with the affected population tend to be underestimated and understaffed. 166 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Participatory strategies still are aimed more at avoiding conflict than at including stakeholders and displaced communities in decisionmaking and in sharing responsibilities central to the project and to people's lives. Developing a strategy that decentralizes decision- making and finding new alternatives related to urban renewal, develop- ment, and sanitation continue to be important goals. Principal Difficulties in Planning and Implementation Planning Beyond many variations in past resettlement planning, in most cases we can identify the same basic pattern: . Projects are conceived by the development agency in charge of preparation and design. . Resettlement policy is established by each agency based on local circumstances rather than on national social policy. * Social and economic information on the affected population is un- reliable, incomplete, and usually based on secondary sources. * The resettlement site is selected based on the availability of public land rather than on what is desirable for the displaced population. . Prevailing state or municipal housing policy is automatically adopted regardless of its adequacy for involuntary resettlement. * Consultation with the affected population is rarely conducted prior to project approval and generally is limited to a small group of local authorities and state agencies. The typical planning process for urban resettlement faces many dif- ficulties. Three of the most important, explained more fully below, are lack of an official resettlement policy at the national or state level, inade- quate definition of project subcomponents at the time of appraisal, and lack of social and economic analysis of adverse resettlement conse- quences within the project preparation process. Lack of an official resettlement policy that defines institutional re- sponsibilities, rights of those displaced, financing mechanisms, land ac- quisition, compensation, indemnification, and titling is one of the most common problems in resettlement planning. No Latin American or Caribbean country has a national policy for involuntary resettlement caused by development projects, nor are we aware of any comprehen- sive sectoral resettlement policy except for the one in the Colombian electricity sector. URBAN RESETTLEMENT IN LATIN AMERICA 167 Inadequate definition of subcomponents, especially those involv- ing public works construction, hinders early identification of the scope and nature of involuntary displacement. Problems in resettlement pro- grams are frequently a direct consequence of socially unfeasible designs in the preparation phase. Some subcomponents lack basic design ele- ments, even at times failing to clearly identify the area affected. In the case of slum upgrading and rehabilitation projects, the need for dis- placement may not be known until the project nears completion. This complicates planning of specific elements, defining a feasible frame- work for relocation, and budgeting of human resources and funds. Lack of social and economic analysis in the project planning pro- cess, and perpetuation of an engineering perspective-from which re- settlement tends to be seen as an obstacle to the advancement of the public works projects, instead of an integral economic and social recon- struction program-is probably the biggest obstacle. Physical elements of resettlement, such as land and housing, generally are taken into ac- count during project planning. But the subtle interdependency between the population and its economic environment, and an assessment of so- cioeconomic impacts before, during, and after the actual move, are usu- ally missing, weakly addressed, or little more than a phrase in a document. Implementation Although each project runs into difficulties related to specific undertak- ings or local conditions, it is possible to identify common problems that are the rule rather than the exception. The most important of these are presented below (some will be discussed in more detail later). * Constant change in project design has been one of the most impor- tant factors complicating resettlement planning in the projects studied here. In almost all of the projects the prevailing resettle- ment plan is different from the one approved two years earlier, due to changes in the design of subcomponents, the enlargement or re- duction of the project area, or new decisions about construction technologies made during implementation. These costly and time- consuming changes often stem from inadequate attention to the so- cial context of the project, which imposes constraints that therefore are not discovered until inappropriate designs are implemented. * Resources of local counterparts are often insufficient or not readily available. In some cases local governments have underestimated costs and failed to allocate sufficient resources, causing delays in 168 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT resettlement programs and distrust toward the project. This, too, is often due to incomplete social and economic analyses of the actual situation on the ground. In addition, most projects experience de- lays because previously allocated local funds are, in effect, not available when needed, partly because constituencies for their ex- penditure have not been mobilized. * Frequently the various states and municipalities involved in a pro- ject lack a housing policy or have different housing policies, which complicates coordination and creates disparities in resettlement procedures and outcomes. Conflicts between resettlement strate- gies and environmental or urban planning regulations can also arise when national or local legal restrictions create obstacles to ob- taining environmental permits for urban renewal, slum improve- ment, and construction of new housing units (see section on legal difficulties). . Despite the fact that on paper the resettlement plan is consistent with the project's general construction schedule, in reality, prog- ress on the main project tends to create pressures that compromise the resettlement plan. This pressure sometimes leads to unsatisfac- tory measures such as offering monetary compensation as the only option, using transitional shelters when the new settlement is not finished, or occupying unfinished housing and settlements that lack operational public services. In some cases the water and sanita- tion agency is separate from the resettlement agency. This makes coordination difficult, especially when the resettlement compo- nent is considered a subordinate activity rather than an integral part of the project. e All projects in the Latin American and Caribbean region involve numerous housing, urban, environmental, and public service agencies. This institutional complexity can result in a very slow process of coordination and decisionmaking. * Political changes in local, state, and even national administrations during the course of project implementation cause revisions to pro- jects and resettlement plans, and consequent delays and difficulties in negotiations with affected populations. Policy differences in suc- cessive administrations, the creation of new municipalities (such as in Paranai and Rio de Janeiro), and ongoing institutional reorgani- zation and shifting of responsibilities have generated enormous difficulties. . Inadequate social teams, sometimes without full-time commit- ment, plague many projects. In Sao Paulo, for example, six social workers are responsible for assisting 2,000 families (including both URBAN RESETTLEMENT iN LATIN AMERICA 169 the displaced and others affected). If the work requires a minimum of two visits per family, each social worker must make about 600 visits, not counting the time required for census interviews, com- munity meetings, preparing for the move, and accompanying the families during the move and the adaptation period. . Weak management capability in some municipal administrations is sometimes aggravated by competing interests between different municipalities with differing stakes in the costs and benefits of a project. Centralization of budgeting and decisionmaking plays an important role, as well as lack of experience on the part of sanitation and public works agencies in resettlement of involuntarily dis- placed populations. * Invasion by squatters, both in the areas affected by the projects as well as the new resettlement sites, generally falls into one of two basic categories: homeless families seeking housing opportunities and groups led by political or community leaders trying to gain electoral advantage. In the Sao Paulo project, the population to be resettled in the first seven slums under rehabilitation increased 16 percent in the year between the initial census and the actual move. In Rio de Janeiro (Strada do Paraiso), homeless people invaded 600 units designated for families displaced by the project. In the Yacyreti project the number of families in the urban area affected doubled in four years. While population pressures pose an addi- tional problem and risk for urban resettlement, it is the responsibil- ity of local public authorities, not the project, to control squatters. . Lack of contingency plans aggravate the effects of natural disas- ters. The most significant case of this is the Parana project, in which one of the resettlement sites, Renato Bonilaure, became partially uninhabitable due to flooding. As a result, an additional drainage project is required. The Rio de Janeiro and Yacyreta projects also have faced various emergencies during implementation. Such problems result in the social science teams constantly "putting out fires," which necessarily disrupts the resettlement schedule. Legal Difficulties BRAzIL: LEI DE PROTECAO DOS MANANCIAIS. Perhaps one of the major legal difficulties confronted by the sanitation and rehabilitation pro- jects in Brazil has been the requirements imposed by the Lei dos Mananciais, which regulates use of the space around bodies of water and springs. The law basically prohibits construction of human set- tlements within 50 meters of a body of water, which in many cases has 170 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT impeded construction of housing developments already approved by the Bank and the borrower. Legal restrictions include rules on lot size and housing density that conflicted with housing plans for low-income families. In Sao Paulo, the scarcity of lots complying with the regulations has led to proposals to amend the law, or to the search for semideveloped lots with licenses approved before the new law in order to meet the bud- getary limitations of the implementing agency. In other cases land is available only at the city outskirts. In this case, two of the municipalities involved in the project, Itapecerica da Serra and Embu-Guacu, fall en- tirely within areas affected by the law, thus prohibiting construction of new housing within the town for the displaced families. LACK OF LAWS ON EXPROPRIATION AND PUBLIC UTILITY. The absence of laws for expropriation and decrees of public utility have created spe- cial difficulties for resettlement operations, making it difficult to take possession of areas necessary for project construction and for relocating the displaced population. In the case of Yacyreta, the resettlement pro- grams have faced innumerable legal obstacles (particularly on the Para- guayan side) due to nonexistent and/or inoperative processes for expropriation under declaration of public utility. INADEQUATE LEGAL PROVISIONS FOR OCCUPANTS. No country in Latin America or the Caribbean has adequate provisions for dealing with cases of squatters: families that occupy public lands without legal title to those lands or the structures built upon them. Therefore there are no agreed rules for protecting the rights of families that have been living in an area for many years, even generations, prior to the government's de- cision to utilize that particular piece of land. As a consequence, in reset- tlement planning these families are often "invisible," and provisions are not made for their relocation to alternative sites and occupations. Economic Difficulties UNDERESTIMATING RESETTLEMENT COSTS. Failure to understand the nature and complexity of the resettlement process has caused chronic underestimation of its cost. The only resettlement costs usually included in the budgets of projects involving involuntary displacement are re- lated to acquisition of land, construction of housing, and compensation for material losses. The 1993 LAC Resettlement Review showed that cost estimates for resettlement in Bank operations were 54 percent too low on average, and in many cases final costs were 300 to 400 percent higher URBAN RESETTLEMENT IN LATIN AMERICA 171 than initial estimates. Urban resettlement projects in Latin America are not an exception in this regard. The original resettlement plan for the Parana project estimated a unit cost of $2,500, even though the current price of a finished low-income housing unit in Curitiba was $7,500. In the Yacyreta Hydroelectric Project, resettlement costs doubled in the three years since the original estimate. Evidently not all costs were con- sidered. Below is a list of some of the typical activities whose costs need to be recognized from the outset: Consultation and participation . Field studies and gathering of census information Design of housing projects * Land acquisition and construction of housing and service infra- structure Investment in expanding or improving existing services to meet the new demand created by the resettled population Transactions with the communities receiving the displaced popu- lations * Indemnification and compensation for material losses and adverse impacts on economic activities Moving the displaced population Social assistance and income restoration programs * Title and registration . Administrative costs and salaries of social researchers, workers, and related specialists * Incremental costs due to urban land speculation * Transaction costs between the institutions and agents involved . Cost contingencies for miscellaneous activities not previously anticipated. Furthermore, the rise in urban land and housing prices between the planning stage and the time of purchase generally tends to aggravate the insufficiency of local counterpart funds. In Sao Paulo, the costs esti- mated by the Housing and Urban Development Company had to be ad- justed 69 percent to meet prevailing prices in the market for low-income housing. Finally, administrative costs due to the institutional complexity of these projects are generally greater than predicted. Very frequently, arriving at agreements between institutions, creating interinstitu- tional coordinating committees, and overcoming the geographic and bureaucratic dispersion of the agencies involved leads to unforeseen costs. 172 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT UNIT COSTS OF RESETTLEMENT. Table 4.4 presents the unit costs f or re- settlement in eight active and in-preparation Bank-financed projects in Latin America and the Caribbean. The average unit costs are approxi- mately $10,200. However, the characteristics of the solutions offered (such as location, lot and house size, services, and conmmunity facilities) vary, and actual costs range from a low of $1,700 for buying a flimsy, ex- isting house in the same or another slum, to a high of $33,450 for build- ing finished single-family units complete with services, community facilities, and socioeconomic assistance programs for urban, lower- middle-class families in Argentina. Table 4.5 shows the total cost of the resettlement program as a per- centage of total cost in seven projects. As mentioned, the real total cost of resettlement is higher due to the fact that the figures do not include all factors associated with resettlement. Necessary costs are much higher in projects that entail massive relocation and large urban upgrading com- ponents; for example, in the Sao Paulo project and the Rio Flood Table 4.4. Resettlement Costs per Family Cost perfamily Country/project (USS) Projects under implementation Brazil-Parana Water Quality and Pollution Control Project 6,800 Brazil-Minas Gerais Water Quality and Pollution Control Project Contagem 6,780-7,300 Belo Horizonte 7,372-12,643 Brazil-Sao Paulo Water Quality and Pollution Control Project 1,700a-20,515 Brazil-Rio Flood Reconstruction and Prevention Project 6,502-13,980 Argentina-Yacyreta Hydroelectric Project II 10,105-17,122 Paraguay-Yacyreta Hydroelectric Project II 17,000-22,000 Brazil-Integrated Urban Transport Project Belo Horizonte 7,200 Recife 6,300 Projects in preparation Colombia-Santa Fe 1 Water Supply and Sewerage Rehabilita- tion and Private Sector Development Project 13,492 Brazil-Minas Gerais Urban Development (Ipatinga) 10,199 a. This is the maximum cost that is invested in the purchase of a slum house where in- tervention takes place. URBAN RESETTLEMENT IN LATIN AMERICA 173 Table 4.5. Cost of Resettlement as a Percentage of Total Cost Cost of Total cost resettlement Resettlement (millions of (millions of as a percentage Project US$) US$) of total cost Rio Flood Reconstruction 300 45.0 15.0 Programa de Saneamiento- Minas Gerais 307 15.0 4.8 Programa de Saneamiento- Parand 223 14.4a 5.4 Programa de Saneamiento- Sao Paulo 260 70.0 27.0 BH-REC Metropolitan Transportation 204 12.3 6.0 Yacyreta Hydroelectric Project 8,500 218.6b 2.6 Santa Fe I Water Supply 641 0.2 Note: The costs of resettlement of various projects only include the housing replace- ment programs. Acquisition of land, compensation, and administrative costs are not in- cluded. a. Through December 1993. b. Total cost including semiurban, urban, and rural resettlement. reconstruction project, resettlement costs rose to 27 percent and 15 per- cent of total project costs, respectively. In most cases, however, resettlement costs are of little significance in relation to total project costs, and their effect on profitability is minimal (table 4.5). Economic studies carried out by the Bank on megaprojects that entailed resettlement have demonstrated the low sensitivity of the rate of return in relation to a significant increase in resettlement costs. A recent calculation made for Yacyreta showed that an increase of 500 per- cent in the cost of resettlement would be required to drive the internal rate of return below 12 percent (World Bank 1996). RESETTLEMENT FINANCING. In most projects, local counterpart re- sources for resettlement have been insufficient due to underestimation of resettlement costs, changes in financing priorities due to changes in public administration, and changes in project design and expan- sion of the area to be affected. Furthermore, the funds that do exist and have been appropriated by the borrower are not always available when 174 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT needed due to bureaucratic delays in moving the funds to the imple- menting agencies. Greater World Bank involvement in resettlement fi- nancing could help solve some of these problems. Although Bank resources designated for resettlement have in- creased significantly in recent years, the financial participation of the Bank is still weak. In accordance with existing guidelines, the Bank can- not finance some important resettlement costs, such as land acquisition, salaries of government functionaries, compensation for goods other than land, or already-built housing units (Ninio 1995). Except for these items, the Bank can finance all investments related to resettlement, sub- ject of course to the Bank's financing norms and prevailing procedures for bidding and contracting. Initially, governments tend to have little interest, if any, in borrow- ing money for resettlement and prefer to finance resettlement opera- tions by using local budgetary resources. However, these resources often are limited, and local agencies often employ a "mechanism" to de- liberately understate the costs of resettlement. Other times, the budget- ary allocation proves to be utterly insufficient in midstream, and resettlement is stalled. The Bank has other financial support instru- ments specifically directed toward project preparation through the Pro- ject Preparation Facility (PPF). Generally, the projects with resettlement in the portfolio for Latin America and the Caribbean have not used such financing.5 It would be interesting to investigate the reasons why, and find ways to facilitate access to preparation funds, specifically for pro- jects with resettlement. Table 4.6 shows current Bank participation in resettlement financ- ing of projects under implementation in Latin America and the Carib- bean. Most of these projects did not include external resettlement financing when they were approved, but once the projects were under implementation (and at the request of the borrowing governments), the Bank expanded its participation by financing construction of housing units, infrastructure, and consultant salaries to carry out independent studies and evaluations. The most notable case is the Rio Flood Recon- struction Project, in which the Bank financed 60 percent of the resettle- ment program. Obviously, we cannot postulate a direct correlation between ex- ternal financing and the success of resettlement programs. But we can empirically document that in practice the availability of external re- sources creates a positive response and increases the commitment by governments to deliver local counterpart financing and thereby facil- itate implementation of the program in accordance with agreed schedules. URBAN RESETTLEMENT IN LATIN AMERICA 175 Table 4.6. World Bank Participation in Financing Resettlement Bank Project participation Comments Parana Water Quality Yes Until now the state of Parana has been and Pollution Control responsible for 100 percent of resettle- ment costs. At the borrower's request, the Bank will finance the infrastructure and earthworks as well as water and sewage services for the lots. Sao Paulo Water Quality Yes The state of Sao Paulo, with the support and Pollution Control of the municipality of Sao Paulo, is re- sponsible for 98.7 percent of the reset- tlement costs. The Bank is financing the services infrastructure in the housing developments (1.3 percent of the total construction costs for the housing). Rio Flood Reconstruc- Yes The Bank is financing 60 percent of the tion and Prevention resettlement costs (state component), and the federal government is financ- ing 40 percent. Minas Gerais Water Yes Until now, the state of Minas Gerais and Quality and Pollution the municipality of Contagem have Control been responsible for 100 percent of re- settlement costs. At the request of the state government, the Bank will finance the housing construction component. YacyretA Yes The Argentine government is responsi- Hydroelectric II ble for resettlement financing, but the Bank is financing 6 percent of the total cost of the plan, including urban and rural relocation. Belo Horizonte and No The implementing agency will be re- Recife Metropolitan sponsible for 100 percent of the resettle- Transportation ment costs. Santa Fe I Water Supply No The implementing agency will be re- and Sewerage Rehabili- sponsible for all resettlement costs. tation and Private Sector Development RECUPERATION OF INVESTMENT IN RESETTLEMENT. Resettlement re- quired by development investments need not be economically viable on its own, or subject to criteria of investment recovery-it is a project cost. However, it is expected that resettlement goals will be achieved at the 176 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT optimal, if not the lowest, cost possible. In massive resettlement cases and those with high resettlement costs, all implicit costs should be iden- tified and included as part of the overall project cost. Both the costs and benefits of resettlement should be considered in the cost-benefit analysis of the main project. It is certainly possible to re- flect the cost of resettlement and socioeconomic reestablishment of the displaced population in the price charged for services provided by the project. In this way the benefits of a project would pay for the resettle- ment necessary to implement it while ensuring more sustainable devel- opment. This clearly goes beyond mere cost considerations, and it involves a broad range of issues in the conception, design, and imple- mentation of resettlement programs within the context of development and urban rehabilitation. These ideas should be explored by the social teams, along with economists, project managers, and urban planning specialists. Relocation Impacts on Income-Generating Activities It is widely recognized that the resettlement process does not end when the population has been moved to the new site (Partridge 1990). In fact, the work of ensuring the community's sustainability has just begun: ad- aptation to the new site and new housing, rebuilding social networks and the structure of the community, new relations with the host popula- tion, recovery of prior family income level, management and mainte- nance of new services, and so on. These challenges go far beyond the mere physical relocation of the population. Successful budgeting, financing, recouping of resettlement invest- ments, and sustainability are closely tied to the central issues of urban development itself. In other words, social and economic rehabilitation programs are the backbone of sustainable urban resettlement. Location of the new settlement is a key element in restoring productive activities and ensuring income opportunities and access to services. In resettling the urban poor it is important to move them the least distance possible from their original location. This is problematic, however, since many of the difficulties of urban resettlement are rooted in the relative scarcity of land and housing (especially for low-income groups), restrictions on land use, and the burden on municipal finances. The impact of resettlement on economic activities has not been seri- ously considered and monitored in most Bank-assisted projects in Latin America. For many years, until recently, income restoration programs generally were not included in project design. This was due to the pre- vailing belief that resettlement of the urban poor does not seriously URBAN RESETTLEMENT IN LATIN AMERICA 177 disrupt economic activities, and that with time families adapt and find other ways to subsist. This erroneous belief persists, in part, because of the scarcity of data on the correlation between economic activities and place of residence. One factor to consider in assessing the potential economic effect of resettlement is the degree to which families gravitate toward particular neighborhoods and how their economic activities are linked to that spe- cific area. This will help to identify impacts on home-based businesses or survival strategies based on the informal sector and to design income restoration strategies at the new site. Among the families relocated by the El Salvador Earthquake Reconstruction Project, for example, field studies documented the decline or even disappearance of such activities as selling tortillas and tailoring that depended on proximity to the busi- ness generated by government workers in downtown offices. In many of the Latin America and Caribbean urban resettlement projects it has been possible to obtain adequate lots near the current set- tlements, but in other cases budget limitations have forced implement- ing agencies to select sites relatively far away. In the projects reviewed, distances to the new sites range from less than 1 kilometer for most of the Yacyreta settlements to 60 kilometers for one of the communities dis- placed by the Parana project. But distance is not the only factor that poses a risk to the displaced families' income, economic opportunities, and survival strategies. Some other important factors are: * Negative effects on the semirural survival patterns developed by poor families living in areas of urban expansion, especially activi- ties that complement family income and nutrition, such as raising animals, maintaining home gardens, scavenging, and garbage re- cycling * Family income erosion due to new financial burdens, such as addi- tional transportation costs or having to pay for services that were previously free or subsidized - Additional expenditures due to the rupture of community ties and interdependence that previously guaranteed services, such as child care and neighborhood security. As mentioned earlier, there are no systematic and reliable data re- garding the impact of completed Bank resettlement programs on family income in Latin America, or on local or home-based businesses. This problem is compounded by the general scarcity of research on informal sector economic activities, which are usually the largest source of 178 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT income for residents in marginalized areas. Even in cases where postrelocation surveys have been carried out (as in Rio Flood) there has not been systematic monitoring of the economic welfare of the resettled families. Usually only a narrowly focused sample survey is carried out to determine the level of satisfaction of relocated families with housing and new services. In the Minas Gerais project (municipality of Contagem) a survey was conducted recently in the new settlement of Camp Alto, located 20 kilometers from the old site, in which 68 families were resettled during 1993-94. The survey showed that 39 percent of the economically active population still were working in industries near the former site of the community, and 17 percent were self-employed in activities near the former site. In total, 56 percent depended on the physical and economic environment of the former site for their subsistence. Although the social team in charge of the resettlement concluded that there has not been a rupture of economic activities and that existing difficulties are mi- nor and transitory, the survey signals potential problems that require attention: * Of those surveyed, 58 percent cited problems related to distance, transportation, and scarce demand for labor in the area of the new settlement. * Thirty-five percent said that lower wages and higher costs for ser- vices in the new settlement affected their family income. * The community complained that insufficient education and health facilities forced them to use other schools and hospitals, thereby in- creasing transportation costs. The El Salvador Earthquake Reconstruction Project is an exception to the usual absence of economic welfare monitoring. At the Bank's re- quest a quick evaluation was carried out on the risks of impoverishment for families resettled in the Villa Italia District, 15 kilometers from the original settlement. The study showed that economic conditions had worsened for 6 percent of the resettled families because they had lost ac- cess to their usual sources of employment in the city's center, and also because of higher costs for transportation and services. A support strat- egy was designed and implemented for these families, with the support of a nongovernmental organization, consisting of a rotating fund for fi- nancing family microenterprises. It should be recognized, however, that each case is different and that distance is not necessarily a negative factor. For example, in the case of Parque Libertad in the municipality of Caixas, the residents chose to be URBAN RESETTLEMENT IN LATIN AMERICA 179 resettled in the Bento Riviero Danta housing development in Rio de Ja- neiro. This was far from their former site but had the advantage of being closer to thejob market. Early consultation with affected families and par- ticipation in decisionmaking were key factors to the success of this case. Finally, an independent evaluation and monitoring system has been set up to assess income restoration and recovery of prior living con- ditions in Yacyreta. Monitoring will cover a postrelocation period of one and a half years. New Challenges and Practical Recommendations The resettlement operations mentioned here, and many others, clearly reflect the progress made by development agencies in understanding, planning, and implementing projects that involve involuntary displace- ment of urban populations. Some of the most notable advances and pos- itive results are briefly mentioned below. Avoiding unnecessary displacement. In all the projects borrowers have made serious efforts to minimize involuntary displacement, often by searching for better design options. Table 4.7 summarizes some cases of design optimization that significantly reduced dis- placement. Table 4.7. Project Optimization to Minimize Displacement Reduction in Countryl number of project Optimization actions people displaced Brazil- The project (state component) has been opti- 1,300 Rio Flood mized through changes in the layout of ditches and water treatment services and the adoption of new construction technologies to minimize the impact on affected neighborhoods Brazil- The project has been improved by changing the 1,000 PROSAM, initial approach of moving the population Parana away from bodies of water and into areas with access to services (to prevent contamination). Instead, appropriate technology was intro- duced to improve water and sanitation infra- structure at the existing settlements. Colombia- The project has been optimized by changing 1,800 Santa Fe 1 some of the areas to be affected. 