Annex 2. Political History of Burkina Faso 144 Annex 3

\v/TPQOCC Lone Iqq&

Settlement and Development Public Disclosure Authorized in the River Blindness Control Zone Case Study Burkina Faso

Della E. McMillan, Jean-Baptiste Nana, and Kimseyinga Savadogo Public Disclosure Authorized

.9 I ~ ~ ~ ~ ~ , Public Disclosure Authorized \ I~ ~ *.

I -___ I Public Disclosure Authorized RECENT WORLD BANK TECHNICAL PAPERS

No. 136 Green, editor, CoconutProduction: Present Status and Prioritiesfor Research No. 137 Constant and Sheldrick, An OutlookforFertilizer Demand, Supply, and Trade,1988189-1993194 No. 138 Steel and Webster,Small Enterprises under Adjustment in Ghana No. 139 Environment Department, EnvironmentalAssessment Sourcebook, vol. I: Policies,Procedures, and Cross-SectoralIssues No. 140 Enviromnent Department, EnvironmentalAssessment Sourcebook, vol. HI:Sectoral Guidelines No. 141 Riverson,Gaviria, and Thriscutt,Rural Roadsin Sub-SaharanAfrica: Less(ms from WorldBank Experience No. 142 Kiss and Meerman, IntegratedPest Management and African Agriculture No. 143 Grut, Gray, and Egli, ForestPricing and ConcessionPolicies: Managing the High Forestof West and CentralAfrica No. 144 The World Bank/FAO/UNIDO/Industry Fertilizer Working Group, Worldand RegionalSupply and DemandBalances for Nitrogen,Phosphate, and Potash,1989/90-1995/96 No. 145 Ivanek, Nulty, and Holcer, ManufacturingTelecommunications Equipment in Newly Industrializing Countries:The Effectof TechnologicalProgress No. 146 Dejene and Olivares, IntegratingEnvironmental Issues into a Strategyfor SustainableAgricultural Development:The Case of Mozambique No. 147 The World Bank/UNDP/CEC/FAO, Fisheriesand AquacultureResearch Capabilities and Needs in Asia:Studies of India, Thailand,Malaysia, Indonesia, the Philippines,and theASEAN Region No. 148 The World Bank/UNDP/CEC/FAO, Fisheriesand AquacultureResearch Capabilities and Needs in LatinAmerica: Studies of Uruguay,Argentina, Chile,Ecuador, and Peru No. 149 The World Bank/UNDP/CEC/FAO, Fisheriesand AquacultureResearch Capabilities and Needs in Africa:Studies of Kenya,Malawi, Mozambique, Zimbabwe, Mauritania, Morocco, and Senegal No. 150 The World Bank/UNDP/CEC/FAO, InternationalCooperation in FisheriesResearch No. 151 The World Bank/UNDP/CEC/FAO, TropicalAquaculture Development: Research Needs No. 152 The World Bank/U1NDP/CEC/FAO, Small-ScaleFisheries: Research Needs No. 153 The World Bank/UNDP/CEC/FAO, Small PelagicFish Utilization:Research Needs No. 154 Environment Department, EnvironmentalAssessment Sourcebook, vol. III:Guidelines for EnvironmentalAssessment of Energyand Industry Projects No. 155 Belot and Weigel, Programsin IndustrialCountries to PromoteForeign Direct Investment in Developing Countries No. 156 De Geyndt, ManagingHealth Expenditures under National Health Insurance: The Case of Korea No. 157 Critchley, Reij,and Seznec, WaterHarvestingfor Plant Production,vol. II: CaseStudies and Conclusionsfor Sub-SaharanAfrica No. 158 Hay and Paul, Regulationand Taxationof CommercialBanks during the InternationalDebt Crisis No. 159 Liese,Sachdeva, and Cochrane, Organizingand ManagingTropical Disease Control Programs: Lessonsof Success No. 160 Boner and Krueger, The Basicsof Antitrust Policy:A Reviewof Ten Nationsand the European Communities No. 161 Riverson and Carapetis, IntermediateMeans of Transportin Sub-SaharanAfrica: Its Potential for ImprovingRural Traveland Transport No. 162 Replogle, Non-MotorizedVehicles in Asian Cities No. 163 Shilling, editor, BeyondSyndicated Loans: Sources of Creditfor DevelopingCountries No. 164 Schwartz and Kampen, AgriculturalExtension in East Africa No. 165 Kellaghan and Greaney, Using Examinationsto ImproveEducation: A Study in FourteenAfrican Countries

(List continues on the inside back cover) WORLDBANK TECHNICALPAPER NUMBER 200 SERIESON RIVERBLINDNESS CONTROL IN WESTAFRICA

Settlement and Development in the River Blindness Control Zone Case Study Burkina Faso

Della E. McMillan, Jean-Baptiste Nana, and Kimseyinga Savadogo

The World Bank Washington, D.C. Copyright i 1993 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A.

Technical Papers are published to communicate the results of the Bank's work to the development community with the least possible delay. The typescript of this paper therefore has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. Any maps that accompany the text have been prepared solely for the convenience of readers; the designations and presentation of material in them do not imply the expression of any opinion whatsoever on the part of the World Bank, its affiliates, or its Board or member countries concerning the legal status of any country, territory, city, or area or of the authorities thereof or concerning the delimitation of its boundaries or its national affiliation. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, 27 Congress Street, Salem, Massachusetts 01970, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of Publications, which contains an alphabetical title list (with full ordering information) and indexes of subjects, authors, and countries and regions. The latest edition is available free of charge from the Distribution Unit, Office of the Publisher, Department F, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, US.A., or from Publications, The World Bank, 66, avenue d'Iena, 75116Paris, France.

ISSN: 0253-7494

Della E. McMillan is a research affiliate to the Department of Anthropology at the University of Florida. Jean-Baptiste Nana is the Director of Cabinet d'Etudes d'Appui et de Recherches pour le Developpement in Ouagadougou, Burkina Faso. Kimseyinga Savadogo is the Chair of the Department of Economics at the University of Ouagadougou in Burkina Faso.

Library of Congress Cataloging-in-Publication Data

McMillan, Della E. Settlement and development in the river blindness control zone case study, Burkina Faso / Della E. McMillan, Jean-Baptiste Nana, and Kimseyinga Savadogo. p. cm. - (World Bank technical paper, ISSN 0253-7494; no. 200. Series on river blindness control in West Africa) Includes bibliographical references. ISBN 0-8213-2381-4 1. Land settlement-Burkina Faso. 2. Agriculture-Economic aspects-Burkina Faso. 3. Onchocerciasis Control Programme in West Africa. 4. Onchocerciasis-Burkina Faso. I. Nana, Jean-Baptiste, 1951- II. Savadogo, Kimnseyinga,1955- . m. Title. I. Series: World Bank technical paper ; no. 200. V. Series: World Bank technical paper. Series on river blindness control in West Africa. HD1018.Z63M35 1993 333.3'16625-dc2O 93-9380 CIP Foreword

This volume is a case study of land settlement mated 300,000 cases of the disease by the end in Burkina Faso prepared as part of the Land of this decade. Settlement Review (LSR), a regional study of In addition to its substantial health benefits, land settlement in West Africa. The objective of one of the primary justifications of the pro- the LSR was to examine the phenomenon of gram has been the substantial development land settlement resulting from the successful potential of sparsely populated onchocerciasis control of onchocerciasis (riverblindness) by endemic river valleys. It is estimated that 25 the Onchocerciasis Control Programme (OCP). million hectares of tillable land is being opened The LSR analyzes the process of land settle- up by onchocerciasis control and rapid migra- ment, both spontaneous and government spon- tion to these river valleys is already occurring. sored, and recommends operational policy OCP is not directly involved in the socio-eco- guidelines for promoting viable and sustain- nomic development of these areas, but the able settlement-related activities in the areas Committee of Sponsoring Agencies (UNDP, where onchocerciasis has been controlled. FAQ, WHO, the World Bank), the statutory OCP is widely acknowledged as the most body that oversees OCP, has encouraged de- successfulon-going disease control p m i velopment in the OCP area by executing two successrul on-go1g dbsease control f-program regional studies: the first, Socio-economic De- Africa. Started in 1974 by a group of 7 West Af- velopment Studies in the Onchocerciasis rican countries and 9 donors, including the Control Programme Area, looked at the devel- World Bank, the program has expanded to 11 opment potential of the onchocerciasis freed countries, includes 22 donors and protects 30 areas and the second, the LSR, examined the million people. In most of the original program extent and impact of new settlement in the area the disease has been virtually eliminated OCP river valleys. and lasting control of the disease throughout The LSR, prepared by the Institute for Devel- the extended program area is expected to be se- opment Anthropology, Binghamton, New cured around the year 2000. The health impact York, consists of in-depth case studies of Burki- of the program has been significant. Nine mil- na Faso, Ghana, Mali and Togo, as well as lion children born since the beginning of the shorter reviews of settlement activities in the program are free of any risk of contracting the other OCP countries (Benin, Cote d'Ivoire, disease. Over 1.5 million people who were Guinea, Guinea-Bissau, Niger, Sierra Leone, once seriously infected are now disease free and Senegal). The Burkina Faso case study is and the program will have prevented an esti- being published as part of the riverblindness

iii technical paper series because Burkina Faso ment. Other volumes in the series are the Final has experienced more settlement in the on- Report from the LSR and an overview of the chocerciasis-controlled areas than any other structure and history of OCP. country. The government of Burkina Faso has been active in promoting settlement over an 18 Katherine Marshall year period and this study provides important Director insights into the possibilities for and limits to Sahelian Department government action in support of land settle- Africa Region

iv Preface

The socio-anthropological study by Della Mc- The Burkina Faso case study was selected for Millan, Jean-Baptist Nana and Kimseyinga Sava- publication because this is the country where dogo on settlement strategies in Burkina Faso, new settlement has progressed most during the published in the present volume, is one of the ba- first 18 years of onchocerciasis control. The study sic, in-depth country case studies of the Land Set- is also distinguished by the fact that it benefitted tlement Review carried out in 11 countries in from previous baseline data (Murphy and Sprey, West Africa in the Onchocerciasis Control Pro- 1980;McMillan 1983)that could be restudied ; it gram area. The overall report of the Land Settle- includes a rare longitudinal case study of settlers ment Review was published in a seperate report who immigrated to the Volta river planned settle- in the present publication series.1 ments between 1974and 1979.3 The extraordinary achievement of reducing The significance of this research extends be- and virtually eradicating river blindness in vast yond the topic itself. Today many of the oncho areas of West Africa, where for decades people control areas are being reoccupied by adjacent have suffered terribly from this disease, has re- populations as well as by immigrants from out- opened these territories to expanding human set- side the immediate basin region. Unassisted and tlements. The Land Settlement Review, carried unguided, spontaneous or anarchic settlement, out by a research team of the Institute for Devel- many result in the development of local systems opment Anthropology (Binghamton, NY), with for agricultural production that have negative support from several sponsoring agencies includ- long term consequences for people and natural ing UNDP and World Bank, has explored the po- resources, ranging from declining productivity tential for such settlements and has formulated and real income levels to outright environmental recommendations to interested governmentsand deterioration and the eventual abandonment of agencies as to the need and approach for assisted once-productive lands. The land settlement resettlement strategies. view and its country studies strive to identify ap- Thecountrystudypresentedinthisvolumeby proaches and strategies that will result in Della McMillan and her associates enables the production systems that are sustainable and that reader to examine, for one country, the wealth of contribute to the viable socioeconomic develop- empirical material on which the analysis and rec- ment of the surrounding region. The dedicated ommendations of the settlement review are field researchers of the Burkina Faso case study based, and the field research methodology have not only reconstructed a convincing image through which it was gathered and processed. of the web of complex social processes going on in the areas they studied, but have also made a 1 See Settlement and Development in the River Blindness Control Zone, by Della McMillan,Thomas Painter and Thayer 2. See Murphy, Josette and Sprey, Leenert 1980. Scudder, 1992,Washington, DC, The World Bank. 3 McMillan,Della, 1983,1986,1987a,1987b.

V valuable contribution by distilling lessons of wid- study case give a much firmer basis for the next er relevance for future settlement policy generation of planning to respond to the new The propositions outlined in the Land Settle- challenge and opportunities created by the suc- ment Review and in the Burkina Faso in-depth cessful control of river blindness.

-Michael M. Cemnea SeniorAdviser Sociologyand SocialPolicy T7heWorld Bank

vi Acknowledgments

The Burkina case study was funded by the Unit- compares settlement processes in the entire ed Nations Development Program (UNDP) and eleven-country region. In March 1990, the three designed and implemented by three senior team senior researchers for the Burkina Country Case members. Kimseyinga Savadogo, Chair of the Study met in Ouagadougou to reanalyze the Department of Economics, University of Ouaga- Burkina data in light of the analysis presented in dougou, was responsible for the design, testing, the Draft Final Report. The present document is and analysis of the economic questionnaires at all the result of this second analysis. It draws heavily of the four study sites (see Savadogo 1989a,b,c,d on the results of the Final Report and, indeed, and Savadogo and Sanders 1989). Jean-Baptiste quotes it in several places. Nana, a sociologist and Director of the Cabinet Guira Frederic, Contr6leur des Enquetes at the d'Etudesd'Appui et de Recherchespour le Developpe- AVV,was responsible for interviewing 60 AVV ment (CEARD) was responsible for coordinating households that had been included in the earlier research at the Niangoloko, Solenzo, and AVV AVV Statistical Service Survey. He is also respon- (Amenagement des Vallees des Volta) Rapadama sible for several special studies including inter- sites (see Nana 1989a,b,c,d,e). Della McMillan, views concerning the Linoghin market and UPI anthropologist, Institute for Development (Unite de Planification) village groupements. Anthropology (IDA), was responsible for re- Moustapha Ouedraogo was responsible for the search in the AVV planned settlements at economic survey and supplementary special Mogtedo, Mogtedo-Bombore, and Linoghin, and studies at Kompienga; he also conducted two at Kompienga. McMillan was also responsible for special substudies, one on gold mining in the UPI writing the first draft of the national report based region and the other on pastoralists living in the on the site reports, sector analyses, and other re- Gadeghin pastoralist area. Guiraud Jean-Baptiste search conducted by the team (McMillan 1989). conducted the interviewing for an extensive mi- The first version of the Burkina Faso country gration study and economic survey in the So- case study, along with the draft versions of the lenzo region as well as interviewing farmers for country case studies for Mali, Togo, and Ghana, the Rapadama study. Fatou Kabore, Boena Salif, and a review of settlement activities in the other and Tinga Ouedraogo worked on the economic seven Onchocerciasis Control Programme (OCP) and market surveys for the AVV-UDs(Unites de countries, provided the basis for the Onchocercia- Developpement)of Mogtedo and Mogtedo-Bom- sis Control Programme, Final Report of the Land bore, as well as special studies of the town of Settlement Review: Settlement Experiencesand De- Mogtedo and the Mogtedo market. Sommaila velopment Strategies in the OnchocerciasisControl Savadogo, sociologist, UP1, worked with us in Programme Areas of West Africa (McMillan, developing a special study of settler dropout in Painter, and Scudder 1990). The Final Report the Linoghin, Mogtedo, and Mogtedo-Bombore

vii AVV-UDs.The UD chiefs of Mogtedo, Bombore, been possible without the help of the UD chiefs at Rapadama, and Linoghin assisted us with the de- Mogtedo, Mogtedo-Bombore, Linoghin, and Ra- velopment of longitudinal data sets on produc- padama. The Prdfetat Mogtedo and the Mogtedo tion and settlement patterns in their own groups CR (Revolutionary Committee) extended their of planned settlements. support and collaboration to the longitudinal re- This research could not have been conducted search we conducted on the Mogtedo market. At without the active support of the OCP and various a personal level we would like to thank Arthur- national and regional agencies that have offered Felix Yanogo, Benjamin Tabsoba, and the hard- their assistance. In the Solenzo region we are in- working farmers of Mogtedo V3 who have made debted to the Director of the CRPA du Mouhoun it possible for us to continue the longitudinal case for giving us access to many of their data sources study these ten long years. as well as occasionally extending help with data The Burkina case study and the Land Settle- collection.We are especially appreciative of the ef- ment Review (LSR) have benefited immensely forts of Adama Traore, who also heads the PNGT from the strong support we received from Dr. (Programme National de Gestion des Terroirs Vil- Ebrahim Samba, Director of the OCP, and his lageois) pilot projects for the Solenzo region. The staff. For more than a year we found encourage- Pr6fetof Solenzo is to be thanked for the adminis- ment and logistical support from our many OCP trative assistance he provided. colleagues, and a home away from home in the Our work in Niangoloko would not have been OCP Economic Unit, under the direction of Mr. possible without the unfailing support we re- Jean-Baptiste Zongo. Mr. Michel Pare, Mr. Patrick ceived from the Pr6fet.We also wish to extend our Chopin, and Mr. Ayina are thanked for their as- thanks to the staff of the Comoe CRPA and the sistance with the myriad administrative and lo- multiple forest use project at the Classified Forest gistical details we faced in doing the research. at Toumousseni. Mme. Kabore, the OCP librarian, worked to track Our research at Kompienga was conducted down much of the bibliography cited here; her ef- with strong support from the dam construction forts are invaluable. Without technical assistance authority (Direction Generalede la Maftrise d'Ou- on microcomputers from Mr. Soumbey, in the vragede la Komnpienga[DGMOK]) as well as SON- OCP Biostatistics and Information Systems Unit, ABEL (SocieteNationale d'Electricitedu Burkina), we would never have been able to analyze the the national electricity authority. We are especially massive amounts of data collected. Mr. Saidou indebted to Pascal Zagre, former General Director Gadjaga, of the Biostatistics and Information Sys- of the Kompienga Project; Guira Antoine, the tems Unit, helped with the analysis of certain DGMOK sociologist, now with the Ministry of data sets; Mr. Daribi and Mme. Simpore, with Education; and Jean Ouedraogo, DGMOK sup- data entry. plies chief, now attached to the new Bagreproject. Jeanne Cornell and Elizabeth Albert, graduate M. Emmanuel Nikiema, Director-General of students at the University of Kentucky, provided the AVV,has been extremely supportive of our research and production assistance on the earlier work throughout the country. We are especially version of the Burkina report as well as transla- indebted to him and his staff for the excellent tion and typing on other country reports. Jeanne overview they gave us of development in the dif- Harlow played a major role in the initial struc- ferent zones. Mme. Celestine Bere, Director of the tural editing of the first draft of the Burkina re- AVV Suivi-Evaluation Department, has served port, and is the primary person responsible for courageously as Burkina's representative on the the design and reproduction of maps and other national onchocerciasis committee for many graphics for this report and the Final Report. years. We appreciate the detailed comments we Melanie Brandt drew some of the maps. Debbie received from Mme. Bere and We are also very Anderson, computer systems analyst, did an ex- grateful to the logistical support provided by the cellent job of carefully verifying the base data last two directors of the AVVDirection des Etudes, and running some of the important data sets-in- Programmationet Controle(DEPC). On the ground cluding land measurements and input use at So- we were aided tremendously by the strong sup- lenzo, the Kompienga census data, the FulBe port we received from the Director of the AVV- economic studies, and the expenditure and reve- UP1 office at Zorgho, and his staff. Our research nue survey. For final editing of this monograph in the AVV planned settlements would not have we are indebted to Vivian Carlip at IDA, for the viii initial draft, and Carol Lauriault at the Center for Bado, Nana Timbila and Dominique Kaunda. We African Studies, University of Florida, for this would also like to thank Josette Murphy of the monograph. At IDA, Vivian Carlip received as- World Bank for her excellent detailed comments. sistance from Sylvia Horowitz, Curt Grimm, Kim Murphy's baseline study of the AVV was the Munson, and Dianne Anson. inspiration for the restudy that is central to the We would like to extend thanks to Michael Burkina analysis. Finally, we would like to thank Cernea and Scott Guggenheim of the World Bank our spouses, Colleen McCracken, Rasmata Nana and to Nicole Glineur, Tijan Sallah, and Beth and David Wilson for the patience they have Skinner of the World Bank Oncho Unit. We would shown in supporting us through the research. also like to thank Mr. Claude R. Delapierre, the We are grateful to Bruce Benton and John Elder Resident Representative of the Burkina World of the World Bank Oncho Unit who encouraged Bank Mission, and his staff including Celestin us to revise the initial report for publication.

ix Acronyms

ALE Adult labor equivalent AVV Autorite des Amenagements des Vallees des Volta (VoltaValley Authority) CEARD Cabinet d'Etudes d'Appui et de Recherches pour le Developpement CILSS Comite Permanent Inter-etats de Lutte contre la Secheresse dans le Sahel CMRPN Comite Militaire de Redressement pour le Progres National CNR Conseil National de la Revolution CR Revolutionary Committee (Comite Revolutionnaire) CRPA Centre Regional de Promotion Agropastorale CSA Committee of Sponsoring Agencies CSFA Conseil Superieur des Forces Armees CSP Conseil de Salut du Peuple DAF Direction Administrative et Financiere DEPC Direction des Etudes, Programmation et Controle DGMOK Direction Generale de la Maitrise d'Ouvrage de la Kompienga DMR Departement du Monde Rural FAO Food and Agriculture Organization of the United Nations GNP Gross national product GRN Gouvernement Renouveau National GTV Gestion des Terroirs Villageois GTZ Gesellschaft fur TechnischeZusammenarbeit (German Development Agency) ICRISAT International Crop Research Institute for the Semi-Arid Tropics IDA Institute for Development Anthropology IMF International Monetary Fund INERA Institut d'Etudes et des Recherches Agricoles INSD Institut National de la Statistique et de la Demographie LSR Land Settlement Review NGO Nongovernmental organization NPK Nitrogen-phosphorus-potassium (fertilizer) OCP Onchocerciasis Control Programme OFNACER Office National des C&reales ONAT Office d'Amenagement des Terroirs ORD Organisme Regional de Developpement PATECORE Projet Am6nagement des Terroirs et Conservation des Ressources dans le Plateau Central x PNGT Programme National de Gestion des Terroirs PNGTV Programme National de Gestion des Terroirs Villageois RDA Rassemblement Democratique Africain SOFITEX Societe des Fibres Textiles SONABEL Soci6te Nationale d'Electricite du Burkina UD Unite de Developpement (Development Unit) UNDP United Nations Development Programme UP Unite de Planification (Planning Unit) USA1D United States Agency for International Development WHO World Health Organization

Note: The exchange rate in July 1989 was approximately $1 US = 315 CFA francs.