180 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT . Resettlement plans and policies. All the projects have resettlement plans generally consistent with Bank policy and with the norms and regulations of each country. * Taking into account the views of the affected population. Consultation and participation by the communities directly affected is a funda- mental part of all resettlement operations currently in progress. In all cases, direct involvement of the population affected is a key ele- ment in the implementation phase. . Participation from the earliest stages. Recent projects have shown progress in conducting consultation and participation from the first stages of planning, thereby contributing significantly to the so- cial feasibility and acceptance of the project. * New initiatives. Significant efforts have been made by the interdisci- plinary technical teams of the projects to provide sound technical, economic, environmental, and social solutions to mitigate the im- pacts of the forced displacement. In several projects, innovative in- struments have been developed for the replacement of housing, such as letters of credit, options to purchase, resident aid, contract- ing out services to third parties, and so on, which are more in har- mony with the reality of major urban centers in Latin America. * Growing institutional capacity. The technical capacity and accumula- tion of experience on the part of the social teams in charge of reset- tlement has slowly permeated and fortified the capacity of the borrowing institutions. * Interinstitutional coordination. Important progress in inter-institu- tional coordination through adoption of formal agreements has permitted a clearer designation of responsibility among agencies involved in resettlement. • Best practices. The process of analyzing and understanding past re- settlement experiences, while still weak, has yielded important les- sons that have been incorporated into current resettlement plans. * Financial support. At the borrowers' request, the Bank has aug- mented its participation in financing resettlement and rehabilita- tion plans, greatly increasing cash flow availability. * In summary, many resettlement projects have turned involuntary displacement from a solely negative experience into a source of positive benefits, such as: * Security and stability for the families removed from the areas at risk * Better housing, health, and sanitary conditions * Secure property rights * Increased value of family assets and access tobasic services. URBAN RESETTLEMENT IN LATIN AMERICA 181 Notwithstanding the progress achieved, there are still many prob- lems that must be studied and resolved to improve the planning, man- agement, and social outcomes of urban resettlement programs. The conclusions and practical recommendations that follow will help to overcome the aforementioned operational problems. Preparation Design of urban projects should be optimized and should incorporate resettlement as an integral component from the very outset. Many diffi- culties result from failure to optimize project design and from viewing resettlement as an obstacle rather than an integral part of urban devel- opment and rehabilitation. Last-minute changes in the location and im- pact area of projects, for example, or the accommodation of civil works (and other components) to the detriment of resettlement has sometimes meant that populations are resettled into housing units without finished services, water, and electricity (Rio Flood, Yacyreta). It also can result in the use of temporary shelters that do not meet even minimal conditions of hygiene and habitability. In Rio de Janeiro, due to the pressure imposed by progress in the main civil works, about 100 families were temporarily placed in rented houses before being moved to Missoes and Campo do America. Such measures usually involve high social costs, uncertainty, and discontent that could have been avoided if explicit designs and optimal technical solutions had been clearly defined before project approval or at the ear- liest possible stage. EARLY RESETTLEMENT ARRANGEMENTS. Early preparation of institu- tional and financial arrangements and particularly the timely definition of mechanisms for participation appear to be critical prerequisites for successful resettlement operations. SOCIAL ASSISTANCE AND ECONOMIC RECOVERY ACTIVITIES. These ac- tivities should be contracted to governmental or nongovernmental enti- ties specializing in this type of program. It is especially important to design support programs for microenterprises and home-based eco- nomic activities. Such programs have proved to be effective tools in re- storing income as well as facilitating the adaptation process and rebuilding of communities. ADOPTION OF A RESETTLEMENT POLICY AT THE NATIONAL, STATE, AND SECTORAL LEVELS. This is particularly important in countries with 182 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT numerous projects involving resettlement, because designing policies project by project is inefficient and time-consuming and can lead to in- consistencies that complicate project management. The Bank could offer technical assistance and help coordinate inter-institutional dialogue. STAFFING PROJECTS WITH ADEQUATE SOCIAL SPECIALIST TEAMS. Along with engineers, economists, and planners, the central coordinating unit of the project should include from the outset a basic team of social spe- cialists to evaluate the project and design of the resettlement plan. Be- fore project approval the Bank and borrower should agree on a minimum number of social specialists to contract during the implemen- tation phase of the project. RESETTLEMENT AS A PLANNING PROCESS AND NOT JUST AS DOCUMENT PREPARATION. Designing the resettlement plan as phases included in the legal agreement-with clauses synchronizing civil works with resettle- ment progress, and authorization requirements for beginning construc- tion of main civil works at each phase-will permit flexible management without unnecessary pressure on social programs and the pace of the work. Just as the resettlement plan needs to be flexible, so too must the civil works be responsive to the vicissitudes of the resettlement process, such as negotiation with the affected population, acquisition and preparation of new sites, availability of housing, new legislation, and political interference. All changes in the project should be submit- ted to the Bank for review and approval. CARRY OUT A SOCIAL ASSESSMENT (SA) AND ANALYSIS OF STAKE- HOLDERS. The SA will help identify the social actors affected by the pro- ject and ensure their participation in the planning and design of resettle- ment options as well as the distribution of associated costs and benefits. The findings of the SA will be a fundamental tool in designing socially feasible projects. They also should be a factor in defining criteria for dis- tributing costs and benefits (including the displaced population), finan- cial analysis and investment decisions, optimal use of project resources, and the price of services produced by the project. MORE THAN JUST REPLACEMENT HOUSING. Urban displacement and resettlement requires an integrated visionnot only of housing issues but also of economic and social impacts. Resettlement programs should in- clude not just housing agencies, which possess enormous experience with low-income housing, but also other institutions responsible for ad- dressing socioeconomic problems associated with forced displacement. URBAN RESETTLEMENT IN LATIN AMERICA 183 ADEQUATE ECONOMIC AND CULTURAL DATA. During preparation, in- formation should be included about the social, economic, and political networks-the modus operandi-of the so-called poverty belts or slums. Likewise, urban planners and government agencies should know about the survival strategies, solidarity networks, and systems of power in those settlements. Indicators of poverty should include additional data such as literacy, marginalization, and people's representation in decisionmaking institutions. Resettlement plans should include pro- grams for restoring productive activities. In selecting the resettlement site and planning income rehabilitation measures, it is important to as- sess the degree to which the economic activities of displaced families are linked to the community's original location. IMPROVED ESTIMATION OF THE COSTS, FINANCING, AND TIME FRAME FOR RESETTLEMENT. Underestimation of costs and the resulting unavail- ability of resources have often led to poorer housing quality or to pro- viding prepared lots rather than finished houses. There also is a tendency to underestimate the time necessary for implementation, par- ticularly for acquiring and preparing resettlement sites, obtaining li- censes, and carrying out renewal and rehabilitation projects. Bank resettlement programs in Latin America are typically 10 to 24 months behind schedule. RESETTLEMENT FINANCING. Bank participation in resettlement fi- nancing has been a key element in the success of some operations, while reliance solely on local counterpart financing has contributed to delays and problems in others. It is strongly recommended that the Bank and borrowers study the pros and cons of increased financing from external sources and facilitate the availability of sufficient local funds. INSTITUTION STRENGTHENING FOR IMPLEMENTING AGENCIES. Mea- sures should also be taken to strengthen the commitment and participa- tion of local governments in project planning and development. Specifically, the Bank should promote, and the borrower should be com- mitted to, the sharing of experiences between projects and countries. A RANGE OF ALTERNATIVES FOR THE AFFECTED POPULATION. This helps address the heterogeneity of displaced groups, recognizes their right to weigh different options and choose the most appropriate one, and permits more flexible management of the resettlement operation. The broader the range of alternatives for replacing land, housing, and services, and for restoring standards of living, the better the chance of 184 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT reducing negative impacts and having a successful outcome. Special emphasis should be given to preparing viable social and economic alternatives for families in extreme poverty, even if that implies large subsidies. CONTINGENCY PLANS. Projects in areas subject to frequent natural disasters such as flooding and earthquakes should have contingency plans for emergency evacuation of the displaced population. The plan should establish explicit institutional responsibilities and economic measures, and the corresponding economic resources should be con- templated within the project budget. Impleementation Immediately after project approval, counterpart funds for resettlement should be transferred to a special account to guarantee timely availabil- ity of resources for the implementing agency. An interdisciplinary team should be in charge of resettlement during the entire implementation phase and should include, where possible, architects and specialists in so- cial science, community participation, and social engineering. The princi- pal civil works and resettlement activities should be developed in parallel, phase by phase, through a special project management unit. It is critical that the resettlement team be part of the project management unit. Housing solutions involving costs to the resettled population should be studied carefully. Requiring those displaced to bear part of the resettlement cost, particularly when no other alternative is offered, can have serious negative effects on precarious family incomes. Such ar- rangements should be rigorously analyzed and monitored to avoid fur- ther impoverishing already marginalized populations. Plans involving monetary compensation should be carefully evalu- ated and monitored and should include safeguards to prevent families from falling into poverty. Difficulties with housing availability and the advance of civil works make cash compensation an attractive option for project planners. While there is no systematic tracking of resettlement outcomes, experience throughout the world has shown that the cash payment alternative does not guarantee replacement of housing or pre- vious living conditions. Indemnification in money may be an adequate option for groups with higher incomes and greater capacity to self-relocate, but it should not be an "easy out" for planners and man- agers to avoid finding viable solutions for vulnerable low-income families. In general, there are no effective regulations or measures to control invasions by squatters in the project area or in the resettlement sites. One URBAN RESETTIEMENT IN LATIN AMERICA 185 tactic is to start the civil works as soon as possible after the population is moved to prevent new squatters from taking over the abandoned site. Before resettlement takes place, the affected communities themselves can often help control invasions, in both the original site and the future site. Once a census has been conducted, and as long as there are no sig- nificant delays, the project area should be "closed" and agreement should be reached with residents that newcomers will not be entitled to resettlement or cash compensation. A systematic monitoring program should be planned and put into place immediately after the census has been conducted. Indicators to measure restoration of social networks and economic activities, adapta- tion to the new environment, and improvement in living conditions should be clearly defined and routinely incorporated into all resettle- ment processes. Quarterly monitoring reports should include specific recommendations for fine-tuning policies, modifying housing alterna- tives and assistance, and avoiding emergencies in future phases of the resettlement. Special attention should be given to field monitoring of economic activities, employment, and family income when the new site is far from the original settlement or job market, or is located in the urban outskirts or in a different municipality. Resettlement should be supervised, and the implementing agency should demand the necessary technical assis- tance during implementation. Supervisory support from the Bank must be planned to allow time for evaluating problems and designing solu- tions together with the agency. In addition to the traditional indicators of success for urban resettle- ment-replacement of housing, quality of housing, access to services, improvement in quality of life, and degree of satisfaction-the comple- tion report should evaluate whether the risks of impoverishment have been counteracted and to what degree economic activities have been re- covered using such indicators as: * Effects on home-based employment and economic activities * New economic activities * Evolution of family income * Cost of basic family necessities in the new place * New expenditures in services * Value of family assets * Changes in sociocultural behavior * Sustainability of the new settlement. Post-project evaluation of results is essential. With the exception of Yacyreta, there are no independent ex post evaluations of urban resettle- 186 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT ment programs in the Latin American and Caribbean region. Few pro- jects have been systematically evaluated; most reviews are only frag- mentary. This is one of the reasons why decisionmaking related to resettlement policy is largely based on speculation and subjective inter- pretations-and why there has been an endless repetition of failures. It is especially important to include analysis of the living conditions of people who received monetary compensation and to adapt prevailing policy accordingly. The ex post evaluation of each resettlement should begin after the physical transfer of the families and establishment of ser- vices is completed. A detailed report of the results of resettlement should always be part of the project's completion report. Acknowledgements The author especially wants to thank Laura Mancini (Rio de Janeiro) for gathering the basic information, much of which was used in this chap- ter, as well as for her contribution to our understanding of urban reset- tlement procedures in Brazil; the resettlement coordinators-Paula Pini (Sao Paulo), Soraya Melgaco (Minas Gerais), and Angela Dias (ParanA)- for the information provided and their contributions to the develop- ment of the resettlement component in the urban renewal projects; the Yacyreta Department of Complementary Works for its contribution to the analysis of urban resettlement experiences in hydroelectric projects; Kathy Johns for her constant and valuable collaboration and assis- tance in the preparation of the study; Scott Guggenheim and William Partridge for their critical contributions to the ideas expressed in this chapter; Michael Cemea for his assistance and guidance; and Peter Brandriss for his help in editing the text. Notes 1. Most of the data utilized here are drawn from sources (such as statistics and field reports) covering several years, through December 1995. By the end of 1996 there were 11 more urban operations entailing involuntary displacement in the Bank's Latin American portfolio, and additional projects were initiated in 1997. This chapter does not purport in any way to be a status report on these projects or a comprehensive review to date of their performance, because like all other pro- jects, performance in these projects varies with better or weaker results in differ- ent implementation periods. Rather, this paper intends to highlight the specific nature and most typical issues of urban projects as a subcategory of projects en- tailing resettlement, using the data about the context and content addressed in these projects to illustrate successes, problems, and challenges. URBAN RESETTLEMENT IN LATIN AMERICA 187 2. Sources: Rio de Janeiro, Program Expectative-Ativa 1993; Sao Paulo, CDHU report March 1995; Minas Gerais (Contagem), SUDECOM report August 1993; Parana (Curitiba), PARDES-COMEC reports 1, 2, and 3, 1994. 3. The objectives of the Bank's involuntary resettlement policy are set forth in Operational Directive 4.30 (World Bank 1990). 4. The economic and social vulnerability of each family is assessed through sociocultural and economic field research, and relocation is offered only to those with high vulnerability, especially those facing a serious risk of impoverishment and social trauma. 5. See World Bank Operational Directive 8.10 on the Project Preparation Fa- cility (World Bank 1994). References The word "processed" describes informally reproduced works that may not be commonly available through libraries. Burki, S. J., and S. Edwards. 1995. Latin America after Mexico: Quickening the Pace. Washington, D.C.: World Bank. Cernea, Michael M. 1989. 'Metropolitan Development and Compulsory Popula- tion Relocation: Policy Issues and Project Experiences." Regional Development Dialogue 10:4. - 1992. Urban Environment and Forced Poptulation Relocation. World Bank Working Paper 152. Washington, D.C.: World Bank. _. 1995. "Urban Settlements and Forced Population Displacement." In Hari Mohan Mathur, ed., Development, Displacement, and Resettlement: Focuts on Asian Experiences. New Delhi: Vikas Publishing House. - 1997. "The Risks and Reconstruction Model for Resettling Displaced Populations." World Development 25:10: 1569-87. Davidson, Forbes, Mirjam Zaaijer, Monique Peltenburg, and Mike Rodell. 1993. Relocation and Resettlement Manual: A Guide to Managing and Planning Reloca- tion. Rotterdam, Netherlands: Institute for Housing and Urban Development Studies. Lomnitz, Larissa Adler. 1977. Networks and Marginality: Life in a Mexican Shanty- town. New York: Academic Press. Mejia, Maria Clara. 1996. "Economic Recovery of Brickmakers in the Yacyreta Project: Argentina and Paraguay." Presented at the conference "Recon- structing Livelihoods: New Approaches to Resettlement," Oxford University, United Kingdom. Processed. Ninio, Alberto. 1995. "Reassentamento Involuntario em Projetos Financiados Pelo Banco Mundial: Requerimentos y PrAticas." World Bank, Washington, D.C. Processed. 188 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Partridge, William. 1990. "Involuntary Resettlement and Development Pro- jects." Journalfor Refuigee Studies. October. Peattie, Lisa Redfield. 1970. The View from the Barrio. Ann Arbor: University of Michigan Press. Perlman, Janice. 1982. "Favela Removal: Eradication of a Lifestyle." In Art Hansen and Anthony Oliver-Smith, eds., Involuntary Resettlement and Migra- tion. Boulder, Colorado: Westview Press. United Nations-World Bank. 1993. "Guidelines for Managing Urban Relocation and Resettlement." World Bank, Washington, D.C. Processed. World Bank. 1990. Operational Directive 4.30 "Involuntary Resettlement." Washington, D.C. Processed. .1993. "LAC Resettlement Review." Latin America Environment Unit Dis- cussion Paper 152. Washington, D.C. Processed. - 1994. Operational Directive 8.10 "Project Preparation Facility." Washing- ton, D.C. Processed. . 1995a. "LAC Annual Review on Portfolio Performance-Resettlement." Latin America and the Caribbean Regional Office, Technical Department, En- vironment Unit. Washington, D.C. Processed. . 1995b. "LAC Regional Remedial Action Plan-Resettlement." Latin America and the Caribbean Regional Office, Technical Department, Environ- ment Unit. Washington, D.C. Processed. . 1996. "Resettlement and Development: The Bankwide Review of Projects Involving Involuntary Resettlement, 1986-93." Environment Department Paper 32. Washington, D.C. Processed. 5 Testing the Risks and Reconstruction Model on India's Resettlement Experiences Lakshman K. Mahapatra Editor's This chapter offers a vast synthesis of resettlement Note experiences in India, centered on the economic risks and the socially adverse impacts of forced displace- ments, as well as on the main approaches for over- coming risks and impoverishment. To build this synthesis, Mahapatra employs a distinct theoretical framework-the im- poverishment risks and livelihood reconstruction model-and reviews empirical findings from numerous field investigations of resettlement in India over the past four decades. Within the conceptual and methodological framework adopted for his analysis, the author builds a broad socioeconomic panorama of resettlement processes in India. One by one he documents and emphasizes the loss of income-generating assets, employment, and common property; the declines in health levels and nutrition; and the loss of education and of social and human capital. Because the analysis of reconstruction processes found less empirical evidence than expected on successful recovery approaches, the author out- lines the key directions in which such efforts should be further car- ried out. The study also discusses the paucity of economic research on resettlement in India, in contrast with the vast anthropological and sociological literature. Unaddressed equity issues are highlighted, while neoclassical economic approaches and narrow cost-benefit analysis are criticized. Mahapatra concludes by outlining the unfinished tasks in the practice and study of displacement and rehabilitation and by sug- 189 190 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT gesting priorities for each. Those tasks include the elaboration of gov- ernmental policies for resettlement and for the effective reestablishment of income and livelihood, strengthening institutions for resettlement operations, supporting the work of NGOs, and fo- cusing more economic research on resettlement in India. In this study I am concerned with the economic and social content of in- voluntary population displacement in India, since it is induced pri- marily through infrastructure development. My interest in the economics of resettlement stems from my concern with global humanis- tic issues in the Indian context. These concerns have impelled me to en- gage in about three decades of independent anthropological studies of resettlement and rehabilitation. Therefore, I propose to undertake a comprehensive reexamination of the empirical evidence available from India's many resettlement operations, funded domestically or jointly with various donors, and to draw from this synthesis some key theoreti- cal, policy, and operational conclusions. To this end I will employ and rely upon two main research tools: *A conceptual model that encompasses both the risks of impover- ishment through displacement and the strategies for reconstruc- tion (Cernea 1990,1996b, 1997) . The rich thesaurus of empirical evidence and interpretations on resettlement generated by Indian researchers over the past four decades. The theoretical model that I will test against empirical evidence is the "impoverishment risks and livelihood reconstruction model" devel- oped by Michael Cernea. This model, itself a recent research product, is based on, and summarizes, findings on resettlement on a worldwide scale. The interest of India's resettlement researchers and practitioners in the model proposed by Cernea is substantial, as shown by three most recent books on resettlement published in India constructed around the analytical model of impoverishment risks (Ota and Agnihotri 1996, Pandey and associates 1998, Mathur and Marsden 1998). Thus it makes sense to apply this new model to, and explore in depth its validity for, the full scope of Indian experiences. In turn, the factual source material on India represents probably the richest body of empirical data on resettlement available in any single country so far and can therefore provide grounds for either confirming TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 191 or negating the proposed risks and reconstruction model. This vast store of seemingly diffuse findings on resettlement, accruing from numerous sites and produced by many different researchers, needs to be synthe- sized, unified, and structured along some key variable. The question is whether, or how, the theoretical model can assist in unifying these data. I will also review the main theoretical elaborations in economic and so- cial research developed by India's resettlement scholars-not just the empirical findings-and examine the risks and reconstruction model against this body of theoretical interpretations Though these may seem ambitious goals for a single study, the need to advance conceptualizing and modeling in Indian (and world) reset- tlement research is the best justification for at least a first step in such an undertaking. Our intention is also to nudge more Indian researchers- especially economists (and not just anthropologists)-into recognizing the seriousness of the economic risks inherent in resettlement and to call on them to elaborate in more detail the economics of reconstruction and rehabilitation. We trust that others will follow up, add to, and improve on what we can achieve through this study. Estimates of India's displaced population from all projects over the past 50 years (1947-97) vary, but the total may be as many as 25 million since independence (Fernandes 1994 for the 1951-90 period only, with my extrapolation to 1997). For water resources development alone the world has already constructed 34,798 dams, of which India has built some 3 to 4 percent (Times of India 1997). India's Central Water Conmmission in March 1990 recorded 465 major dams of 30 meters and higher, while dams of 20 meters and above numbered 1,174 (Fernandes 1991). This clearly indicates the macroeconomic scale of displacements in India. India is not the only country, of course, in which massive displace- ments take place. Every year since 1990, about 10 million people have been displaced involuntarily all over the world in the wake of infrastructural development projects alone (World Bank 1994, Cernea 1997). Of these, some 6 million have been displaced by urban develop- ment and transport programs. Conceptual Frameworks and Theoretical Perspectives Clarifying the key concepts from the outset will be useful. The voluntary migration of people, either within the nation state or outside, must be distinguished from involuntary displacement, where people are forced to leave their dwelling places-under physical threat, if not actual applica- tion of force, in the event they do not vacate. Resettlement in the Western literature encompasses the processes of displacement and reestablishment. In India, however, the usual practice 192 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT is to differentiate between resettlement and rehabilitation. Resettlement in India is taken to mean simple relocation after physical displacement from the original habitat. In turn, rehabilitation is defined as "grafting a community at a new place and nurturing it to ensure its steady and bal- anced growth" (Joshi 1987). Elsewhere, the concept of rehabilitation may be interpreted rather narrowly as referring to a household, or a sec- tion of the targeted groups, or even an individual such as an orphan or a displaced person who is physically handicapped. But the basic thrust re- mains: rehabilitation focuses on achieving sustained development for displaced people. Cemea pointed out that the distinction made in India captures the fact that Rehabilitation .. . refers to the fate of the displaced people after relocation and to the reconstruction of their patterns of socio- economic organization .. . In India, a country tested by much population displacement and distinguished by a vast social sci- ence literature about it,. .. two distinct concepts are used for the two post-displacement phases of this process: resettlement and rehabilitation (R&R). The Indian legal and sociological litera- tures uniformly emphasize that "rehabilitation" does not occur automatically just after relocation ... Indeed, resettlement may oc- cur without rehabilitation, and unfortunately it often does (Cernea 1996a, author's italics). We emphasize further that resettlement and rehabilitation are two in- terrelated processes that form a continuum, partly overlap, and should not be seennecessarily as sequential "stages." On the contrary: tobe suc- cessful, rehabilitation processes may be, and sometimes should be, initi- ated before physical displacement of people takes place. As a sociological concept risk can be defined as "the potential that a certain course of action will trigger future injurious effect-losses and de- struction" (Cernea 1997). Specific economic risks and other risks of im- poverishment are embedded in all displacements. Stages of Resettlement and Rehabilitation Displacement-development dynamics havebeen studied and conceptu- alized by some scholars in terms of stages or phases, sequentially. Scudder identifies four stages in what he calls a "dynamic model of set- tlement processes," built upon the three-stage framework of Robert Chambers (Scudder 1985). Scudder's four stages are (1) planning and settlement recruitment, (2) transition, (3) potential economic and social development, and (4) handing over and incorporation. When they oc- cur, the four stages cover at least one generation of the displaced people, TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 193 beginning with the planning of resettlement after displacement. But Scudder notes frequent exceptions to this less-than-general model be- cause "not all projects pass through all stages or pass through them in the order outlined here ... A steady movement through the four stages is the exception rather than the rule" (Scudder 1985). Scudder's model was formulated initially to reflect voluntary settle- ment schemes with four stages, then was extended to involuntary reset- tlement, despite the very different circumstances that characterize these two processes. Voluntary settlements, for example, rarely have tran- sitional resettlement colonies, which are unavoidable in many projects involving resettlement and rehabilitation under involuntary displace- ment in developing countries. There are numerous other structural, cultural, and political differences between these two types of socioeco- nomic processes, which create inconsistencies if they are included in the same model. Scudder's modelhas proved tobe of limited value in the Indian con- text. As learned from predisplacement and postdisplacement studies in India, notably the remarkable resettlement in Orissa State during 1950-90 (Pandey and associates 1998), these four stages may go hay- wire, because there is no synchronization between government efforts and the initiatives taken by the ousted persons to plan and invest for their own rehabilitation. The process, as it unfolds on the ground, may differ significantly from the progression suggested by Scudder's four stages. This is especially so when land is not provided by the govern- ment or the project and the displaced persons themselves have to scout for land and assets to restore their livelihood. In most cases the land provided by authorities is not fully reclaimed, is rocky, or is upland with no irrigation facility. It is no surprise then that there are large-scale reverse flows of ousted persons from the govem- ment-provided colonies to their old village sites-or to whatever re- mains above flood level, or to sites along the rim. As to resettlers' organization, posited by Scudder to belong to the second stage, or tran- sition, we found that in many predisplacement villages ousted persons had organized during or even before stage one, in order to fight for higher compensation or better resettlement sites or amenities. The Impoverishment Risks and Reconstruction Model Displacement involves a long, complex process of human interactions. Each case has its own story rather than occurring in standard, sequential stages: its own economy, ecology, sociocultural disruptions (class and ethnic tensions), technological and bureaucratic constraints, and pat- terns of recovery and reconstruction. 194 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT The most frequent end result is that rich farmers, bureaucrats, and people with power become the "haves," at the expense of the "have-nots," who overnight become even more deprived-landless, jobless, homeless, and poor, perhaps begging for their food, vulnerable to increased morbidity and mortality, socially disoriented, and politi- cally powerless. These processes of impoverishment and social disinte- gration, and the means of overcoming them through the rebuilding of a life-support system, are central issues and challenges in involuntary dis- placement and reconstruction. Starting from this key point, and based on a bold comparative anal- ysis of many resettlement experiences around the world (especially in the developing world, including China and India), Michael Cemea has formulated and proposed a complex and comprehensive model of "im- poverishment risks and reconstruction of livelihood." This theoretical model has evolved over two decades of analysis, monitoring, and evalu- ation of involuntary resettlement. As a conceptual framework it has been outlined in several studies, first in a Harvard paper (Cernea 1990) and then in several other studies (World Bank 1994; Cernea 1995,1996b, 1998, and, in a comprehensive and refined form, Cemea 1997. Further in this study, references will be made to the 1997 version). The strength and methodological novelty of this model derives from its integration of impoverishment risk analysis with recovery analysis. This essential integration also equips the model with opera- tional capacity as a toolfor action, in addition to its explanatory or de- scriptive function. This is the main reason the model seems eminently suitable for understanding the forces, factors, and processes at work in the displacement-rehabilitation-development dynamics in India. Cemea's model identifies the eight most important dimensions of impoverishment: Landlessness (expropriation of land assets) - Joblessness (even when the development project ostensibly creates some temporary jobs) * Homelessness (loss of not merely the physical house, but of the fam- ily home and cultural space, with resulting alienation and "placelessness") * Marginalization ("downward mobility"-socially and psychologi- cally, as well as economically) * Increased morbidity and mortality (especially among the weakest seg- ments of the population) * Food insecurity (low daily calorie intake, malnourishment) TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 195 * Loss of access to common property (such as forests, bodies of water, and wastelands, which substantially supplement the food and in- come of lower-income groups) . Social disarticulation (loss of social, economic, and moral support among kinsmen and members of community networks, leading to social anomie). These basic risks have varying intensities, depending on local con- ditions and on the nature of the projects (Cernea 1997). It is very important that the model also captures the key pro- cesses that could counteract the impoverishment risks and lead to the economic and social reestablishment of the displaced. These pro- cesses-based on land and employment, on restoration of social ser- vices for health and education, and on community reconstruction-will be examined later in the chapter. Whether these reconstructive pro- cesses occur is a matter of governmental policy, resource allocation by project initiators, and the initiative of resettlers themselves. The model is convincing and practical precisely because it offers a conceptual framework designed not only to explain but also to trigger reconstruc- tive processes and policies. Cernea's model is seminal, too, in that it is not a closed or finite framework. Appropriately, it offers scope for further thinking and con- ceptual development. Cernea rightly notes that the "eight impoverish- ment hazards are not the only ones that result in processes of economic and social deprivation, but rather the most important ones" (Cernea 1997). The model can accommodate and include other risk variables, for instance, to which I would add: edutcation loss among displaced children. In a recent study Cernea agrees and notes, "Indeed, relocation often in- terrupts schooling and some children never return to school. After dis- placement, as a result of losses in family income, many children are drafted into the labor market earlier than what would otherwise occur" (Cernea 1997). A refreshing aspect of this model is that Cernea is concerned not only with economic impoverishment, but also with the loss of social and cultural endowments. The model contends that during displacement people lose natural capital, manmade (physical) capital, human capital, and social capital. Cernea then builds a reconstruction model aimed at enabling and helping displaced people 'to restore their 'capital' in all its multifaceted forms" (Cernea 1997). The components in this reconstruction model would therefore re- quire not only adequate compensation for property losses, but also 196 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT adequate rebuilding of the displaced people's income-generating capacity and livelihoods, particularly in terms of the four lost "capitals" (World Bank 1990,1994; Cernea 1997). What is needed, Cernea argues, "is a change in concept and method predicated on treating resettlement operations as opportunitiesfor development, as development projects in their own right. This includes risk mitigation but goes beyond it to construct a new socioeco- nomic basis on which resettlers' livelihoods can first be restored and then improved, so that their 'income curve' could exceed predisplace- ment levels" (Cernea 1995, 1997; Shi and Hu 1994). The eight components of the reconstruction process-land-based reestablishment, reemployment (through either self-employment or jobs created by or for the project), house reconstruction, social inclu- sion (status improvement), better health care, adequate nutrition, res- toration of community assets (creation of new common property resources), and community reconstruction (social reintegration)-are designed to reverse eight impoverishment risks. To these one could add education restitution as a follow-up to the addition of the impoverish- ment risk of education loss. Cernea correctly adds that "it is important to repeat that the risks and reconstruction framework emphasizes their in- terdependence and synergy" (Cernea 1997). Cernea's model, unlike Scudder's, is not tied to a rigid time frame or invariable stage sequence, nor is it weighed down by assumptions (sometimes unwarranted, arbitrary, or merely rhetorical) of social change, modernization, and irreversible progress. The Social Actors in Resettlement An essential complement to the overall model of resettlement is the identification of the key social actors who participate in this process. Based on the Indian experience with involuntary displacement and var- ious field observations, my own view is that in the context of Indian soci- ety it is possible to differentiate the following social actors: • Development agencies: state sector, private sector, and joint sector, especially in India, where the private sector is promoted, sup- ported, and facilitated by the state sector in joint ventures because of its financial or technical inputs. * Project beneficiaries: specific (for example, landowners in the com- mand area, industrial enterprise, mining company), general (for example, the regional development, national biodiversity, or de- fense capability), and intermediaries (for example, ancillary indus- tries, traders, service sector). TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 197 * Development operators: planners and executives at the implemen- tation level (technocrats and bureaucrats in the field). * Development facilitators: planners and policymakers (politicians, technocrats, bureaucrats), NGOs and activists who mobilize af- fected people, social analysts and resettlement researchers, peo- ple-sponsored institutions (for example, oustees' organizations, associations of affected ethnic groups, organizations of more vul- nerable groups such as women). . The adversely affected: these include persons, families, and vil- lages immediately affected, both directly (for example, village displacees) and indirectly (individuals or groups upstream or downstream), and persons or groups affected because of their particular vulnerability (scheduled tribes, scheduled castes, women, children, the handicapped, service providers, artisans, the landless). It also includes those secondarily affected, such as host communities. The above listing differs marginally from the general set of actors identified by Cernea (1996a), because in this study we are taking into ac- count India's stratified social structure. But we similarly emphasize that all the key actors must deliberately participate in risk prevention and re- construction management. Essential in this respect is achieving adequate communication be- tween planners and policymakers, on the one hand, and resettlers, on the other. When communication processes are absent or break down, the result is what Cernea calls "negative participation," that is, "active opposition movements against development" (Cernea 1997). The common experience in India is that development operators and development facilitators-especially in the state sector, private sector, and joint ventures-usually overlook or underestimate the indirect im- pacts of the project on affected or vulnerable groups and the secondary impacts on host communities. Indian researchers, such as Upendra Baxi, have correctly and fre- quently deplored the Indian pattern of "development without partici- pation," using as particular examples projects that cause displacement. Baxi writes, "No consultation with the people in general or the affected constituencies is thought necessary in any single one of the following matters: the site of dams; size of dams; allocation of resources, including international aid; design, including safety designing; epidemiological impact analysis; contracts for construction; flow of benefits to certain classes/sectors; displacement; rehabilitation!" (Baxi 1989). As a matter of fact, there is no genuine consultation with the affected people even on 198 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT such cultural issues as the desirability of a project or the selection of re- settlement sites after displacement. Interestingly, the World Bank has sought in recent years to improve participation in resettlement by requiring its borrowers (since 1993) to make resettlement plans publicly available (in draft form in the project area) for review and comment by local NGOs and other local institu- tions and organizations. Encouraging and securing the active participa- tion of the people involuntarily displaced in solving resettlement problems is essential-more important than fostering participation in many other types of projects (Cernea 1995). And participation of the host communities may similarly bring about a smooth transition to com- munity reintegration, thus nipping conflicts in the bud. The Economics of Resettlement: Missing Research in India Economists have studied poverty and impoverishment in developing countries, including India. But there is little writing by Indian econo- mists devoted to the specific types and processes of impoverishment arising because of expropriation and displacement for development projects. There has been, however, a lot of thought, debate, and writing on the cost of irrigation projects and their cost-benefit ratio. In 1964 a spe- cial committee appointed to suggest ways and means of "improving the financial returns from irrigation projects" recommended that an eco- nomic benefit criterion be adopted in place of the financial productivity criterion for a selection of irrigation projects. The committee recom- mended a cost-benefit ratio of at least 1.5 for project approval, or more, if possible, to leave room for unanticipated cost escalations. When he eval- uated the Tehri Dam by balancing its benefits and costs, economist Paranjpye noted that the cost-benefit ratio tries to weigh societal bene- fits and societal costs. "But since no distributional weights have been at- tached, the [cost-benefit] ratio does not tell us the distributional impact of this investment; though we get to know what benefit [can be ob- tained] at what cost, we cannot ascertain "whose benefit" at "whose cost" (Paranjpye 1988). Cost-benefit analysis does not aim to achieve balanced income distribution, or regional balance, or other forms of dis- tributive equity. Only those costs or benefits that are amenable to mone- tization are included in cost-benefit ratio calculation. Paranjpye correctly noted, "The problem of rehabilitation of displaced persons is merely represented as a statistically insignificant value measured in terms of monetary compensation" (Paranjpye 1988). He goes on to assert that this leads to wrongly treating the demands of rehabilitation TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 199 as just an "obstacle" to be brushed aside from the path of "national development." Solid, extensive research by Indian economists on the economics of resettlement is sorely missing. It is surprising that, over several decades, when economists have undertaken special studies like the economic survey of Orissa State (Misra 1960), or produced reports on the benefits of Hirakud irrigation (Government of Orissa 1968), or when the Na- tional Council of Applied Economic Research prepared (in 1977) the Perspective Plan of Economic Development of Orissa from 1974 to 1984, they have shown no intellectual interest in the many forms of impover- ishment experienced by project-affected persons. Nor are current main- stream economists concerned with the economic and financial dimensions of reconstructing displaced persons' livelihoods. It is as if the development projects they study did not result in the displacement of any people. By contrast, India's noneconomist social scientists, primarily an- thropologists and sociologists, but also some jurists and political scien- tists and many environmentalists, have given due importance to the forms of impoverishment, before and after expropriation, of pro- ject-affected persons. An important center of these studies in India has been the Department of Anthropology, Utkal University, where my col- leagues and I began to look into these issues as early as the mid-1970s. Another noteworthy center is the Indian Social Institute, New Delhi, where Walter Fernandes (sociologist) and his colleagues have taken up, both academically and as NGO activists, impoverishment and recon- struction issues. Important research has been carried out by the Surat Center for Social Studies and particularly by researchers associated with the Tata Institute for Social Sciences in Mumbai. B. K. Roy-Burman, the eminent and most senior anthropologist in India, who carried out the pi- oneering displacement and impoverishment study on the persons dis- placed by the first post-independence steel mega-industry at Rourkela (Roy-Burman 1961), has recently compared the rehabilitation provi- sions of the 1950s with those of the Narmada Sardar Sarovar project pol- icy. He notes that "the rehabilitation pattern as prescribed at Rourkela in the 1950s was certainly much more liberal and comprehensive and was inherently community-oriented" (Roy-Burman 1995). The lack of adequate economic information on resettlement is both a cause and a consequence of the absence of economic research on reset- tlement. Like many other developing countries India has been bedev- iled by a lack of reliable and transparent official information. The area to be acquired for a development project, and its structure by type of ten- ure and use, are often not made public in a clear manner. As to the 200 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT proportion in the overall population of scheduled tribes, scheduled castes, or vulnerable groups such as the handicapped, artisans, landless laborers, service providers, and orphans, the information either is not available or differs from one official agency to another (Mahapatra 1990). In the official documentation for projects only rarely can one find data about resettlement costs beyond compensation paid to the pro- ject-affected families. No rehabilitation expenses are computed and re- corded (Government of Orissa 1991). The official data we examined from the IDA-funded projects in Orissa (1960-83) did not mention any expenses incurred apart from compensation paid, although some land-for-rehabilitation grants were given for house construction. The lack of full economic information is, in part, attributable to the mindset of the officials in charge, who try to keep secret the information base on rehabilitation. They regard it as sensitive, with the potential of leading to agitations from the population and NGOs. Only since the late 1980s has there been better generation and sharing of the information base, thanks to international funding agencies' demands on the Indian government, as well as to pressures brought to bear by local NGOs. The insufficiency of economic debate and analysis also obscures the profound issues of equity and sustainability that arise when develop- ment programs result not only in beneficiariesbut also in new cohorts of impoverished people. I have treated this set of issues elsewhere, under the heading "Development for Whom?" (Mahapatra 1991). It is gratify- ing to see the issues of equity and sustainability with respect to resettle- ment taken up recently by an increasing number of scholars, including Indian and foreign students of India's resettlement experiences (for ex- ample, Fisher 1995, and on sustainability Morse and Berger 1992, Dreze, Sampson, and Singh 1997). A scholar with the reputation of M. S. Swaminathan wisely observed that "development which is not equita- ble, will not be sustainable in the long term" (1996). And in what they called a "class-benefit analysis," Singh, Kothari, and Amin noted, "In virtually every project it is seen that the primary costs are being paid by the poor and the tribals, while the benefits are flowing to big farmers and the urban elite. Those who are displaced by such projects are usually too poor and politically weak to safeguard their own interests" (1992). Mathur, in turn, deplores the exclusion of the poor: "In sharing ben- efits from development, the poor lag behind the rich. The rich are quick to seize opportunities which open up with the inauguration of develop- ment projects. The new jobs go to those who have the requisite skills, but their lack among the poor is common" (Mathur 1995). Among the dis- placed, "the affluent groups do not lose so completely. They are in a better position to adjust to change. In fact, they prefer to resettle TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 201 themselves with resources of their own rather than endure governmen- tal administration of their lives. For them uprooting is certainly less traumatic (Mathur 1995). Many other field researchers in India corrobo- rate this observation (Nayak 1986, Baboo 1992). In neoclassical theory the problem of social costs has been dismissed with the rationale that in some instances remedial measures have been taken either by the government or by private organizations. Such social costs are treated as the inevitable price of economic development, and the social losses are justified as exceptional cases or minor disturbances. We can only say that such neoclassical interpretation is clearly anachronistic today. Yet this view still guides practice and is not explic- itly rejected by modem Indian economics. The net result is that painful types of social and human costs are not taken into account and effective remedial measures for rehabilitation are in most cases not taken. More- over, involuntary displacement in the coming decades in both rural and urban areas in India is expected to continue on a considerable scale. Consequently, issues of equity and of the exclusion of persons displaced by development are bound to become even more pressing and galling in the near future (Reddy 1996, Cernea 1993). Impoverishment Risks and Social Exclusion We turn now to the typology of impoverishment as identified in Cernea's model of impoverishment risks and examine its eight elements directly, one to one, against empirical evidence from predisplacement and postdisplacement studies in India. This evidence, culled from re- search carried out by many Indian scholars, is only a fragment of the much richer evidence we found in their original studies. Landlessness Loss of land sets the level of impoverishment in involuntary displace- ment. The model correctly defines causality: "Expropriation of land re- moves the main foundation upon which people's productive systems, commercial activities, and livelihoods are constructed. This is the prin- cipal form of de-capitalization and pauperization of displaced people, as they lose both physical and man-made capital" (Cernea 1997). In India, land is often lost, not only by those who are ousted, but also by people who are not physically displaced. When land lost reduces a landholding to an uneconomical size, impoverishment sets in, though households affected in this way are often denied the assistance that would be accorded to a displaced household. Landlessness also brings about change in occupation, and in the ability to hold assets (like 202 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT livestock). This reduces the food supply and the resources base used to secure other necessities. An Indian peasant, for example, often sells part of his production to meet other needs, including payment of taxes and gifts of food (Pandey and associates 1996, 1998; Mahapatra 1997). The state-owned industry and mining sectors have even experi- enced clear regression in rehabilitation policy. In the late 1950s, when the first steel mega-industry was established at Rourkela, the govern- ment of India provided the displaced with land for land rehabilitation, in addition to compensation and a resettlement colony (Roy-Burman 1961). Later, in 1967, the T. N. Singh formula promised one job per dis- placed family in industries and mines, but without any land for land re- habilitation. As most of the displaced belonged to tribal or to other largely illiterate, vulnerable groups and did not possess the skills re- quired, only low-paid, unskilled jobs, if any at all, were made available to them, and most of these were temporary. Later, with the reduction of unskilled jobs because of mechanization, the Bureau of Public Enter- prises abandoned this scheme (Femandes and Asif 1997). The Constitution of India, Section 31 A(1), provides for "payment of compensation at a rate which shall not be less than the market value thereof" (India 1993). But the procedures involved in setting compensa- tion levels and the payment of compensation before