Contents

Foreword iii Preface v Acknowledgements vii Acronyms x Abstract xix

1. Introduction 1

2. Overview of the OCP River Basins 6 Location and Natural Resource Potential 6 Historic Settlement Patterns 7 Early Development Planning for the OCP River Basins 7 Settlement and Immigration Patterns: 1974-1990 7 Other Physical, Demographic and Social Constraints 8 Projected Immigration Trends: 1990-2000 11 Potential Policy Implications 12

3. Types of Settlement and Settlement Dynamics 14 Immigration to the Study Sites as Examples of Settlement Types 14 Settlement Dynamics at the Study Sites 21 The Concept of Settlement Stages 23 Policy Implications of Interrelated Social and Economic Stages 2S

4. Linkages between Local Planning and National Agricultural Policy 28 Producer and Input Price Policies 28 Local Impact of Policies 29 Implications for Future Policy 34

5. Land Tenure 38 Le Regime Coutumier-Indigenous Land Tenure Systems 38 Le Regime Reglementaire-The Concept of Individual and State Ownership 39

xiii Agrarian and Land Tenure Reorganization, 1984 40 Early Impact of the Land Tenure and Agrarian Reform Act 40 Indirect Consequences of Reform for Settlement-Related Development 42 Policy Implications 43

6. Production at the Household and Community Levels: The Centrality of Diversification 45 Household Organization of Production at the Study Sites 45 The Centrality of Diversification within Household Production Systems 48 Factors That Affect the Development of Off-farm Employment at the Study Sites 53 Household Level Factors that Affect the Development of Off-farm Employment 55 Linkages among Diversification, Agricultural Intensification, and Sustainability 55 Settler "Step" Migration from the AVV to Kompienga 56 Policy Implications 58

7. Farming Systems 60 Farming Systems in Areas of Unassisted Spontaneous Settlement: Niangoloko and Kompienga 60 Efforts to Promote Higher Yielding, Sustainable Farming Systems 62 Factors Affecting the Development of Sustainable Farming Systems 68 Macro-level Factors that Contribute to or Detract from Intensification 70 Policy Implications 71

8. Importance of Markets and Service/Administrative Centers 74 Settlement-Related Development and Markets 74 Settlement-Related Development and Markets in the Niangoloko and Kompienga Areas 76 Markets and Settlement-Related Development 79 Factors That Affect Market Dynamics 81 Policy Implications 85

9. Integration of Hosts, Settlers, and Pastoralists 87 The Importance of Integration 87 Integration of Hosts with Settlers 87 Integration of Hosts and Settlers with Local Pastoralists 89 Agriculturalist-Pastoralist Relations over Time 89 Factors that Affect Integration Dynamics 91 Policy Implications 92

10. Knowledge and Management of the Natural Resource Base 95 The Dilemma of Spontaneous New Lands Settlement 95 Early Plans for and Response to the AVV Model for Sustainable Agriculture 96 Early Plans for and Response to Zoning for Agropastoral Areas 96 Early Plans for and Response to Zoning for Classified Areas 99 Alternative Models for Village Land Management: PNGT 100 Land-Use Planning in Connection with Forest and Wildlife Preserves 102 Policy Implications 102

I1. National Management Institutions 106 Creation of the Autorite des Amenagements des Vallees des Volta (AVV), 1974 106 Early Emphasis on Planned Settlement as a Strategy for Development of the Valleys 107 Reorientation of National Planning for the OCP River Basins, 1983-1990 110 Assisted Settlement Programs Incorporating Spontaneous Settlers in Sponsored Settlements 112

xiv Current Reorganization of the AVV, 1990 116 Lessons Learned and Potential Applications 118

12. Local Participation and Institutions 122 Centrality of Village Extension Groups in Crop Extension Programs 122 Early Development of Settler Organizations at Kompienga 127 Policy Implications 127

Bibliography 129

Annex 1. Methodology of the Land Settlement Review In Burkina Faso 138 Annex 2. Political History of Burkina Faso 144 Annex 3. Supplementary Tables on Consumer Prices and Price Indexing 146 Annex 4. Supplementary Tables on Livestock Production 151 Annex 5. Supplementary Tables on Farming Systems 154

Tables 1.1 The Land Settlement Review Case Studies 4 2.1 Burkina's Major River Basins and Area Estimated to Be Directly Affected by the Onchocerciasis Control Programme 6 2.2 New Land Cultivated in Major River Basins, 1975-1983 8 3.1 Classified Forests in Comoe Province 21 3.2 A Stage Model of Settlement-Related Development Planning 23 4.1 Official Price Per Kilogram of Fertilizer (In CFA) and Percent Subsidized 28 4.2 Official Prices for Cotton and Cotton Inputs at the AVV, 1979-1989 29 4.3 Price Ranges for White Sorghum at the Study Sites, 1988-1989 30 4.4 Estimated Cash Returns to One Hectare of Cotton or Sorghum, 1988/89 30 4.5 Average Cotton Yields, Production, and Area Planted, in the Solenzo Sector, 19762/73-1988/89 33 4.6 Comparison of Net Agricultural Income in the Older AVV Planned Settlements, 1977-1988/89 per Unit Labor 34 5.1 Years of Cultivation of Collective Fields, Solenzo, 1988/1989 42 6.1 Income Sources by Zone and Income Stratum, 1981-1985 Averages 48 6.2 Mean Income from the Sale of Handicrafts and Trade, 1979 and 1988/89 51 6.3 Average Non-farm Income from Different Sources for Individuals Reporting Income from the Source, 1988-1989 52 6.4 Percent of Women Indicating Income from Off-farm Employment and Livestock Production and Average Income from these Sources for the Entire Sample, 1988/89 52 6.5 Average Income by Income Source for Male Heads of Household and Women in the Study Sample, 1988/89 53 7.1 Mean Input Costs for the Male Head of Household for Different Ethnic Groups, Kompienga, 1988/89 61 7.2 Crop and Livestock Production per Adult Labor Equivalent (ALE), Kompienga, 1988/1989 62 7.3 The AVV Index of Adult Labor Equivalents (ALE) by Sex and Age 63 7.4 Recommended AVV Rotation 63 7.5 A Comparison of the AVV's Recommended Crop Area with the Actual Area Planted in 1979 and 1988 64 7.6 Crop and Livestock Production and Production Expenses per Adult Labor Equivalent (ALE) in AVV Planned Settlements, 1988/1989 65 7.7 Number of Fields that Received Different Doses of NPK Fertilizer at Solenzo 65

xv 7.8 Crop and Livestock Production per Adult Labor Equivalent (ALE Solenzo 1988/89) 66 8.1 Settlement and Development in the Town of Mogtedo, 1960-1989 74 8.2 Dry Season Attendance and Number of Market Stalls at the Linoghin Market, 1975-1989 76 8.3 Number of Miners and Merchants at the Mogtedo Area Gold Sites, April 1990 83 8.4 Reported Net Income Earned from Gold Mining for 15 Miners 84 11.1 ActivitiesoftheAVV, 1973-1981 107 11.2 Cost to the AVV of Settling One Family according to the BEI-Agrer Report 109 11.3 A Comparison of the Actual and Projected Rates of Planned Settlement at the AVV, 1973-1979 110 11.4 Method of Calculating Carrying Capacity in the AVV-Assisted Settlement Program at Rapadama 115 11.5 Census of AVV Areas of Intervention during 1987-1988 118 11.6 Financing of the AVV, 1972-1988 120 A.1 Site and Field Reports Completed as Part of the Land Settlement Review in Burkina Faso 138 A.2 Sample Size of the Economic Surveys Conducted in Burkina Faso, 1988-1989: Number of Households 139 A.3 Follow-up of Farmers Included in the 1979 Statistical Service Survey 140 B. 1 IMF Consumer Price Index, Burkina Faso 147 B.2 Price Changes in Important Products Purchased by Farmers, 1979-1989, and Percent Change 147 B.3 Price Ranges of the Major Crops at the Study Sites, 1988-1989 148 B.4 Retrospective Prices for Major Products, the Linoghin Market, 1988-1989 148 B.5 Retrospective Study of Prices for Livestock, 1979-1989 149 B.6 Cereals Food Balance 149 B.7 Official Producer Prices and Prices Used To Calculate the Value of Production in the 1979 AVV Statistical Service Survey 150 C.1 Number of Catle and Sheep/Goats Lost to Disease and Theft, Five-Year Period 152 C.2 Rates of Losses of Cattle and Sheep/Goats by Death and Theft, on a Yearly Basis 152 C.3 Expected Annual Livestock Income for the Heads of Household, by Region 153 D. 1 Division of Cereal Production between Privately and Cooperatively Worked Fields, Solenzo 1988/1989 156 D.2 Mean Input Costs for the Male Heads of Household with Different Levels of Farm Technology, Solenzo 1988/1989 156 D.3 Average Livestock Holdings for Male Heads of Household at Different Levels of Technology, and Expected Annual Livestock Income per Adult Labor Equivalent (ALE), Solenzo 1988/1989 157 D.4 Distribution of Cattle Ownership, Solenzo 1988/1989 157 D.5 Characteristics of Settler Households with Different Levels of Investment in Livestock, the AVV Planned Settlements 1988/1989 158 D.6 Mean Number of Cattle, Goats, and Sheep, and Mean Anticipated Income, the AVV Planned Settlements 1988/1989 159 D.7 Livestock Ownership for the Study Households in Mogtedo, Mogtedo-Bombore, and Linoghin, the AVV Planned Settlements 1988-1989 159 D.8 Subdivision of Cereal Production between Privately and Cooperatively Worked Fields, Kompienga 1988/1989 159 D.9 Distribution of Livestock for Male Heads of Household at Kompienga, 1988/1989 160

xvi Figures 1.1 The Onchocerciasis Control Programme 2 2.1 Gains in Cultivated Area by Sponsored and Spontaneous Settlers. 1974-1983 9 3.1 Types of Govemment Intervention and Immigration to Burkina's OCP River Basins 15 3.2 Sponsored and Spontaneous Immigration in Connection with the AVV Planned Settlements at Linoghin, Mogtedo, Mogtedo-Bombore, and Rapadama 16 3.3 Year of Immigration for Male Heads of Immigrant Households, Solenzo Subsector, 1989 18 3.4 Year of Immigration for Settlers in the Kompienga Basin and Town of Kompienga, 1989 19 3.5 Mossi and Fulani (Pastoralist) Immigration to the Kompienga River Basin 20 3.6 Year of Immigration to the Niangoloko Subsector for Male Heads of Immigrant Households, 1989 22 4.1 Total Cotton Production in the AVV Planned Settlements 31 4.2 Total Fertilizer Purchases in the AVV Planned Settlements 32 6.1 Subdivision of Net Crop Value of Income per Household, the AVV Sponsored Settlements, 1979-1989 47 6.2 Total Number of Cattle Listed in the AVV Extension Records 50 6.3 Rate of Credit Reimbursement, AVV Planned Settlements 57 7.1 Average Household Field Area Farmed Collectively, Net Crop Income per Adult Labor Equivalent (ALE) for the Major Technology Groups at Solenzo, and Comparison of Net Crop Income for the AVV and Solenzo 67 7.2 Location of the Study Villages in the Solenzo Subsector: Daboura, Kie, and Dar-es-Salaam 69 8.1 Location and Relative Importance of Existing and Abandoned Markets Frequented by Settlers Living in the AVV Sponsored Settlements at Mogtedo, Mogtedo-Bombore, and Linoghin, 1988/89 75 8.2 Location, Relative Importance, and Date Created and Relocated of Markets in the Solenzo Subsector, 1989 77 8.3 Location and Type of Markets in the Southem Kompienga Basin, 1987/88 78 9.1 Integration of Hosts, Settlers, and Pastoralists 88 10.1 Projected Installation of AVV Planned Settlements and Other Types of Land-use Planning in the Nakainbe and Nazinon (ex-White and ex-Red Volta) River Basins, 1974 97 10.2 Average Land Area Per Unit Labor for Different Technology Groups at the AVV and Solenzo, 1988 98 10.3 Year of Immigration for Settlers Living in the Villages Surrounding the Classified Forest at Toumousseni, 1989 101 10.4 Interrelated Research and Interventions: the PNGTV Approach 104 11.1 Organigram of the AVV Administrative Structure, 1974-1982 108 11.2 Organigram of the AVV Administrative Structure, 1982-1989 111 11.3 The AVV Planning Unit I (AVV-UP 1) 113 11.4 Relationship of ONAT (ex-AVV) to the National PNGT/GTV 117 12.1 Returns from Farmer-Managed Cotton Markets Returning to AVV Settlement Groupements (CFA) 124

Maps IBRD 23841 Population Density, 1975 IBRD 23842 Major River Basins and Areas Covered by AVV Programs IBRD 23843 Agricultural and Transhumant Migration Patterns IBRD 23844 Classifications of Land According to Agricultural Potential and Areas Designated as National Forests and Parks IBRD 23845 Agricultural Migration from the Mossi Plateau, 1975

xvii IBRD 23846 Case Study Research Sites IBRD 23852 Classification of Provinces in Terms of Their Forestry Situation IBRD 23854 Location of AVV Planned Settlements Created 1974-1984

xviii Abstract

The Onchocerciasis (River Blindness) Control major roads and/or dynamic markets and ad- Programme (OCP) was initiated in 1974 to con- ministrative centers. trol river blindness in a seven country area in Strong similarities in the settlers' social and West Africa (Benin, Burkina Faso, Cote d'Ivoire, economic adjustment were observed at all study Ghana, Mali, Niger, and Togo). Later, the pro- sites. These responses, and the associated con- gram was expanded to include Guinea, Guinea- flicts over the allocation of land, grazing, water Bissau,Senegal, and Sierra Leone. Today, 18 years and forest resources, followed a predictable pat- after the inception of control, the OCP is widely tern. This report examines a wide range of factors regarded as one of the most effective, regional in the micro and macro policy environment that health programs ever launched. River blindness affect settler responses to different problems at is no longer cited as a public health threat in the each settlement stage and their effect on sound original control zone and many of the river basins natural resource management. are undergoing active resettlement. Especially important in this study were the na- Eighty percent of the territory of Burkina Faso, tional policies relating to crop subsidies, market- or 235,000 km2, falls within the area covered by ing boards, transportation infrastructure and the OCP.Of this total area, 41,000km 2 was classi- agronomic research. The study shows how na- fied as uninhabited and uncultivated in 1975.In tional policies which fledas anninhbitd unultiate in 975 In basin developvmentexacerbated targeted investment pre-existing, in river in- an effort to control the projected immigration to teionaloinetiesandbled trapid, u n- thes these new lands andandthenewland the underpopulated nderopulted riveriver trolledter-regional spontaneous inequities immigrationand led to rapid,from theuncon- less basins, the Burkina government created a special well endowed plateau areas. Heavy investment national agency, the Volta Valley Authority or in the development of an appropriate cotton AVV,and gave the agency complete control of package for the high potential south increased 30,000 km of underpopulated river basins. In the attractiveness of settlement, while there is no theory this mandate was to supersede all pre-ex- equivalent technical package for rainfed agricul- isting claims to land. ture in the less favored center and north. The pace of spontaneous settlement quickly It is clear from the research here that govern- outgrew the ability of the AVVto finance and cre- ment efforts to promote sustainable land use ate a sufficient number of sponsored settlements. practices by official mandates are unlikely to suc- In addition, there were sizeable intra and interre- ceed. More likely to be successful are interven- gional differences in the rate of new lands settle- tions similar to the new ProgrammeNational de ment. Within any subregion, the villages that Gestiondes Terroirs(PNGT), a program for village attracted the most immigration were those near land management which reinforces the capacity

xix of local inhabitants to regulate access to land and inhibit the development of positive relationships. other resources. The PNGT model combines zoning (to delineate Farmers' perceptions of their long-term oppor- land most appropriate for rainfed crop cultivation tunities for developing diversified non-crop from land to be used for livestock, or managed sources of employment were key factors in deter- and natural forests) with a strong base of local vil- mining settlers' long-term investment strategies lage land management committees. and their willingness to invest in more labor in- The volume and complexity of the tasks intensive crop production practices. These research volved in this type of integrated development results have implications for the future design of planning have led many governments through- river basin planning which heretofore has fo- out the world to advocate autonomous manage- cused almost exclusively on intensifying crop ment structures like Burkina's AVV.The history production. of the AW highlights the fact that such agencies New land settlement and the growth of region- may allow the government to organize a complex al markets and administrative centers are clearly assistance package much more quickly and to at- intertwined. Early attempts by the AW to rein- tract high levels of donor funding. However, au- force the development of market centers were of- tonomous agencies tend to suffer from high costs ten misplaced. For this reason it is recommended and administrative inflexibility. that aid for infrastructure development to rein- The Burkina case study provides ample evi- force new market and administrative centers be dence for success in areas where new land settle- phased over a longer (10-15year) period. ment has been assisted and the high environ- The complex interweaving or synergism be- mental, social and economic costs of unassisted, tween settlers and the hosts who provided them spontaneous settlement at higher population den- access to land and between livestock owners and sities. This is a rich history that deserves to be herders needs to be well understood. Government thoroughly analyzed before embarking on a new policies which favor one group over the other can generation of follow-up planning.

xx Introduction

The Onchocerciasis (river blindness) Control Pro- portion of the original settlers-or their offspring- gramme (OCP) is one of the most successful may relocate at this point in the settlement cycle. health programs ever launched in Sub-Saharan To develop more intensive crop and livestock Africa (Figure 1.1). Today, eighteen years after production practices, settlers require assistance control began in 1974, the disease is no longer ci- (Scudder 1981,1984,1985,1991;Scudder and Col- ted as a public health threat in the original control son 1981;McMillan, Painter and Scudder 1990). zone and many of the river basins are undergoing This azistance includes public and private in- active resettlement (OCP 1985,1986; Remme and vestment in roads and bridges, along with the Zongo 1989;Younger and Zongo 1989). provision of crop and livestock research and ex- In the wake of this success, however, there is in- tension services. One of the key questions that creasing concern about the environmental sustai- face policy planners today is: what form should nability of crop, livestock and forestry practices this assistance assume? in the newly settled areas. A ten year assessment This report describes the experience of Burkina of settlement trends revealed that most immigra- Faso with new lands settlement and follow-up tion is spontaneous or unassisted, with little or no planning in its underpopulated river basins affec- settler access to basic infrastructure, and econo- ted by onchocerciasis control. The study is of spe- mic or social services (Hervouet et al. 1984; OCP cial interest because Burkina Faso is the only one 1986). By far the major criticism of spontaneous of the 11 countries covered by control where the settlement is that settlers tend to farm the largest of govemmentries shown a sustained te area possible, with little attention to sound mana- ational government has shown a sustained inte- gement of soil, forest or water resources.1 rest in area specific planning for the underpopu- If immigration rates are low, extensive cultiva- lated valleys. In addition, new lands settlement tion and livestockpractices are ecologicallyand so- has been most extensive in Burkina Faso. This cially sustainable; population densities are low and planning includes a number of experimental vil- reserve lands remain available. Invariably, howe- lage land management, agropastoral and forestry ver, problems arise as population densities increase projects. The analysis of this experience provides and it becomes difficult to allow for periods of fal- valuable insights for future planning in the entire low sufficient to restore soil fertility.Higher densi- 11country control zone. ties are associated as well with increased This research was conducted as part of the se- competition for potable water, grazing areas and cond socioeconomic study coordinated by the fuelwood. If new land that can be cleared and farm OCP's Committee of Sponsoring Agencies (CSA). land remains easy to acquire, and if the economic The goal of the socioeconomic studies is to help and social costs of abandoning old land are low, governments develop appropriate planning mo- then, as population densities increase, a sizeable dels for the OCP river valleys.

1 /5 1_ MAURITANIA b

*t '\*% / \r -j(---- _ i SENEGAL )J THE GAMBIA _51 ~~z<- -, B- A i~~~ 7 : fi:::: 77: ~i 7 ~~~~~~~~~M~ AL ~ I ~ ~ T N I *-E! *'0A"NAWALI

GUINEA- w - _,. , BURKINAFASO / bp~BISSAUx ~ - SOLENZO oauAWPAj

L, o *TIENFAtA! O- NDRAOAAMA SELINGUE 0DIOILA KOENG tJ G UI N.E A YANFOLILAFJK0Q

0n ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~TNIANGOLOKO S RON AREARR J4GAf RED

O cOe1anTeLEME?\TOGOj) Ocean LEN GHANAT.OfW

A~~~~~~~~~~~~~~~~~~t c t iI a n __VPR D'JIVOTI R E z

K' d ' . JJ>'

SI ALIBERIA MO PA

< (> a~~~~~~~~~~~~~~~Abidjan y ~ THE SAH-ELIANAFRICA . e 5HEONCHOCERCIASIS CONTROL '- < 1~ xo0 3 PROGRAMMEAND CASESTUDY. SITESIN BURKINAFASO, GHANA, MALI, AND TOGO a3 ORIGINALPROGRAM AREA NATIONALCAPITALS _ ZONEINITIALE CAPITALES PROGRAMMEDE LUTTECON TRE i E EXTENSIONAREAS - ITRONAdTESIILilONALF ONDRESm L 'ONCHOCERCASEET LOCALISAT/ON 7E--7:X:ZONESDEXTENSION thtsi b DESSITES D'ETUDE AU BURKINA, 0 SCATSsEStTrUDY SITES .h d E bi w wGHANA, MALI, ET TOGO SITESD'ETOE 6 ri15° I V 0 R Uno The first study, which is referred to as the Pre- In assisted settlement, one or more governmental paratory Phase Study,2 provides an inventory of or nongovermnental organizations (NGOs) pro- existing physical and socioeconomic data for the vide basic services and infrastructure for sponta- OCP areas in the eleven countries currently cove- neous settlers who have moved or will move to a red by control. The final report includes develop- site on their own. Areas where migration is occur- ment proposals for OCP zones in each of the ring (or has occurred) with little infrastructure seven initial OCP countries and makes recom- and few economic and social services are classi- mendations for further in-depth studies as part of fied as "unassisted" or spontaneous. the development planning process (Hunting The first case study includes three groups of Technical Services 1988a,b,c,d). planned settlements (Linoghin, Mogtedo, and The second study, which is referred to as the Mogtedo-Bombore) created by the Autorite des Land Settlement Review, was designed to provide Amenagements des Valleesdes Volta (Volta Valley the national governments with recommendations Authority or AVV)in the Nakambe (ex-WhiteVol- and guidelines on how best to facilitate develop- ta) river basins east of Ouagadougou (Figure 1.1). ment in areas of new lands settlement based on an A less intensive study to examine the early results analysis of settlement experiences to date and, of an experimental program to incorporate sponta- where relevant, selected experiences elsewhere. neous settlers into village land management As part of the Land settlement Review,in-depth groups and into coverage by the extension services case studies of land settlement were conducted at based in the neighboring planned settlement at sixteen sites in Burkina Faso, Ghana, Mali, and Rapadama was also included. In addition, we also Togo (Figure 1.1; Table 1.1). A less intensive re- conducted a small number of interviews among view of settlement, based on interviews and ana- pastoralists recently settled in the new AVVagro- lysis of existing data, was conducted in C6te pastoral zone at Gadeghin, immediately adjacent d'lvoire and Guinea. A third level of the Land to the Rapadama planned settlements. Settlement Review, which entailed critical analy- Solenzo, the second study site, is located in the sis of existing documentation, was conducted for southwest Mouhoun River Basin (ex-Black Vol- the remaining OCP countries: Senegal, Guinea- ta). Like the AVV-UPI, Solenzo is located in an Bissau, Sierra Leone, Niger, and Benin. The major area that was highly endemic for onchocerciasis conclusions and recommendations of the compa- before 1974. Solenzo has higher rainfall and bet- rative Land Settlement Review are described in ter soil fertility, however. In addition, the Mou- the Land Settlement Review Final Report which houn Basin and Solenzo region were already appears as a separate volume in this series. experiencing an increase in agricultural in-migra- The Burkina case study is based on a mixture of tion in the late 1960sdue to a successful regional quantitative and qualitative research at four stu- program to promote intensive cotton production. dy sites and a series of complementary special The third and fourth case study sites-Kom- studies on market systems, gold mining, the pienga and Niangoloko-are located in the extre- agropastoralist zone at Gadeghin, settler turno- me southwest and southeast of Burkina. Both ver in the government sponsored settlements, areas were highly endemic for onchocerciasis and local and national institutions (see Annex I prior to 1974. Despite the highest rainfall in the for a detailed discussion of the research methods country, both areas have experienced far less new used at each site). The four case study sites inclu- lands settlement since 1974 than the more nor- de representative examples of the different settle- them river basins. Immigration to the Kompien- ment "types" found in the wider control area. ga basin has increased since 1983, due to the These settlement types - sponsored, assisted creation of the country's first hydroelectric dam. spontaneous, and unassisted spontaneous-re- The present report is divided into 12 chapters flect different levels of government and other and 5 annexes. Chapter 2 provides a brief over- agency intervention in the settlement process. view of the river basins including their natural In sponsored settlement, a government agency resource potential and settlement patterns- is involved in almost every phase of the develop- past, present, and projected. Their important role ment program including surveying and prepa- as livestock pasture and passageways are also ring the land, constructing basic infrastructure, described. creating new community organizations and desi- Descriptive information on the sites is presented gning and administering agricultural extension. in Chapter 3, where the point is made that we

3 Table 1.1 The Land Settlement Review Case Studies Country; nameof site; Settlementtype; natureof sample; approx.date of settlementonset Locationin country BURKINAFASO 1. Solenzo (3v:36hh+19v) Spontaneous (1960s) Kossi province 2. Niangoloko (22v) Spontaneous (1982) Comoe province 3. Kompienga (1v:35hh) Dam-related planned Gourma province resettlement and spontaneous (1983) 4. Volta Valley Authority (AVV-UP1) a. Survey Restudy Linoghin (6v:20hh) Planned (1973) Oubritenga province Mogtedo-Bombore Planned (1979) Ganzourgou province (7v:20hh) Mogtedo (6v:40hh) Planned (1974) Ganzourgou province b. Case Study Restudy Planned (1975) Ganzourgou province Mogtedo V3 (lv:20hh) c. Rapadama (+7) Assisted (1987) Ganzourgou province

GHANA 1. Red Volta Valley Cyclical,spont. Upper East Region and Plateau (2v:30hh) (late 19th cent.) . 2. Fumbisi-Yagoba-SooMankarigu Assisted (1985) Upper West, Upper East, and ("Overseas") Northern Region (4v:30hh) 3. Damongo Settlements Planned (1950s) Northern Region (3v:30hh) 4. Tono IrrigationScheme (2v:30hh) Planned (1980s) Upper East Region

MALI 1. Yanfolila (5v:30hh) Spontaneous;(1970s) Third Region Assisted (1985) (Sikasso) 2. Selingue (3v:29hh) Dam-related Third Region planned and spontaneous (Sikasso) (late 1970s) 3. Diofila(4v:30hh) Spontaneous (1960s) Second Region (Koulikoro) 4. Finkolo(3v:30hh) Wage workers in workers' villages Third Region (Sikasso) at tea plantation (late 1960s) 5. Tienfala (3v:9hh) Spontaneous, by railway workers Second Region (from early 1900s;continuing (Koulikoro) 6. Manantali (14v:70hh) Dam-related planned First Region (Kayes) (1986/87)

TOGO 1. FED-Agbassa(3v:30hh) Planned (1972) Kara Region 2. Mo Plain (3v:30hh+6v) Spontaneous (1970s) Central Region Note: Numbers in parentheses that follow site names denote the number of villages at each site where household interviews were conducted, followed by the number of households in the site sample. Numbers preceded by a plus sign denote the number of additional villages at the site where leaders and other community members were interviewed. Information on Manantali was drawn from earlier research dealing with that project (Horowitz, Koenig, Grimm and Konate forthcoming)

4 observed strong similarities in terms of settler res- * settlers' social and economic integration ponse patterns at all four sites. Moreover, these si- with indigenous inhabitants and area pastoralists milarities seemed to be equally true in areas of (Chapter 9); planned and assisted spontaneous settlement. * the long-term environmental sustainability These responses were predictable and reflect the of the settlers' local production systems (Chapter manner and results of settlers' adjustments in their 10); and new social, ecological, and economic settings. * the effectiveness of local and national man- The chapters that follow (4-12) look at a wide agement structures (Chapters 11 and 12). range of factors in the micro and macro policy en- Annex 1 provides a brief overview of the re- vironment that affected settler responses to diffe- search methodology in Burkina. Annex 2 gives an rent problems at each settlement stage. These overview of Burkina's political history. Annexes factors include: 3-5 group supplementary tables. * national price and subsidy policies that determine the prices settlers and hosts receive for Notes their crops and livestock (Chapter 4); * settlers' and hosts' perceptions of their local I See McMillan, Painter and Scudder 1990;OCP 1986;Hunt- land tenure(Chapter rights 5). TechnicalServices 1988a, 1988b, 1988c, 1988d; Van Raay land tenure rights (Chapter 5); ~ingand Hilhorst 1981; Weitz, Pe'ley and Applebaum 1978;An- * local perceptions of long-term opportuni- gel 1985;Remy 1973;Becker 1985;Bharin 1981;Couty, Mar- ties for developing more diversified, non-crop chal, Pelissier, Poussi, Savonnet and Schwartz 1979; Dollfus 1981; Raison 1979, 1981, 1985; Nicolai and Laserre 1981; sources of employment (Chapter 6); Remy 1968,1973,1975,1981. * the profitability and risk associated with Rm 98 93 95 91 for c rcialsoodand ex-h 2. The Preparatory Phase Study was conducted by Hunting new opportunities for commercal food and ex- Technical Services, Ltd. (U.K.), in collaboration with Organi- port crop production (Chapter 7); sationet Environnement(France). - settlers' proximity to and/or degree of social and economic integration with area markets (Chapter 8);

5 ~~2

Overview of the OCP River Basins

Location and Natural Resource Potential characterized by a single rainy season and a dry season that can last from six to eight months. Eighty percent of the territory of Burkina Faso, The river basins are generally considered to or 235,000 kM2 , falls within the area of the On- have greater potential for rainfed agriculture chocerciasis Control Programme (Figure 1.1) than the adjacent plateau areas. Moreover, the (Hervouet et al. 1984:1). Of this total area, river basins include almost all of the land 41,000 km2 (17 percent) of river basin land was deemed to have potential for large scale irrigat- considered uninhabited and uncultivated be- ed farming. The water resources and vegetation fore the OCP started in 1975 (ibid.) (Table 2.1; of the valleys have long been important to West Map 23841).1 Africa's livestock industry-both for grazing The river basins straddle two broad climatic and for transportation from the landlocked cat- zones: a northern sudanian region with an average tle producing areas of Niger, Mali, and Burkina annual rainfall of 600-800 nmm,and a southern suda- to the coast (Map 23843). nian region with an average annual rainfall of 800- These basins have also been a vital source of fu- 1,200 nmm(Map 23842). Both agrnclimatic zones are elwood and construction materials for Burkina's

Table 2.1 Burkina's Major River Basins and Area Estimated to be Directly Affected by the Onchocerciasis Control Programme River Basin Area (km2) Valleys of the Nakambe (ex-White Volta)and Nouhao 8,250 Valleys of the Nazinon (ex-Red Volta) 3,900 Middle Valley of the Mouhoun (ex-BlackVolta)/and Valley of the Bougouriba 6,100 Poni 2,150 Sourou 1,000 Valley of Kompienga 4,000 Valley of the Sissili 3,000 Valleys of the inferior and superior effluents of the Mouhoun 19,000 Total 47,400 Source:AVV 1985d:6.

6 growing cities. As for other resources, only a of these epidemics and wars by destroying the fraction of the basins' potential for fishing and hy- scattered, low-density settlements that were left droelectric power has been tapped. One large hy- in their wake (Hervouet 1978). droelectric dam with plans for limited irrigation In addition, earlier studies may have overesti- has been constructed at Kompienga. Construction mated the total area of valley land that was avail- of another, with greater potential for irrigation, is able in the sense that it was not settled and was of beginning at Bagre. A third large dam is planned sufficiently high quality soil (i.e., not covered by for the future at Noumbel along the Ghana fron- rock debris, iron pans, and gravel at the soil sur- tier. The recent discovery of gold at several small face, or of a very shallow profile) (Berg et al. sites in the Nakambe and a large commercial gold 1978:14).Many of the valleys are subject to fre- site at Fara-Poura in the province of Sissili have quent flooding and/or suffer from excessive created interest in the area's potential for mining. moisture (ibid.:15). The heavy vertisols that are In the 1950s,the basins' sparse population den- characteristic of much of the area are also difficult sities made it possible to designate vast areas as to work with only a hand-held hoe (ibid.). classified forests and wildlife reserves. Today, the The presence of other historic constraints to set- highest concentration of Burkina's protected tling lands in the region, however, does not di- wildlife and forest areas are found in the river minish the importance of onchocerciasis in valleys (Map 23844). deterring settlement. It was assumed, therefore, that the rate of spontaneous settlement to the val- Historic Settlement Patterns leys would increase once spraying to curb the disease began. In the past, the poverty of Burkina's central plateau region has contributed to high rates of out- Early Development Planning for the migration.2 The best known of these emigration OCP River Basins patterns in the mid-1970s was that of young men leaving to work as labor migrants in the more de- In an effort to control the projected immigration veloped coastal countries like C6te d'lvoire.3 A into the underpopulated river basins, the Burkina second population movement, which took on government created a special national agency, the momentum after the 1968-1973drought, was that Volta Valley Authority or AVV (Autorite des of Mossi settlers moving off the central plateau to Amenagementsdes Vall6esdes Volta).By presidential the less-populated fertile non-Mossi regions of decree, the AVV was given complete control of al- the west (Map 23845).4 most 30,000km 2 of the least populated river basins Despite low population densities and higher where new lands settlement was expected to occur rainfall in the deserted river valleys of the (Map 23842).In theory this mandate was to over- Nakambe and Nazinon, and substantial ride all preexisting claims to the land. population pressure in the neighboring pla- The early AVV program focused on the design teau, almost none of this earlier emigration was and implementation of groups of model settle- directed toward those valleys prior to 1975. Set- ments in the northern river basins with good ag- tlers did occupy lands in the vicinity of the ricultural potential and a few specialized Mouhoun, but not in the more infected areas industrial or irrigation projects. The first settle- immediately adjacent to the river. ments were created in the basins of the Nakambe While onchocerciasis is considered to be an im- and Nazinon (Map 23842). portant factor in the lack of settlement in Burki- na's river basins, it is by no means the only factor. Settlementand Immigration Patterns:1974-1990 Certain regions are characterized by a Precam- brian basement complex, known for its poor aqui- Spontaneousvs. SponsoredAgriculturalist fer qualities that makes drinking water a problem Settlement (Berg et al. 1978:14). A series of epidemics that ravaged the White and Red Volta Valleys in the The national government and donors were cor- 1940s(preceded by nineteenth century slave raid- rect in predicting that the successful control of on- ing and warfare) contributed to the depopulation chocerciasis would increase existing patterns of of certain areas. Contemporary research suggests rural-rural migration toward the south. Where that onchocerciasis simply completed the ravages they were wrong was in their expectation that

7 spontaneous settlement to the area could be di- The percentage increase in new land clearance rected and controlled via a centralized program between 1974 and 1983 was lower in the for sponsored settlement. Bougouriba and Comoe-Leraba Basins (Paris By 1979, it was already clear that the pace of 1980,1983;Clanet 1983).These areasare more dis- spontaneous settlement had quickly outgrown the tant from population centers in the central plateau ability of the AVV to finance and create planned where there are fewer opportunities for commer- settlements in sufficient number. By 1983, studies cial cotton production was less widespread. showed that more than 80 percent of the increase Settlement patterns within the same basin and in cultivated land in Burkina's river basins could under identical ecological conditions varied ac- be attributed to spontaneous settlers (Figure 2.1). cording to ethnic groups. For example, the Bissa This percentage would have been even higher had and the Nakana showed large land gains, as op- the research included the Mouhoun and Comoe posed to the Dagara, Birifor,Wiile, and Lobi (Her- basins, where the AVV was not active. vouet et al. 1984).5 Different rates of settlement within the same SpontaneousAgriculturalist Settlement river basin were also strongly connected with the presence of targeted government investment Although the efficacy of larviciding was the same such as the AVV planned settlements, regional for all the river basins in Burkina, there have been programs to promote commercial cotton produc- sizable intra- and interregional differences in the tion, paved highways, and special projects like rate of new lands settlement since eradication the Kompienga dam and the AVV industrial started in 1974. For example, settlement rates wood project at Wayen (Hervouet et al. 1984).In north and south of latitude 110 differed signifi- general, the villages with the highest rates of cantly (Table2.2), while within the northern river immigration were those nearest roads and/or dy- basins, colonization proceeded much more rapid- namic markets and administrative centers. ly along the Mouhoun and its tributaries (a six- fold increase) than in the Nakambe and the PastoralistImmigration Nazinon in the central south (a threefold increase), in large part because of the potential for A large portion of immigration to the valleys in- commercial cotton production in the Mouhoun volves pastoralists and their cattle herds. Even basin (Table 2.2; Hervouet et al. 1984:4). before onchocerciasis control, the river basins were an important part of the transhumant migration patterns of FulBe from the central plateau. During the rainy season, herders would Table 2.2 New LandCultivated in Major River take their animals to the valleys to move the ani- Basins, 1975-1983 mals away from agriculturalists' fields. As agri- Landunder culturalist settlement increased, a growing cultivation number of FulBe were attracted to the valleys to Basin 1975(%) 1983(%) work as herders for farmers who have livestock. Other FulBe settled in mass around the manmade North of 110 and natural water resources. A fourth very recent Nakambe trend involves the forced repatriation of allegedly (ex-WhiteVolta) 9.2 29 Burkinabe pastoralists from Ghana and, to a Nazinon lesser extent, from C6te d'Ivoire. Many of these (ex-RedVolta) 8.1 28 repartriated FulBe were attracted to the valleys. Mouhoun 3.0 15.7 (ex-BlackVolta) Other Physical, Demographic and Mouhountributaries 4.1 24.7 Social Constraints Southof 11° Bougouriba 9.9 13.1 This analysis of migration trends highlights the Comoe-Leraba 5.2 15.9 fact that migrants consider far more than the availability of new land in their selection of settlement sites. The OCP areas are frequently isolated, lack supplies of potable water, and are affected by

8 Figure 2.1 Gains in cultivated area by sponsored and spontaneous settlers, 1974-1983

Hectares (thousands) 20

15 __

Nakambe.Nazinon Bougouriba I Sponsored = Spontaneous

Source:Hervouet, Clanet,Paris and Some 1984:16/see AnnexMf.

9 diseases other than onchocerciasis that threaten cal formations make it necessary to drill 40-50 human and animal life. High rates of immigra- meters for drinking water. Prior to the creation of tion may also create social conflicts over land ten- the AVV the sponsored settlements, there were ure rights (see Chapter 4). All these factors affect few permanent water points. As a result, sponta- the settlers' standard of living as well as the via- neous settlement remains largely confined to ar- bility of diversified production systems based on eas adjacent to sponsored settlements with crop, livestock, and off-farm employment. In ad- borehole wells. Even with sophisticated machin- dition, the close proximity of many of Burkina's ery, however, the AVV was not always successful river basins to national borders makes them high- in drilling new wells. The woods around the AVV ly vulnerable to economic and social problems in sponsored settlements are littered with the bases adjacent countries. of wells that either did not work or have dried up. More recently, many of the profitable wells that Isolation were drilled when the planned settlements were first created in 1974-1977are drying up; the origi- With the exception of a few areas near major cities nal pumps are wearing o-it and need to be re- and towns, the river basins tend to be isolated. placed-pumps that can cost up to 500,000CFA. This isolation takes several forms. Because of At the same time, demand for water has increased their historic lack of population, Burkina's river because large numbers of spontaneous agricul- basins tend to be distant from prefecture and sub- turalists and pastoralists have settled around the prefecture administrative centers and their asso- edges of the sponsored settlements, and settlers ciated health care, education, and extension have increased the size of their livestock herds. services. Many areas had no major regional mar- The AVVsettlers have managed, in some cases, to kets prior to 1974,and thus were poorly integrat- pool money for repairs and replacement pumps. ed into national marketing systems. The cost of repairs and replacement is steep, how- Because of the same historic lack of settlement, ever, and generally must be covered by small vil- the basins typically have few highways and ac- lages of 25-50 families, since attempts to get cess roads. For example, unpaved roads linking nonsponsored settlers to contribute to the upkeep major production centers in the Mouhoun basin, of the wells have met with only limited success. like Solenzo, with urban population centers at Bobo-Dioulasso and Ouagadougou exist, but are Other Health Problems difficult to travel. Public transportation between Solenzo and Ouagadougou could take two days A wide variety of human and animal diseases before biweekly bus service was established. besides onchocerciasis are associated with OCP Even now bulk transport is expensive and affects river basins. This includes a strong overlap be- producer prices. As recently as 1983,the 100 kilo- tween the areas affected by onchocerciasis and meter trip from the eastern provincial capital of bovine trypanosomiasis (animal sleeping sick- Fada N'Gourma to the present site of the Kompi- ness). One consequence of the river Gasons' low enga dam in Burkina's southeastern Kompienga population densities is that health hazards are basin could take an entire day. not always known. For example, a mysterious Bridges are another problem. During the rainy virus contaminated the water supply in one of season, when waters rise, rivers may isolate entire the early AVV planned settlements at Mogtedo areas. The road linking Fada with Kompienga is in 1978. The original village site was abandoned, still impassable to normal vehicles fo several days but not before the entire village population and after a major rain. If bridges wash out, as they did the extension staff found hands and arms cov- at the Mogtedo planned settlements in 1977, the ered with painful skin growths. travel distance to major markets may be tripled. New settlers in the river basins are not familiar with local medicinal plants and healers. They are Lackof PotableWater also vulnerable to new risks. The hard work associated with clearing a new farm can weaken them, Lack of potable water appears to be one of the making them susceptible to various illnessesor in- principal reasons for sparse preexisting settle- jury. Snakesand wild animals are another problem. ment in the area occupied by the AVV planned Two farmers in the economic survey were bitten settlements to the east of Ouagadougou. Geologi- during the 1989-1990agricultural season. Although