displacement have led to impoverishment and destitution, rather than equitable restitution and reconstruction. One reason is that only immovable properties with title deeds or other valid evidence of absolute ownership entitled one to compensation, while many tribal groups own land customarily, with- out formal title deeds. Also, for many historic and current reasons land is undervalued in land records. The "market value," which is the stan- dard for compensation under the Constitution as well as under the Land Acquisition Act, is very much on the low side. Compensation paid based on the capitalized value of the annual production of the land, as determined by the law courts, or based on the land's replacement value, would have been more realistic and fair to the displaced (Mahapatra 1996a). For trees the market value for compensation has been fixed abso- lutely arbitrarily, without any reference to the market. This has been studied in depth by Pandey (Pandey and associates 1996, 1998). Payment of compensation in cash to tribal people, whose economy is largely nonmonetized, has invariably left most of the money in the pockets of brokers, bootleggers, gamblers, or confidence tricksters (Mahapatra 1994; Pandey and associates 1996,1998; and Fernandes and Raj 1992, among others). Typically, "the money lenders cooperated with the local leaders in trying to convince the displaced persons to accept monetary compensation instead of land-based resettlement," for obvi- ously self-serving purposes (Stanley 1996). And when compensation TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 203 was paid very late, sometimes not even after 15 years, as in Natedi vil- lage under the Mahanadi Coalfields Project (Pandey and associates 1996,1998), the displaced persons' movable assets were used up even as their land was inexorably swallowed up by the mine. Furthermore, India's experience has confirmed time and time again that without replacement value as compensation, it is not possible for the landowning displaced persons and other project-affected persons to buy land. And landlessness has resulted not only in destitution, as it pushed more people below the poverty line (Reddy 1996), but also in en- during trauma and listlessness in many cases (Nayak 1986). It even caused some women to suffer mental disorientation because of their utterly helpless condition (Das, Das, and Das 1996; Baboo 1992). This trauma is compounded when joblessness, another grievous loss, is added to it. Joblessness Loss of employment by wage earners caught in the turmoil of forced dis- placement represents, as the model we examine correctly posits, a fun- damental risk and, unfortunately, a widespread reality in India. And the risk of joblessness is particularly difficult to counteract, even if a job is secured. When rehabilitation is land-based, most members of a dis- placed family are likely to remain gainfully involved in work. But when rehabilitation is job-based, as in mining and industry projects, typically only one member of a family is given a job, while the other members are likely to remain unemployed. The landless, and other displaced persons who are not eligible for a job, are completely helpless. "The unemployed landless may lose in three ways: in urban areas, they lose jobs in indus- try and services, or other job opportunities; in rural areas, they lose ac- cess to work on land owned by others (leased or share-cropped) and the use of assets under common property regimes" (Cernea 1997). After the employment boom caused by the construction phase of a dam or industry project, employment "severely drops toward the end of the project. This compounds the incidence of chronic or temporary job- lessness among the displaced population" (Cernea 1996b). Although Cernea finds some hope in the new enterprises established by farmers after displacement, as in China, he warns, "New enterprises have high failure rates: often they are either unable to pay salaries for months in a row, or they soon go bankrupt leaving resettlers without either a job or land" (Cernea 1997). Evidence from many displacements in India confirms this analysis. From the Ukai Project of Gujarat, Mankodi reported that, although two new areas of activity provided some new employment (cooperative 204 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT marketing of milk and fish), this did not add up to sufficient employ- ment. "Apart from these, there simply was nothing for the people to do. This forced unemployment generated its own social problems" (Mankodi 1992). Mankodi found a high incidence of social ills (drunken- ness, gambling) attributed to the imposed idleness. Unemployment and underemployment push displaced people to engage in seasonal migration, interstate migrant work, or bonded or child labor. Researchers have found that, lacking other income sources, large numbers of women, children, and even adult men collect firewood from India's shrinking forests to sell it in urban centers (Baboo 1991, Femandes and Raj 1992, Mahapatra 1994, Mohanty 1983). This is often a last resort, before they fall into starvation. A study of the Rengali project (Ota 1996) clearly revealed the disruption in employment. Before dis- placement Rengali villagers got work for 290 days, but after resettle- ment work was available for only 119 days, on average. A startling failure in employment rehabilitation is seen in the Vindhyachal Super Thermal Power Project (first phase) of the National Thermal Power Corporation (NTPC), with 2,330 affected families, of which only 1,298 (56 percent) could be traced, only 272 (or 21 percent) of those traced were rehabilitated with a job or self-employment (shops). "NTPC has not paid any serious attention to restoration of livelihood or training for appropriate skills improvement" (Thangaraj 1996). M. S. Swaminathan (1982) wisely observed that famines in India are not famines of food but famines of work, with 7 million or more people entering the job market every year. Involuntary displacement, leaving many formerly productive people unemployed, only aggravates this sordid situation. Homelessness The family "home" is far more than a mere house; it enshrines and en- riches life. The loss of a family dwelling is tantamount to the loss of cul- tural space, which weakens identity and, ultimately, overlays economic loss with cultural impoverishment. "Loss of housing and shelter may be only temporary for many displacees, but for some homelessness re- mains a chronic condition. In a broader cultural sense, loss of a family's individual home is linked with the loss of a group's cultural space, re- sulting in alienation and deprivation" (Cernea 1997). To substantiate the multifaceted character of impoverishment, Cernea relied considerably on empirical evidence from India. Indeed, such evidence is abundant. The emergency housing centers or tempo- rary relocation camps frequently used in India as fallback solutions (by projects that have not prepared housing alternatives in time) tend to TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 205 make homelessness chronic rather than temporary and do not provide an acceptable "cultural space." Poor people consider it more important to invest their money pri- marily in land rather than in house construction. It is thus contrary to people's strategies when project authorities require construction of a house up to a certain level before they will pay the next installment. I heard complaints about this from displaced people resettled in govern- ment colonies under the Hariharjore Project in Sonepur district in 1995. The poor people often eat up the house construction advance while pre- paring their land for cultivation. Pandey's studies (Pandey and associ- ates 1996, 1998) also revealed a substantial decline over the years in the area allotted by projects in Orissa for homestead plots after resettle- ment-from 0.33 acres on average to 0.08 acre. Under these conditions there will be homelessness in a relative sense, in relation to need, and in the next generation. Marginalization The risks and reconstruction model conceptualizes marginalization as occurring "when families lose economic power and slide on a 'down- ward mobility' path: middle income farm-households do not become landless, they become small landholders; small shopkeepers and crafts- men downsize and slip below poverty thresholds. Many individuals cannot use their previously acquired skills at the new location and hu- man capital is lost or rendered useless. The coerciveness of displace- ment also depreciates the image of self" (Cernea 1997). Being uprooted often is perceived as a drop in social status in India. It can trigger a psychological downturn, a loss of confidence in society and the self, and it may be perceived as a form of injustice. The relation- ship among the model's building blocks is obvious, as marginalization may be seen also as a byproduct of the other processes of impoverish- ment, especially in the economic and social spheres, namely, landlessness, joblessness, homelessness, and social disarticulation. At NTPC in Orissa "those who are aged or those who could not get jobs in the project have no other sources of earning and become marginalized. Again those who got jobs in the project have to manage the expenditure needs of their whole family based on a single income" (Pandey and associates 1996, 1998). Nayak reported that the displaced of the Rengali Dam Project had suffered public status devaluation by be- ing dubbed budiloka, or "people of submerged areas." This is a deroga- tory term implying that the ousted people are not respectable. Daughters and sons from these families do not find bridegrooms and brides outside the budiloka category easily. This designation encompasses 206 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT all the caste groups and other classes of farmers in Rengali. The stigma attached to the displaced of Rengali has affected their self-esteem and sense of group dignity. This is an extreme case of social marginalization (Nayak 1986, Baboo 1996). Many young couples from among those displaced by Upper Krishna's Narayanpur reservoir migrated to Mumbai, Goa, and Ratnagiri for mining, agricultural wage labor, or construction work. Mi- gration of families seeking work as construction labor to places with hard living conditions, such as Mumbai, hurts children and women most severely. Most child-and-women-related services like health care, pre- school, ration shops, schools and other important basic services are place specific. Construction work involved constant moves from site to site. During monsoon months the families were forced to return to the villages to work as agricultural laborers in the sowing season, only to move out in September. The whole process deprived women and children of services meant for them. Displacement pushed men and women into repetitive, unrewarding, seasonal migration for construction and other temporary work, and it marginalized not just them, but subse- quent generations as well (Parasuraman 1996). Increased Morbidity and Mortality The serious risk to health associated with displacement has been much discussed recently in India. Empirical evidence testifies to the pervasive- ness of relocation-related illnesses and outbreaks of parasitic and vec- tor-borne diseases. Pandey's studies (Pandey and associates 1996, 1998) of the ITPS resettlement colony near Hirakud reservoir in Orissa revealed a high incidence of malaria and water-borne diseases, such as diarrhea and dysentery. Toxic effluents from the ITPS plant caused water pollu- tion in the reservoir, leading to skin diseases and other illnesses. In a com- prehensive study of morbidity, Savitri Ramaiah (1996) compared those who had been displaced in eight projects (three irrigation projects, two coal-based power projects, one hydroelectric power project, two coal mining projects) and those who had not (control group), totaling 110,000 families from rural and tribal areas. The findings showed that: . The health status of a large number of affected people who were cultivators was adversely affected. * A higher prevalence of malaria, fluorosis, guinea worm, and schis- tosomiasis was attributable to reservoir and irrigation projects, TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 207 which increase the subsoil water level and provide breeding places for mosquitoes and other vectors. • The incidence of acute and chronic respiratory illnesses, as well as gastrointestinal disturbances, was higher among the project- affected persons of the Singrauli area as a result of decreased vege- table intake. * The prevalence of acute illnesses was higher in seven resettlement groups, and of chronic diseases in five resettlement groups; the se- verity of acute illnesses among children up to six years old was higher in six resettlement groups (Ramaiah 1996). Effects on mental health as a consequence of displacement have not been seriously studied in India, except for the clinical and ethnographic observation of traumatic and stressed behavior. Byron J. Good, through Harvard-sponsored research, has taken this up. He notes, "Given the ex- traordinary importance of place, of attachments to the most minute de- tails of a community's environment, grief should be an expected part of adaptation... .When dislocation and resettlement are experienced as unjust, when compensation is viewed as inadequate or unfairly distrib- uted, bitterness and a sense of 'relative deprivation' may prolong griev- ing, even for generations" (Good 1996). Nayak highlighted the same kind of grieving, brooding, and bitterness (Nayak 1986). Food Insecurity "Forced uprooting increases the risk that people will fall into chronic undernourishment, defined as calorie-protein intake levels below the minimum necessary for normal growth and work, and food insecu- rity.... Undernourishment is both a symptom and result of inadequate resettlement" (Cernea 1997). As if in illustration of this risk becoming reality, Ota (1996) found re- cently that after 12 years of resettlement in colonies and villages under the Rengali irrigation and power project in Orissa, the displaced still ex- perience acute food insecurity. People are often half-starved during the three-to-four-month lean season, when wage work is also scarce. This is primarily because the (paddy) food grains produced have drastically declined from an average of 16 quintals per family to just 7 quintals. The land allotted is of poor quality and has not been adequately reclaimed by the project. Moreover, the nontimber forest produce, an abundant supplementary food resource before displacement, is no longer accessi- ble. Similar situations have been observed in many other resettlement locations (Nayak 1986, Baboo 1992). 208 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Children in displaced families are particularly victimized by food insecurity. Ramaiah's research compared nutritional data for children in seven resettlement groups with data for children in control groups. Findings revealed that "mean energy intake among children below six years was lower in five resettlement groups," while the mean energy in- take of female children 12 to 35 months old was lower in six resettlement groups. The prevalence of underweight children below six years was higher in four resettlement groups, and the prevalence of underweight female children 12 to 41 months old was higher in four resettlement groups. The "mean nutritional status score was poor in six resettlement groups, with calorie intake inadequate for up to 50 percent children be- low six years in all projects. More than 60 percent of children below six years were underweight in all seven resettlement colonies, as compared with the control groups (Ramaiah 1996). Of course, this inequity is com- pounded by skewed intrahousehold food distribution patterns. Ramaiah found that "in a majority of projects up to 50 percent house- holds had calorie intake adequate at the household level but inadequate at the child level, indicating that the ability to buy adequate food for the family did not ensure equal distribution of food among all household members" (Ramaiah 1996). Food insecurity in terms of inadequate calorie-protein intake is en- demic in certain parts of India, as demonstrated by Rath's findings of undernourishment (less than 2,250 calories) (Rath 1996). We note that undernourishment is particularly prevalent in those Indian states where much involuntary resettlement has taken place in recent decades: Gujarat (74 percent of the minimal required intake), Andhra Pradesh and Maharashtra (79 percent each), Orissa (70 percent), Karnataka (69 percent), Bihar (65 percent), and Madhya Pradesh (61 percent). Food insecurity due to forced displacement is still neglected, in my view, by Indian scholars and NGOs active in predisplacement and postdis- placement contexts. Loss of Access to Common Property This component of the impoverishment process is probably the most ig- nored by governments. Yet for poor people, particularly those without assets, loss of access to common (nonindividual) property assets (forest lands, bodies of water, grazing lands) results in significant deterioration in income and livelihood. And loss of common property assets is usu- ally not compensated under government relocation schemes. Cernea derives his argument for this feature of the risks model primarily from empirical evidence in India. Indeed, this evidence fully supports the model. In various semi-arid regions of India between 91 and 100 percent TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 209 of firewood and 66 and 89 percent of poor households' grazing needs are supplied by lands held under common property regimes. Another important common property asset in India is burial grounds. Food, nutrition, and income patterns in India make gleanings from common property lands essential for many. Gleanings range from food, fuel, and raw materials (for tools, construction, and medicines) to forest produce that can guarantee income. Examples include roots; yams; fruits; flowers; honey; edible resin; sago palm (for food and wine); birds and wild animals; deadwood and brush for fuel; fodder; thatching grass for making house roofs; bamboo and wood for building; herbs for medicines; and forest produce sold for cash, such as oil seeds, me- dicinal fruits, firewood; leaves for making leaf platters and brooms; resin leaves for making country cigars and cigarettes (bidi); detergents and preservatives; raw materials for artisans-rope fiber, carpenter's materials, charcoal (for blacksmiths); leaves for matmaking; bamboo and cane for basket making; potter's clay and ochres; and bark and plant dyes for textiles. Fernandes and Raj (1992) observe, A comparison with the past . .. shows that even in recent pro- jects, the availability of forest produce was better before the dis- placement/loss of land of the respondents than after it. Today many displaced and affected people, particularly those from Koraput, continue to depend on it. But no provision is made by project authorities to replace the resource that has been lost.... Consequently, like many others who have been deprived of their livelihood without alternatives, they have to resort to de- structive measures like cutting trees for sale as firewood. It is this dynamic of impoverishment that boomerangs back on the environment and compounds the costs of displacement, with repercussions on a much wider societal scale. Social Disarticulation The social-anthropological approach to resettlement compels us to ex- plore not only how displacement threatens individuals with impover- ishment, but also how society may be affected as a whole, in its structure and fabric. Integrating the previously discussed risks at the societal level, Cernea writes: "Forced displacement tears apart existing commu- nities and structures of social organization, interpersonal ties, and the enveloping social fabric. Kinship groups tend to get scattered. Life-sustaining informal networks of mutual help, local voluntary asso- ciations, and self-organized service arrangements are dismantled. The 210 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT destabilization of community life is apt to generate a typical state of ano- mie, crisis-laden insecurity, and loss of a sense of cultural identity" (Cernea 1997). Nayak's research has contributed to identification and analysis of the parameters of social disarticulation centering on and within the kin- ship system. He found that payment of compensation money to the fa- ther as head of the displaced family has led to "altercations between father and married son and between married brothers. ... Consequent upon this estrangement in the family ... father and son and also joint brothers have chosen to resettle in separate villages.... The biradari group relationship and the lineage tie at the level of 'cas te group' [have] been broken off, for the families have been spread out in different reset- tlement sites. . . . Now the displaced persons are not being properly served by the traditional service-caste men; they have been relegated to lower social status position" (Nayak 1986). A distinct trend has devel- oped to contract endogamous marriages within resettlement of dis- placed oustee populations. Kinsmen do not interact or help whole- heartedly anymore, as the impoverished displaced persons cannot reciprocate (Nayak 1986). Mahapatra (1960) showed how poor tribal villagers depended on their kinsmen during crises and also at the time of marriages and funer- als, which required deployment of substantial resources, beyond the ca- pacity of a family. Kinsmen also included ritual kinsmen, who sometimes belonged to other ethnic groups in the region. Mahapatra calls this interdependence among kinsmen "social insurance" which corresponds to the "social capital" in Cernea's conceptualization. In one example of this mutual support among kinsmen, a Hill Bhuiyan family providing for a marriage (in 1956) was given what was, in that economy, an astounding array of goods: 580 kilos of rice, 7 kilos of biri pulses, seven goats, two bolts of cloth, and one pot of rice beer. These goods, in addition to the manpower and ritual skills required at the time of the wedding, constituted the fruits of social insurance of- fered with the understanding that the family would reciprocate later on (Mahapatra 1960). Patnaik (1996) contrasts the social networking, dignity, self-assur- ance, and command over their resources that characterized the people of Paraja before their displacement by the Upper Kolab Project with the sense of inferiority, insult, impoverishment, and powerlessness that overwhelmed them when they shifted to resettlement colonies and found themselves subservient to the host tribal group and to the Bhatra and Hindu caste groups of the locality. As the Paraja were beefeaters and killed cows, they were felt to be ritually impure, bringing TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 211 defilement to the community. In all senses the displaced of Paraja were completely unwelcome immigrants into the area. They had already lost their social base and power and dignity during the displacement, and now they were denied a social anchor and articulation with other groups in the resettlement area, which might have been extended to other ethnic groups more acceptable to the Bhatra and Hindu castes. In this inhospitable climate, and because of the impoverishment of their better-off Paraja kinsmen who had lent them money or food grains at low interest, the Paraja now had to depend on an exploiting sundi (dis- tillers and moneylenders) group, which had no compunction about im- poverishing them further. In light of such experience, anthropologists and other social scien- tists, as well as the World Bank guidelines, plead for resettling tribal groups as village communities in an effort to maintain the power of their community life, social network, social capital, and social insurance. With this power, the displaced may better withstand the cultural, eco- nomic, and power shocks in the new environrnent after relocation. The lower castes, especially the scheduled castes, historically have been dominated by the higher castes. But for the educated, younger gen- erations in the weaker classes this has become intolerable. Hence dis- placement and the consequent opportunity for a resettlement of one's own choosing offered an avenue of escape from the constraints and pre- determination of caste. One thus finds that castes belonging to a lower position without dominance in a village "have chosen sites together so they can avoid recreating the confines of their low status position" and establish themselves in resettlement villages as members of relatively unstratified caste groups (Behura and Nayak 1993). Thus there was so- cial disruption in the resettlement phase, when considered from the per- spective of the old village, but there was also an unprecedented opportunity for regeneration of dignified community life. Cernea speaks of the significance of social disorganization in the overall impoverishment scenario: "Overall, if poverty is not only an ab- sence of material means or basic services such as shelter, work, food, health, or education but also powerlessness, dependency, and vulnera- bility, then the disorganization of communities and the loss of reciproc- ity networks are significant factors in aggravating poverty" (Cemea 1997). Loss of Education The children of the tribal groups displaced by the IDA-funded Salandi Major Irrigation Project in Orissa (around 1965) suffered a serious set- 212 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT back in education when displacement schools were not provided at the relocation sites. Only after about 10 years were schools established in the government resettlement colonies. During those years, the children mostly worked for the family, either collecting and selling firewood or earning daily wages (Mohanty 1983). Fernandes and Raj (1992) found that out of 184 children only 57 (31 percent) were going to school and out of 90 girls only 10 (11 percent) were in school-and this was after the schools had been established for nearly 20 years. The children in nearby villages, by contrast, were receiving regular education either in govern- ment schools or in village-sponsored schools (Mahapatra 1996a, Mohanty 1983). Beyond India itself, calculations suggest that if 10 million people are displaced worldwide every year, mostly in Africa, Asia, Central Amer- ica, and South America, there could be at least 3 million children up to 14 years old, or at least 2 million school-age children, affected by relocation (calculating with a ratio lower than for India's growing population). The sample-based registration system in India puts children up to 14 at 36 percent and children aged 5 to 14 at 22.9 percent of the total popula- tion in 1992 (IIPS 1995). Conservatively estimated, well over 1 million primary school-aged children of displaced and project-affected families will be affected. The educational lag is already heavy in developing countries compared with levels in industrialized countries: 66 percent of school-age children are expected to reach grade five in Central Amer- ica and the Caribbean, 50 percent in South Asia, 48 percent in South America, and 48 percent in Sub-Saharan Africa (UNICEF 1994). To this we have to add the loss in human capital that results when school-aged children must go to work to help out in displaced families who have mi- grated to cities (Parasuraman 1996) or who have been relocated in vil- lages and tribal areas (Fernandes and Raj 1992; Baboo 1991, 1992; Mohanty 1983; Mahapatra 1996a). If a child is withdrawn from school and compelled to work and loses the opportunity to receive a general education, it is a loss with repercussions not only for that child but for fu- ture generations of the displaced. With this in mind I suggest that Cemea's solid theoretical model be expanded to include education loss as a major impoverishment risk. Children of the displaced and other project-affected persons in India and around the world face this severe risk and threat to their develop- ment. Indeed, Cernea's recent discussion of the model acknowledges that "relocation often interrupts schooling and some children never re- turn to school. After displacement, as a result of losses in family income, many children are drafted into labor market earlier than what would otherwise occur" (Cernea 1997). TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 213 Reconstructing the Livelihoods of Resettlers Perhaps the main attribute of the impoverishment risks and reconstruc- tion model is that it analytically points to resettlement operations as op- portunities for development. Its internal logic places reestablishment processes under the same analytical lens as the risks of impoverishment in order to model the key dimensions of reconstruction. Thus the model's logic supports the proposition that involuntary re- settlement either should be programmed as development operations in themselves or should not be deliberately caused to begin with. We now ask whether the recovery variable and processes of the general model are substantiated by the empirical findings of India's re- settlement research. The issue is economic, social, and cultural rehabili- tation of resettlers' lives. We discuss the elements of reconstruction by grouping the economic dimensions (income generation), the social di- mensions (social services and welfare), and the community restoration dimensions. These are not a series of elements but rather the complex pieces of a single mosaic that must become the new socioeconomic basis of resettlers' existence. The pieces are internally linked and integrated, and as we shall see, countering each risk has a cumulative and synergis- tic effect in mitigating other risks. Countering Landlessness and Joblessness India's resettlement experiences, despite the numerous cases of impov- erishment without rehabilitation, nonetheless reveal significant in- stances of adequate postdisplacement recovery, including land-based recovery. Cernea