10 both survived, one was hospitalized for a week, Projected Immigration Trends: 1990-2000 then spent several months in painful recovery at home. It is symptomatic of the priority placed on in- In general the case study research shows a slacken- creased health risks, that one of the first indepen- ing off of migration to the less isolated northern dent activities undertaken by new settlers in the river basins near Solenzo and the AVVsponsored AVVwas to organize a group meeting to learn who settlements that have been actively resettled since knew different cures for snake and scorpion bites the early 1970s.Because of mounting population and other illnesses (McMillan 1983). densities, new immigrants have greater difficulty in obtaining good agricultural land. As a result, Proximity to InternationalBorders the factors that tend to motivate later immigrants to an area are often linked to the growing opportu- Many of Burkina's OCP areas are in politically nities for non-farm employment in the region or sensitive border areas. Some, like Kompienga, the desire to rejoin family members who immi- border more than one country (Ghana, Benin,and grated earlier. Togo). The Leraba and Comoe basins are in the One of the first indications of settlement sat- frontier zone near Cote d'lvoire; the Po and uration is a reorientation of pastoralist immi- Nouhao basins are close to Ghana. gration. Recent research by the National On the positive side, when international bor- Village Land Management Program (PNGTV ders coincide with paved or all-weather high- or Programme National pour la Gestion des Ter- ways, they generally give rise to prosperous trade roirs Villageois, now known as PNGT or Pro- centers at the border crossings, as is the case at Po gramme National pour la Gestion des Terroirs) and Niangoloko. On the negative side, proximity supports our own findings that pastoralist im- to an international border introduces a new level migration is being redirected from the north- of uncertainty into national, regional, and local ern basins of the Mouhoun, Nakambe, and development plans. Livestock disease control Nazinon, to southern basins like the Leraba, programs in Kompienga, for example, have been Comoe, and Kompienga-basins that did not plagued with a fluctuating disease control envi- show a significant increase in forest clearance ronment because of the uneven animal disease or new lands settlement during the first de- control standards of their neighbors in Ghana, cade of control. Benin, and Togo. Although aggregate immigration rates have Border areas also receive the brunt of econom- been low in the southern river basins, we can ic or social changes in the adjacent countries. For predict that the rates of new lands settlement in example, the expulsion of pastoralists from these areas will increase substantially over the Ghana flooded the town of Kompienga and Ko- next decade. Our best macro-level evidence for mpienga basin. This sudden influx of pastoral- this trend is the 1985 census which shows that ists and their animals put additional strain on the the southernmost provinces are-with the ex- area's indigenous land tenure systems and re- ception of the province surrounding Ouaga- sources, which were already under pressure be- dougou-the provinces that are experiencing cause of immigration associated with the the highest in-migration. Our best micro-level Kompienga dam. evidence are our two regional surveys and the The recent downturn in the C6te d'Ivoire research by Agrotechnik (1989) which show a economy has had a similar broad impact on the sharp increase in immigration to the regions Niangoloko region. Growing numbers of mi- surrounding the Kompienga and Niangoloko grants are returning to Burkina and settling in research sites. Especially important for Burkina the Comoe and Leraba basins rather than re- will be the reduced opportunities for employ- turning to the plateau. This economic malaise ment in Cote d'Ivoire. also contributed to armed conflicts between About one million Burkinabe were living in pastoralists and agriculturalists, and forced Cote d'lvoire in 1985. With the decline in cocoa thousands of pastoralists, alleged to be Burkin- prices, short-term labor migration is no longer abe, to flee across the border, some without the profitable option that it used to be. Thus we their animals. can expect large numbers of Burkinabe to return to Burkina. It would be naive to think that many of these immigrants-especially those

11 who remained for long periods of time in Cote Need to DevelopBetter Methods of Predicting d'Ivoire-would return to the impoverished ImmigrationFlows Mossi Plateau. We can also predict that the reduced opportunities abroad will encourage National censuses tend to be conducted only ev- young farmers and pastoralists to develop ery decade. Yet dramatic changes in the rate of commercial opportunities in the OCP river ba- new lands settlement and land use patterns can sins, as well as to augment the labor resources be set in motion in less than five years. There is of their extended families. While new wage la- a strong need, therefore, for planning agencies bor opportunities may be created in Ghana, it is such as the PNGT to develop more cost effective unlikely that these opportunities will result in tools for assessing settlement trends. any sort of dramatic reorientation of foreign Much can be learned from interviewing farmes. wage labor flows, at least in the short run. We found their perceptions almost always to be correct-and generally far in advance of our own. Potential Policy Implications The settlers at Kompienga, for example, predicted that contrary to government thinking a high per- To summarize, while lack of opportunity for centage of the immigrants to the town would re- development in Burkina's densely populated main after the inauguration of the dam. This is Mossi Plateau may be the basic force driving what has happened. The settlers we interviewed immigration to the valleys, many factors affect also said that many of the agriculturalists would the choice of a settlement area. These factors in- immigrate to the Bagre site during the 1989-1990 clude placement of infrastructure, possibilities dry season but return to Kompienga to farm in for livestock, type and level of extension servic- 1990, as they expected the indigenous Bissa near es, opportunities to develop commercial non- Bagre would be far less welcoming to immigrant farm income producing activities, and the workers and farmers than the indigenous Gour- presence of relatives or other individuals from mantche near Kompienga. In contrast to the the same home area, as well as the availability Gourmantche, Bissa in the Bagre area generally of "new lands." live in large, high density settlements where good Within each subregion, the villages that at- agricultural land may already be in short supply. tracted disproportionate numbers of agricul- Have again their predictions are correct. turalist immigration were those in which we Very useful quantitative information can be found: (1) available land; (2) concentrations of gathered through interviews with village migrant immigrants from a single area of origin; and, (3) and host leaders concerning the name, approxi- opportunities for engaging in commercial activ- mate size of household, and length of residence for ities. The very highest rates were observed in immigrant agriculturalist households. This infor- areas that included these three characteristics in mation was more difficultto acquire for immigrant addition to opportunities to engage in commer- pastoralists. Nevertheless, the analysis of these fig- cial cotton production. ures can help policy makers to quantify inter and These predictable patterns have a number of intra-regional settlement trends. policy implications. Selectionof Priority Sites for Need to ConsiderOther Constraintsto GovernmentAssistance Developmentbesides Onchocerciasis In view of the marked preference of settlers for mi- Policy planners need to consider a whole range grating to less isolated areas near roads, preexisting of factors that may impact upon or constrain de- market-i, and administrative centers, we recom- velopment of an area, including isolation, health mend that national governments seriously consider risks, and lack of water. Failure to consider these putting their initial priority on developing sites al- factors, or the high risk and uncertainty associ- ready receiving immigrants. (McMillan, Painter ated with location near a major border, can re- and Scudder 1990). duce the efficiency of development planning. In Specialprograms will be needed to develop high a worst case scenario, lack of attention to such potential areas like the Kompienga and Bagre dam factors can lead policy makers to create projects basins or isolated areas in the Comoe and Leraba ba- that may actually endanger peoples' lives. sins where wecan predict that future settlement will

12 occur.Careful planning will be necessary,however, Notes to insure sustained funding for some of the necessary urban functions (health services,reliable water 1. The exceptions to this pattern are a few large Bissa com- installation and maintenance programs, reasonable munities on the Nakambe (Hervouet 1977) and the Dagara villages on the Mouhoun (Paris 1980, 1983). In studying assurances for animal and human health) that are these anomalies, researchers discovered that onchocercia- ordinarily fulfilled by proximity to preexisting in- sis does not generally pose a threat to gr2ups with popula- digenous populations and/or population centers. tion densities above 50 persons per km (Hervouet et al. dignoup1984; Remme and Zongo 1989).

Need for Guided 2. Burkina Faso has one of the highest rural population Assisted Settlemnentin densities in West Africa, more than half of it concentrated Saturated Areas in the central plateau (Map 23841). The majority of farmers on the plateau were and continue to be dependent on exten- The site reports provide strong evidence to support sive rainfed farming. Because of high population densities, most plateau farmers have been forced to shorten the fal- PNGT research showing that certain villages in low cycle necessary to restore soil nutrients lost in cultiva- high potential river basins are already becoming tion. This has had a negative effect on soil fertility and saturated. Land management programs like the yields and increased the farmers' vulnerability to periodic and lengthy drought (see, for example, de Wilde et al. 1967; PNGT might therefore consider targeting invest- Barrett et al. 1981; Jaeger 1983; Broekhuyse 1974, 1982a, ment in nearby less populated villages. This sort of 1982b; Rey 1980; Nagy, Sanders, and Ohm 1987). "guided" assisted settlement is likely to attract re- 3. Songre 1973; Deniel 1967; Gregory 1974;Coulibaly et al. cent migrants who have lower quality holdings 1980. away from the more saturated area villages and to 4. Conde 1978; Capron and Kohler 1975; Remy 1973; Mar- give these older settlements more latitude in reor- chal 1975; Lahuec 1970; Lesselingue 1975; Ancey 1974; ganizing villageganizing land villageuse patterns. Again, howev- ~~~~~~~~QueantIzard andand Izard-Hertier Rouville 1969; 1958;Kohler Izlard-Hertier 1968, 1972; and Benoit Izard 1973a,1958; er, we would caution that this type of targeted 1973b; Bakyono 1989;Terrible 1979. investment will necessitate the same sort of careful 5. The lack of interest by certain ethnic groups in recoloni- consideration of other reasons why the less con- zation of basins adjacent to their home areas was attribut- gested villages mnightnot have been settled exten- ed to their historic pattern of loose settlements and sively before encouraging people extensive cultivation practices, both of which had made to move there. them vulnerable to onchocerciasis infection in the past (Hervouet 1980).

13 ~~3

Types of Settlement and Settlement Dynamics

The types of new lands settlement taking place sins, however, have more alluvial soils and high- in Burkina's river basins can be characterized ac- er initial soil fertility than the plateau region. cording to the degree of intervention by govern- Prior to the creation of the first AVV planned ment or other agencies as they assist the settlements in 1973, there was little immigration settlement process (Figure 3.1). to the region. Since planned settlement started, the area around the sponsored settlements has at- Immigration to the Study Sites as Examples of tracted large numbers of spontaneous agricultur- Settlement Types alists and pastoralists as well as the sponsored migrants (Figure 3.2). The case study sites represent a cross section of In January 1988, the AVV started an experi- important settlement and government interven- mental program to incorporate spontaneous set- tions in different ecological and climatic zones tlers into village land management groups and (Map 23846). into coverage by the extension services based in the neighboring planned settlement at Rapadama OlderSponsored Settlements and an Experimental (see Chapter 11). A less intensive study to exam- Assisted SettlementProgram in the AVV-UPI ine the early economic and social responses to this program was conducted. We also conducted The first case study site includes three groups of a small number of supplementary interviews planned settlements created by the AVV in the among pastoralists recently settled in the new Nakambe river basin east of Ouagadougou: Li- AVV agropastoral zone at Gadeghin, immediate- noghin, Mogtedo, and Mogtedo-Bombore. Since ly adjacent to the Rapadama planned settlements. 1983, the planned settlements have been administered as Planning Unit One (UP1) under the new Older SpontaneousSettlements and an Experimental model of decentralized administration that char- Assisted Settlement Programat Solenzo acterized the AVVafter 1982 (see Chapter 11). With an annual rainfall of 700 mm, the AVV-UPI Like UPI, Solenzo, the second study site, is locat- is considered to have only medium potential for ed on the northern fringe of the area- that twas intensive rainfed farming. That is to say, the river highly endemic for onchocerciasis before 1974. basins have the same constraint of low and vari- The area has higher rainfall and heavier soils that able rainfall as the neighboring central plateau, are better adapted to cotton than the AVV spon- from which the settlers presently living at UP1 sored settlements. Solenzo is a recent northern ex- planned settlements immigrated. The river ba- tension of Burkina's southwest "cotton boom" that took off in the 1960s.Since 1970,the Solenzo

14 Figure 3.1 Impe of government intervention and immigration to Burkina's OCP River Basins

-C Rairued

Sponsored

Irrigated

Incorporated w/ sponsored

Incorporated w/regional commercial cropping programs

Settlement Types Assisted

In connection w/ classified forests

Assisted pastoralist zones

{Indigenous populations LSpontaneous

Indigenous and outside populations

15 Figure 3.2 Sponsored and spontaneous immigration in connection with the AVV planned settlements at Linoghin, Mogtedo, Mogtedo-Bombore, and Rapadama

AVV Linoghin AVV Mogtedo

Number of families Number of families 80 300

60 A _ 240 _ _

40 Spontaneous 160 __

40 _ _ 8 __ __ _A

20 ___,_ _ A 80 =_ __= /| r--VSponsored \op- O Sponsored 0 Spontaneous '. 66 72 78 84 89 66 72 78 84 89 Year of immigration Year of immigration

AVV Rapadama AVV Mogtedo-Bombore

Number of families Number of families 300 300 2 Sponsored

1 Sponsored 100 II____ 100 _ Spontaneous

0 1 ____0 1Spontaneous 66 72 78 84 89 66 72 78 84 89 Year of immigration Year of immigration

Incomplete data after 1985. Source:AVV 1985c.

16 subsector has experienced rapid spontaneous set- dam site dramatically increased immigration to tlement (Figure 3.3), relatively high crop yields, the river basin in the mid-1980s (Figure 3.4). A and relativelyhigh input use (fertilizer,insecticide). second, smaller wave of immigration involved This continuous stream of immigration, cou- Mossi migrants to the villages adjacent to the pro- pled with the extensive cultivation practices used posed site and new access road (Figure 3.5). Un- by most migrants, is putting stress on area soils. like the indigenous Gourmantche and earlier Planners are especially concerned about the immigrant Yarse, the new Mossi immigrants steady decline in soil fertility related to decreases tended to settle closer to the rivers and "bas in soil organic matter. More intensive cultivation fonds" (Figure 3.5). A third wave involved the practices, using manure and the reincorporation "sponsored" immigration of skilled construction of crop residues, have enjoyed limited success. Lo- workers from Europe, Canada, and other parts of cal authorities believe that because the migrants Africa, as well as the unsolicited spontaneous im- had insecure land rights, they tended to farm the migration of large numbers of unskilled workers largest area possible rather than to adopt more in- to the dam site location. The labor migrants were tensive soil conservation practices (CRPA1988). accompanied by merchants and farmers hoping In an attempt to intensify the high yielding cot- to supply them food and services. ton production systems, the CRPA created three Most of the workers left before or immediately pilot village land management projects in 1988 after the official inauguration of the dam in April (CRPA1988,1989d, 1990a; Kafondo 1987).The pi- 1989.Their exodus coincided with a fourth immi- lot projects' objectives are to help farmers create gration to the former dam construction site, which viable local institutions (village land manage- is now Kompienga town (Figure 3.4). In contrast ment committees) for determining access to and to earlier immigrants, these settlers tended to be management of village land resources with a agriculturalists who were attracted by the area's minimal amount of outside intervention from the prospects for irrigated dry season farming. In Au- government. The program hopes to accomplish gust 1989, the town of Kompienga numbered this goal through a combination of voluntary re- 3,239, not counting civil servants. Ninety-eight striction of the total area that farmers are allowed percent of the inhabitants had immigrated since to cultivate, promotion of extension themes, and 1985 (Figure 3.4).Only 63 (15 percent) of the heads stricter regulation (though not restriction) of the of household indicated that they had worked for terms under which new migrants may occupy the dam, and 310 (75 percent) of the household village lands. Preliminary planning for the pilot heads reported that either agriculture or livestock assisted settlement program began in 1988. One production was their primary activity. of the pilot projects is at Daboura, one of two Solenzo villages in the economic survey. Recent SpontaneousImmigration to River Basins with LargeConcentrations of ClassifiedForests Recent SpontaneousSettlement in Connectionwith the Creationof a DevelopmentProject The fourth case study site is in Comoe Province in the extreme southwest along the C6te d'Ivoire . The Kompienga region provides a graphic illus- The area is traversed by the Comoe and Leraba tration of the way that creating a special project river basins and has many similarities to the Ko- (such as an industrial wood plantation, a major mpienga basin. Before 1974,both the Comoe and road, or a hydroelectric dam) can increase on the Leraba were highly infected with onchocer- spontaneous settlement. Until recently,spontane- ciasis. Here, as at Kompienga, the indigenous ous immigration to the Kompienga River Basin people traditionally shunned cultivation of the has been sparse. Living in small, scattered, low low lying, riverine areas. Despite some of the density settlements, the indigenous Gour- highest rainfall, and soils with relatively good ag- mantche have generally shunned cultivation in ricultural potential, the Comoe and Leraba basins the fly ridden, low lying marshes and river ba- have experienced little immigration in compari- sins. Until preparations for the Kompienga dam son with the other valleys. Especially important began, the area commercial centers or production for the LSR is the fact that Comoe Province has opportunities more classified forests than any other province in Construction of a road linking the provincial the country (Table 3.1). capital at Fada N'Gourma with the proposed

17 Figure 3.3 Yearof immigration for male heads of immigrant households, Solenzo subsechdr,1969

Number of Families 1,000

800 ______

600 ______

400__ _ _ _

200

1973 1976 1979 1982 1985 1988 and before

Source:Nana 19B9c.

18 Figure 3.4 Year of immigration for settlers in the Kompienga Basin and town of Kompienga, 1989

Number of Families 225

Basin

:11

75 _ _ _ _ _ Town I

0~~~~~~~~ .

197 19718 9896 18

Year

Source:Agrotechnik 1988 and McMiflan 1989

19 FIGURE3.5 IBRD23847

M A L I

NIGER

ToF\o-Ngourma BURKINAFASO < - TOuagadougou Areo of / urD- \ mop,,-R loKouar-6 I 4 BENIN 71 Cc)TE D'IVOIR~~~~Ei GHANA \°C

BURKINA FASO MOSSI AND FULANI . .,--~) / \ r (PASTORALIST)IMMIGRATION TO 3003 THE KOMPIENGA RIVER BASIN 13 ° MIGRATION PAYSANNE (MOSSI ET °0 o 3 ; GOURMANTCHE) ET PASTORALE o0 o0 z3 \FOULBE)( DANS LE BASSIN VERSANT 0Komin-Ya ° j3 N ni DE LA KOMPIENGA 3 o 3 0°o ° °E * GOURMANTCHE

.* o 3o D3 3 . YARSE NM 0 D ° ° % t: J | * MOBA

o ° to~O - - -) f 'IMMIGRATION MOSSIAND O U 0~1 D GOURMANTCHE MIGRATIONMOSSI ET o ° '\ D °*fi0 GOURMANTCHE Oa jp-'\\0 w .a: \ .: s,: . ; IMMIGRATIONPEUL OR FULANI \Y031 MIGRATIONFOULBE

Ouargaye ! 0o 0 , o 0n30 n -Source: a Agrotechnik, 1989

* 3 [ 01

)°s"°oo'tD 0/4 D~~~0 Kope : Pm

10 2 -0 1303 6omnebaa 03 00 0 0100000~~~~~~~~~~~~~~=, 0am

\0~_. */ - , , Sanga0 0\,~~~~~00 ~ ~ / ~~ Tin~dangou 0 3 0:013 0 03 3/93

003 El A A

0 10 20 KILOMETERS ------

This -op ho. b-n pr _pcr.dby rhe World 8-oo crc/I e-9lUoly f.. *., ,h. oonc.o.orc ofr odsfor, ioder /re ,rrerrol no of TheWorld BonhG Th,d.-nmrorioos osed od /ho doroan or rhrs * ._ cop do.ot i-mplyon rh prr of ThoWorld Book G-p. ony ,odgr; -O on_h.fbgol,roro.oforyrorrroryoronyerdorserrr orho.eTone of G suchbo-ndorJ.1

JUNE 1992 Until recently, the major force driving immi- projects, the Classified Forest at Toumousseni, gration to the Niangoloko subsector was transhu- was included as a study site. mant pastoralists. Agricultural immigration to the area has increased, however, since 1983, most- Settlement Dynamics at the Study Sites ly because Burkinabe imnmigrantshave been returning from Cote d'lvoire as a consequence of Our research revealed strong similarities in set- that country's growing economic problems (Fig- tler response patterns at all four sites. These re- ure 3.6). These returning migrants are primarily sponses were predictable and reflect the manner responsible for the pronounced increase in ani- and results of settlers' adjustments to their new mal traction and cotton cultivation that has social, ecological, and economic settings. emerged in certain areas of the Leraba since 1983. Initially the settlers at all the study sites fo- Although immigration to the Niangoloko subsec- cused primarily on reinstituting their basic so- tor is still weak compared with other sites includ- cial and food production systems. In the older ed in the study, spontaneous agriculturalist and settlements at Solenzo and AVV, most settlers pastoralist immigration will undoubtedly contin- shifted to a more dynamic phase of economic ue. The main impetus for this is likely to be a and social development after three to five years. steady increase in the number of Burkinabe re- It is during this later stage that the settlers were turning from Cote d'Ivoire. We also anticipate generally most willing to experiment with new that a significant portion of the pastoralist and ag- types of crop technology and income earning riculturalist immigration to the river basins of opportunities. This stage was also characterized Kossi, Mouhoun, Bougouriba, and Houet Prov- by: inces will be redirected to this area as the more (1) increasing problems with loss of organic northern basins become saturated. matter (at Solenzo) and erosion (at the Development planners must accommodate AVV) in village cropping areas because and, if possible, control the depredations of the of extensive cultivation and livestock expanding herds and farmer populations on the practices with little attention to preserv- area's classified forests. As one solution, new ing soil fertility despite the actions un- sources of income from renewable forest resourc- dertaken by their respective extension es are being developed. These multiple use forest services; development projects are intended to encourage (2) reduced access to new uncleared fields local people to take a more active interest in the because of the expansion of the first gen- management and preservation of the forests. One eration of settlers' fields and the attraction of these experimental "assisted" settlement of additional pastoralist and agriculturalist immigrants to the areas; (3) increased competition between agricul- Table 3.1 Classified Forestsin Comoe Province turalists and pastoralists, hosts and set- Name Area(hectares) tlers, over pasture and water for livestock Toumousseni 2,500 because of a substantial increase in the Bounouma 1,300 size of both the pastoralists' and agricul- Beregadougou 5,00 turalists' herds; Badolo 500 (4) a shift in settlers' goals because of the Boulon 1,200 emergence of a second generation of Diefoula 8,500 adult settlers with goals and resources Dida 7,500 different from those of their parents; Gaoandougou 12,000 (5) termination of first generation funding Konkogo 25,000 for special programs that initially attract- Koflande 30,000 ed settlers to these areas; Logonegue 23,000 (6) greater replacement and maintenance Niangoloko 700 costs of first generation infrastructure Source Volta 100 (wells, roads); and Total 117,300 Source: Nana 1989a.

21 Figure 3.6 Yearof immigration to the Niangoloko subsector for male heads of immigrant households, 1989

Number of Families

100

801 l _ _ _ _ _ l _ _ _I_ _

60 1 1 _ _ _ _

40 l-

20 _ _ _ _

1973 1977 1981 1985 1989 and before

Year

Source: Nana 1989a.

22 (7) mounting pressure from adult males and Table 3.2. A Stage Model of Settlement-Related females for the creation of new ways to Development Planning generate private income. The period was also characterized by increased Stage One Planning conflict over the allocation of land, water, and for- Stage Two Initial Infrastructure Development, est resources. Predictable sources of conflict in- Recruitment, and Installation cluded livestock damage to settler crops as herd Stage Three Adaptation (Transition; Settling In) sizes increase, and jealousy created by widening Stage Four Economicand SocialDevelopment gaps between new economic groups. Stage Five Handing Over and Incorporation The way settlers responded to the resolution of these different problems depended on a wide Source:Scudder 1981,1985 range of factors in the macro and micro policy environment. These are discussed in succeeding more diversified, success-ful settlers to leave chapters. (see Chapter 6).

The Concept of Settlement Stages Stage Two:Initial InfrastructureDevelopment,

The concept of settlement stages has been used Recruitment,and Installation to describe similar sorts of interrelated social Ideally, major construction of roads, wells, and and economic change in older settlement other community infrastructure does not start schemes throughout Latin America, Africa, and Asia Asia~(Nlo(Nelson ~ 1931973; ChamberChambers 1969;Scudde1969; Scudder ninguntil problemafter planning at this iS stage completed. involves A keydecisions plan- 1981, 1985) (Table 3.2). The policy significance nigpolmathstgenvvsdcsos of tecne iTat it2canThelpolicy mgvake about how best to develop basic infrastructure. Of the concept IS that It can help polpcu makers The AVV and Solenzo experiences highlight the to anticipate the problems and opportunities need for planners to think ahead 10 or 15 years that are likely to occur at each stage. in planning infrastructure, to a time when pop- StageOne: Planning ulation densities in the surrounding areas will be much higher. In view of the difficulty of Comparative analyses of successful assisted and predicting future immigration trends, it is rec- sponsored land settlement projects worldwide ommended that infrastructure development be highlight the critical importance of advance phased in over a longer period of time, like fif- planning (Scudder 1985). During the planning teen years. l U . U ^ v ~~~~~~~~Theconcept of "recruitment" refers to a variety stage, a broad range of important decisions are made regarding site selection, management of deliberate or implicit policy decisions that af- structure, settler recruitment, crop extension fect the types f settlers and settler households that and basic infrastructure. Planning are attracted to a project. During this stage it is ex- programs, antrl rrsrcue.llnm also considers the extent to which members of tremely important that policy-makers consider how certain attractive features, such as food aid, the host poculat.on will be mcluded m the set- may encourage immigration by families that themAVVper iuamight not be able to survive once special subsi- The AVV experience highlights the crucial im- dies are removed. portance of this stage. Early recruitment policies diearmoved. haadrmti ipctonte evlpmnto Policy-makers should also remember that the settledraintitutions 10-15 ya daer McMioant long-term social security of the settlers rests on 1983). Other initial decisions about land tenure their extended family networks created by birth 1983. Oheriniialdecsiosaoutlan teure and marriage, and that these networks are rein- have had far ranging effects. In particular, hos- and marriae, tantattese ne areren- tile feelings created by tenure policies blocked fre by at terns tedexpere at effective integration of the settlers into regional th at Moghvoa Mog t erensug- market systems, a fact reduced the potential re- ietthat hglevels of governmnentintervention ginaefecs ofgvrmn.n oo net in settler selectionand installation can actually re- mgitnineffectsponsgored ments bydionduingesth tard the process of community reintegration. This ment in sponsored settlements by inducing the semtobtrewthrhepjctflwda seems to be true whether the project followed a

23 strategy of recongregating home village groups Stage Fouir:Economiiic antd Social Development (the case in certain villages at Mogtedo) or deliber- ately tried to mix them (the case at Mogtedo-Bom- The adaptation phase ends when enough settlers bore). Whenever possible, immigrants should be shift from a conservative stance to a dynamic, allowed to self-select the particular social groups open-ended stance, hence initiating a new stage with which they choose to immigrate and live. of economic and social development (Scudder Our experience with spontaneous settlement at 1981, 1985). This transition usually occurs once Kompienga and Solenzo suggests that, left on their food securitv has been established and the set- own, immigrants will typically choose to move in tlers begin to feel "more at home" in their new en- small extended family groups of two or three relat- vironment. ed brothers and then intermarry, over the next two One of the clearest signs that AVV settlers were decades, with other immigrants from the same moving toward a more dynamic period of eco- home area. The only area where we saw pro- nomic and community development was the in- nounced intermarriage between hosts and mi- crease in new conflict situations whose origins grants was at Linoghin, where both hosts and could be traced to greater social and economic di- migrants were from the same home area. Marriage versity within the population. The transition was linkages were extremely important in defining new also signaled by a reinstatement of some of the economic and social alliances at all of the sites. traditional harvest and circumcision ceremonies, Sociologist Some (1992, personal communica- and a greater willingness and ability to resolve tion) notes that it is increasingly common to find internal economic and social problems without new settlements in the valleys that are created by relying on the extension agent. In addition, the a religious leader and his disciples. In this case, period was associated with greater ease and fre- the religious settlement creates a new set of social quency of dealing with local civil and traditional norms particular to itself. authorities. A comparison of these stages with the econom- Stage Three: Adaptation (Settling In) ic results that were being obtained in other areas of the program suggests that important interrela- The concept of a third stage of adaptation or set- tionships exist between community development tling in refers to the complex period of adjust- and settlers' willingniess to modify the proposed ment during which the settlers adapt to their new extension program. This is also the period when economic and social environment. This integra- certain settlers increased their interest and ability tion is likely to proceed more quickly if the mi- to invest in such nonproject areas as extensive grants do not constitute a rival social unit that is livestock herding and off-farm trade. perceived as a danger for the social order of the One of the more dramatic indications of diver- indigenous people. Indigenous populations seem sification at the Solenzo and AVV sites was an in- more willing to accept individuals or small fami- crease in the number of large and small livestock, ly groups. Larger groups are usually more diffi- as well as a parallel increase in the number of con- cult to assimilate (personal communication, J. flicts related to livestock depredations. The desire Some, March 1992). for diversification (Solenzo and AVV) and in- During stage three settlers tend to be strongly creased investment in new labor saving technolo- focused on the reinstatement of basic subsistence gy (Solenzo) was linked to increasing demands levels, and the re-creation of wider social linkages by the first generation of settlers' wives and chil- with other settlers and hosts. Under such circum- dren for semi-independent sources of income or stances it is understandable that their first priori- for redistributing revenues produced on the fam- ty is to establish food security, rather than to ilies' collectively worked fields. experiment with new cash crops or the develop- Success among first generation settlers at So- ment of new, more diversified income sources. lenzo and in the AVV sponsored settlements led During this stage, settlers at all the sites typically to a substantial increase in spontaneous immigra- placed a high value on social cooperation with tion of agriculturalists and pastoralists to the area one another and with hosts.' This phase of adjust- surrounding the settlements. This tended to ex- ment typically lasts three to five years. pand local markets and to offer other opportunities for exchange. At the same time, it increased competitiort for scarce forest, water, and pasture

24 resources and exacerbated the problem of crop village in 1988 (after an average of 13 years of liv- damage by livestock. Opportunities for conflicts ing in the sponsored settlement), only two of the between hosts and settlers also increased as the male household heads continue to work regular- settlers' wealth increased and their immigration lv in the families' cooperatively worked fields. began to appear to be increasingly permanent. Most leadership positions in the village Revolu- This stage of economic and social development tionary Committees (CRs) and in village land was also associated with increased stratification management committees at Solenzo are filled by among settler and host households at the Solenzo the younger brothers and elders sons of the first and AVV sites. At each site, only a small percent- generation settlers. The AVV planned settlements age of households was highly successful. At the experienced a similar sort of generational change. other extreme, 20 to 30 percent of the settler Althoughtthefatermayremainthetitularheadof households at both the Solenzo and AVV sites the family in dealings with the project younger were considerably poorer than the rest (see Chap- brothers and sons made many of the key product- ters 6 and 71). Their poverty could be measured in ton decisions by 1989. terms of the CFA value of their crop production, The village groupements at both the Solenzo and the small size of their livestock herds, and, in a AVV sites have taken over many of the functions few of the worst cases, their need to sell livestock once fulfilled by the project administration, such in order to buy food grains. At the A\/V, these as marketing their cotton and managing the input poorer households tended to be older couples warehouses. In general, however, this transfer to without adequate access to labor or capital, or farmer managed institutions has been much households that had to adjust to the loss of a ma- slower in the sponsored than in the assisted set- jor part of their labor force through death or mi- tlement areas (See Chapter 12). gration. At Solenzo the poorer households tended to be new immigrants who were still ad- Policy Implications of Interrelated Social and justing to the new area or, in villages wlhere new Economic Stages land was already in short supply, late arrivals who may have been given less and/or poorer The concept of settlement stages has important quality land for settlement. In any case, the needs implications for future development planning. of each category-the wealthier and the poorer- were quite different. Need for a Longer Time Perspective in Project Evaluation Stage Five: Handing Over and Incorporation Project evaluations conducted after only three Scudder (1985) emphasizes that no type of gov- to five years are likely to be carried out too early ernment intervention is considered com-iplete un- to find evidence of the wider regional impact of til the activities have been taken over by a second a development intervention. Such premature generation of settlers and the operation of many evaluations can lead national governments and project specific activities have been takeni over by donors to either underrate or overrate the long- local, regional, and national authorities. term development potential of new lands settle- One of the striking features of new lands settlement. It can also encourage donors to cut off ment in the older settlements at Solenzo and the funding for projects at key points in the settle- AVV was the speed with which a new generation ment and development cycle. A myopic focus of settlers-who were teenagers or children when on the first five years of settlement can also en- their parents immigrated-have assumed vari- courage national governments to ignore the ous leadership positions in the settlements. The need to adapt extension programs to the evolv- decision of the first generation "millionaire" ing needs, concerns, and farming systems of the farmers to immigrate to Kompienga was spear- settlers. headed by the older sons and younger brothers of the first generation "millionaires" who were frus- Need for PhiasedAssistance to Anticipate trated at what they perceived as limited opportu- Evolving Needs nities for advancement. It is symptomatic of the generational change that of the eight households Settler and host needs and concerns are very who immigrated to Kompienga from one AVV different at different settlement stages (Table

25 3.2). During the initial installation and adapta- Difficultyof AdministeringSupplementary Food Aid tion period, for example, settlers are primarily concerned with the reinstatement of basic food NGO and government assistance for new lands security. Government assistance during this pe- settlement tends to consume large quantities of riod should focus on providing access to pro- administrative and extension talent in the distri- duction supports (extension, inputs, credit) that bution of food aid. While food aid may be neces- will enable settlers to establish satisfactory con- sary in some cases, especially where displacement sumption levels. The intermediate goal of assis- is forced or where settlements are in isolated, alien tance should be to bring the adaptation phase to surroundings, the importance of food aid should an end as quickly as possible. not be overemphasized, either in the advertise- If government assistance succeeds in facilitat- ment or the design of project interventions. ing the development of higher yielding produc- New settlers at the AVV received one year of tion systems in the third stage of settling in or full rations and a second year of half rations from adaptation, then it typically creates a group of foreign donors. In theory, the food aid allowed settlers with very different expectations, re- settler households who might otherwise not have sources, and goals at the fourth stage of eco- been able to afford to do so to migrate to the nomic and social development. If "captured" by project. The promise of food aid, however, seems new technology and opportunities for diversi- also to have attracted an unusually high percent- fication, the wealthier, more entrepreneurial age of poor households, headed by elderly male farmers can be a force for economic change and and female farmers with few other workers who development. If not captured, that is, if settlers were unable to survive once the food rations were do not see adequate opportunities for local in- withdrawn. These households were among the vestment in the development of crops and in more than 50 who immigrated to the Mogtedo off-farm employment, some of the wealthier sponsored settlements between 1974 and 1977 settlers and settlers with alternative opportuni- but left within two years of coming to the project. ties in their home areas are likely to move. We understand that Linoghin experienced a sim- The problems with increasing competition for ilar high initial dropout during the initial 1973- pasture, water, and agricultural land at this lat- 1977 settlement period. Unfortunately this is not er stage are predictable. Zoning to reserve reflected in our data, which only examine settler certain areas exclusively for pasture offers one dropout between 1979 and 1989.One of the other potential policy option for reducing agricultur- negative impacts of food aid was to create a high alist-pastoralist conflict. These zones must level of dependence on the extension agent and therefore have sufficient pasture and water, that retarded the early development of dynamic which is not, for example, the experience at local institutions. Gadeghin. Policies that increase land security In project villages where other conditions for the indigenous hosts and migrants in the were favorable, the first generation of settlers early stages of development can reduce settler- showed a strong willingness and ability to host conflict and facilitate a peaceful integra- sponsor new immigrants. Several of the average tion of the economic and social interests of the income migrants to Kompienga from the AVV two groups. settlements at Mogtedo had their transportation The primary roles of government at this lat- expenses paid for by wealthier AVVsettlers. er stage-when incomes are on the rise, popu- Settlers who had migrated to the AVVfrom 1975 lation densities are increasing, and settlers are to 1977 gave substantial amounts of sorghum to feeling more "at home"-should be to contin- new settlers who arrived at the neighboring group ue to promote the types of appropriate tech- of sponsored settlements at Mogtedo-Bomboreaf- nology and regulated land use management ter 1979.The new settlers would also sometimes be that encourages farmers to invest in sustain- helped with clearing their fields. Several of the ability. Another role of government should be new settlers' wives spent one week visits in the to assist with economic diversification by pro- older settlements-the ostensible reason being that viding information relating to new opportuni- they could thus learn about the local wild vegeta- ties, and to facilitate linkages with outside bles that they needed to gather and dry for sauces. markets. Almost every one of the sponsored settler households at Mogtedo V3 has hosted at least one (and

26 more often two) farm families for all or part of one than reinstating their production systems at higher agricultural season before the family acquired its levels than they had before immigrating. The chief own AVVfarm. We estimated that this sort of gift exception to this would be when resettlement was exchange of grain and hosting new immigrant forced, as in repatriation from a foreign land or in households accounted for as much as one-third to response to dam related flooding-circumstances one-half of the grain produced in one Mogtedo in which the "normal" process of host and early planned settlement in 1983 (McMillan1987a). Life settler sponsorship are likely to be inadequate. histories of the Solenzo settlers showed that the early spontaneous settlers in this region also Note helped later immigrants by providing food, housing, and supplementary labor on their fields. 1. Although settlers at the AW placed a high value on so- Thus, although government agencies often fo- cial interactions, they were generally not able to interact easily with local inhabitants due to host resentment over the cus on food aid in the early settlement stages, AW's land tenure policies. settlers generally perceive this as less important

27 ~~4

Linkages between Local Planning and National Agricultural Policy

Many development projects in general, and deve- repayments are deducted from the money received lopment projects in the OCP river basins in parti- at harvest sales. Credit for rainy season crops has cular, are planned as if the delineation of a project generally not been available on any significantscale zone renders them immune to the impact of natio- for crops other than cotton. Extension services and nal policy and wider socioeconomic issues. Al- crop research programs also remained strongly fo- though isolated projects may be immune in the cused on cotton production until the mid-1980s. short run, they must be able to survive within a Because of the efficient operation of SOFITEX, broader national and regional context once special farmers receive a guaranteed market and gua- funding is withdrawn. Failure to consider this can ranteed price (95 CFA/kg in 1988; see Table 4.2) lead planners to advocate land management prac- for their production. Since 1984 these prices tices and institutions that are not realistic except in have risen (55 CFA/kg in 1979 versus 58.7 in the short-term. "real" CFA/kg in 1988, adjusted for inflation Especially important in areas of new lands sett- with the INSD [Institut National de la Statistique lement are national policies relating to subsidies, et de la D'mnographie1consumer index [Table 4.1; marketing boards, and transportation that affect the prices farmers receive for their products and pay for yield increasing inputs and technology like fertilizer, insecticide, tractors and plows (Mc- Table 4.1. Official Price Per Kilogram of Millan, Painter and Scudder 1990). These prices Fertilizer (in CFA) and Percent Subsidized influence the net income that farmers earn from their agricultural enterprises. Year CFA Percent 1980 35 63 Producer and Input Price Policies 1981 40 64 1982 55 54 Burkina's agricultural policies between 1982 and 1983 62 49 1987followed a general trend in WestAfrica toward 1984 78 40 removal of input subsidies and government inter- 1985 90 36 vention in markets. Although many marketing 1986 114 20 boards remain, the only one that effectivelyinterve- 1987 91 5 ned in the study areas was SOF1IEX (SocietddesFi- 1988 96 (cash) 0 bres Textiles)-the cotton parastatal. SOFITEXhas 106 (credit) extended short-term credit to farmers for animal Sources.World Bank1985:31; BF Proj. Engr. 1988/89:72,74; traction equipment, fertilizer, and insecticides; IFDC1987:63.

28 Annex 3, Table B-1]).1 The higher prices recei- peak periods, it is typically only in small quanti- ved, however, have been offset by the steady ties and in response to specific needs. rise in fertilizer and insecticide prices, which In contrast, both price and sure market access has accompanied reduced subsidies (Table 4.4). are key problems for cereal production in the As a result, the possible cash returns for one hec- Southwest. Interseasonal price fluctuations are tare, at the same level of yields and input use, especially dramatic in the western grain surplus were 12 percent lower in 1988 than they were in areas like Solenzo. High levels of production 1979 (Table 4.4). and difficult, expensive transport over badly One serious policy problem is the low profita- scarred, unpaved roads have meant that the me- bility and substantial price fluctuation of cereals dian low price for white sorghum-still the between and within years (Sanders 1990). Prices principal food crop consumed and the principal for cereals, which are not supported by the go- food crop grown by all but a very small minori- vernment, can fluctuate widely within a single ty of mini-tractor farmers-was 35 percent be- year, between years, and between regions. Pro- low the median low price in the AVV, and 45 ducers typically receive the highest prices in the percent below the median price at Kompienga. period just before harvest (August), with me- Median prices for white sorghum in the Solenzo dium prices at the end of the dry season (March), region were 28 percent below AVV and 60 per- and lowest prices during and immediately after cent below Kompienga town; even the highest harvest (December). Average and better-off far- prices in the study were 20 percent below those mers in the AVV typically "play the market," sa- recorded for the AVV farmers, and 40 percent ving their major cereal sales for the time when below Kompienga (Table 4.3). Price differences outside demand and local prices are highest. A between the regions were less extreme for red large percentage of the AVV grain that is sold sorghum and millet, which are grown in smaller when prices are low is grain produced on men's quantities.2 and women's private fields. When grain is sold We recorded a higher median price for maize from the collectively worked fields under the su- than for sorghum at Solenzo (Annex 3, Table B- pervision of the household head during the non- 3)-a fact that reflects state intervention to

Table 4.2 Official Prices for Cotton and Cotton Inputs at the AVV,1979-1989 (in CFA)

Fertilizer/kg Insecticides/liter Year Cotton/kg" AMV Min. du Commerce' AVVa Min. du Commercet' 1974 40 35 - 115 - 1975 40 35 - 115 -

1976 40 35 -115

1977 55 35 - 115 - 1978 55 35 - 360 - 1979 55 40 - 400 - 1980 55 40 - 400 - 1981 62 40 - 400 - 1982 62 60 - 400 - 1983 70 60 - 450 - 1984 90 97 - 450 - 1985 100 97 90 450 500 1986 100 120 114 825 800 1987 95 120 91 1,710 1,595 1988 95 106 86 1,260 1,260 1989 95 114 105 1,386 1,582

- Not available. a. AVV, Credit Agricole. b. Ministere du Commerce, Burkina Faso.

29 Table 4.3 Price Ranges for White Sorghum at the Study Sites, 1988-1989

Economic survey' Retrospective surveyb Location Low Medium High August December March August AVV 44 47 50 83 25 42 66 Solenzo 28 34 40 Kompienga 50 60 70 a. Savadogo and Sanders 1989:38; Annex III, Table 2a. b. Guira, F 1989; Annex III, Table 2b.

Table 4.4 Estimated Cash Returns to One Hectare of Cotton or Sorghum, 1988/89 White sorghum Cotton AVV Solenzo 1979 1988 1988 1979 1989 1989 NPK Fertilizer (CFA) 5,550 -15,750- - - - Insecticides (CFA) (12 1/ha) 4,800 -15,120- _ _ Costs/ha (CFA) 10,350 30,870 30,870 - - - Kg/ha 1,000 1,000 800 1,000 1,000 1,000 Gross/ha (CFA) 55,000 95,000 76,000 62,000 47,000 34,000

Price/kg (CFA) (55) (95) (95) (62c) (47c) (3 4d) CFA Net (nominal) 44,650 64,130 45,130 62,000 47,000 34,000 CFA Net (real: adjusted forinflation 1988=100) 72,249 64,130 45,130 100,323 47,000 34,000

- Not available. a. Above average yields for cotton at AVV-UPI, 1988 (1,000 kg/ha). b. Average yields for cotton at AVV-UPI, 1988 (800 kg/ha). c. Average price based on a two-year market survey conducted by the Statistical Service, Murphy and Sprey 1980:58. d. Annex Ill, Table 2a. secure stable maize prices for the tractor farmers. crops and on the use of intensive cultivation At the time of our study, the cotton parastatal methods like prescribed rotations. SOFITEXhad a special arrangement with Burki- Faced with: (1) the huge labor demands for cot- na's national brewery whereby the latter agreed ton (1,520adult labor equivalent3 labor hours per to purchase the farmers' entire harvest for a gua- hectare as opposed to 394 weighted labor hours per ranteed price of 60 CFA per kilogram. This poli- hectare for white sorghum and 507 weighted labor cy encouraged the tractor farmers to practice the hours per hectare for maize in 1979 [McMillan recommended crop rotation while it also helped 1983:397]); (2) rising production costs for cotton; them reimburse their project credits. In contrast, and (3) past experience with uneven, unreliable the price available to the animal traction and ma- yields in the northern climatic zone, the AVV far- nual farmers who had no access to this guaran- mers have gradually shifted to a more diversified teed market could fall to as low as 25 to 28 CFA system that includes commercial cereal produc- per kilogram, which is 53 to 58 percent below the tion, livestock sales and, when possible, trade. The- guaranteed price for tractor farmers. se localchanges in settler production patterns have resulted in a steady decline in total cotton produc- Local Impact of Policies tion and fertilizer use in each of the AVVplanned settlements (Figures 4.1 and 4.2). On Area Planted in Different Crops Cotton production at the AVV was generally highest in the first three to five years of planned The net income per unit land and labor for diffe- settlement, when the settlers were reimbursing rent crops at available crop and input prices has their initial medium-term credits to the project. had a decisive impact on the settlers' choice of Despite a net decrease in total cotton production

30 Figure 4.1 Total cotton production in the AVV planned settlements

Kilograms (000s) 1,200

I ~ ~ ~ ~ ~ I

800 .