rightly stated that "placing displaced people .. . back on cul- tivable land or in income-generating employment/self-employment ac- tivities is the heart of the matter in reconstructing livelihoods" (1997). The case reviewed below, some of which I studied firsthand, confirms the risks and reconstruction model's reliance on land or employment, more than any other variable, as the key to recovery. We have seen that not all resettlement risks are equal in their inten- sity and effects; similarly, not all remedies for recovery are equal. Cen- tral strategies that provide an economic basis for recovery through land or employment are best, and Indian experience confirms that if these are followed, the other variables of the reestablishment process are strongly and positively influenced. In earlier projects, like the Machhkund Hydroelectricity Project in south Orissa, the mostly tribal resettlers were given a generous amount 214 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT of land: 8 acres of largely fertile land, on average, in Malkangiri, where forests and common property resources were abundant and there was no conflict with host people (Mahapatra 1990). Moreover, the displaced could move there with many of their relatives from the same village or from villages near the displacement site. Hence the people affected were able to reknit their cultural and social fabric rather comfortably in a new location. Femandes and Raj also wrote of this case as a "successful reha- bilitation" (1992). It is nonetheless difficult to confirm that the displaced have achieved development. To assess a situation as "successful," the linear scale of a single component of recovery-in this case reconstruc- tion of livelihood-is not sufficient; we must consider the other recon- struction variables, including education. Thangaraj correctly noted that "Orissa is a state that very seriously implemented the land for land option and provided either irrigated or unirrigated land of various sizes to displaced persons. But the quantum of land provided gradually declined over a period" (Thangaraj 1996). Responding to protests by a politically organized group of persons dis- placed by the Rengali Project in the early 1970s, the government of Orissa decided to allocate to each displaced family, whether land-owning or landless, up to 6 acres. The terms were either 3 acres of reclaimed irrigated land or 6 acres of reclaimed unirrigated land, free of salami (land price), along with compensation. Besides this, each dis- placed family, including the landless, was granted half an acre of land as homestead land. The practice of giving land to the landless, on par with land grants to those who had lost land, enabled not only recovery, but also develop- ment of the landless people's livelihood. The provisions went further than the Maharashtra Act of 1976 and the Narmada Water Disputes Tri- bunal award for the displaced of Sardar Sarovar. If providing a job is a criterion for success in employment-based re- habilitation, NALCO-Angul has done well (Samal 1995). One hundred percent of the persons displaced from their land have been given NALCO jobs. The others have been employed either by NALCO or by its contractors (see also Reddy 1996). Reservoir-related employment has also been, at times, a success. Mahapatra (1996a) found that persons dis- placed by a reservoir organized a cooperative of fish workers, consisting of 287 members drawn from 24 villages. This fishing cooperative was originally initiated by project authorities. Most of the cooperative's members were scheduled tribes and displaced persons who had fished occasionally in the Salandi River before reservoir impoundment. Now fishing has become an occupation for them. Thus some of the pro- ject-affected persons also became project beneficiaries. TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 215 Planning for human resource development-by assessing the growth potential of relocation sites, including host communities, over the next 10 years or more-can permit appropriate skills development for the next generation, enabling people to take up new types of jobs or pursue avenues of self-employment. Movingfrom Hlomelessness to House Reconstruction Resettlement housing can take two forms. In one the project authority provides a resettlement colony or village with some infrastructure, civic amenities, and developed plots; most often it is up to the resettler to con- struct a house, while in relatively fewer instances built-up houses are provided. In the other the resettler decides to relocate to a place of his or her choosing and to construct a house there. In the Upper Indravati Pro- ject, for example, the displaced selected for their resettlement 560 differ- ent locations in small clusters (Ravindran 1996). The provision of substantial financial assistance for house building that typifies government of Orissa irrigation projects makes it possible to construct a good house, but some projects offer only a pittance-such as the Rs2,000 given by Coal India (Mahanadi Coalfields in Talcher) or the Rs3,000 given by NTPC for the construction of temporary sheds-and a good house cannot be built. Housing after resettlement is usually improved for the poorer classes, who most often receive a house plot of their own. There is strong evidence of the importance displaced people put on replacing housing. Fernandes and Raj (1992) looked at the spending pat- terns and found that when displaced persons in Orissa spent their com- pensation money, they gave house construction third priority. Indeed, "improving shelter conditions is one of the relatively easier achievable improvements in reconstructing resettlers' livelihoods, even though it is still far from occurring widely" (Cemea 1997). Ota (1996) showed that in the Rengali project, only 64 percent of families had their own house be- fore displacement, while after resettlement 93 percent of the displaced families had houses. More families (53 percent) had two-room houses compared with before displacement (37 percent); 10 percent of families now had three-room houses (7 percent before displacement); and there were even more four-room houses (5 percent postdisplacement, 3 per- cent predisplacement). Homelessness most certainly has been coun- tered by house reconstruction. In India we have examples of urban resettlement of slum dwellers in such large cities as Mumbai, Calcutta, and Madras (Reddy 1996). Squat- ters and pavement dwellers in Mumbai have been offered the 216 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT replacement cost of their meager structures and, most important, a free housing plot of 25 mi2; shops and other small business establishments have been given 16.75 m2 free. Besides this compensation, house own- ers, lessees, and lessee tenants can buy floor space of up to 70 mi2 at the actual cost of construction (Reddy 1996). In the resettlement colonies of the Ramial project in Orissa (Mahapatra 1997) the houses have been constructed at the displaced per- sons' expense, but they are invariably of better quality than the houses left behind in their old villages. All have homestead land of 0.3 acres, which leaves space for kitchen gardens. Experience indicates that for effective house reconstruction in projects where house plots, but no house-build- ing grants, are given, it is essential to make house-building loans from fi- nancial institutions available to displaced persons who become residents of officially recognized villages. Reconstructing Living Communities Economic recovery must be complemented by the social reconstruction of living communities. Three strategies are interdependent as compo- nents of social reintegration: * Moving from social disarticulation to community reconstruction - Moving from marginalization to social inclusion - Moving from expropriation to restoration of community assets. Widespread underestimation of these sociocultural and psycholog- ical processes still persists among both social researchers and planners in India. Cernea rightly argues that addressing each of these partly over- lapping dimensions of reconstructing livelihoods is mutually reinforc- ing and can achieve synergistic effects. Social reintegration depends not only on the relationships and inter- dependence among the resettlement colony or cluster groups and the neighboring communities, but also on the relationships and network building within the resettlement colony. Resettlement colonies are most often heterogeneous, with families drawn from various villages, ethnic groups, and religions. Establishing a new leadership structure, facilitat- ing colony interactions, and tackling community tasks all require a lot of adaptation, accommodation, and mutual trust, which can only be built up over time, withstanding the bumps and setbacks. Analyzing the social reintegration processes in the rehabilitation colonies of the Ramial irrigation project, Mahapatra (1997) argued that there were three levels of integration to aim for in this specific case, TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 217 which brought together colonies with heterogeneous ethnic composi- tion. The displaced had re-created the wards of their old villages, with more or less ethnic homogeneity, and had resettled largely en bloc since 1979-80. The three levels of integration Mahapatra cites are intracolony community regeneration, pancolony social reconstruction, and pan- colony host villages' social rearticulation. These three levels dovetail and reinforce one another. In seeking to carve out a niche in the social ecology of the adopted resettlement region, the displaced seek to retain a strong sense of the social and moral world they came from (intracolony community regeneration). Mahapatra found eightbases for positive bonding in support of this first level of social reintegration. One of these involved committees in each colony, constituted to represent all the ethnic groups in the colony, entrusted with the responsibility of maintaining harmony and peace as well as the traditional intercaste/intertribe division of secular and ritual labor. (For more details about this social rearticulation case, see Mahapatra 1997). Interestingly, in this context many displacement villages in the Talcher Coal Mining Project and in other project areas have been found to have functioning, traditional village leadership, with village councils acting in indifferent coexistence with the modern, statutory panchayat (multivillage) leadership systems (Pandey and associates 1996,1998). In the Madhuban Nagar rehabilitation colony under the lb Valley coal mining project, the predisplacement village structure has been trans- planted, with the old village committee still functioning and attending to postresettlement village issues. The old leaders are still village spokesmen for problem solving and decisionmaking with regard to re- habilitation claims and options (Pandey and associates 1998). Similarly, discussing the displaced of Hirakud, Baboo observed that "it was gath- ered that people usually wanted to go enbloc to a particular locality.... In a few villages the villagers did not want to go along with a particular cast groups, if the latter was crooked and exploitative, or if they were a divided lot" (1996). In the Rengali project, people adopted group-based strategies, and these have been "the resettlers' most effective means for adapting to the new settlements, as well as the most positive and endur- ing response to the involuntary displacement stimulus" (Behura and Nayak 1993). Thus we find that displaced people themselves had tried hard to re-create as much as possible harmonious groupings of resettlers in new rehabilitation sites, in both Hirakud and Rengali and in other pro- jects. However, in those host villages where land was purchased by the displaced but there were no kinsmen (that is, by the standards of 218 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT endogamy practiced by some castes or tribes), mistrust or estrangement over resource sharing (especially, common property resources and civic amenities) was likely, and harmony difficult to achieve. From Hunger and Morbidity Risks to Adequate Nutrition and Better Health Care Resettlers' nutrition levels and health care will depend largely on their economic recovery (through land or employment). But in the short run sudden disruption in food supply and adverse health effects of resettle- ment require specific and immediate action, even before economic re- covery, in order to prevent increases in morbidity and mortality rates. In Upper Indravati Project about 96 percent of the displaced have been enabled to purchase up to 1.25 acres of irrigated land or up to 2.50 acres of unirrigated land. Moreover, the resettlement and rehabili- tation unit is to provide irrigation, productive assets, and linkages with other sources of finance. There is, in this case at least, reason to think that food insecurity will be substantially reduced. In other projects as well, where the lands allotted or purchased have been developed over the years, there is hope of mitigating shortages in food supply. Ramaiah's study revealed that 70 percent of resettler households achieved adequate energy intake at the household level. Also, it was clear that more and more of the displaced were taking modern medi- cines and that there was a greater awareness of government programs. These localized improvements are few and far between, however, and what is needed is a generalized effort in this area. Of some significance, though still an isolated case, were the measures taken in Orissa in 1994, extending medical care to pregnant women in the villages to be evacu- ated. These measures taken by the government of Orissa required that medical facilities "be provided at least for a period of 15 days in the resettlement area. ...The hospital/dispensary... programmed to func- tion in resettlement areas" was to be set up immediately (Ramaiah 1996). The evidence we found about effective concerted strategies for resettlers' health protection and nutritional security in India is limited, and it is our conviction that many innovative approaches are necessary in these areas. From Educational Loss to Educational Restitution We have emphasized the significance of education loss by displaced school-age children as a risk of resettlement. What is Indian experience, if any, with remedies for this type of impoverishment? How can such loss of human capital be preempted or counteracted? TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 219 Two examples-drawn from opposite ends of the vast spectrum of education scenarios in India-may capture the essence of the problem. The first involves the resettlers from Machhkund. Parents in this village were promised at the time of their displacement that a new school for their children would be rebuilt soon. But the promised school was built and opened by the state only after 30 years! During this time an unac- countable number of children were deprived of the education due them by law. Thus they became impoverished by default, and the community was denied any growth in human capital. The second example, equally eloquent, involves educational recov- ery. In the Ramial resettlement colonies the residents themselves acted to prevent educational impoverishment from hurting their children. Without project or government support, the resettlers organized and funded a high school, over and above the limited educational facilities the government had provided. The lesson is clear, in our view: human capital growth can easily be stunted for long years if education is overlooked in India's ap- proaches to resettlement. But the remedy is not to be sought in the government alone. Reestablishment-and even improvement-of regular education services can be accelerated if resettlers mobilize their collective will and resources and act jointly to protect the inter- est of their children. Institutional Mechanisms for Rehabilitation Although resettlement and rehabilitation is a pervasive, and mostly problematic, process all over India, the government of India has no insti- tutional mechanism at the central level to address it. There is neither a national policy nor constitutional prescriptions. As of 1985 there was perhaps only one state, Maharashtra, that had a department of rehabili- tation to look after the resettlement and rehabilitation of persons dis- placed by irrigation projects. By the late 1980s and mid-1990s some states, Gujarat and Orissa among them, had established directorates of rehabilitation to deal with the rehabilitation of only those persons displaced by irrigation projects. There is still no coordinated institutional arrangement to take care of re- habilitation matters for projects in wildlife conservation or in the en- ergy, industry, or mining sectors. Most Indian scholars studying resettlement issues are convinced that integrated institutional mechanisms to oversee rehabilitation must be set up at the national and state levels. Gujarat has set up a resettle- ment and rehabilitation organization directly under the state's chief minister, to make it more effective. 220 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT The World Bank must be given credit for requesting and insisting that the government of India and the state governments create rehabili- tation machinery for irrigation and other projects supported by the World Bank in Orissa, Gujarat, Madhya Pradesh, and a few other states. Acts and Policies Are Not Enough In overpopulated India, where agriculture is still the most important oc- cupation and source of livelihood, power, and status in rural areas, land is extremely scarce. So-called government land, often encroached upon by both the haves and have-nots, has not been easily available for occu- pation, even for allotment by the authorities for resettlement purposes. After passage of the Forest Conservation Act of 1980, the forest land could not be used to resettle displaced communities. The land that was available could be allotted on a land-for-land basis to displaced persons who owned land. And since 1977, in response to sustained protests and agitations by the nontribal displaced (for example, in the early 1970s in Orissa's Rengali project) the landless could get the same rehabilitation package as given to land-owning families in Orissa. There was no government policy on resettlement. Decisions were made through government resolutions on an ad hoc basis. Between 1973 and 1983 as many as 24 such resolutions were adopted in response to pe- riodic agitations by displaced persons and political leaders. Similarly, where the market value of land had to be established because land sale records in a project-affected village or its neighboring villages were not available, this was done by government resolution. This greatly helped the displaced in the Upper Kolab and Upper Indravati projects, in the tribal, interior areas. The displaced in the Rengali project were promised land in 1973, and many of them eventually received reclaimed irrigated land or land in the command area. All of this only highlights the point that government policy or no, if the government of the day is forced to make decisions in the interest of displaced persons, it can do so. Tribal and other vulnerable groups, iso- lated in remote interior areas and without political support, cannot wrest such decisions from the government, however, unless, as in the Sardar Sarovar project, it becomes a national issue. The multilateral funding agencies, and the World Bank in particu- lar, have been instrumental in bringing about a sea change in many practices, regulations, and acts governing the resettlement and rehabili- tation of displaced persons (see Patel and others 1991 on the history of the Sardar Sarovar project). Indian scholars, activists, and NGOs have been active in this process since the 1970s, and in the 1980s they formu- lated policy guidelines (Fernandes and Asif 1997, Mahapatra 1991, TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 221 Fernandes and Thukral 1989). Yet only in response to the added pres- sures brought to bear by resistance and international agencies has the government of India taken up formulation of national policies on reset- tlement and rehabilitation (Indian Ministry of Rural Development 1994, Indian Ministry of Water Resources 1996). My reading of the situation in India is that acts and policies are not enough! A proper humane attitude toward displaced populations, ade- quate provisions for rehabilitation built into development programs, transparent and people-friendly institutional mechanisms, and the active participation of the affected people in promoting their own development are absolutely essential prerequisites for successful rehabilitation. Conclusions: Unfinished Tasks Our detailed examination of India's resettlement experiences confirms empirically and theoretically the validity of the conceptual model of risks and reconstruction as an analytical, explanatory, and strategic tool. Several other policy planks and intellectual contributions on reset- tlement in India have converged with the theoretical model discussed here. Indian NGOs, scholars, environmentalists, and activists have for- mulated and proposed to the government a Draft National Policy on De- velopmental Resettlement of Project-Affected People, which requires an objective assessment of the overall adverse impacts (risks) of resettle- ment on people and their environment and the preparation of rehabili- tation plans that include benefit sharing (Fernandes and Thukral 1989). Such plans would have to be approved by project-affected people in any given case, for their "welfare and development," and would become an inseparable part of the project itself. This approach is called "develop- mental resettlement." The thinking behind these policy proposals is fully convergent with Cernea's model for impoverishment risks and reconstruction of liveli- hoods. This model is to be credited with placing resettlement issues squarely in the context of a development framework. It brings together coherently both the impoverishment risks of resettlement, which are de- monstrably preventable, and the planning, strategy building, and pre- ventive actions necessary to avoid or mitigate these risks and promote the development of uprooted people. Scholars and others in India have contributed much to the measure- ment and analysis of these impoverishment risks. They have been inci- sive and creative in this endeavor. The same cannot be said, however, about the empirical study and analysis of the modes of reconstructing livelihoods in India. This has seldom been done, and on the whole, it has been done insufficiently. 222 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT One reason may be that rehabilitation programs in India, which are still evolving, have rarely had sufficient resources allocated to them. Moreover, it takes at least 10 years to assess the success or failure of a re- habilitation program. Few projects have been studied longitudinally with such hindsight. The strategic recovery activities suggested by the livelihood reconstruction model provide useful guidance, not just for social science research, but for government and corporate (private sec- tor) action. When the policy for resettlement in India is adopted, the hope is that such strategic recovery activities will be followed and tested in a systematic, rather than an occasional, manner. In light of the integrated conceptualization of Cemea's model and of India's own experience with resettlement and rehabilitation, it appears that we in India have just begun to unravel the total reality-to find the best means to gauge problems and develop solutions to them. In this far-from-finished task, India's academic research community, the NGOs, the civil society, state governments and various agencies all have roles to play and contributions to make. As noted earlier, Indian scholars have made significant contribu- tions to compiling, analyzing, and conceptualizing rehabilitation- development scenarios involving all the aspects Cernea developed in his reconstruction of livelihoods model. Now that rehabilitation pro- grams in many projects are more than 10, even 20, years old, there are many opportunities for independent evaluators to learn from the fail- ures and successes. Toward that end Indian scholars should develop in- dicators of success in such areas as: - Trauma management • Compensation, savings, and utilization-land purchase committees • Predisplacement impoverishment and coping activities * Evacuation and resettlement * Economic reconstruction processes: land, jobs, shared benefits. Together, economic rehabilitation, sociocultural reestablishment, and environmental rehabilitation may form the components of an inte- grated approach to enhancing the development of project-affected peo- ple, and indicators for each may be developed. I have attempted to develop overall indicators of success based pri- marily on the ultimate goal of reconstructing livelihood, which implies sustained development not only of thefirst generation of displaced persons, but also of their progeny in the second generation. The analysis of impoverish- ment risks and the formulation of action programs to counter them, ade- quately and in an appropriate time frame, together with substantial TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 223 follow-up and institutions to ensure maintenance and development of infrastructure and amenities, will provide viable means of reaching this ultimate goal. The indicators I would propose for evaluating development within the first 10 years of resettlement (whether in project-sponsored resettle- ment colonies or in self-sponsored resettlement clusters) are as follows: (a) the quality of life in the majority of households, including physical quality of life, is higher than in the old community; (b) vulnerable groups (women, children, the elderly, tribal and backward castes and classes) have adequate provisions; (c) the displaced participate fully and freely in local development programs; (d) the displaced are integrated into the region, treated equitably, and empowered; (e) amity is estab- lished with the host community and a network of interdependence is es- tablished with neighboring communities; (f) the displaced have retained or rebuilt their identity; (g) the resettlement community en- gages in self-regulation, self-management of common property, and ecoconservation. A further task for Indian scholars might be a comparative study of the whole gamut of displacement-rehabilitation-development Indian experiences with experiences in other countries. NGOs and Civil Society The NGOs have lent their voice to radically modify the Draft National Policy for Rehabilitation, as proposed by the government of India (In- dian Ministry of Rural Development 1994). But their role is crucial in bringing about a rapprochement between the displaced in resettlement clusters and rehabilitation colonies and the neighboring villages in host areas, whose infrastructure and common property resources are often shared by the displaced. Since competition, rather than collaboration, characterizes the initial relations between these groups, areas of cooper- ation must be widened. Projects should identify educated, younger persons from among colony residents, as has been done in some NTPC projects, so that they may learn how to look after the affairs of colony life. NGOs may make a lasting contribution to resettlers through capacity and institution build- ing in the community, so that after the project is completed or closed, the people will be able, through their own skills and institutions, to manage the maintenance and development of infrastructure and amenities. If more awareness of the sacrifices required from project-affected people can be built, the attitude of the civil society toward them may change. The civil society-for example, the political parties-can, given 224 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT sufficient reason, persuade authorities or companies to change either dam sites, height, or other project parameters to reduce displacement. The civil society may also put pressure on the government to pay the project-affected people the replacement value of land and other assets, so that they may build their production base elsewhere on their own. Central and State Governments The government of India should heed the warning sounded by the Su- preme Court of India: "No development project, however laudable, can possibly justify impoverishment of large sections of people and their ut- ter destitution" (Supreme Court of India, Civil OriginalJurisdiction, MP No. 16331 of 1982, Lanchand Mahto and others versus Coal India Ltd.). In our view, it is high time to amend the Constitution of India, Arti- cle 31A (Fundamental Rights), by substituting replacement value for the term "market value" (as it occurs in the second proviso), and to replace in the Land Acquisition Act of 1894 "market value" with replacement value, wherever the former is mentioned. With such a step, the single most frequent manner of deprivation and of impoverishment would be kept in better check. It would protect not only the land-owning pro- ject-affected people, but also the landless laborer, whose rehabilitation package is also defined in terms of land or cash in lieu of land, as in Orissa. (For other suggested amendments to the Constitution and the Land Acquisition Act, see Pandey and associates 1998). Such amend- ments would standardize the concepts of rights which could then be brought to bear on the interpretation of other instruments, for example, ILO Convention No. 107 of 1957, ratified by the government of India. A second important legal change would have substantial impact on people's participation in consultation over land acquisition for govern- ment or private projects. For tribal (scheduled) areas, the Panchayats Acts of 1996 (Extension to the Scheduled Areas) has empowered the Gram Sabha, Gram Panchayat, Block-level Panchayat, or Zilla Parishad (at the district level) to be "consulted before making the acquisition of land in the Scheduled Areas for development projects and before reset- tling or rehabilitating persons affected by such projects in the Scheduled Areas." Similar empowerment may be achieved in favor of other, more numerous, panchayats outside the Scheduled Areas, through the gen- eral Panchayat Act. This would immediately put a brake on the often ex- cessive acquisition of land for development projects, which displace many more families than necessary. Similarly, the central government needs to amend the Land Acqui- sition Act of 1894 and require that a certain portion of the ex gratia TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 225 payment (over and above the price paid for land and the solatium of 50 percent of the compensation) be contributed by the responsible agency or private company to be used as a deposit. Such a deposit would yield interest for infrastructure maintenance and development in the resettlement colony after the project closes. Further, as provided in the Modified Draft National Policy for Re- settlement (Indian Ministry of Water Resources 1996), there should be further amendment to the Land Acquisition Act of 1894, to the effect that an independent body shall make a strict assessment of the land require- ments of the project. Firm provisions are needed to make sure that land acquired but not used for 10 years or more shall revert to the state and be distributed to the former landowners without compensation. Indian states should each enact legislation in line with the amend- ments proposed to the Constitution and the Land Acquisition Act. Fur- ther, state govemments should constitute a Department of Resettlement and Rehabilitation, under a secretary, to formulate binding rules and procedures for acquiring land, issuing compensation, establishing eligi- bility and amount of rehabilitation grants, and setting standards for amenities and services in resettlement colonies and clusters. This body should also set standards for infrastructure development in colonies and neighboring host villages, so that the compensation and rehabilita- tion benefits accorded to displaced people will not be discriminatory among sectors (industry, energy, water resources, mining, urban devel- opment) within the same state. The proposed Department of Resettle- ment and Rehabilitation should have sufficient personnel to monitor and ensure that the laws and regulations are being observed by state and private sector programs. Testing the impoverishment risks and reconstruction framework against the vast amount of empirical findings communicated by aca- demic and applied studies of R&R in India has not only confirmed the model itself, but also enabled this author to organize a huge and dispa- rate volume of field data into a broader picture. With the considerable new factual information now "loaded" onto the general matrix of the risks and reconstruction model, we now hope that a structured synthe- sis of concepts andfacts can be constructed from the perspective of pro- cesses in the Indian subcontinent. We hope that this synthesis will give additional impetus to further research, in India and elsewhere, particularly to joint economic and an- thropological studies of resettlement. We also hope that this study would serve advocacy and policy purposes, as an argument for improv- ing resettlement practice and for providing better development oppor- tunities to the downtrodden oustees. 