0 Y

1974 1978 1982 1986 1988

Linoghin -|-lmi'll Bombore li"|Wgrai""i Mogtedo

Source:AVV/UD Extension records

31 Figure 4.2 Total Fertilizer Purchases in the AVV Planned Settlements

Linoghin 140

120

100

1977 1980 1983 1986 1988

Linoghin - nIi-inSI Bombore " ' 1"uU11 Mogtedo

Source:AVV/UDExtension records

32 in each of the planned settlement groups, house- Nagy, Sanders and Ohm 1988; Deuson and holds with larger labor forces have continued to Sanders 1990). grow this crop, fearing, from experience, that after a plentiful harvest, the price for sorghum On Input Utilization might drop. (Farmers reported that the price had been as low as 25 CFA/kg in December-January It is more difficult to discuss the impact of hi- 1988, which is a time period when farmers typi- gher input and output prices on input utiliza- cally need the most cash.) Smaller households, tion. In general, purchased inputs such as with only two or three active workers, usually mineral fertilizer, insecticides, and hired labor, grow less cotton. With a smaller labor force, the- tend to be used only on cotton. At all of the stu- se households are less able to absorb the high le- dy sites fertilizer use is concentrated on cotton. vels of risk associated with commercial cotton Our field research confirms the results of the production in the northern sudano-sahelian ri- CRPA du Mouhoun farm monitoring survey, ver basins. For them, the loss of a key family which shows high levels of fertilizer use on cot- member (through illness or migration) at an im- ton, and lower levels of application on maize portant point in the production cycle (e.g. wee- (Table 7.7). Large areas of maize are grown only ding) can decrease cotton yields. If the crop is by the tractor farmers (Table 7.8). lost, farmers still are obliged to reimburse the Fertilizer at the AVV is used almost entirely on cost of the inputs, which can amount to as much cotton, and this was reportedly at levels far be- as 30,870 CFA for one hectare of cotton (Table low those recommended by the project. In addi- 4.4).4 tion, a large percentage of the sorghum and In contrast to AVV farmers, Solenzo farmers maize fields are not grown in rotation with cot- have remained heavily focused on cotton pro- ton, thus reducing further the benefits of the reduction. Between 1972 and 1988, cotton produc- sidual effects of fertilizer use on cotton. tion in the Solenzo sector has risen from 801 Higher input prices undoubtedly played a tons to 41,500tons (Table 4.5). At current prices, role in diminishing the use of inputs, especially the farmers have little incentive to diversify into for households with smaller labor forces for the commercial production of crops other than whom cotton production was especially high cotton. (The one exception to this is maize, be- risk. Nevertheless, the net decrease in fertilizer cause of the arrangement with the national consumption at the AVVbetween 1979 and 1988 brewery.) High yields and low price and yield is more easily attributed to the lack of an appro- risks are the incentives leading farmers to grow priate technology package than to higher fertili- cotton (Sanders, Nagy and Ramaswamy 1990; zer prices.5 In the higher rainfall Solenzo area where the existing cotton production technology is more profitable, research by the CRPA shows Table 4.5 Average Cotton Yields, Production, only a slight decrease in fertilizer use on cotton and Area Planted in the Solenzo Sector, between 1982-1983 and 1987-1988 (from 152.5 1972/73-1988/89 kg/ha to 140 kg/ha) despite a substantial increase in input prices (CRPA du Mouhoun Average Total Total 1989b:33'. yields production area *9 ' Year (kg/ha)t (tons)b (ha)' 1972/73 544 801 4,037 On Settler Incomes and Patterns of Diversification 1980/81 870 - - 1981/82 849 - 14,440 Producer and input prices affect settler incomes 1982/83 1,043 - - and the profitability of different cropping strate- 1983/84 1,083 - - gies. Average prices used to calculate the value of 1985/86 1,076 - - agricultural production in CFA per kilogram at 1986/87 1,290 - - the AVVin 1979are based on a two year survey of 1987/88 1,598 - - local markets (Annex 3, Table B-7). The prices 1988/89 1,050 41,500 38,,00 used for calculating crop income for the 1988-1989 crop year are based on the median prices that far- - Not available. mers received for their products, based on our a. CRPA 1989b:29. analysis of crop sales over one calendar year b. Nana1989c. year

33 (Annex 3, Table B-3). Both 1988 and 1979 were rely offset by output price increases; (4) declining considered good agricultural years. soil fertility due to an inappropriate technology Average agricultural income recorded for a base for cotton production; and (5) the fact that subsample of the AVV farmers included in the settlers were restricted to a 10-20 hectare project AW Statistical Service Farm Monitoring Survey farm and animal traction cultivation technology. in 1979 is compared with the crop income recor- To offset this decline in net crop income, the AW ded for them in 1988 in Table4.6. When nominal settlers shifted toward a mixed strategy of com- income is adjusted for inflation using the IMF mercial cereal production and diversification into price index, average real income per ALE seems livestock and nonfarm employment. to have declined by 40-50 percent (Table 4.6). This observed decline cannot be attributed to a On FoodSecurity reduction in the amount of land farmed per ALE (Adult Labor Equivalent; cf. Table 7.3). Instead, Current policies are compatible with national the chief reasons appear to be: (1) settlers redu- goals for food self-sufficiencyand increased forei- cing the percent of their total land area planted in gn exchange, at least in the short-run. High levels cotton and increasing the percent of the area de- of cash income and investment in cattle would voted to cereals, which have a lower median pri- presumably allow the Solenzo farmers to purcha- ce;6 (2) the lower recorded yields on cotton (800 se food grains, were they to experience shortfalls. kg/ha vs. 1,000 kg/ha in 1979) and higher recor- Recent research on cotton farmers in a similar eco- ded yields for sorghum (1,000kg/ha vs. 800 kg/ logical zone in Mali suggests that farmers ha in 1979) (Murphy and Sprey 1980; AVV-UP1 growing a high ratio of cotton to subsistence food Unpublished Extension Notes 1989); (3) lower crops were as well, if not better, nourished than profit margins for cotton because average yields those growing greater amounts of food due to decreased and higher input prices were not enti- their greater purchasing power (Dione 1989).

Table 4.6 Comparisonof Net Agricultural Income in the Older AW Planned Settlements, 1977-1988/89per Unit Labor(ALE)

Linoghin Bombore Mogtedo Mogtedo V3 Nominal CFAa Cooperatively worked fields only, 1977 55,700 -62,200- 1978 51,700 -50,550- 1979 157,300 to 66,300) -156,600 to 65,000) _ 1988 61,315 54,988 51,072 55,101 ($204) ($183) ($170) ($184) Privately and cooperatively worked fields 1979 75,566 1988 76,766 64,329 65,742 72,588 Real CFA, 1988=lOOa Cooperatively worked fields only 1979 192,718 to 107,2821 -191,586 to 105,178)- - 1988 61,315 54,988 51,072 55,101 Private and cooperatively worked fields 1979 - - - 122,275 1988 76,766 64,329 65,742 72,588

- Not available. a. Adjusted for inflation using IMF consumer price index for Burkina Faso (1988=100). Both 1979 and 1988 were good rainfall years. b. Figures for 1977, 1978, and 1979 are for AVV settlers in their third to fifth years of living at the project, by which time they were presumably farming the full area that they were authorized to farm under the AVV systemr. For 1979, the first figure refers to farmers in their third year, and the second to farmers in their fourth or higher year. Net income figures for farmers in their first or second years were calculated separately in the Murphy and Sprey report (1980:69) but are not listed here. Sources: Savadogo 1989a; Murphy and Sprey 1980:69; McMillan 1983; IMF 1989:256.

34 These same agricultural policies may, however, regional markets of Linoghin and Mogtedo also work against national goals for longer-term, shows that indigenous market systems are able more sustainable food self-sufficiency. This is be- to redistribute large quantities of grain and li- cause they discourage small farmers (who are wi- vestock from rural producer to urban consumer, thout access to the guaranteed price for maize) given easy access to an urban market on a paved from following the recommended crop rotation, highway. Marketing problems in the valleys are and discourage all farmers from using fertilizer linked to very practical problems of expensive on crops other than cotton. Marketing policies and difficult transportation-rather than to any that provide tractor farmers a guaranteed price tendency toward monopolistic control of rural for their maize production (but exclude animal markets by local traders. Recent collaborative traction and manual farmers) and subsidize trac- research by OFNACER and USAID demonstra- tor purchases and repairs, have facilitated the ted that the high price of transporting crops emergence of a group of highly successful, com- from the high potential river basins was the mercial cotton "millionaires" and inadvertently principal factor accounting for the much lower fueled a race for land. prices in these areas (Sherman, Shapiro and Gil- bert 1987; Dejou 1987). Implications for Future Policy Need to Promote Improved Transportation and Current programs to promote the development of Applied Research more sustainable cropping systems are made more difficult by the substantial price variability Any long-term solution to the development of of the cereals within and between years. Even less more sustainable land use practices must be lin- isolated, lower rainfall areas like the AVV planned ked to a variety of macro policies to improve in- settlements, which have easy access to the Ouaga- frastructure in the valleys, to additional macro- dougou markets, are vulnerable to the low market and micro-level research, and to extension poli- prices that inevitably follow a series of bumper cies that give rise to more diversified production crop years. High prices for mineral fertilizer, sta- systems that are not as hard on valley soils. gnant yields, and low average selling prices com- What would be the effect of improving hi- bine to give farmers little incentive to invest in ghway transportation to the high potential river increasing crop yields, except in areas where there basins of the west? In the short run, it would are opportunities to invest in more diversified doubtless attract grain merchants from Ouaga- agricultural and nonagricultural enterprises. dougou and other consuming areas, bringing pri- With present price policies and transportation ces in the isolated west into greater parity with infrastructure, settlers in the more isolated west other regions. Inter- and intra-annual price fluc- have few alternatives to commercial cotton pro- tuations would be reduced and many settlers and duction. Diversification outside of crop produc- indigenous farm families would be encouraged tion is achieved primarily through livestock and to convert from commercial cotton to commercial trade, but opportunities for local livestock pro- food crop production. The rate of substitution duction are decreasing. between cotton and cereals would also depend Cotton production is less profitable in the drier, on the state's commercialization policies. northern river basins. There this same policy en- While improved transportation would tend to vironment encourages farmers in the less isolated benefit urban consumers and the nation as a who- AVV settlements to convert from the prescribed le, it would not in and of itself resolve the income cropping pattern which focuses on commercial problems of settlers in either the higher or lower cotton production to a more diversified strategy rainfall basins. Even with improved transporta- that mixes commercial cereal and livestock pro- tion, cotton is likely to remain the dominant cash duction with small amounts of cotton production crop in the southwest because of its high guaran- and off-farm employment. teed price. Further, lower national prices would Past experience with cereal marketing boards probably displace the comparative advantage in Burkina Faso suggests that these are not the that is currently enjoyed by settlers living in the preferred route to more secure grain and lives- northern lower rainfall river basins (a process not tock markets. Indeed, the case study research on unlike the gradual shift of agriculture from the the dynamic growth and present volume of the southern states of the U.S. to the midwest) (Viera,

35 Sanders and Habash 1988).Thus expensive inno- cerbated these inequities and led to a substantial vations like roads must be paired with macro and rapid, uncontrolled, spontaneous immigration micro policies to diversify the rural economies of from the less well endowed adjacent plateau both the higher and lower potential river basins. areas. These high rates of immigration-much of Sample programs might include: it driven by poverty as well as new opportunity- (1) research and marketing assistance for have already contributed to the saturation of grain legumes and dry season irrigated some of the higher potential, more accessible river gardening to increase diversifications basins. Parallel programs to develop the home (crops that might complement commer- (sending) areas will be needed to reduce the speed cial cereal production, much as soybean at which the valleys will be settled and to give cultivation combined with cereal pro- planners more time to develop appropriate exten- duction in large areas of the southern sion and infrastructure services. U.S. and Brazil; see Vieira, Sanders and To date, good farmland with flexible tenure ru- Habash 1988); les is still available in the more southerly river val- (2) increase efforts to supply national indus- leys of the Nakambe and Nazinon (ex-White and tries with local products, such as the hi- ex-Red Voltas),as well as in the Kompienga, Lera- ghly successful program whereby ba and Comoe basins. In contrast, the more nor- Burkina's national brewery provides the therly basins are becoming increasingly saturated. Solenzo tractor farmers with a guaran- The earliest sign of river basin saturation is the de- teed market for their maize; parture of the pastoralists as farmers' fields (3) a continuation of current policies to pro- spread, placing additional pressure on pasture mote the processing and sale of locally and water resources. If the first 15 years of settle- produced crop and forest products (e.g. ment are any indication, these less populated ri- karite butter, peanuts, grain and rice) and ver basins will soon be occupied as well. It is livestock, especially small ruminants. therefore important that development planning Studies throughout developing countries point for the basins be incorporated into a national pro- out that as incomes increase, farmers and urban gram that uses targeted investment to develop the people alike want a more diversified diet that in- impoverished home areas-an investment that cludes larger quantities of grain legumes, other should pay off by raising income levels and redu- vegetables, and livestock products. A compari- cing the flood of immigration from these zones. son of the evolving average prices for small lives- Targeted investment in the valleys to the exclu- tock at the Linoghin market since 1979 shows the sion of investment in the settlers' home areas average price for a large goat in 1990 at 188 per- creates a magnet that can attract massive cent more than it was in 1979;a large sheep was amounts of spontaneous settlement. While addi- 200 percent more, a pintade (guinea fowl) 100 tional settlement may be necessary for dynamic percent more, and a chicken 200 percent more regional development, rapid, uncontrolled, spon- (Annex 3, Table B-5). Pigs are said to be the most taneous settlement can quickly overload the ca- profitable of all, with prices increasing from an pacity of existing social, land tenure, extension, average of 4,000 CFAfor a hog in 1979to 16,000in and marketing systems. 1990 (a 300 percent increase). The rapid spontaneous settlement that occurred in the area around the AVVsponsored settle- Need to Link Planningfor Settlement Areas with ments contributed to the sponsored settlers' Developmentof Sending Areas insecurity about their long-term land tenure rights in the area. In contrast to the sponsored sett- National planning also must include targeted in- lers, who were recruited by the AVV from the vestment to develop the settlers' home areas and more densely settled plateau, over 90 percent of to reduce the inequities between these areas and spontaneous settlers around the AVV planned the river valleys (McMillan, Painter and Scudder settlements at Linoghin, Rapadama, Mogtedo, 1990).The OCP river basins already include some and Mogtedo-Bombore were from more distant, of the best agricultural lands in Burkina and the isolated, infrastructure poor and service poor other landlocked Sahelian countries. Targeted in- villages in the adjacent provinces. While these vestment in extension services and infrastructure settlers were attracted by the opportunities crea- at Solenzo and the AVVin the mid-1970s only exa- ted by AVV wells and roads, they were also res-

36 ponding to the lack of opportunity in their home ral household's incomes were spent on food, transportation, villages. A large part of this immigration was and such manufactured items as bicydes, radios, and mopeds. These goods show substantial price increases-on a scale of 200 short-term seasonal migration, i.e., settlers ha- to 400 percent-between 1979 and 1988 (Annex 3, Table B-2). ving little intention of remaining in the area once Accordingly,we consider that the national consumer price index (Annex 3, TableB-1), which calculatesan inflationrate of 62 the productivity of their newly cleared fields be- percent since 1979,seems to be a very conservative means of gan to decline. Nevertheless, these high rates of adjusting the value of net value of production in 1978-1979for spontaneous, uncontrolled settlement in the ad- comparisons with 1988-1989net figure. jacent areas contributed to the atmosphere of in- 2. The results of the survey coincide with information from a security about settlers' long-term tenure rights joint OFNACER (OfficeNational des Cbftles)-USAID (United to land. States Agency for International Development)study of national grain markets in 1983-1984(Sherman, Shapiro and Gilbert To date, however, there is little agreement 1987).The survey showed that, in 8 of 13 months, the producer about how to revitalize the impoverished areas of prices recorded for maize were 12 to 26 percent greater in Oua- origin. One example involves promoting new gadougouthan in the southwestern city of Bobo-Dioulamo. During the same period, the averageprice of white sorghum small-scale water conserving technologies like was 25 to 70 percent higher at Mogtedo and the neighboring the anti-erosive diguettes that have been success- market of Zorgho than at Solenzo.Millet prices during this pe- fully promoted in the provinces of Yatenga and riod were 20 to 90 percent lower at Solenzo. No substantial price differenceswere recorded for red sorghum in the few cas- Sanmatenga. There is a chronic need to develop es where ORD recorded figures for it. The higher median prices this and other types of appropriate technology to that were observed for the AW can be attributed largely to the increase yields and to reduce soil erosion. Ano- settlements' easy access to the major grain market in Ouaga- ther possibility lies in the development of income dougou, the capital. generatinggenerating micro-projects. Some of the most suc- AWV3. Seesystem Table 7.3 of laborand Footnoteweights 2, which Chapterwas 7 used for a discussion to adjust of the cessful projects like the "Six S" projects in Yaten- ga have been launched by non-governmental 4 To reimnbursethese costs could require the cash value of 700 kilogzrams of sozrghum at the low prices (44 CFA/kg) we record- organizations. This type of project can benefit wo- ed immediately after harvests. If such a household were farm- men as well as men. "Six S," for example, has ingthefourhectam recommended for thatlabor forcesize and had lostone hectare'sproduction due to a failed cotton crop, re- started a program for on-farm fattening of small imbursing the cost of the inputs would consume about a third ruminants that is providing women with signifi- of the total cash value of production for the family (Murphy cant revenues. and Sprey 1980:11). 5. Even in 1978-1979and 1979-1980,when fertilizer prices were Notes being subsidized by 60-70percent, average fertilizer use for cotton was 90-120kg/ha for settlersin their third to fifth year at Li- noghin and Mogtedo, but only 20-30kg/ha for white sorghum, 1. Since 1981 the International Monetary Fund (IMF)consumer and 90-140kg/ha for cotton for settlers in their third to fifth price index seems has been constructed from the Institut Na- year at Mogtedo-Bombore (Murphy and Sprey 1980:46).In tionalde laStatistique et de la D*nographie(World Bank 1989).Al- 1978and 1979,the AW recommended 150 kg/ha for both sor- though the index reflectsan urban market basket that includes ghum and cotton. Only 14of the 43 (32 percent) white sorghum fuelwood, cereal, and housing expenses not paid for by rural fields cultivated by the settlers studied at Mogtedo received farmers, certain categories of food, clothing, meat, fish, equip- any fertilizer at aDlin 1975-1980(McMillan 1983:392). ment, education, and leisure are purchased by settlers and indigenous farmers with middle and higher incomes. Other 6. This increased total cereal production. Murphy and Sprey expenses, such as transportation, are probably higher than (1980:78)estimated average cereal production in the AW spon- those paid by poor urban dwellers and lower level civil ser- sored settlements to be 200 kg per resident in 1988.In 1988,the vants. settlers in the AW restudy produced an average of 900-1,100 Our analysis of expenditures by male household heads and kilograms of cereals per ALE or 350-400kilograms per resident other famnilymembers shows that the largest portions of the ru- at Linoghin, Bombore,and Mogtedo (Table7.6).

37 ~~5

Land Tenure

A requisite for successful land settlement is se- forces the capacity of local inhabitants to regulate cure tenure to land and water resources as de- access to village land and other natural resources. fined by both the settler and host populations (McMillan, Painter and Scudder 1990). Land re- Le Regime Coutumier-Indigenous Land sources include not only village sites and arable Tenure Systems land, but also grazing areas. Water resources include potable supplies for people and their do- The "new lands" in Burkina's river valleys may mestic animals as well as fishing rights. have been underpopulated, but they were never If tenure rights are unclear, both settlers and in- unclaimed. digenous inhabitants will be unwilling to invest The first category of indigenous land rights, in the development of more environmentally sus- droit d'appropriation collectif, refers to the collective tainable land use practices. Even if land tenure rights to land that are vested in various lineage rights are spelled out at the national level, settlers segments (generally defined as those extended may continue to feel insecure if these guaranteed from the same father, grandfather, or great- land rights disrupt their long-term collaboration grandfather). with indigenous inhabitants. The second category of land tenure refers to the We recommend, therefore, that policymakers rights of the extended kin group or lineage to a focus on land security, rather than on a more le- body of land. These rights are much less specific galistic inteprettionoflndtnurerighsthan the droit d'appropriation collectif, which refers galistic interpretation of land tenure rights. More- to spcii pice ofln.Telnag'"adrgt over, we argue that recourse to legal mandates to enforceland tenure rights is likely to create more are represented by the chef de terre or earth priest, enforce land tenure rights is likely to create more by virtue of his literal and spiritual descent from problems than it resolves. Settlers are aware of the first occupants of the land. The earth priest's the contradiction between state policies that leg- role includes that of interceding between men islate rights and the reality of peaceful social and and the supernatural powers related to the land. economic incorporation into a new community. A third category of tenure includes individual They are therefore likely to favor more traditional use rights that are accorded either permanently mechanisms for acquiring land, especially when or temporarily to individuals or families who the number of migrants has not surpassed the have rights to droit d'appropriation collectif. These number of indigenous inhabitants in a region. rights are passed from father to son or, in the More apt to be successful are government pro- matrilineal areas of the southwest, through the grams like Burkina's new national program for mother's line. The rights of other groups-i.e., village land management (PNGT) which rein- groups without collective allocation rights such

38 as immigrants from outside the area-are consid- domain lands are considered to be natural re- ered provisional and revocable, or "borrowed." sources and certain resources created by man The outward and visible sign of this symbol of (buildings, infrastructure) that by their nature or the difference between "borrowed" and "inherit- social importance should not belong to any pri- ed" individual use rights involves certain rules vate individual. This category included rivers, regarding land use. For example, one may be for- roads, lakes, river and lake banks, bridges, mili- bidden from planting trees. Enforcement of these tary bases, and public transport. edicts is generally through the social ostracism of Private domain (le domaine prive) was consid- those who violate the rules. The threat of social ered to be a separate category with a distinction ostracism functions as a powerful force since set- between domain affecte (land affected by the stat- tlers know that their long-term economic success ute regarding improved land) and domain nonaf- is strongly linked to their ability to integrate fecte, for unimproved land not affected by the themselves into their new social settings. statute. Once a land loan has been inherited, it is nor- The Burkinabe state inherited this system of mally considered to pass into the category of in- state and individual private and customary land dividual use right (category one) and is no longer tenure rights, and modified it by additional subject to being revoked. The most substantial edicts. A law enacted in 1960 (No. 77/600/AN, 12 body of evidence of land loans being reclaimed is juillet 1960) made the state the final proprietor of in the high density central plateau area, where all land not already registered by that date. This farmers are reluctant to make land loans for peri- was followed by a later act (No. 29/63/AN) in ods of more than a few years (Boutillier 1964; Mc- 1963 that authorized the government to reserve Millan 1980). for the state any land that benefited from state in- Despite certain differences in application that vestment in infrastructure development, and to relate to different patterns of social organization declare as property of the state all underpopulat- (matrilineal versus patrilineal kinship systems, ed lands or lands distant from major population for example), one encounters the same three basic centers. categories of land tenure rights in all of the river These early laws establishing the regime regle- basin populations. The essential characteristic of mentaire provided the legal basis for creating the all three categories is that they do not allow for AVV, as well as Burkina's extensive network of appropriation individuelle des terres, or private land classified forests and protected wildlife areas. In ownership. Instead, all three safeguard the in- theory, the presidential decree that created the alienable collective rights of individual lineages AVV in 1974 gave the agency complete control of or extended kinship groups. This collective sys- all settlement and development planning in the tem does not exclude migrants, as long as the mi- less populated river basins covered by the OCP. grants are willing to respect local customs and This area 2yas legally defined to include almost traditions. 30,000 km -over 10 percent of the total land area of the country. Similarly, the laws creating the Le Regime Reglementaire-The Concept of classified forests were considered to render null Individual and State Ownership and void any previous land tenure claims. By 1979, it was clear that the state could do little Parallel to this system of traditional land tenure, to control the floods of spontaneous settlers mov- the French colonial government introduced the ing into the area surrounding the AWVsponsored legal concept of private and state ownership (re- settlements. It also became increasingly clear that gime rlglementaire). Individuals or organizations large areas of the most accessible fertile river ba- holding such rights have the power to dispose of sins in the western cotton boom area would be oc- property in an absolute manner. The legal basis cupied by spontaneous settlers long before the for the regime reglementaire dates from a series of AVV could design sponsored settlements. These government edicts issued between 1906 and 1956. spontaneous settlers continued to pass through On the basis of these edicts, Burkinabe could customary channels to acquire land. Efforts by claim private ownership of land on which they the government to evict settlers from the AVV had built or made improvements. project lands and classified forests were compli- A further legal distinction was made between cated by legal ambiguities in the national land public domain and private domain lands. Public tenure laws, which recognized state rights at the

39 same time that they validated local rights ac- self-help and collective group financing of village quired through customary channels. development projects (AVV 1985d:10).

Agrarian and Land Tenare Reorganization, 1984 Early Impact of the Land Tenure and Agrarian Reform Act On 4 August 1983 Burkina underwent a complete change in government. The new revolutionary Our field studies show that the majority of spon- government made agrarian and land reform a top taneous settlers at the research sites acquired land priority. An interministerial commission was cre- to farm by making formal requests to customary ated in June 1984 to oversee a complete rethink- authorities. This traditional pattern of acquiring ing of the land tenure and agrarian systems in land appears to be continuing despite the 1984 time for the first anniversary of the revolution. Its land tenure act. In fact, however, even when set- final document, announced less than two months tlers received authorization from the local Revolu- later, has provided the basic background for rural tionary Committee, they typically also requested development planning in Burkina ever since. authorization from a traditional authority, on Two of the central tenets of the interministerial grounds that the land is subject to supernatural report and the resulting Rdorganisation agraire et forces with which it is necessary to remain in ac- fonci?re act involve the nationalization of all land cord. In cases where this authorization was re- rights and the appropriation of usufructuary fused, settlers preferred to move elsewhere. rights to whoever clears land. Access to urban or The continuation of these traditional practices rural land was to be restricted to "those who despite the limited abilitv of the indigenous peo- show a real need, without distinction of sex or ple to expel violators (before land reform), and matrimonial status" (AVV 1985d:18). the ability of the immigrants to bypass local au- A third theme focuses on the need to invest lo- thorities (after land reform), shows the high value cal residents and village leaders with responsibil- that immigrants place on minimizing conflict. ity for the management of land resources. The The settlers are vividly aware that, land title or local Committees for the Defense of the Revolu- not, they have few opportunities for a satisfactory tion (known today as Revolutionary Committees social and economic life in the new area if they or CRs; see Annex 2) were given the power to al- overlook local authorities. locate village lands. The conditions for receiving Many things happen when a settler borrows land were supposed to be formulated by a village land under the traditional land tenure systems. land management committee, under the guid- T he tra ditionl land but sota ance of the local CR. In project areas, like the AVV The migranot acquires not only land but also a or irrigated areas, the organization responsible "friend" to whom he or she is indebted in the new for development has the right to define the condi- community In the valley areas, where land is still tions for land occupation in working with farm- abundant, the "loan" is usuallv a straight grant, ers through the CR structures for popular with little expectation that it will ever be re- democracy (AVV 1985d:11). claimed. By giving a new settler land and, per- A fourth theme deals with regional land use haps, extra food, an indigenous farmer makes a planning. Chapter 2, Article 21, states that rural social investment. He creates indebtedness and land management must consider the integration gains a friend and political ally. Special sacrifices of all the activities of rural life-agriculture, live- or rituals to ratify a land loan provide additional stock, and forestry-taking into account the natu- opportunities to formalize the migrant's incorpo- ral potential of an area. Thus each area should be ration into the new community. In passing subject to a battery of soil and hydraulic surveys through the customary channels to acquire land, as a background to the preparation of a compre- the borrower acknowledges the lineage and vil- hensive land use survey The need for each devel- lage leaders' right to control the use of the village opment plan to designate land for pastoralism land in service to community life. was also emphasized (Chapter 3, Article 24). Settlers moving into the Solenzo villages ac- National ministries-in collaboration with the quired land through the traditional mechanism of new system of popular democratic structures (the land loans. New settlers would address their re- CRs)-were supposed to implement these pro- quest to the indigenous village leaders. Over grams. The commission also promoted the idea of time, the immigrant group's access to "new" land

40 that could be cleared and farmed on the edge of tiated only with isolated spontaneous settlers, or the settlement would gradually be increased. with the leaders of small villages whose fields At Kompienga, 43 percent of the 425 farm fam- were not immediately adjacent to the sponsored ilies in the town purchased their homes and 11 settlement (McMillan 1983). percent rented; most of the other 46 percent ac- By 1979, the fifth year of settlement at Mogtedo quired their homesites without reference to tradi- and Mogtedo-Bombore, relationships between tional authorities: through inheritance (1 farmers living in the sponsored settlements, in- percent); from the Revolutionary Committees (5 digenous farmers, and spontaneous settlers had percent); as grants from the construction compa- become increasingly hostile. The spontaneous ny (6 percent); or without asking anyone (squat- settlers quickly learned that although the AVV ter's rights-16 percent). Only 18 percent went could threaten to expel them, it could do little to through the traditional authorities. Nevertheless, implement such threats because of the legal am- nearly all the farmers in the farming systems sur- biguity of the settlers' land tenure. Although by vey acquired their bush fields by requests to the presidential decree the project had the authority Gourmantche chiefs or another Gourmantche el- to declare a given land area property of the state, der in the community. Twenty-four percent of the the state also recognized indigenous land rights. spontaneous settlers interviewed as part of the The result was confusion regarding the issue of assisted spontaneous settlement program at Ra- whether the AVV had any real power to revoke the padama and 66 percent of the settlers at Linoghin spontaneous settlers rights' to remain on project claim to have passed through the customary au- lands. Spontaneous settlers moved into the areas thorities for the area occupied by the AVV surrounding the sponsored villages, in some cases planned settlements to acquire permission to even onto project fields. Social conflicts between farm (AVV 1988:21). the AVV settlers and the spontaneous settlers In structured interviews with 20 male house- emerged over the more accessible areas with better hold heads at Kompienga and 20 heads of house- land. The conflicts over land mounted as settlers' holds at Solenzo, settlers emphasized their plans to expand their private fields ran into direct hospitable relations with the indigenous inhabit- competition with spontaneous settlers' plans for ants. Hence, the social investment obtained the same fertile, low lying areas. through respecting traditional land loan practices In one graphic case, a spontaneous settler who proved valuable. could claim a customary right to an area of the In contrast, our research on the AVV planned adjacent AVV village used cotton insecticide to settlements showed the high price to be paid for poison the chickens of sponsored settlers. The bypassing traditional mechanisms for acquiring dead chickens were then eaten by the settlers' land. When the first planned settlements were dogs, who also became sick and died. Although created-in 1973 at l1inoghin and 1974 at Mogte- of little economic consequence in itself, the total do-the indigenous inhabitants were not anxious effect of a series of small incidents such as this to join. Consequently, the project was implement- was to increase settler anxiety and reduce the set- ed with little cooperation from indigenous inhab- tlers' sense of security in the new area. itants. By virtue of the 1974 presidential decree, The more long-term economic impact of the all existing land claims were declared void. Thus, project's bypassing of traditional land tenure the sponsored settlers' claims to their AVV farms rules became significant in 1979, when some of were accorded by the project, not by the custom- the wealthier AVV sponsored settlers tried to ary authorities responsible for land allocation. move at least one member of their family into The indigenous farmers' lack of recognition of trade. Despite repeated attempts to encourage the AVV settlers' land rights was a constant the development of markets within the project source of anxiety. Although the project had by- area, only one of these intemal markets survived passed the traditional mechanisms for acquiring more than a few years. The settlers began to real- land, the new settlers tried to create linkages with ize that without the participation of the nearby indigenous groups for assistance with such prac- indigenous farmers and spontaneous settlers, tical problems as determining which plants were they could not create a successful village market. suitable for medicines, identifying suitable burial Efforts by the wealthier settlers to launch trading grounds, and performing certain rituals. For the enterprises in the growing regional center of most part, these linkages were successfully nego- Mogtedo were similarly thwarted. The Mogtedo

41 townsmen were not going to relinquish their themselves and their children-before it could be command of new trade opportunities as they had occupied by the outside immigrants. The result- been forced to relinquish what they considered to ing race for land is reflected in the high percent- be their land rights. To this day, not a single bou- age of fields that have been farmed for less than tique or bar in Mogtedo is owned by an actual or six years (Table 5.1).' former AVV settler. The few AVV settlers that Under the new land law, certain cultivation have become prosperous grain or livestock mer- hamlets (groups of migrants who had moved chants are typically from the same Ganzourgou away from the central village to be closer to their region as the hosts. fields) have requested and received administrative In contrast to the positive to neutral responses autonomy from the traditional villages that first of the settlers at Solenzo and Kompienga, almost granted them the authorization to farm. When mi- all of the 40 male household heads interviewed in grant hamlets break away, the indigenous inhabit- the AVV sponsored settlements indicated that ants consider the separation a profound act of their relationships with the indigenous inhabit- ingratitude. These "breakaways" have in a few ants had been strained. Interviews with 20 study cases provoked physical violence between mi- farmers at Linoghin emphasized that although re- grants and indigenous inhabitants. Moreover, lationships had been very hostile in the past, they there is an increasing tendency for such land dis- were improving. Linoghin farmers cited the par- tericipaiovingspontaneous setters antedindigenou putes to be resolved by the police or another public farmersion the mponaret ed adidgnusettlers administrator. The settlers are well aware that as an important factor created them togethers these conflicts may jeopardize their future claims asanimportantmtogether.rbringing to land and other resources in the region. Indirect Consequences of Reform for Settlement-Related Development The AVV Planned Settlements

Solenzo The early effects of the 1984 reorganization act were quite different in the areas of AVV planned settle- Despite the persistence of many of the formal ment. One result was to strengthen the land tenure mechanisms of the traditional land tenure sys- rights of the AVV sponsored settlers. The law thus tem, the large influx of immigrants in some areas increased the spontaneous settlers' living near the like Solenzo, in combination with the change in AVV planned settlements willingness to be incor- national land laws, has created a concours or com- porated with the sponsored settlers into a single, petition for land. Once the new land tenure laws unified extension and land management program. were announced, many of the settlers with ani- The assisted settlement program that the AVV mal traction and mini-tractors began to clear the started at Rapadama in 1987 was the first attempt largest area possible. The indigenous inhabitants to carry out this type of activity; similar programs then launched a counterattack in an attempt to are now being tried in all of the older AVV planned lay claim to as much village land as possible-for settlements.

Table 5.1 Years of Cultivation of Collective Fields, Solenzo, 1988/1989

Percent of area of fields Percent of number of fields Years Dar/Kie Daboura Average Darn Kie Daboura Average 1 - 3 27 11 21 33.5 33.3 33 4 - 6 24 12 19 27.5 17.5 23 7 - 9 10 34 19 13 28.6 20.5 10+ 38 43 40 26 20.6 23.5 99 100 99 100 100 100 2 Total (000 m ) 2,659 1,646 4,305 Total (Number 69 63 132 of fields) Source: Land Settlement Revieuw.

42 Policy Implications ers were well aware of these limits. Nevertheless, by the time the actual development of infrastruc- Need to Associate Local Populations in ture and social and economic services for the Regulating Land Tenure agropastoral zone was started, more than 200 immigrant agriculturalist families had already occu- Government edicts to promote sustainable land pied the zone. To evict these agriculturalists use practices by mandating more secure land ten- required the intervention of local police and ad- ure rights for settlers are not likely to succeed. The ministrators. Similar problems were encountered original land tenure and agrarian reorganization in the creation of the pastoralist zone at Nouhao. act was less than fully successful because it has Projects to protect the grazing rights of pasto- been insufficiently appreciative of the wider social ralists are unlikely to be successful until the pas- context within which Burkina's local land tenure toralists themselves are effectively organized and systems operate. This local context includes the are empowered by the regional administration to important religious role of the village chef de terre, enforce respect for their boundaries. Equally im- as well as other local customs and the traditional portant to the success of these efforts will be rights of the indigenous people.2 More promising whether the local agriculturalists-especially will be government initiatives that attempt to rein- large agriculturalist households which have di- force the ability of established local institutions to versified into extensive livestock-believe that legislate land tenure rights for immigrants. they will derive any long-term benefits from the zoning effort. Need for Government Intervention to Secure Pasture Rights for Pastoralists Need to Promote Joint Settler-Host Natural Resource Management Administrative intervention will generally be necessary to secure pasture rights for pastoralists. Burkina's PNGT program for village land man- The reason is that there was never a category of agement is a promising model for increasing ten- customary law or a modem droit reglementaire ure security for both settlers and indigenous that recognized pasture rights prior to the Agrar- inhabitants (Chapter 10). In theory, the PNGT ian and Land Tenure Reorganization. Pasture is model promises secure tenure for migrants while considered a natural resource, accessible to all, at the same time protecting the rights of indige- but rights to pasture are always considered sec- nous inhabitants and pastoralists. While this ondary to cultivation rights. For this reason, pas- model is considered to be appropriate for the toralists who have traditionally used the valleys country as a whole, it is especially appropriate for as part of their transhumant cycle are being the valleys where low population densities and pushed further and further south. distance from administrative centers make it vir- We observed a general trend of pastoralists tually impossible to enforce "top down" govern- throughout the valleys becoming increasingly ment land tenure policies that do not have a sedentarized, and demanding more land to farm. strong base of local support. Such cultivation rights are guaranteed under ex- A special set of land tenure problems is present- isting laws, and were generally granted through ed by projects that necessarily involve sponsored the same customary channels used by immigrant settlement-either because of dam related flood- and indigenous agriculturalists to "borrow" ing or the use of sponsored settlements as part of a land. The pastoralists' right to pasture, however, regional planning effort. Special care must be tak- has never been guaranteed. en to involve the local population in the design of One solution has been to designate certain vil- such projects. For their participation to be volun- lage areas for pasture. In addition, the state has tary and not forced, the indigenous inhabitants supported the development of large agropastoral must perceive strong evidence of how such a zones reserved exclusively for pastoralists. Past project will ultimately benefit them. The question experience, however, shows that these land ten- of land rights for subsequent waves of spontane- ure rights will be difficult to enforce. Boundaries ous immigrants who will inevitably be attracted for a large agropastoral zone at Gadeghin were by the new economic opportunities must be dealt traced in the initial planning for the upper Naka- with openly, with the indigenous inhabitants tak- mbe in 1974. Local migrants and indigenous leading the lead in deciding how the land rights for

43 these inhabitants will be handled. Field trips, such given area. We know, however, that most farmers typically as those organized by the AVV to expose Linoghin expanded their initial land grant to encompass a much larger area. Thus, while a three hectare cotton field may be list- settlers to the water harvesting technologies being ed as ten years old, only part of the field may have been developed by PATECORE(Projet Amniagement des farmed for this length of time. Terroirset Conservationdes Ressourcesdans le Plateau 2. While the power and influence of the chef de terre is re- Central)in the Kongoussi region, can be used to portedly decreasing in many areas of the plateau, this does prepare local leaders and village land manage- not discount the central point made here. Efforts to bypass local customs and the rights and roles of traditional author- ment committeeswith what to expect. ities can create a socioeconomic backlash and increase, rather than decrease, local insecurity about land. Notes

1. The recorded "age of field" presented here refers to the period since a particular farm family started farming in a

44 ~~6

Production at the Household and Community Levels: the Centrality of Diversification

Diversification is a goal as well as a product of sources of income is thus not just a luxury, but a successful development in areas of new lands set- necessity that is driven by the nature of the tlement (McMillan, Painter and Scudder 1990). household family unit. As settlers' incomes rise, their desires for manu- For all these reasons, the LSR predicts that the factured goods, services, and a wider range of presence or absence of opportunities to invest in foodstuffs increase. These changes in consump- more diversified crop, livestock, and nonagricul- tion goals and patterns create demands for new tural production activities will be an important products and services, which lead, in turn, to new determinant in settler decisions to immigrate to opportunities for employment. Employment ac- or remain in a given area. tivities may include cultivation of a more diversified package of rainfed and irrigated crops, as Household Organization of Production well as expansion into animal husbandry, small at the Study Sites manufacturing and trade. Diversification is also a product of the success- Most rural households at the study sites are ful reinstatement of "normal" intrahousehold best perceived as overlapping units of produc- production and consumption patterns (McMill- tion and consumption. Household members an, Painter and Scudder 1990). In the absence of collaborate in certain crop and livestock pro- some major change in technology patterns in duction processes, but also engage in agricul- Burkina, settlers know that their long-term secu- tural and nonagricultural activities on their rity is linked to keeping the largest possible num- own. Income earned from the latter is usually ber of people attached to the family farm. totally or at least partially under the individu- Especially important is whether the household al's control. Moreover, there are almost always can retain one or more sons and their families. A different expectations for the spending of pri- critical determinant of the motivation and wel- vate as opposed to collectively produced house- fare of the 51 percent of the population that is fe- hold income. Women, for example, may be male will be the level and type of income they are responsible for purchasing the condiments and able to earn in private crop, livestock, and nonag- sometimes even the meat used in daily sauces. ricultural activities. Since a high proportion of Women may also be responsible for their own this independently produced income is used for clothing and, in a few cases, the clothing and/ condiments and clothing, private female income or schooling of their children. When we talk plays an important role in determining the nutri- about household goals, we are talking about a tional well being and living standards of the en- dynamic balance of these private and collective tire family unit. The development of private concerns.