226 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Acknowledgements The author wishes to thank Michael M. Cernea for his kind interest in and comments on this study and for his critical and valuable sugges- tions regarding the complexities of resettlement programs in India. Thanks are also due to A. B. Ota, anthropologist, Directorate of Rehabili- tation, Water Resources Department, Government of Orissa, for encour- agement and assistance. Rita Ray, Department of Sociology, Utkal University, and R. M. Mallick and N. C. Choudhury, Center for Devel- opment Studies, in Bhubaneswar, Orissa, also have the author's grati- tude for facilitating this work. A debt is also due to Hari Mohan Mathur, anthropologist consul- tant, EDIEN, the World Bank, and P. M. Mohapatra, I.A.S., Director General, Gopabandhu Academy of Administration, Orissa, for their en- couragement. The author thanks his daughter, Sheela Mahapatra, an an- thropologist, for correcting the manuscript; C. Das, for putting the manuscript on diskette; and P. Sahoo, for undertaking the arduous typing of the manuscript. References The word "processed" describes informally reproduced works that may not be commonly available through libraries. Baboo, Balgovind. 1991. "Big Dams and the Tribals: The Case of the Hirakud Dam Oustees in Orissa." Social Action 41(3). . 1992. Technology and Social Transformation: Tile Case of the Hirakud Multi- purpose Dam in Orissa. Delhi: Concept Publishing House. . 1996. "State Policies and People's Response: Lessons from Hirakud Dam." In A. B. Ota and A. Agnihotri, eds., Involuntary Displacement in Dam Projects. New Delhi: Prachi Prakashan. Baxi, Upendra. 1989. "Notes on Constitutional and Legal Aspects of Rehabilita- tion and Displacement." In Walter Fernandes and E. G. Thukral, eds., Develop- ment Displacement and Rehabilitation. New Delhi: Indian Social Institute. Behura, N. K., and P. K. Nayak. 1993. "Involuntary displacement and the Changing Frontiers of Kinship: A Study of Resettlement of Orissa." In M. M. Cernea and S. E. Guggenheim, eds., Anthropological Approaches to Resettlement: Policy, Practice, Theory. Boulder, Colo.: Westview Press. Cernea, Michael M. 1990. Poverty Risksfrom Population Displacement in Water Re- sources Development: Policy and Operational Issues. DDP 335. Cambridge, Mass.: Harvard University, HIID. . 1993. "Anthropological and Sociological Research for Policy Develop- ment on Population Resettlement." In M. M. Cernea and S. E. Guggenheim, TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 227 eds., Anthropological Approaches to Resettlement: Policy, Practice, Theory. Boul- der, Colo.: Westview Press. . 1995. "Understanding and Preventing Impoverishment from Displace- ment: Reflections on the State of Knowledge." Keynote address at the Intemational Conference on Development-Induced Displacement and Impover- ishment, Oxford University. Processed. (See also Chris McDowell, ed., Under- standing Impoverishment. Providence: Berghahn.) . 1996a. "Public Policy Responses to Development-Induced Population Displacements." Economic and Political Weekly June 15). _ - 1996b. "The Risks and Reconstruction Model for Resettling Displaced Populations." Keynote Address, International Conference on Reconstructing Livelihoods. University of Oxford, Refugee Studies Programme. Processed. - 1997. "The Risks and Reconstruction Model for Resettling Displaced Populations." World Development 25(10). . 1998. Impoverishment or Social Justice? In H. M. Mathur and D. Mar- sden, eds., Development Projects and Impoverishment Risks. New Delhi: Oxford University Press. Das, A., V. Das, and B. Das. 1996. "Involuntary Displacement in Upper Indravati Hydroelectricity Project." In A. B. Ota and A. Agnihotri, eds., Involuntary Dis- placement in Dam Projects. New Delhi: Prachi Prakashan. Dreze, J. P., M. Sampson, and S. Singh, eds. 1997. The Dam and the Nation: Displace- ment and Resettlement in the Narmada Valley. Delhi: Oxford University Press. Fernandes, Walter. 1991. "Power and Powerlessness: Development Projects and Displacement of Tribals." Social Action 41(3). .1994. Development-Ind uced Displacement in the TribalAreas of Eastern India. New Delhi, Indian Social Institute. Femandes, Walter, and Mohammed Asif. 1997. Development-Induced Displace- ment and Rehabilitation in Orissa, 1951 to 1995: A Database on Its Extent and Na- ture. New Delhi: ISI. Femandes, Walter, and S. Anthony Raj. 1992. Development, Displacement, and Re- habilitation in the Tribal Areas of Orissa. New Delhi: Indian Social Institute. Fernandes, Walter, and Enakshi Ganguly Thukral, eds. 1989. Development, Dis- placement, and Rehabilitation. New Delhi: Indian Social Institute. Fisher, William F. 1995. Towards Sustainable Development? Struggling over India's Narmada River. Armonk, N.Y.: Sharpe. Good, Byron J. 1996. "Mental Health Consequences of Displacement and Reset- tlement." Economic and Political Weekly (June 15). Government of Orissa, Bureau of Statistics and Economics, Department of Planning and Coordination. 1968. "Report on the Benefits of Hirakud Irriga- tion." Bhubaneswar. Processed. Government of Orissa, Water and Land Management Institute. 1991. "Irrigation in Orissa." A. C. Nayak, ed. Bhubaneswar. Processed. 228 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT India. 1993. The Constitution of India. With Short Notes and Subject Index. 13th ed. Lucknow: Eastern Book Company. India, Ministry of Rural Development. 1994. "The Draft National Policy for Re- habilitation of Persons Displaced as a Consequence of Acquisition of Land: Cabinet Note." New Delhi. Processed. India, Ministry of Water Resources, Central Water Commission. 1996. "Mod- ified Draft National Policy for Resettlement and Rehabilitation of Persons Af- fected by Reservoir Projects." New Delhi. Processed. IIPS (International Institute for Population Studies). 1995. National Family Health Survey, 1992-93. Bombay. Joshi, Vidyut. 1987. Submerging Villages: Problems and Prospects. Delhi: Ajanta Publications. Mahapatra, L. K. 1960. "A Hill Bhuiyan Village: An Empirical Socio-Econonmic Study." Hamburg University, Germany. Processed. . 1990. "Rehabilitation of Tribals Affected by Major Dams and Other Pro- jects in Orissa." In A. P. Fernandez, ed., Report on Workshop on Rehabilitation of Persons Displaced by Development Projects. Bangalore: Institute for Social Eco- nomic Change. .1991. "Development for Whom? Depriving the Dispossessed Tribals." Social Action (July-September). (Original version presented at World Con- gress of Anthropology, Vancouver, 1983). . 1994. Tribal Development in India: Myth and Reality. New Delhi: Vikas Pub- lishers. . 1996a. "Good Intentions or Policy Are Not Enough: Reducing Impover- ishment Risks for the Tribal Oustees." Presented at the World Bank Work- shop on Impoverishment Risks in Involuntary Resettlement, New Delhi. Processed. . 1996b. "Rehabilitation by the Government and the People: A Case Study of Impoverishment of Salandi Major Irrigation Project Oustees." Presented at the Training Seminar on Resettlement and Rehabilitation, Gopabandhu Academy of Administration, Bhubaneswar. Processed. . 1996c. "Society in Orissa." In Mishra and Samal, eds., Comprehensive His- tory and Culture of Orissa. Delhi: Kaveri Publishers. _____. 1997. "Social Rearticulation and Community Regeneration among the Resettled Oustees of Ramial Medium Irrigation Project, Dhenkanal District." Presented at the Training Seminar on Resettlement and Rehabilitation, Gopabandhu Academy of Administration, Bhubaneswar. Processed. Mankodi, Kashyap. 1992. "Resettlement and Rehabilitation of Dam Oustees: A Case Study of Ukai Dam." In E. G. Thukral, ed., Big Dams, Displaced People. New Delhi: Sage Publications. Mathur, Hari Mohan. 1995. Anthropology and Development in Traditional Societies. 2nd ed. New Delhi: Vikas Publishing House. TESTING THE RISKS AND RECONSTRUCTION MODEL ON INDIA 229 Mathur, Hari Mohan, and David Marsden, eds. 1998. Development Projects and Impoverishment Risks. New Delhi: Oxford University Press. Misra, Sadasiv. 1960. Economic Survey of Orissa, vols. 1 and 2. Bhubaneswar: Gov- ernment of Orissa, Finance Departrnent. Mohanty, B. P. 1983. A Socio-Economic Investigation on the Displaced Families of Salandi Riverdam, Hadgarh. Anandapur: Antyodaya Chetana Kendra. Morse, Bradford, and Thomas R. Berger. 1992. Sardar Sarovar: Report of the Inde- pendent Review. Ottawa: Resource Futures International, Inc. Nayak, P. K. 1986. "Displaced Denizens of Rengali Dam Project: A Study of Their Socio-Economic Conditions, Resettlement Problems and Ameliorative Action Measures." OXFAM, Bhubaneswar. Processed. Ota, A. B. 1996. "Countering the Impoverishment Risk: The Case of the Rengali Dam Project." In A. B. Ota and A. Agnihotri, eds., Involuntary Displacement in Dam Projects. New Delhi. Prachi Prakashan. Pandey, Balaji, and associates. 1996. Development, Displacement and Rehabilitation in Orissa, 1950-1990. Bhubaneswar: ISED. . 1998. Depriving the Underprivileged for Development. Bhubaneswar, Orissa: Institute for Socio-Economic Development. Paranjpye, Vijay. 1988. Evaluating the Tehri Dam: An Extended Cost Benefit Ap- praisal. New Delhi: Indian National Trust for Art and Cultural Heritage. Parasuraman, S. 1996. "Development Projects, Displacement, and Outcomes for Displaced: Two Case Studies." Economic and Political Weekly (June 15). Patel, Anil, and others. 1991. "Rehabilitation and Resettlement in Sardar Project: Are the Critics Right?" Social Action 41(3). Patnaik, S. M. 1996. Displacement, Rehabilitation, and Social Change: The Case of Paraja Highlanders. New Delhi: Inter-India Publications. Ramaiah, Savitri. 1996. "Impact of Involuntary Resettlement and Rehabilitation on Level of Living of Project-Affected Persons (PAPs)." Presented at the World Bank Workshop on Impoverishment Risks in Involuntary Resettle- ment. New Delhi. Processed. Rath, Nilkantha. 1996. "Poverty in India Revisted." Indian Journal of Agricultural Economics 51(1, 2, January-June). Ravindran, Latha. 1996. "Resettlement and Rehabilitation: From Policy Guide- lines to Practice-A Case Study." In Theme Papers of the First Training Seminar on Resettlement and Rehabilitation. Bhubaneswar: Gopabandhu Academy of Administration. Reddy, I. U. B. 1996. "New Approaches in Involuntary Resettlement in Urban In- frastructure Projects in India." Presented at the International Conference on Reconstructing Livelihoods. Oxford University. Processed. Roy-Burman, B. K. 1961. Social Processes in the Industrialization of Ron rkela. New Delhi: Census of India. 230 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT .1995. "Rourkela Steel and SardarSarovar Project Oustees: Rehabilitation Policies Compared." Main Stream (March 26). Samal, K. C. 1995. Socio-Economic Impact of NALCO Angul Sector: Involuntary Dis- placement and Adjustment. Bhubaneswar: NC Centre for Development Studies. Scudder, Thayer. 1985. "A Sociological Framework for the Analysis of New Land Settlements." In M. M. Cernea, ed., Putting People First. New York: Ox- ford University Press. Shi, Guoqing and Hu Weison. 1994. "Comprehensive Evaluation and Moni- toring of Displaced Persons' Standards of Living and Production." Hohai University, National Research Centre of Resettlement, Nanjing. Processed. Singh, Sekhar, Ashish Kothari, and Kulan Amin. 1992. "Evaluating Major Irriga- tion Projects in India." In E. G. Thukral, ed., Big Dams: Displaced People. New Delhi: Sage Publications. Stanley, William. 1996. "Machhkund, Upper Kolab, and NALCO Projects in Koraput District, Orissa." Economic and Political Weekly (June 15). Swaminathan, M. S. 1982. "Irrigation and Our Agricultural Future." A. N. Khosla Memorial Lecture. New Delhi. Processed. _____. 1996. "Science and Technology for Sustainable Food Security." Indian Jouirnal of Agricultural Economics 51(1, 2). Thangaraj, Sam. 1996. "'Impoverishment Risks in Involuntary Resettlement: An Overview." Presented at the World Bank Workshop on Impoverishment Risks in Involuntary Resettlement. New Delhi. Processed. Times of India. 1997. "Snapshot World: Four Countries with the Most Dams." June 28, 1997. UNICEF (United Nations Children's Fund). 1994. The State of World's Children, 1994. India ed. New Delhi: Oxford University Press. World Bank. 1990. Operational Directive 4.30. "Involuntary Resettlement." Washington, D.C. Processed. _.____ 1994. "Resettlement and Development. The Bankwide Review of Pro- jects Involving Involuntary Resettlement, 1986-1993." Environment Depart- ment. Washington, D.C. 6 Sharing Project Benefits to Improve Resettlers' Livelihoods Warren A. Van Wicklin III Editor's Any attempt to develop an economics of recovery Note and development, as this volume advocates, would certainly have to reexamine the rationale of benefit distribution patterns in the projects that involve forced resettlement. These patterns are too often too skewed, divid- ing the affected people into net gainers and net losers, with some sharing the gains and some only the pains of development. Contrary to what the vast majority of resettlement studies do-namely, reporting the deprivations people suffer through dis- placement-the study by Warren Van Wicklin explores the circum- stances under which displaced people also can participate, through deliberate project design, in sharing the gains from the projects that caused their relocation. Such studies are seldom done. The author has reviewed a large number of projects under im- plementation between 1987 and 1997, using both the ex ante and the ex post project documents. The purpose is to identify if, and how, the stream of benefits generated by the project can be channeled to resettlers as well, rather than relying only on up-front project finan- cial allocations for resettlement. A typology of benefits is con- structed, all resulting from project outputs rather than inputs. This typology includes both in-kind benefits (such as access to electric- ity) and various forms of access to newly created revenues based on project experiences from the Philippines, China, India, Mexico, In- donesia, and many other countries. One, partly unexpected, finding of the study is that there is a significant number of projects in which displaced people also share in some of the project benefits, in one form or another. This finding is important in itself. But even more revealing is the demonstration 231 232 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT that projects of various types, and in different sectors, contain within their own fabric options to direct project benefits to the resettlers, either during project implementation or afterwards. Ef- fective ways and cases of channeling project-created benefits to- wards reconstructing resettlers' livelihoods are discussed, citing projects in Indonesia, China, and Latin America. Yet this potential to rechannel resources remains unused in many projects. This is both inequitable and unjustifiable. Often, such potential is overlooked because of trivial, not structural, rea- sons: lack of commitment and concern of project decisionmakers, poor methodologies of economic and financial analysis at the pro- ject planning level, poorly designed resettlement components, a lack of creativity and inventiveness in using resources available to the project, and so on. Van Wicklin emphasizes, however, that such apparently trivial reasons also have deeper roots: the absence of na- tional policies and of legally compelling regulations that would make it mandatory for developers from the public or private sector to deliberately open access for those displaced to the project's stream of benefits. National policies for making benefit sharing a standard would always reach further and deeper than only discrete, project-specific initiatives to promote such approaches. A public debate is going on about whether restoring resettlers to their prior standards is sufficient or if they must necessarily be assisted to reach, sooner rather than later, improved levels. Surely, development is insufficiently served when only restoration is achieved, and restoration of prior poverty levels is not a great accom- plishment, to say the least. This only highlights the importance of the distributional issues raised in this study, for it is only through much more generous benefit sharing, that improving rather than only restoring resettlers' livelihood will become possible. The World Bank's resettlement policy emphasizes benefit sharing and recommends to all countries that undertake projects entailing re- settlement to adopt and apply in practice benefit-sharing policies and procedures. Benefit-sharing and forward-looking developmental ap- proaches (rather than only asset compensation) are the most construc- tive and progressive dimensions in resettlement policy. Toward such goals, the study that follows provides policy- makers and planners with an inspiring inventory of good practices, able to mitigate to a significant extent the sacrifices of those af- fected by development-induced displacements. Policies and pro- jects that are not predicated on benefit sharing misuse the potential for resettlers' recovery: they have much to leam from policies and projects that do this effectively. SHARING PROTECT BENEFITS 233 One of the most difficult aspects of reconstructing the livelihoods O and incomes of displaced people is finding the financial resources necessary for this complex endeavor. As the other chapters in this vol- ume make clear, the costs of resettlement and rehabilitation are often not internalized in project budgets. Typically, initial project budgets tend to provide little more than compensation payments for land acquisition. Too often this undercompensates the displaced, who may lack legal title to their land or sustain losses other than land. It also underestimates the actual replacement cost of land in situations where there is a sudden in- creased demand for it and a shrinking supply. As the recent OED study on involuntary resettlement associated with eight large dam projects found, "Land prices spiral beyond all reasonable budget limits in situa- tions with large numbers of resettlers taking unrestricted cash packages and competing for the same limited land pool" (World Bank 1998d). Because resettlement costs tend to be underestimated and externalized (see Cernea, this volume, chapter 1), corrections in current practice are necessary to recognize and generate the necessary financial resources for rehabilitating displaced people. This chapter argues that sharing project benefits with resettlers is an effective-yet underutilized-approach, providing numerous examples illustrating its merits. Arguments for Sharing Benefits Empirical evidence informs that compensation alone is usually insuffi- cient to relaunch resettlers on a sustainable income path (World Bank 1994g, World Bank 1998d). Proper socioeconomic reestablishment re- quires more than paying the fair market value of the condemned land. Reconstructing livelihoods implies reestablishing the income flows that the displaced people had prior to resettlement. New income streams, or investments that can produce these income streams, must be developed. As a rule of thumb, appropriate resettlement budgets on a per family ba- sis need to be a multiple (four or more) of the lost income streams (World Bank 1994g). Therefore successful resettlement requires financial re- sources commensurate to this difficult task. In every single project the means to generate them need to be identified or created. One way to help alleviate financial constraints on resettlement bud- gets is to deliberately design projects so that involuntarily displaced people receive some of the benefits created by the projects displacing them. This means that in addition to allocating resources from the up-front project budget for the purpose of resettlement and reconstruc- tion, the stream of benefits created by the project should also be tapped to provide direct benefits and resources for resettlers. This is one of the 234 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT most important issues in the economics of involuntary resettlement op- erations, though so far it has received scant attention in the economic lit- erature. Besides the economic and financial rationale, there are also major equity arguments for sharing benefits. Giving up one's home and land is much more than an economic transaction: it has profound social, cul- tural, and psychological ramifications. Communities and neighbor- hoods, with their mutual support and labor exchange networks, are often broken up and dispersed as a consequence of forced relocation. Cultural losses include ties to the land, ancestral sites, indigenous knowledge of local production systems, and other forms of cultural her- itage. Among the psychological costs and consequences of relocation are isolation, alienation, anomie, substance abuse, higher rates of prosti- tution and divorce, and other signs of social disarticulation (Pandey and associates 1996). Since resettlers are called upon to make sacrifices for the general good by giving up their habitats and their income- generating assets, it is only equitable that they should share in the bene- fits of the project displacing them. There is also a political rationale for projects to share benefits with resettlers. Typically, the displaced and the project beneficiaries are largely separate groups. Resettlers bear the burden for someone else's benefit. Displaced people are frequently the "losers" while "project bene- ficiaries" are the "winners" of development (Cernea 1997). Many benefi- ciaries live at a great distance from the project area, while many in the project area are the first to suffer the adverse effects, and therefore often oppose the project. Recent research on resistance to involuntary resettlement docu- ments the extent to which people will go to avoid displacement and im- poverishment (Cernea 1995c, Gray 1996, Oliver-Smith 1996, and Posey 1996). Lack of benefits from the project commensurate with the sacri- fices the displaced people make is part of the rationale for their resis- tance and opposition to the projects. Various communities threatened with resettlement and many nongovernmental organizations (NGOs) have mounted vigorous public campaigns against the Itaparica (Brazil), Narmada Sardar Sarovar (India), Arun III (Nepal), Nam Theun (Laos), Pak Mun (Thailand), Kedung Ombo (Indonesia), Bakun (Malaysia), and Katse (Lesotho) dams, among others. In turn, political opposition tends to increase the cost of projects and to delay their benefits. It even threat- ens their sustainability (Fisher 1995). Learning from extensive experi- ence, World Bank strategies for promoting participation and public involvement encourage governments to develop collaborative relation- ships with those to be displaced and to be more sensitive to their needs and rights. One important form of participation is to make resettlers SHARING PROJECT BENEFITS 235 direct beneficiaries of the project so that they gain a vested interest in the project, instead of having only an adversarial relationship. Sharing project benefits is not just a way to gain acceptance of dis- placement, or to generate more resources; it provides an opportunity to do resettlement better. Yet many projects have not availed themselves of this opportunity. Part of the reason has been a narrow-or lack of-un- derstanding of the economics of involuntary resettlement. This is mani- fested in the design of rehabilitation packages that are not linked to the benefits and resources generated by the main project. This is often due to the weak economic analysis and projections characteristically per- formed for involuntary resettlement components, compared with the design of voluntary resettlement projects where resettlement is inte- grated into regional development strategies. As Eriksen has docu- mented (this volume, chapter 3), not only are involuntary resettlement components more often poorly prepared, they are not fully integrated into the economic design of the project and are seen as "add-ons." The Bank's original goal with resettlement was to mitigate adverse impacts on resettlers. Since the World Bank's initial policy on involun- tary resettlement was first promulgated, in 1980, increasing attention has been given to procedures and mechanisms for achieving this mitiga- tion. The Bank's first published policy and operational paper on resettle- ment (Cernea 1988) analyzed many of the practical pitfalls leading to resettlement failure, including inadequate financial resources, and sug- gested strategies and methodologies to avoid them. In the Bank's subse- quent formulations of its original policy (World Bank 1986, Cemea 1988, World Bank 1990), what was previously implicit in the Bank's policy was made explicit. Indeed, the new text states that all involuntary resettle- ment should be conceived and executed as development programs, with resettlers provided sufficient investment resources and opportunities to share in project benefits (World Bank 1990). Thus since the mid-1980s the concept of sharing project benefits with resettlers was repeatedly formulated, explicitly and normatively (Cemea 1995a). The Bank went on record as encouraging governments to use benefit sharing as a strategy for mobilizing financial resources for relocation, in order to improve resettlement performance. Rather than emphasizing only the mitigation of losses, the revised policy took a forward-looking stance and stated that projects should be deliberately designed in ways that would directly benefit resettlers, in addition to the economic rehabilitation they might receive. Some Bank-assisted projects had already adopted benefit sharing, and in 1990 provisions were made to translate this "good practice" into a "main- stream" approach to resettlement. This coincided with some of the first literature focusing on impoverishment risks, recomrnending stronger, 236 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT proactive strategies to counteract, and not merely mitigate, such risks (Cernea 1990). Indeed, the norm of sharing project benefits with resettlers requires radically improving the project design and planning process, because benefit sharing has to be planned for at the outset. Given this require- ment, this chapter inquires to what extent development projects plan to, and actually do, share benefits with resettlers, and how they do it. To answer these questions, we reviewed a large set of Bank-assisted projects that involved involuntary resettlement during the past 10 years; that is, projects under implementation between 1987 and 1997. The in- formation is derived from staff appraisal reports (the principal project document at the time a project is approved), implementation or project completion reports (the principal project document when Bank financ- ing of a project is completed), and performance audit reports (another postproject document prepared for about half of all Bank-assisted pro- jects). Itis possible that some projects mighthave shared projectbenefits with resettlers, either by design or otherwise, without the main project documents reporting on this, but it is unlikely that this happened often. We complemented written sources with interviews with project task managers and others knowledgeable about the project. The author supplemented the desk review with field research carried out during 1996-97 under OED auspices at seven projects in Asia and Africa (World Bank 1998a-g). Only about half of the projects reviewed were approved after 1990, when sharing project benefits with resettlers was explicitly recommended for the first time. Thus, the purpose of this chapter is not to ascertain whether or not specific projects shared project benefits with resettlers, but to identify methods of sharing benefits. T'he approach used for this analysis was to determine whether there was benefit sharing of direct outputs of the project's main components. For example, if the project was an urban upgrading project creating new housing as one of its core outputs, and resettlers were given priority for this new housing, then the resettlers would be seen as sharing directly in the benefits. Similarly, if an irrigation dam moved displaced people into the "command area," that is, onto land where they would receive water from the reservoir or the irrigation canal that displaced them, they would be directly sharing in the benefits. A remarkable new practice is that some development projects are specifically designed to rehabilitate resettlers displaced by a separate in- frastructure project. For example, in 1994 the Bank approved the first "pair" of projects that treated the resettlement component of infrastruc- ture building as a full-fledged, freestanding project: this pair was the China Xiaolangdi Dam Project and the Xiaolangdi Resettlement Project SHARING PROJECT BENEFITS 237 (box 6.1). Another, more recent example of a pair of projects is the India Coal Sector Rehabilitation Project and the India Coal Sector Environ- mental and Social Mitigation Project, approved in 1997. In such cases all the outputs of the resettlement project go to the resettlers (and their hosts), so none of those outputs is considered benefit sharing. The issue is to determine whether and how the benefits of the associated infra- structure project are shared with resettlers. National Policy Provisions for Sharing Benefits National policies with provisions for benefit sharing have far broader impact than discrete, project-specific decisions to promote such ap- proaches. National policies provide legal frameworks and make benefit sharing standard, rather than leaving it to the discretion of each project planner. So far the most progress made in putting into practice national policies mandating benefit sharing has been in the energy sector. Some hydroelectric and thermal projects allocate a percentage of their electric- ity sales revenues to resettlers and local administrative units, in re- sponse to legal and policy frameworks that were adopted to institutionalize this practice. Among the projects using this approach are the Lubuge, Yantan, Shuikou, and Ertan hydroelectric projects in China; the Rio Grande Hydroelectric Project in Colombia; and the Leyte-Cebu and Leyte-Luzon geothermal projects in the Philippines. Of course, this approach is used in other, non-Bank-assisted projects in those countries. Brazil developed a specific policy for benefit sharing with people af- fected by energy sector projects. Brazilian Law 7990/89 mandates that 6 percent of electricity sale revenues are to be returned to affected munici- palities. This created a significant source of funds in the Itaparica projec that will, among other things, permit the relocated rural population to receive a higher grade of irrigation technology and the relocated urban population to obtain upgraded services (Serra 1993). Colombia developed a similar policy on sharing the benefits of hy- droelectric projects with locally affected people. Law 56/91 of 1991 re- quires that 4 percent of annual electricity revenues generated by a hydroelectric plant be allocated to municipalities in that plant's area. Specifically, 2 percent of the revenues are for reforestation and other en- vironmental mitigation activities, and 2 percent for social infrastructure such as schools, roads, rural electrification, and so forth. In the first year expenditures are for improvements in the area closest to the reservoir. As the needs of those most immediately affected are met, additional rev- enues are shared more broadly. 238 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Box 6.1 Financing Separate Projects for Resettlement In 1994 the Bank helped design and approved the first pair of "twinned" projects where the resettlement component of an infrastructure project is separated into a parallel, stand-alone project. The China Xiaolangdi dam and reservoir project will displace about 182,000 people. The dam will provide substantial irrigation, hydropower, and flood control ben- efits. To ensure that resettlers also benefit, a a parallel and distinct $571 million Xiaolangdi Resettlement Project was created, with its major ob- jective being "to restore and improve the livelihoods of 154,000 resettlers and 300,000 host people affected by the construction and op- eration of the Xiaolangdi multipurpose dam, and to minimize the ad- verse effects of their social adjustment to their new environments" (World Bank 1994c). The resettlement project will construct all infra- structure including housing for 276 villages and 10 towns for the resettlers. It will also develop 11,100 hectares of land, of which 7,000 will be irrigated, principally by the reservoir. A total of 252 small indus- tries and mines will be relocated, and the workers will move with these industries. An additional 84 industries will be established, employing some 20,500 resettlers. All the financial and human resources dedicated to this project will benefit resettlers and their hosts. Already the benefits of "twinning" are clear. This project is one of the most detailed resettlement efforts ever designed. The budget, timetable, and resettlement plan are unparalleled in their detail. Thus far not only the planning but the implementation is successful. Incomes for the first group of 2,000 resettlers exceeded their previous incomes by 10 to 60 percent 14 months after relocation. Independent monitoring found that more than 95 percent of the first 7,000 resettlers were satisfied with their relocation sites and the facilities provided. In 1996 the Bank approved a full-scale project that would address resettlement problems for an infrastructure project that had already closed, but where the resettlement component was not satisfactorily completed. The India Coal Sector Environmental and Social Mitiga- tion Project is the twin project to the Coal Sector Rehabilitation Project (World Bank 1996a). The social mitigation project is undertaken to re- establish incomes not only for people displaced by 25 coal mines being financed by the current coal sector rehabilitation project, but for peo- ple displaced by four mines in previous World Bank-financed coal mining projects. Economic rehabilitation for displaced people re- volves around training and employment with the coal companies themselves . SHARING PROJECT BENEFITS 239 China has applied the principle of sharing project benefits with resettlers to the greatest number of projects. Practically all of the recent Bank-assisted projects in China include the applicable Chinese legal frameworks for sharing project benefits. This applies not only to resettlers but to host commrunities. To make this possible, many incen- tives are included for the receiving population. Chinese legal frameworks for benefit sharing were first developed in large reservoir projects. In June 1981 the Ministry of Electric Power is- sued the "Circular on Collection of Reservoir Maintenance Fund from Power Generation Revenue of Hydropower Stations." This act estab- lished a "reservoir maintenance fund" for each reservoir to (a) compen- sate those relocated for both lost livelihood and, if needed, lost means of production, after completion of the reservoir; (b) maintain reservoir structures; and (c) maintain structures for drinking water supply, irriga- tion, and transportation used by the relocated. This fund, one yuan (Y) per thousand kilowatt-hours (kWh), is collected from the hydropower revenue of power stations under the Ministry of Electric Power. In July 1986 the General Office of the State Council issued the "Cir- cular on Transmitting the Report of Paying Close Attention to Reservoir Resettlement" prepared by the Ministry of Water Resources and Electric Power. This document stipulated that (a) additional resettlement funds for reservoirs begun in 1986 or later should be allocated in the total capi- tal construction costs, and (b) the additional funds would have three sources-hydropower revenue, water tariffs, and ministry or local gov- ernment budgets. In 1991 the State Council issued the "Regulation on Compensation of Land Requisition and Resettlement for the Construction of Big and Medium-Sized Water Resources and Hydropower Projects." It states that the support period for resettlement would be the first five to 10 years after completion of resettlement. Recently the hydropower reve- nue tariff was increased from Y0.001/kWh to YO.0045/kWh. This ap- proach provides significant revenues for benefit sharing and codifies the method in the legal policies of the governing national authorities. There are at least two major advantages to this sort of benefit shar- ing. First, because it takes the form of a revenue stream, it is very flexible in providing financial resources for whatever is needed. Revenues have been used for community facilities, training and job creation, and stimu- lation of the local economy. Second, tapping future revenue streams en- sures postproject financial resources for sustaining resettlement. Resettlement is rarely completed at the same time as project completion. As Eriksen's chapter in this volume suggests, resettlement requires a system establishment phase of four to six years and then a system 240 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Box 6.2 Enabling Reservoir Displacees to Benefit from Irrigated Command Areas Several dam projects in India, where large numbers of people have been displaced, shared direct project benefits with resettlers by mov- ing displaced people into the newly irrigated areas. These include the Andhra Pradesh II and III, Gujarat Medium II, and Maharashtra III Ir- rigation projects. In the Orissa Water Resources Consolidation Project, "over 75 percent of displaced and affected people would be direct beneficiaries of new irri- gation provided under the project. As with other beneficiaries of new irri- gation, their incomes would thus be expected to increase significantly under the project" (World Bank 1995j). The Tamil Nadu Newsprint Project is very relevant as a second-phase project that remedied resettlement problems from an earlier project. As the implementation completion report explained: Although expansion of the [Tamil Nadu] paper mill did not in- volve any new displacement of the population, the operation of the first mill had adversely affected 102 families in an adjacent village as a result of seepage from the plant's effluent lagoon and settling of coal dust from the plant. Resettlement and rehabilita- tion was therefore required. Tamil Nadu Newsprint and Papers Ltd., with the support of farmers residing in villages adjacent to the mill site, implemented a unique lift irrigation system to fully exploit the treated effluent discharges from the plant. The scheme irrigates about 1,400 acres of formerly arid land and po- tentially another 3,000 to 4,000 acres, allowing local farmers to now harvest two or even three crops a year compared to only a single crop previously (World Bank 1996c). Many reservoir projects outside India also share irrigation benefits with involuntary resettlers, such as the Brazil Ceara Water Resources, China Daguangba Multipurpose, Lubuge Hydroelectric, and Iran Irrigation pro- jects. Plans for the China Daguangba Project estimated that "approxi- mately half of the [displaced] people will move to new settlements which will be provided irrigation from the canals [constructed by the project]. New paddy fields will be irrigated on reclaimed lands now utilized by the affected people for rainfed crops." The Brazil Ceara Water Resources Pro- ject staff appraisal report noted that: "Those families which cannot be relo- cated on the same properties will be re-established in planned resettlement areas situated in the immediate vicinity of the reservoir, granted land tenure (independently of whether they possessed tenure of their former lots prior to resettlement) and direct access to water for agri- cultural production purposes" (World Bank 1995a). SHARING PROJECT BENEFITS 241 stabilization phase of at least an additional five years. Guaranteed reve- nue streams during this longer time horizon help facilitate successful stabilization and sustainability of reconstruction. A Typology of Project-Level Benefit Sharing Projects create several kinds of benefits that can be directed to the dis- placed. For example, the outputs of a multipurpose dam that can be shared by resettlers include water for irrigation or drinking; electric- ity from hydroelectric plants; fishing rights in the reservoir; flood control; inland, waterborne transportation; and revenues from sales of electricity and water. Other types of infrastructure project benefits include improved roads and other transportation systems, health clinics, schools, and so on. Besides facilities, projects offer employ- ment directly generated by the project, either temporary or perma- nent. Urban redevelopment projects provide housing, or serviced sites or plots for building, so that resettlers improve the quality of their housing and benefit from more secure tenure arrangements and community facilities. The following typology of benefits, with project examples, is by no means exhaustive, but it shows the range of experiences in the set of projects we examined. Furthermore, when only one example is cited for a project, that does not mean that other forms of benefit shar- ing did not take place in that project. Of course, benefit sharing is usu- ally only one of the provisions for economic rehabilitation, and each project also requires other measures to restore incomes and recon- struct livelihoods. Irrigation Dams-whether for irrigation, hydropower, drinking water, flood con- trol, inland transportation, or a combination of these-have tradition- ally been the largest cause of displacement in Bank-assisted projects. The World Bank resettlement review found that 63 percent of people be- ing involuntarily resettled in ongoing projects were being displaced by dams (World Bank 1996g). Many of the large Bank-financed irrigation projects in India, which as a group have displaced more than a million people in the past 15 years, have moved at least some of the displaced into command areas so they benefit directly from the project (box 6.2). With the increased yields arising from irrigation, these people can earn more than they did before they were displaced. 242 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Electricity Since many of the dams generate hydroelectric power, the opportunity is usually created to provide involuntary resettlers with electricity at their new sites. Hydroelectricity is often added to an existing power grid, but local connections are also technically feasible and cheaper if they are incorporated into project design from the outset. The Yantan, Shuikou, and Ertan hydroelectric projects in China did this. Thermal power projects also can provide electricity to resettlers. The China Zouxian Thermal will provide electricity to resettlers from the transmis- sion component of the project, and the India Farakka II Thermal project has already done this (World Bank 1996e). Several projects earmark a portion of their electricity revenues for the benefit of those displaced by the dam. In the Philippines' Leyte- Luzon Geothermal Project, "1 [percent] of the gross revenues from pro- ject electricity sales will be distributed among the province, the barangay, and the local community" (World Bank 1994f). This will re- sult in significant funds for the area, and the resettlers and their hosts will be among those benefiting. The China Lubuge Hydroelectric Project set aside one yuan for every thousand kilowatt-hours generated by the Lubuge plant to estab- lish a reservoir maintenance fund, 80 percent of which is used by local governments for reforestation, fishery, and agricultural development activities. Twenty percent is retained by the power station for reservoir cleaning, maintenance works, and so on. The China Ertan, Yantan, and Shuikou hydroelectric projects also earmarked a portion of the revenues from electricity sales to benefit resettlers. Yantan Hydro "provides 1 yuan per 1,000 kilowatt-hours from power plant revenue towards de- velopment in the reservoir area, for the life of the project," generating a substantial financial flow (World Bank 1995c). Fishing One income-producing resource that can be allocated to resettlers is fishing rights in the newly created reservoir. Often these benefits have been allowed to go to outsiders, but a much better approach is to reserve these benefits, primarily or exclusively, for those displaced by the reser- voir. This is a good example of transforming the cause of displacement or hardship into a source of new employment and income. Reservoir fisheries have proved to be a more successful means of sharing benefits and reconstructing livelihoods than expected. In the Aguamilpa reservoir, part of the Mexico Hydroelectric Project, fishing represented only 4.1 percent of productive activities among the SHARING PROJECT BENEFITS 243 displaced people prior to the project (1989), but by 1995 that figure had risen to 60.8 percent (Johns 1996). The Cirata and Saguling reservoirs in Indonesia provide an even more dramatic example: cage aquaculture employed 7,500 families directly and another 21,000 in secondary enter- prises, through production of 24,500 tons of fish a year, compared with the 10 tons produced annually in the original river fishery. Final produc- tion revenues are estimated to be seven times the value of the lost pro- duction from the submerged farms (box 6.3). In the 1998 OED review of eight large dam projects, six projects developed reservoir fisheries (World Bank 1998a-g). Yet it was found that only the two reservoirs in China developed fisheries into anything close to their real potential, using a combination of fish cages, netted areas of tributaries, fish ponds, and oyster beds. It appears that dissemination of the adequate technologies to aggressively stock the reservoirs with fish and shrimp could generate very substantial benefits for resettlers and surrounding populations. Transport Reservoirs create new possibilities and needs for water transport ser- vices, both for passengers and cargo. In both the Shuikou and Yantan hydro projects in China, previously less accessible villages are now on the shores of 100-kilometer-long reservoirs. Displaced people took ad- vantage of the opportunity to establish ferrying services, with the sup- port of local resettlement authorities. In the China Wanjiazhai Water Transfer Project, some affected farmers have seen the opportunity in the transport business and have opted for self-employment; 76 displaced people will be engaged in either private or collective transport business (World Bank 1997b). This phenomenon has been observed at other res- ervoirs in India and Africa. When displacement creates a need for new services or products, proactive resettlement planners examine these needs to see if displaced people can provide them and thereby create new income sources. Employment The benefit generally valued most by resettlers is wage employment, which yields a steady income flow and a sense of control over their lives. Even temporary employment by the project can provide much-needed income during the transition period, as well as skills, experience, and other benefits. The most feasible form of employment is in the activity the project is supporting. Several projects have given displaced people priority for 244 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Box 6.3 Diverting Benefits Away from Resettlers Using reservoir fisheries for the economic rehabilitation of displaced people provides an enlightening example of benefit sharing with resettlers. Indo- nesia's Saguling dam (completed in 1984) and Cirata dam (1987) displaced more than 115,000 people living in the Citarum River basin in Western Java. Initial planning made overoptimistic assumptions about resettlers' willing- ness to join the transmigration program and the suitability of cash compen- sation. Studies carried out shortly after displacement found that incomes for farmers resettled on land had declined 40 to 50 percent. Many displaced families, however, with support from a program launched by the government of Indonesia, the World Bank, and ICLARM (International Center for Living Aquatic Resources Management), took advantage of the newly created reservoirs to develop an innovative fisher- ies enterprise using fish culture technologies designed especially for small farmers. By 1992 cage aquaculture systems in the two reservoirs employed 7,500 people, produced 8,000 tons of fish-up from 10 tons produced in the original river fishery-and provided 25 percent of the total fish supply entering the Bandung district, an area with about 3 million people. By 1996 total aquaculture production was 24,500 tons. Total production could in- crease further if appropriate methods were utilized. Indonesia's Institute of Ecology reports that landless resettlers hired for cage aquaculture earn more than rice field workers in nearby areas. By 1996 fishery revenues from the two reservoirs exceeded $24 million a year, almost five times the $5.1 million a year lost by the regional economy from submerged rice fields. Cage aquaculture created an additional 21,000 jobs for resettlers in secondary small enterprises, such as fish feed farms, cage maintenance, and marketing. Current estimates suggest that final produc- tion revenues will level off at about $34.5 million a year, nearly seven times the value of the lost rice production from the submerged farms. A recent follow-up study of Cirata found that 59 percent of the resettlers believed they were better off than before displacement. Both reservoirs have be- come sources of economic growth for surrounding communities. However, the sustainability of benefit sharing in these projects is now at risk. The gains from cage aquaculture are so attractive that they draw nonresettlers. By the end of 1996, 52 percent of the cages were owned by competing nonresettlers who had paid resettlers or bribed officials for cage permits. These "outsiders" also had wrested almost complete control over marketing of the fish produced through aquaculture. Poor location of cages contributes to water pollution due to waste feed and nitrogen dis- charges. Furthermore, industrial discharges upstream cause fish kills. Therefore social, political, technical, economic, and environmental factors must be closely monitored and adaptations must be made to preserve resettlers' priority access to these benefits. Source: Costa-Pierce and Soemarwoto (1990), Costa-Pierce (1998). SHARING PROJECT BENEFITS 245 employment with the project-temporary, permanent, or both. For ex- ample, in the Pakistan Ghazi-Barotha Hydropower Project, "WAPDA [the Water and Power Development Authority] will give priority in em- ployment to the affectees in the construction phase and in filling open- ings in all job categories in the permanent operations and maintenance staff" (World Bank 1995n). In the Karnataka Power Project, "KPC [the power company] has absorbed 43 project affected persons in its regular workforce" (World Bank 1995i). In the Vietnam Power Development Project, the agency managing construction and operation of the thermal plant will present displaced people seeking employment to the contrac- tors. Electricity of Vietnam has recently started a two-year residential course for power plant operators, and suitable displaced people have been accepted (World Bank 1996h). Numerous coal mining and thermal power projects in India offer jobs to the people they displace, when feasible, at the rate of one job per household (box 6.4). These jobs are highly prized and often are passed on as inheritances within families. Difficulties arise when not all the dis- placed can be employed in the new plants and alternative options must be found. For example, the Jharia Coking Coal Project "provided em- ployment for 247 project-affected people. At this point, a deadlock has been reached since the remaining 551 project-affected families refuse to vacate their houses on the project site unless BCCL [Bharat Coking Coal Ltd.] offers employment to every person above 18. BCCL, however, has a significant overstaffing and low productivity problem. The company, therefore, decided not to resettle the remaining families and to reduce the mining area." Production reached only 24 percent of the target (World Bank 1996b). Many projects in China have deliberately allocated jobs to involun- tary displacees. This has been facilitated by China's rapid economic ex- pansion and the shift in the labor force from agriculture to industry. Local authorities can grant preference to resettlers, who are often eager to receive the higher and more reliable incomes found in industry. The Sichuan Power Transmission Project will give resettlers priority for em- ployment as transmission line maintenance workers and temporary construction workers (World Bank 1995b). This provides employment both during the construction phase and after the project is complete. The Shanghai Environment Project will recruit displaced people as opera- tion and maintenance staff at the pumping station for which the displacees' land is being acquired (World Bank 1994b). The Southern Jiangsu Environmental Protection Project is retraining farmers to work at the wastewater treatment plants that are displacing them. The com- pany implementing the water supply component of the Shandong Envi- ronment Project has agreed to provide some construction jobs to the 246 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT Box 6.4 Coal Companies That Employ People Whose Land They Acquire Some of the coal mines and thermal power plants constructed under pro- jects have offered jobs to some of the people displaced when their land was taken to be used for mines, plants, and ash ponds. Coal India has a com- pany policy of giving resettlers preference for unskilled and semiskilled jobs. "Considering that wages in the coal industry are 8 to 10 times the pre- vailing minimum wage, most project-affected people opt for employment with the coal companies" (World Bank 1995f, 1995h). Coal company jobs are seen as much more reliable sources of income, compared with the va- garies of smallholder agriculture. Several projects in India have used this method of benefit sharing. In the Dudhichua Coal Project "NCL [Northern Coalfields Ltd.] has employed 225 out of the 378 project-affected people" (World Bank 1995f). The Coal Mining and Coal Quality Improvement Project provided employment to 1,828 members of the 968 families displaced by the Gevra mine, employing 363 people directly and 1,500 indirectly (through contractors) of the 3,303 people displaced by the Sonepur-Bazari mine (World Bank 1997c). The Talcher Thermal Power Project employed 251 people directly and another 350 on a permanent basis under maintenance contracts for the main plant, the coal handling plant, lighting, sanitation and township maintenance, housekeeping, and so on. All 207 of the entitled families in the Chandrapur Thermal Project were provided one job per family with the state electricity board (World Bank 1995e). In the Second Farakka Thermal Power Project "449 PAPs (project affected people) have been employed by NTPC [National Thermal Power Corporation]. Another 165 PAPs opted to become contractors for NTPC. In addition, about 2,075 PAPs obtained temporary jobs with the contractors" (World Bank 1996e). Other countries also use this strategy. The Indonesia Suralaya Thermal Project will offer displaced families employment with the state electricity company. Both the Leyte Cebu and Leyte Luzon Geothermal projects in the Philippines contained plans to employ resettlers in the geothermal or power plant, which would not only increase their present income, but pro- vide them with training, health care, and retirement benefits (World Bank 1994e, 1994f). These are highly prized positions, and resettlers are eager to receive such opportunities. affected persons, as well as technical assistance for the construction of piped water systems for the affected villages, thereby combining jobs and infrastructure services for resettlers (World Bank 1997a). The Sec- ond Red Soils Area Development project will offer one permanent job per household in the construction and operation of the agroprocessing enterprise acquiring their land (World Bank 1994a). In the Lubuge SHARING PROJECT BENEFITS 247 Hydro Project 100 resettlers were assigned to reservoir maintenance and other development activities by the local governments. Most of these projects are displacing relatively small numbers of people, thus fa- cilitating reestablishment through employment opportunities with the project. The India Upper Krishna II and Maharashtra Composite III Irriga- tion projects employed displaced people as laborers in the construction of canals in developing the command areas (World Bank 1998g). In the Maharashtra case, many of the displaced people resettled in the com- mand areas, thus linking their temporary employment and source of permanent means of livelihood. Temporary employment with the pro- ject during the critical reestablishment phase should be offered even when the people being resettled cannot receive permanent employment from the project. Of course, temporary employment is not a substitute for complete, long-term reconstruction of livelihoods. Additional in- vestments are therefore necessary. Quite often projects that require preparation of relocation sites allo- cate that work to the displaced people themselves. This may have a dou- ble benefit: paying wages to the displaced during a particularly vulnerable period in the resettlement process, and helping ensure that preparation at the new site reflects the preferences of the resettlers. The India Kamataka Power Project mandated that displaced families do all the manual work in the land preparation process and be paid a daily wage. The China Ertan Hydroelectric Project is paying a salary to family members working on development of the resettlement sites. Although this labor is necessitated by the resettlement operation, and thus is not strictly a project output, it is work that could be given to outsiders, so paying the resettlers is directing a project resource to them. Housing Many urban-upgrading projects build new housing, so it makes sense that those displaced by a project should be among the first to receive new housing generated by that project. The Mozambique Urban Reha- bilitation Project planned to allocate 380 core units to be built in the first phase of the project to house families moved out of decaying (but high- quality) apartments that will be rehabilitated. The Ethiopia Second Addis Ababa Urban Development Project, though displacing only a few households, also planned to provide them with newly constructed housing units that are part of the project. Many urban projects provide serviced plots or a core unit (walls and a roof). The owners can then build or complete houses according to their own preferences and resources. Resettlers often use part of their 248 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT compensation money for this purpose and almost always can obtain better housing than they had before the project. This is the case in the Brazil Rio Flood Reconstruction and Prevention Project. The component that provides serviced sites for low-income families will create 11,000 serviced lots, of which 8,280 are to be allocated to families relocated be- cause of the project's civil works (drainage, slope stabilization, and re- forestation) components. Similar examples come from the Middle East and North Africa. The Jordan National Urban Development Project will give relocated house- holds first preference for newly serviced plots in either the infill or over- spill areas adjacent to or near the upgrading scheme. In the Tunisia Third Urban Development Project, about 600 of the 1,600 newly serviced plots would be used to rehouse families that have been displaced to re- duce overcrowding or as a result of necessary demolition in the project areas. The Tunisia Fourth Urban Project calls for families whose houses are demolished because of the infrastructure works to receive first prior- ity for lots. And in the Morocco Land Development Project for Low-Income Families, bidonville dwellers are offered such highly desir- able, subsidized lots that no one refuses in favor of remaining in the shantytown. The project executes the resettlement operation as de-densification to alleviate overcrowding in the slum areas and to pro- vide improved services. Some projects do not restrict resettlers to choosing only from among the housing or serviced plots created by the project. Instead, they offer vouchers or other in-kind credit toward housing of the resettler's choice. Although such projects add to the stock of housing, those specific houses or apartments may not go to resettlers. The project relies more on a free market solution. Resettlers get the chance to express their own preferences by obtaining credit toward any new housing they select, not just project-constructed housing. This method has been used success- fully in several urban projects in China, most notably in Shanghai (World Bank 1996g). Access to Improved Infrastructure Besides housing or serviced plots, urban development projects often provide those they displace with access to improved infrastructure. The Sierra Leone Freetown Infrastructure Rehabilitation Project plans to set aside two tracts of government land in the rehabilitated Kroo Bay area in order to provide those relocated with access to water supply, solid waste disposal, upgraded markets, roads, and footways. The Kenya Ex- port Development Project gave people displaced by the export process- ing zone (EPZ) titled plots next to the EPZ so that "they will be in the SHARING PROJECT BENEFITS 249 future connected [by] the same water and sewerage mains as the EPZ worker housing sites" (World Bank 19951). In the Guinea Second Urban Project, sites near the infrastructure works are rehabilitated for reset- tling displaced people who willbenefit from improved servicing and ac- cess to employment opportunities. Transportation and relocation costs are also kept to a minimum. Some urban projects provide benefits that are aimed at the entire af- fected population, not only at resettlers. The China Liaoning Urban Pro- ject is providing clean water to all project area residents, including resettlers. The China Henan Provincial Highway Project-as is true in many transport projects in China-is paying land compensation fees and resettlement subsidies to county, township, or village govern- ments, which use them collectively for area development. In many re- settlement projects in China, benefits that are provided to host communities receiving resettlers wind up being shared with resettlers. In turn, nonresettlers often share a part of the land they have been using (through redistribution) with families who have lost their land com- pletely. Everyone thus shares the costs (through land adjustment) for the greater good of all. Some of the preceding examples, especially in urban rehousing pro- jects, demonstrate how projects can be designed with resettler benefits as a central objective. In the Korea Ports Development and Environment Project, which is developing a new industrial-commercial complex, owners of displaced seafood eateries are offered the opportunity to par- ticipate in the project and reestablish their small shops as restaurants once the new complex is completed (World Bank 1994d). In this way the main project creates opportunities for those it displaces. There are many other examples of project benefit sharing. The ones presented are sufficient to illustrate the validity of the strategy and the diversity of approaches. These should be a source of inspiration for pro- ject planners and decisionmakers to expand the use of benefit sharing in forthcoming projects. Replicating Good Practices in Benefit Sharing for Income Recovery Judging from the evidence presented above, we can conclude that shar- ing project benefits with resettlers is a relatively recent practice. None- theless, our survey of projects receiving World Bank assistance over the past 10 years has revealed dozens of projects adopting a benefit-sharing approach to reconstructing resettlers' livelihoods. The projects doing this most frequently are those constructing dams and reservoirs, irriga- tion systems, coal mines, and urban infrastructure. However, almost 250 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT any kind of project can share benefits with resettlers if an orientation to- ward this strategy exists and there is the political will to implement it. Projects that combine multiple forms of benefit sharing hold the greatest promise for restoring and improving resettler living standards. Although the use of such approaches became more frequent after the Bank instituted explicit policy support, benefit sharing is still far from standard practice. Some projects utilize a single method of benefit shar- ing, when several methods could have been used in combination. If more than the restoration of resettlers' prior livelihood levels is pur- sued, and if improvements over these levels is to be achieved through resettlement with development, strategies based on deliberate bene- fit-sharing designs become imperative. Of great concern should be the consistent evidence that benefit shar- ing is seldom used in projects financed domestically without assistance from international donors. This chapter focused almost entirely on pro- jects the Bank partially finances in order to signal some of their best prac- tices. The need for generalizing them is acute. A broad review of resettlement (World Bank 1996g) found that the Bank is currently in- volved in relatively few projects, accounting for less than 3 percent of all the people displaced worldwide by development projects. Only several countries so far have been willing to establish resettlement policies with standards that match those recommended by the World Bank. Even a large country, such as India, with hundreds of domestically financed projects causing displacement, has not issued a national policy on invol- untary resettlement. As it has been well argued, the absence of policy responses to the perverse effects of forcible displacement is a cause of development dis- tortions in many developing countries that must be overcome (Cernea 1995a, 1995b). Worldwide, the majority of resettlers are displaced in projects not subject to World Bank resettlement policies or to policies with similar standards. The available documentation shows that the majority of those resettlers do not share in project benefits, largely be- cause national policies do not yet direct planners to pursue this goal and because private entrepreneurs rarely have such objectives. T'he sectoral and national policy approaches to benefit sharing de- scribed near the beginning of this chapter demonstrate the advantages of systematizing the practice. Many of the project examples came from projects that were subject to national and sectoral policies, illustrating the synergy of project-specific and policy-mandated approaches. For governments to use the Bank's resettlement policy in Bank-financed projects but not in domestically financed projects is short-sighted. Simi- larly, to adopt Bank policy and then to ignore the critical dimension of benefit sharing is an unfortunately limited and restrictive adoption of SHARING PROJECT BENEFITS 251 the policy (Cernea 1995a, 1995b). Benefit-sharing and developmental approaches are the most forward-looking dimensions in resettlement policy. Projects that are not predicated on benefit sharing have much to learn from projects that are. Benefit sharing generates resources beyond those already budgeted for resettlement. The provision of direct project outputs and benefits gives resettlers a positive, vested interest in the project, because beneficiaries and resettlers both benefit. Most impor- tant, resettlement is more effective and sustainable when an ongoing stream of benefits can be tapped for resettlers. Resettlement budgets are often insufficient to achieve lasting rehabilitation, but benefit sharing can provide continuous streams of benefits that are more se- cure and sustainable. Although this chapter focuses primarily on positive examples of benefit sharing-in order to document the advantages of this approach in improving resettlers' livelihoods-the important converse message is that cost and harm will result if this approach is not utilized. The chap- ter contains a few examples of projects that failed to share benefits or did so ineffectively. Many more examples are available and could have been included. During the research we came across dozens of projects that could have shared benefits but failed to do so. All too often the differ- ence between satisfactory and unsatisfactory resettlement performance was the failure to share benefits. This is symptomatic of the larger failure to seize resettlement as a development opportunity. This was mainly due to a failure of imagination or resolve on the part of project planners to "put resettlers first" when they targeted project benefits. Most significant are the cases where project authorities adopted benefit sharing after initial resettler resistance to the project. As benefit sharing yielded tangible benefits to resettlers (and project authorities) and increased their incomes and living standards, resettlers gained an interest in the project. An adversarial relationship was transformed into one of mutual benefits. Managers of these resettlement operations be- came effective champions of a more enlightened approach to resettle- ment planning and implementation. Our conclusion is that most projects cannot afford to treat benefit sharing as merely an act of goodwill; sharing benefits is necessary for good resettlement. The potential for this strategy is well established, even though interesting questions remain. For example, how to treat benefit-sharing issues in the economic and financial justification of pro- ject investments? How much additional resources can be raised? How to channel both project funds and project benefits accruing to resettlers to- ward new investments by the resettlers themselves? Which benefit- sharing methods are most successful, efficient, and sustainable? And 252 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT which ones do resettlers themselves value the most? Much more re- search is needed if we are to become adequately informed about this im- portant resettlement strategy, which could improve performance in a wide range of projects. Acknowledgements The author would like to thank Michael Cernea for initiating the writing of this chapter, and Gloria Davis for providing time in my work pro- gram to draft this chapter. Michael Cernea offered extensive sugges- tions and ideas during numerous drafts, which significantly improved the quality of the chapter. Dan Aronson, Daniel Gibson, Maninder Gill, and Scott Guggenheim provided valuable comments on an earlier paper that were incorporated into this chapter. Youxuan Zhu, Maria Clara Mejia, and Maria Teresa Serra explained to me the Chinese, Colombian, and Brazilian laws (respectively) that allocate hydroelectric revenues to people displaced by hydropower reservoirs. Robert Picciotto, Edward B. Rice, and Andres Liebenthal reviewed the chapter and contributed to my understanding of the economic and financial issues. References The word "processed" describes informally reproduced works that may not be commonly available through libraries. Cernea, Michael M. 1988. Involiuntary Resettlement in Development Projects: Policy Gutidelines in World Bank-Financed Projects. Technical Paper 80. Washington, D.C.: World Bank. . 1990. Poverty Risks from Population Displacement in Water Resouirces Devel- opment. Development Discussion Paper 355. Cambridge, Mass.: Harvard In- stitute for International Development. . 1995a. "Public Policy Responses to Development-Induced Population Displacement." Economic and Political Weekly 31 (24): 1515-23. . 1995b. "Social Integration and Population Displacement: The Contribu- tion of Social Science." International Social Science Journal 143 (1): 91-112. . 1995c. "Understanding and Preventing Impoverishment from Displace- ment: Reflections on the State of Knowledge." Journal of Reffigees Studies 8 (3): 245-64. .1997. "The Risks and Reconstruction Model for Resettling Displaced Pop- ulations." World Development 25 (10): 1569-87. Costa-Pierce, Barry. 1998. "Constraints to the Sustainability of Cage Aquaculture for Resettlement from Hydropower Dams in Asia: An Indone- sian Case Study." Journal of Environment and Development 7 (4). SHARING PROJECT BENEFITS 253 Costa-Pierce, Barry, and Otto Soemarwoto, eds. 1990. Reservoir Fisheries and Aquacultuirefor Resettlement in Indonesia. Manila: ICLARM. Fisher, William F., ed. 1995. Toward Sustainable Development? Struggling over In- dia's Narmada River. Armonk, N.Y.: M. E. Sharpe. Gray, Andrew. 1996. "Indigenous Resistance to Involuntary Relocation." In Christopher McDowell, ed., Understanding Impoverishlment: The Consequences of Development-Indutced Displacement. Oxford, U.K.: Berghahn Books. Johns, Kathryn. 1996. "Final Report of the Resettlement Program of the Aguamilpa Hydroelectric Project." World Bank, Washington, D.C. Processed. Oliver-Smith, Anthony. 1996. "Fighting for a Place: The Policy Implications of Resistance to Development-Induced Resettlement." In Christopher McDowell, ed., Understanding Impoverishment: The Consequences of Develop- ment-Induced Displacement. Oxford, U.K.: Berghahn Books. Pandey, Balaji, and associates. 1996. Development, Displacement, and Rehabilitation in Orissa 1950-1990. Bhubaneswar, India: Institute for Socio- Economic Development. Posey, Darrell A. 1996. "The Kayapo Indian Protests against Amazonian Dams: Successes, Alliances, and Unending Battles." In Christopher McDowell, ed., Understanding Impoverishment: The Consequtences of Devel- opment-Induced Displacement. Oxford, U.K.: Berghahn Books. Serra, Maria Teresa Fernandes. 1993. "Resettlement Planning in the Brazil- ian Power Sector: Recent Changes in Approach." In Michael M. Cernea and Scott Guggenheim, eds., Anthropological Approaches to Resettlement: Policy, Practice, and Theory. Boulder, Colo.: Westview. World Bank. 1980. Operational Manual Statement 2.33. "Social Issues Asso- ciated with Involuntary Resettlement in Bank-Financed Projects." Wash- ington, D.C. Processed. . 1986. Operations Policy Note 10.08. "Operations Policy Issues in the Treatment of Involuntary Resettlement in Bank-Financed Projects." Washing- ton, D.C. Processed. . 1990. Operational Directive 4.30. "Involuntary Resettlement." Washing- ton, D.C. Processed. . 1993a. "Early Experience with Involuntary Resettlement: Overview." OED Report 12142. Operations Evaluation Department, Washington, D.C. Processed. _.1993b. "Early Experience with Involuntary Resettlement: Impact Evalua- tion on India Karnataka Irrigation Project." OED Report 12132. Operations Evaluation Department, Washington, D.C. Processed. _. 1993c. "Early Experience with Involuntary Resettlement: Impact Evalua- tion on India Maharashtra Irrigation II Project." OED Report 12133. Opera- tions Evaluation Department, Washington, D.C. Processed. ._ 1994a. "China: Second Red Soils Area Development Project." Staff Ap- praisal Report 12394. Washington, D.C. Processed. 254 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT . 1994b. "China: Shanghai Environment Project." Staff Appraisal Report 12386. Washington, D.C. Processed. _ . 1994c. "China: Xiaolangdi Resettlement Project." Staff Appraisal Report 12527. Washington, D.C. Processed. . 1994d. "Korea: Ports Development and Environment Improvement Pro- ject." Staff Appraisal Report 12998. Washington, D.C. Processed. .1994e."Philippines: Leyte-Cebu Geothermal Project." Staff Appraisal Re- port 11449. Washington, D.C. Processed. . 1994f. "Philippines: Leyte-Luzon Geothermal Project." Staff Appraisal Report 12568. Washington, D.C. Processed. - 1994g. "Resettlement and Development: The Bankwide Review of Pro- jects Involving Involuntary Resettlement, 1986-1993." Environment Depart- ment, Washington, D.C. Processed. _ . 1995a. "Brazil: Ceara Urban Development and Water Resource Manage- ment Project." Staff Appraisal Report 12527. Washington, D.C. Processed. . 1995b. "China: Sichuan Power Transmission Project." Staff Appraisal Re- port 13468. Washington, D.C. Processed. . 1995c. "China: Yantan Hydroelectric Project." Implementation Comple- tion Report 14773. Washington, D.C. Processed. =_. 1995d. "Colombia: Rio Grande Multipurpose Project." Project Comple- tion Report 14663. Washington, D.C. Processed. . 1995e. "India: Chandrapur Thermal Power Project." Project Completion Report 15246. Washington, D.C. Processed. =. 1995f. "India: Dudhichua Coal Project." Project Completion Report 13938. Washington, D.C. Processed. .1995g. "India: Gujarat Medium II Irrigation Project." Project Completion Report 14768. Washington, D.C. Processed. . 1995i. "India: Karnataka Project." Implementation Completion Report 15184. Washington, D.C. Processed. - 1995j. "India: Orissa Water Resources Consolidation Project." Staff Ap- praisal Report 14888. Washington, D.C. Processed. _.1995k. "Indonesia: Kedung Ombo Multipurpose Dam and Irrigation Pro- ject." Project Completion Report 14636. Washington, D.C. Processed. _.19951. "Kenya: Export Development Project." Project Completion Report 13886. Washington, D.C. Processed. . 1995m. "Pakistan: Ghazi-Barotha Hydropower Project." Staff Appraisal Report 14587. Washington, D.C. Processed. . 1996a. "India: Coal Sector Environmental and Social Mitigation Project." Staff Appraisal Report 15405. Washington, D.C. Processed. SHARING PROJECT BENEFITS 255 . 1996b. "India: Jharia Coking Coal Project." Project Completion Report 15238. Washington, D.C. Processed. . 1996c. "India: Renewable Resources Development Project; Tamil Nadu Newsprint and Papers Limited Papermill Expansion Compo- nent." Implementation Completion Report 15619. Washington, D.C. Processed. . 1996d. "India: Second Andhra Pradesh Irrigation Project." Implementa- tion Completion Report 15708. Washington, D.C. Processed. - 1996e. "India: Second Farakka Thermal Power Project." Project Comple- tion Report 15243. Washington, D.C. Processed. - 1996f. "India: Upper Indravati Hydropower Project." Implementation Completion Report 15676. Washington, D.C. Processed. . 1996g. "Resettlement and Development: The Bankwide Review of Pro- jects Involving Involuntary Resettlement, 1986-1993" (second edition). Envi- ronment Department, Washington, D.C. Processed. - 1996h. "Vietnam: Power Development Project." Staff Appraisal Report 14893. Washington, D.C. Processed. _- 1997a."China: Shandong Environment Project." Staff Appraisal Report 16065. Washington, D.C. Processed. . 1997b. "China: Wanjianzhai Water Transfer Project." Staff Appraisal Re- port 15999. Washington, D.C. Processed. - 1997c. "India: Coat Mining and Coal Quality Improvement Project." Im- plementation Completion Report 16518. Washington, D.C. Processed. - 1997d. "India: Coal Sector Rehabilitation Project." Staff Appraisal Report 16473. Washington, D.C. Processed. - 1997e. "India: Third Andhra Pradesh Irrigation Project." Staff Appraisal Report 16336. Washington, D.C. Processed. . 1998a. "Recent Experience with Involuntary Resettlement: Brazil- Itaparica." OED Report 17544. Operations Evaluation Department, Washing- ton, D.C. Processed. . 1998b. "Recent Experience with Involuntary Resettlement: China- Shuikou (and Yantan)." OED Report 17539. Operations Evaluation Depart- ment, Washington, D.C. Processed. . 1998c. "Recent Experience with Involuntary Resettlement: Indonesia- Kedung Ombo." OED Report 17540. Operations Evaluation Department, Washington, D.C. Processed. _ . 1998d. "Recent Experience with Involuntary Resettlement: Overview." OED Report 17538. Operations Evaluation Department, Washington, D.C. Processed. 256 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT ___. 1998e. "Recent Experience with Involuntary Resettlement: Thai- land-Pak Mun." OED Report 17541. Operations Evaluation Department, Washington, D.C. Processed. . 1998f. "Recent Experience with Involuntary Resettlement: Togo- Nangbeto." OED Report 17543. Operations Evaluation Department, Wash- ington, D.C. Processed. . 1998g. "Recent Experience with Involuntary Resettlement: Upper Krishna (Karnataka and Maharashtra)." OED Report 17542. Operations Eval- uation Department, Washington, D.C. Processed. About the Contributors Michael M. Cernea joined the World Bank in 1974 as its first in-house so- ciologist and worked as the Bank's Senior Adviser for Sociology and So- cial Policy beginning 1997. He has carried out social research and operational project work in many countries in Asia, Africa, the Middle East, Europe, and Latin America. Mr. Cernea has a Ph.D. in sociology and social philosophy and has taught and lectured in universities in Eu- rope and the United States. He is Honorary Professor for Resettlement and Social Studies at Hohai University in Nanjing, China, was ap- pointed Visiting Scholar at Harvard University, and is a Mem- ber-correspondent of the Academy of Sciences, Romania. For his research and applied work, he received the Solon T. Kimball Award for Public Policy and Applied Anthropology, offered by the American An- thropological Association, and the Bronislaw Malinowski Award, con- ferred by the Society for Applied Anthropology "in recognition of scholarly efforts to understand and serve the needs of the world through social science." He has written and edited numerous books and articles on development, social change, population resettlement, social forestry, grassroots organizations, and participation, including Putting People First: Sociological Variables in Development (1985, 1991), Anthropological Approaches to Resettlement: Policy, Practice, Theory (ed. with Scott E. Guggenheim, 1993), Social Assessments for Better Development (ed. with Ayse Kudat, 1997), and Resettlement and Development, vols. I and II (in Chinese, 1997 and 1998). John H. Eriksen (1939-February 1999). (As this volume was prepared for publication, we received the sad news of John Eriksen's untimely death, in February 1999. All the co-authors of this volume express their sympa- thy to John's family for their, and our, tragic loss). John Eriksen was a founder and principal Executive Officer with Ithaca International Limited, an international development consulting firm. He consulted extensively for the United Nations Agency for 257 258 THE ECONOMICS OF INVOLUNTARY RESETTLEMENT International Development (USAID) and the World Bank throughout Asia and Africa. Prior to his international consulting career, he was a country director in the Peace Corps and a regional agricultural econo- mist for USAID. He obtained his M.S. and Ph.D. in agricultural econom- ics from Cornell University. John Eriksen published numerous studies on agriculture, livestock, natural resource management, and related fields. Lakshman K. Mahapatra. Emeritus Professor and former Head (1967 to 1989), Department of Anthropology, Utkal University, Bhubaneswar, Orissa, India; Visiting Professor at Hamburg University; Visiting Fellow of the Ford Foundation in Indonesia; former Vice-Chancellor of Utkal and Sambalpur Universities; former Director of Nabakrushna Centre for Development Studies, Bhubaneswar. Awards: Swami Pranava- nanda Award, University Grants Commission, 1994; R. P. Chanda Cen- tenary Memorial Medal, Asiatic Society, 1995. At present, Member, In- ternational Commission on Anthropology in Policy and Practice, and Member of International Commission on Anthropological Aspects of Global Environmental Change; President, Anthropological Society, Orissa. Main publications: over 150 scientific papers published in India or abroad; several books, including Swidden Cultivation in Asia (co-author); Tribal Development: Myth and Reality, Vikas Publishers, 1994; and Resettlement, Impoverishment, and Reconstruction in India, Vikas Publishers, 1999. (Address: 16, Satyanagar, Bhubaneswar 751007, Orissa, India). Maria Clara Mejia is a resettlement specialist in the Environment Unit for the Latin America and Caribbean Region of the World Bank and has worked in many countries throughout Latin America. Prior to joining the World Bank, she was Head of the unit of planning and socioeco- nomic studies in the Environment Office of the National Electricity Au- thority of Colombia. Ms. Mejia was trained in economics and energy planning at the Universidad de Medellin, Colombia, and the Instituto de Economia Energetica in Bariloche, Argentina. Her M.A. in anthropol- ogy is from Catholic University in Washington, D.C. Ms. Mejia has writ- ten several papers on social and environmental aspects of development projects in Latin America. David W. Pearce is Professor of Environmental Economics at University College, London, and Associate Director of the Centre for Social and Economic Research on the Global Environment (CSERGE). He has been Personal Adviser to the United Kingdom Secretary of State for Environ- ABOUT THE CONTRIBUTORS 259 ment, Chairman of the United Nations Economic Commission for Eu- rope (ECE), Economics Group on Acid Rain, and a Member of the Scientific and Advisory Panel of the Global Environment Facility (GEF). He has been a consultant to the World Bank, European Commission, In- ternational Union for the Conservation of Nature, Arab Fund, Overseas Development Administration, United Nations Conference on Trade and Development, International Labor Organization, and the World Health Organization. Mr. Pearce obtained his M.A. from Oxford Uni- versity and is a recipient of the United Nations Global 500 Award for services to the world environment. He is the author or editor of more than 40 books, including Blueprint for a Green Economy (1989), Economics of Natural Resources and the Environment (1990), Sustainable Development (1990), and World Without End (1993). Warren A. Van Wicklin III is a social scientist in the Operations Evalua- tion Department at the World Bank. Prior to joining the World Bank, he was a consultant to USAID, the United Nations Centre for Human Set- tlement, and various government agencies and foundations. He has conducted fieldwork and evaluations in several countries in Asia and in Latin America. Mr. Van Wicklin earned his Ph.D. in political science at MIT. His publications include several articles on the contribution of beneficiary participation to development project effectiveness and pro- ject evaluation. 44j THE WORLD BANK 1818 11 Str-cet, NAX". 'Mashlngton, D).C. 20433, UJ.S.A. I I 11 1F ION 1 (2(02) 4 2 1 2 34 I: (202~~~ 4,7-6391 I 1 [ .m: 64145 \\1'11I DI)ANIK mci 24842.3 \W0RII)IIANIK W oRI 1I) WI \I I VII: 1.1 htt)://www.worldbanik.org,/ I 1 \11: book("'worldbanik.org COVER DESIGN: THE MAGAZINE GROUP ISBN 0-8213-3798-X