45 The case study revealed great diversity among son several months beyond the time when cereal the sites in terms of household social organiza- crop activities would normally cease. The aver- tion, land use patterns, patterns of production, age number of weighted labor hours per hectare goals, and constraints. Within the same site, im- for cotton were 386 percent as high as for white portant differences were apparent among ethnic sorghum, and 300 percent as high as for maize groups and even among different geographical (1,520weighted hours per hectare for cotton ver- areas of the same ethnic group. Household orga- sus 394 hours for white sorghum, and 507 hours nization also varied according to length of resi- for maize [McMillan 1983:397]). Female farmers dence at a site. were earning a smaller amount of money from Private crop production, for example, was livestock and off-farm enterprises than women in more important at some sites than at others. An the settlers' home village. At least part of the ex- estimated 20 percent of the gross cash value of planation was, once again, the higher demands crop production of AVV settler households was for women's labor for cultivating cotton; other produced on individual private fields. Female factors included the greater distances from mar- Mossi settlers from areas like Kongoussi and kets, which made it more difficult for settler Koupela were generally less involved in private women to engage in trade. crop production than female settlers from areas At the AVV,this decline in semi-autonomous like Kaya and Ouahigouya.1 sources of income was at least partially offset by Private crop production was less important the fact that expenses customarily under wom- among both the indigenous Bobo and the immi- en's responsibility became the responsibility of grant Mossi at the western Solenzo site. Although their husbands. The male household head, for ex- 21 percent of total cereal production for animal ample, paid all the expenses for family clothing traction farm families at Dar-es-Salaam/Kie was and school fees, as well as purchasing sauce con- from private fields, less than 2 percent of the total diments and meat. cereal production was private for animal traction It was anticipated that we would see a gradual farmers at Daboura (Annex 5, Table D-1). No pri- reinstatement of private crop and livestock activ- vate crop production was recorded for either ities by women and males who are not recog- males or females in the large tractor farming nized as official household heads. The households. contribution of the settlers' private fields has in- Even allowing for our having underestimated deed increased to about 20 percent of net agricul- the phenomenon, it is clear that private crop pro- tural income. Although greater than ten years duction is less important as a percentage of total ago, when the figure was 11 percent (before sub- crop production in the cotton boom west than in tracting the costs of animal traction), this is still the AVV.One possible explanation has to do with less than in the settlers' home village, where an the much higher labor demands on the collective- average of 36 percent of the total CFA value of ly worked cotton fields. A similar decrease in pri- crop production or gross crop income in 1979was vate crop production was observed in the early from private fields (Figure 6.1). days of AVV (Figure 6.1). Research comparing Private crop production was, on average, less Mossi settlers who had been living in an AVV important for both Mossi and non-Mossi settlers sponsored settlement for periods of three to five at Kompienga than at the AVV (Annex 5, Table D- years with families still living in the settlers' 8). The aggregate figures, however, mask signifi- home village showed that, as a whole, the settlers cant differences between the different ethnic cultivated a much smaller percentage of their to- groups. In general, although Mossi women set- tal area as private fields (McMillan 1983; Figure tlers at Kompienga were less actively involved in 6.1). This was attributed, in part, to the higher la- private crop production than female settlers at bor demands for cotton, which was grown almost the AVVor at home, they were nevertheless more exclusively on fields worked cooperatively by the actively involved than women settlers from the entire family. The same study showed that male other ethnic groups. The wives of the indigenous settlers typically worked an average of 58 percent Gourmantche, for example, engaged in only lim- more unweighted labor hours, and female set- ited private cropping and trade. Their participa- tlers 44 percent more, than in the settlers' home tion in collective family farm labor was also much village (McMillan 1983:139),in large part because less than that of Mossi women, and generally the cotton harvest extended the agricultural sea- confined to a narrower range of activities such as

46 Figure 6.1 Subdivision of net crop value of income per household, the AVV sponsored settlements, 1979-1989

AVVSponsored settlements, 1979 Settlers'home village in Kayaarea, 1979

Wivesof malehousehold head 4% Unmarriedchildren Wivesof 5% * household\/ _male head Marriedsons, brothers, 14% their wives,children nmarriedchildren 1.5% 5% /Married sons, brothers, their wives,children 12% Elders(Mothers, fathers and uncles) 5%

AVVSponsored settlers, 1988

ollectivelyworkedX 80%\

Privatelyworked 20%

Source:McMillan 1983. Savadogo,Sanders and McMillan1990.

47 planting and harvesting. In contrast, the Mossi 85), Reardon, Delgado, and Matlon (1992) show women were involved at every stage of crop pro- the importance of diversification for both poor duction except tree cutting and heavy bush clear- and rich households in all three of the country's ance. Yana women were also much less involved major agroclimatic zones (Table 6.1). Especially in private crop production, but were actively in- important for this study was the observation that volved in local trade. local nonfarm activities, including food process- The wives of FulBe herders do not work in ing, cottage industry, and commerce, constituted fields. Women participate in the family's live- over a third (38 percent) of total income in the stock herding, however, and have primary re- higher rainfall, Guinean zone that is most charac- sponsibility for daily milking, milk processing, teristic of the OCP river basins. and the sale of milk and milk products. In most Agricultural immigrants at all the sites were cases the women pastoralists also "own" or have initially concerned with reinstating their basic exclusive rights to a limited number of cattle and food systems at their former-or higher-levels. small livestock. While a detailed study of herding The newly established production systems were practices is not available, it appears that women generally quite extensive, with emphasis on are responsible for some herding and feeding and planting the largest area possible in basic food most milking of the male household head's ani- grains. Once subsistence was insured, settlers mals as well as their own. Women are also expect- typically shifted to a wider range of income pro- ed to compensate the younger household males ducing activities. who herd their animals with income earned from In 1979, AVV farmers stated that moving at the sale of milk and milk products. least part of the family into noncrop concentrated employment was one of their goals for the next The Centrality of Diversification within ten years. When 78 randomly chosen households Household Production Systems were asked what they would do if they had a large sum of money at their disposal, 40 percent The propensity of settler households to diversify mentioned investing in livestock and 45 percent economic activities as incomes rise characterizes wanted to buy merchandise or cereals to trade households throughout Burkina Faso (Reardon, (Murphy and Sprey 1980:73).In 1989, increasing Matlon and Delgado 1988). Using income data investments in livestock and commerce was the from four rainfed harvest years (1981/82 to 1984/

Table 6.1 Income Sources by Zone and Income Stratum, 1981-1985Averages (in percent of total income)

Local, Non-local, Total Crop Livestock non-farm non-farm Other incomein income income income income Foodaid transfers CFA/ALE Sahelian Poorest 1/3 66 17 6 5 3 4 22,566 Richest 1/3 36 17 35 9 1 1 70,545 Overall 48 15 34 11 2 1 42,205 N = 45

Sudanian Poorest 1/3 82 5 8 2 0 3 15,660 Richest 1/3 69 6 16 5 0 3 48,616 Overall 74 6 14 3 0 3 29,295 N = 44

Guinean Poorest 1/3 57 18 22 2 0 3 29,886 Richest 1/3 31 19 48 1 0 0 95,629 Overall 43 17 38 1 0 1 55,261 N = 47

Source: Reardon, Delgado, and Matlon (forthcoming).

48 most frequently stated five year goal of male through disease and theft (Annex 4, Table C-1). household heads at all of the sites. These figures, divided by stock at interview time, were used to calculate the annual rates of loss per LivestockProduction year (Annex 4, Table C-2). The rate of loss is an important factor in a farmer's investment strate- In the absence of a fundamental change in the nagy (Annex 4, Table C-3). Theft became a major tional banking system (like the creation of local problem in the AVV sponsored settlements. Ani- Caissespopulaire d'epargne in certain regions of the mals are also subject to a variety of livestock dis- country), livestock is likely to remain the principal eases. The returns to livestock were calculated in source of investment for all rural households. Small two ways: (a) one, a "low" return scenario, which livestock were especially important for women. considered all feed costs by imputing a value to Increased immigration to the Niangoloko region nonpurchased food, and high losses due to death after 1974 was associated with an increase in the and theft; (b) the second, a "high" return scenar- number of area cattle. Extension records at the AVV io, did not consider imputed costs, but only cash showed a similar increase in the total number of operating costs and low losses (Savadogo 1989c). cattle and oxen in each of the groups of planned set- The low scenario indicates that livestock income tlements (Figure 6.2).No doubt this figure underes- came between CFA 8,500 per ALE (AVV) and timates the actual scope of the phenomenon, given CFA 18,000 CFA per ALE (Solenzo).When the the settlers' reluctance to report the exact size of high scenario was considered, expected livestock their herds. Almost all of the settlers arrived with- income per ALE increased to 14,000CFA in the out large livestock or draft animals. After the first AVVand 25,000CFA at Solenzo. In both cases, the year they were required to purchase two animals returns in Kompienga were lower. on credit from the project. The fifth year of settle- As livestock densities increase, so does the ment was an important turning point as most set- difficulty of supervising animals so that they do tlers completed the final payment on their animal not damage neighboring farmers' fields. These traction equipment. The year usually coincided two factors increase the risk and limit the op- with the sale of their first set of fattened animal trac- portunities for additional income growth tion animals. The combination of ending their first through animal husbandry. Opportunities are medium-term credit and selling off the first set of especially constricted for smaller, limited re- fattened oxen presented farmers with a large block source families, who lack labor as well as the of cash to invest. Many farmers chose to reinvest willingness and ability to gamble on major live- this in further development of non-draft livestock. stock holdings. If farmers have fewer than five cattle, the cattle The high labor demands, and in some cases are generally kept on-farm. Under these circum- high risks, of large livestock production help to stances, the major cash expense for livestock is explain why cattle ownership was unevenly dis- veterinary care, with some payment for supple- tributed across the AVVpopulation. With rare ex- mentary feeding like salt. Larger herds were ceptions, all AVV settlers were required to boarded with local FulBe. The actual cash and purchase a pair of oxen after their first year in the kind costs of boarding cattle are difficult to esti- project. In 1988-1989the average number of cattle mate since many of the costs are conceived as per head of household was 4.2; the average num- "gift" exchange between herder and boarder (see ber of goats and sheep, 11.0 (Annex 5, Table D-6). Chapter 9). Pastoralists may also have the right to These averages mask the fact, however, that a certain percentage of the calves born while the about 25 percent of male household heads (who animals are under their supervision, for example have been living in the AVV planned settlements one young cow every two to four years. When for periods of 10-15years) have no oxen or cattle taken together (including the price of the calves at all (Annex 5, Table D-7). These are farmers given to the herder divided over three years), the who, for one reason or another, had either lost or cash and kind costs of herding livestock can run sold their animal traction animals, and did not over 100,000CFA per year for farmers with large purchase a new pair. herds (over 40) of cattle and oxen. Our study suggests that cattle and oxen owner- Animal husbandry can be risky as well as cost- ship is more uneven at Solenzo (Annex 5, TableD- ly. To evaluate risk, farmers were asked to give 2). At Solenzo, the households with large herds the number of cattle, goats, and sheep lost tend to be those of tractor owners or sedentarized

49 Figure 6.2 Total number of cattle listed in the AVV extension records Number of cattle 1,600

1,200 _ _ _ _

1976 1980 1984 1988 1989 _u Mogtedo-Bombore *** Linoghin Mogtedo

Source:AW UD Extensionrecords.

50 FulBe, with a higher concentration of animals in monthly take-home pay of lower level civil ser- the older settlement at Daboura than in the newer vants. One successful young merchant, son of settlement at Dar-es-Salaam/Kie. an AVV settler at Mogtedo, was earning a The livestock figures at Kompienga are not re- monthly net income of between 30,000 and ally representative because settlement in the 50,000 CFA from a prosperous, active regional area is relatively recent (Annex 5, Table D-9). In trading enterprise in 1987. We estimated the net addition, the high concentration of population income from the store owned by a settler in the in the town makes it difficult to keep livestock study sample at Linoghin to be 300,000 to on the farm; hostile relations with the FulBe, 350,000 CFA during the 12 month study period many of whom were expelled from Ghana in in 1988-1989. 1988, also have made the local FulBe more reluc- Only a small number of the study farmers tant to board settlers' cattle. were engaged in these sorts of lucrative, year round off-farm activities. The majority earned Nonfarm Production Activities much smaller amounts from the sale of consumer goods and food products, production and Settlers generally perceive off-farm production marketing of foodstuffs (including beer), and as a condition for sustainable income growth. local and nonlocal agricultural and nonagricul- Grain merchants operating with working capi- tural wage labor. For this reason, the analysis tal (estimated at 500,000-1,000,000 CFA for larg- distinguishes the first category of settlers as er traders) at Mogtedo and Linoghin could having "secondary occupations" and analyzes easily average monthly net profits of 30,000- their income patterns separately (Table 6.2). In- 50,000 CFA in addition to earnings from crop come earned from small scale commercial and enterprises. This amount is well above the wage activities was less important as a source of

Table 6.2 Mean Income from the Sale of Handicrafts and Trade, 1978 and 1988/89

Male head of household First Second Developing Entire year year farm Full farm Average family 1979 Linoghin' - 1,899 6,588 13,834 6,227 Mogtedo-Bombore' 495 6,627 19,670 13,791 9,715 - Mogtedo V3b Case Study With secondary occupation(2) - - - - - 193,792 Without secondary occupation(6) - - - - - 17,375

1988/89 Linoghin With secondary occupation (2)' 375,000 Without secondary occupation (18) 9,461 Bombore With secondary occupations (2) 122,500

Without secondary occupations (1 8 )d 1,921 Mogtedo (20) - - - 19,695 Mogtedo V3 With secondary occupations (1) 250,000 Without secondary occupations (19)e 18,118

-Not available. ()indicates number of households in this category. a. Source: Statistical Service Survey Research on subsample of 78 farm families in the farm monitoring program (Murphy and Sprey 1980:73). b. McMillan 1983:417. c. Includes pension of retired soldier and estimated annual income from one farmer's boutique. d. Does not include large incomes eamed by two butchers. e. Does not include income from one farmer with secondary occupation.

51 Table 6.3 Average Non-farm Income from Different Sources for Individuals Reporting Income from the Source, 1988-1989 Linoghin Bombore Mogtedo MogtedoV3 Solenzo Kompienga Malehead of household 1. Unprocessed agric. products 64,120(l) - 42,000(1) - - 23,167(3) 2. Processed agric. products - - - - 24,000(1) 8,425(2) 3. Forest and water products - - - - - 87,000(1) 4. Meat and animal by-products 28,500(1) 122,500(2) 38,650(3) 25,000 - 107,667(3) 5. Commerce 300,000(1) - 50,250(2) 14,000(3) - 81,000(3) 6. Handicrafts 21,300(2) 6,070(6) 1,950(1) 6,950(2) - 53,167(3) 7. Gold - 2,000(1) 36,166(3) 27,721(7) - - 8. Services 4,000(1) - 5,000(1) 23,700(2) 41,250(2) 50,000(1) 9. Agriculture Labor - 20,000(1) - - 10. Nonagriculture Labor 450,000(1) - _ 250,000 - 11. Mill - - - - 550,000(1)

One womanper family 1. Unprocessed agric. products 9,050(2) - - 74,200(2) - 10,437(10) 2. Processed agric. products 26,043(13) 3,833(5) 3,186(7) 6,642(14) 29,300(6) 6,361(11) 3. Forest and water products 30,050(2) - - - - 3,000(1) 4. Meat and animal by-products - - 37,500(1) 5. Commerce 2,200(1) 6,000(1) - _ 350(1) 6. Handicrafts 7,650(1) 8,150(1) - - _ 17,000(1) 7. Gold 500(1) 19,69102) 13,914(16) - - 8. Services - - 3,500(1)

- Not available. () indicates how many men and women provided non-zero numbers. Twenty men and twenty women were interviewed in each of the AVV planned settlements; 36 men and 36 women were interviewed at Solenzo. Means are taken over these numbers. Source: Savadogo 1989a.

Table 6.4 Percent of Women Indicating Income from Off-farm Employment and Livestock Produc- tion and Average Income from these Sources for the Entire Sample, 1988/89(per woman)

Off-farm income Net livestock incomea N Percent CFA Percent CFA Linoghin (20) 80 21,331 20 201 Bombore (20) 35 1,690 45 486 Mogtedo (20) 60 14,805 45 1,887 Mogtedo V3 (20) 95 23,392 55 3,865 Solenzo (36) 17 6,833 0 0 Kompienga (25) 64 9,774 8 442 N indicates sample size. a. Net income on livestock includes deductions for animal losses by death and theft. The low figures reflect both this and the fact that only a few women in each sample had sold livestock at the time this initial analysis (August 1989) was conducted. Source: Savadogo 1989a,c.

52 Table 6.5 Average Income by Income Source for Male Heads of Household and Women in the Study Sample, 1988/89(FCFA)

Net crop Net livestock Net off-farm production production (w/osecondary Site (perALE) (per ALE) occupations) Total Malehead of household Linoghin 61,315 21,897 9,461 92,673 Bombore 54,988 6,797 1,921 63,706 Mogtedo 51,072 14,764 19,695 85,531 Mogtedo V3 55,101 5,415 18,118 78,634 Solenzo manual 30,965 2,445 2,958a 36,368 traction 78,784 8,494 2,958 90,236 tractor 294,422 44,802 2,958 342,182 Kompienga 78,784 8,522 39,495 126,801

One wonan per family Linoghin 19,336 201 21,331 40,868 Bombore 11,145 486 1,690 13,321 Mogtedo 16,976 1,887 14,805 33,668 Mogtedo V3 17,385 3,865 23,392 44,642 Solenzo 12,687 0 6,833 19,520 Kompiengab 12,633 -442 9,774 21,%5 Note:Net crop and livestock production are calculated per ALE (adult labor equivalent) for the collectively worked fields and supervised livestock under the responsibility of the male head of household. These herds and fields are generally used for the main food and cash needs of the family unit. a. The study did not distinguish off-farm income by gmups with different technologies. b. Ethnic groups differed substantially in terms of women's economic activities. Sources:Savadogo 1989c;Savadogo and Sanders 1989. cash income or of total income for the male years after the first planned settlement was household head than it was for women (Tables created at Mogtedo in 1974,a considerable num- 6.3, 6.4, and 6.5),3 married sons and brothers, ber of settlers (five in just one of the six villages we and unmarried male and female children studied) had profitable secondary occupations. attached to households. This trend did not last, however, and today the bloc of Mogtedo, and the adjacent bloc of Mogte- Factors that Affect the Development of do-Bombore have no permanent stores and only a Off-farm Employment at the Study Sites few settlers can be classified as having secondary occupations. Several factors explain the early rise We found substantial differences among the AW and fall of profitable secondary occupations. planned settlements in both the total level and the These include the termination of construction of sources of off-farm income. These activities were planned settlements at Mogtedo-Bombore in affected by settlers' proximity and "social access" 1981;gradual depletion of the surrounding wood to area markets, as well as by the presence of alter- resources and an associated drop in income from native opportunities in commercial agriculture. commercial woodcutting; the substantial decrease Gold mining was an important factor affecting in cotton production and conversion to commer- off-farm employment at one of the study sites. cial cereal production that is sold in the larger regional markets away from the villages; and a The AVV PlannedSettlements at Mogtedoand gradual decrease in the number of salaried exten- Mogtedo-Bombore sion workers who provided farmers with a steady market for their services and products. EARLY INCREASEIN AND THEN DEMISEOF PROFIT- As early as 1979,the most prosperous AW set- ABLE SOURCESOF NONFARMINCOME. Within five tlers at Mogtedo and Bombore were planning to

53 invest in trade and special services for the larger The AVV Planned Settlements at Linoghin Mogtedo market that was 20 to 30 kilometers from their villages and located on the paved A different scenario was recorded in the AVV highway linking the area to Ouagadougou. These sponsored settlements at Linoghin largely be- early efforts to diversify into area trade were ac- cause the settlers developed a dynamic market on tively blocked by the indigenous inhabitants, the paved road to Ouagadougou. The importance however. Although they were from the same of these activities is not reflected in the interviews Mossi ethnic group as most settlers, the indige- with official heads of households. Instead, it is nous inhabitants resented the AVV land tenure more accurately reflected in the higher income policies. In the six AVV villages at Mogtedo, only levels of men who were not household heads two AVV settlers' children and one male house- (married sons, brothers, and unmarried children) hold head who had immigrated from outside the and of women. Even the 300,000 CFA in annual region, had official places in the Mogtedo market income recorded by one household head (Table in 1989, and only one of the original settlers' sons 6.3) was actually earned by a married son. The fa- had ever developed a successful regional trading ther simply put up the money for the initial in- enterprise. The successful merchant has since left vestment. the Mogtedo area to develop trade in another, Eighty percent of the 20 women interviewed in more hospitable area. The one exception-a pros- the farming systems survey at Linoghin recorded perous grain merchant now living in the Mogte- income from off-farm employment, none of it do sponsored settlements-is a native of the from gold (Table 6.4). Most of this income derived surrounding region, and only came to the project from selling cooked food, homemade beer, and in 1988. gathered forest products. The average off-farm DISCOVERYOF GOLD. Gold was discovered near income for all the women interviewed at Li- the AVV planned settlements at Mogtedo in late noghin was at about the same level as income re- 1987. Since then, panning for gold has become an corded at Mogtedo V3-the village where important revenue source in certain villages. In women gained the most from the direct and indi- addition to the direct benefits of the gold, the rect benefits of the 1988-1989 gold rush (Tables 6.3 flood of gold miners created a market for meat, and 6.5). Linoghin is also the village where exten- water, and raw and cooked food. The largest gold sion programs to promote women's weaving mine and gold market is at Nobsin, near villages have been most successful. Although it was a dis- V3 and V4 in the AVV sponsored settlements at mal failure for most of the first ten years that it Mogtedo; a second smaller gold site and gold was promoted, a combination of increased de- market is located near village V3. Lesser sites are mand for handwoven cloth, proximity to Ouaga- being prospected throughout the region. dougou, and increased free time has made hand Twenty-eight of the 40 women that we interweaving very popular. 4 In 1989, 21 female spon- viewed at Mogtedo earned income from gold- sored settlers, 8 female spontaneous migrants, mining. The range of recorded income was and 2 female hosts belonged to the Linoghin wo- large-from a few thousand to over 50,000 CFA men's weaving cooperative (personal communi- (Table 6.3). The gold mines also created new op- cation, F. Guira, DEPC, AVV,April 1990). portunities to sell food products and water. The net result, which is certainly underestimated in Solenzo our figures, was a substantial increase in the income of women farmers in the Mogtedo villages Only three household heads in the Solenzo vil- nearest the sites. Some of the indirect effects of lage reported earning off-farm income (Table 6.3). this higher income could be seen in the improved Among the women, the most noticeable source of quality of women's clothing on Muslim and off-farm income was from sales of homemade Christian holidays, in the volume of new metal beer, from which six women earned an average of dinnerware, and in women's purchases of locks 29,300 CFA ($98) (Table 6.3). and other improvements for their houses. No We suspect that these figures underestimate similar increase was observed in women's in- the importance of off-farm income in the village come from nonfarm employment in the more iso- economy. A high proportion of the income is lated Bombore planned settlements except in V2 probably earned by the married and unmarried and V5, villages that had functioning markets. brothers and sons we did not interview. The exist-

54 ence of several privately owned grain mills, well year round secondary occupation, at least one stocked, sturdily built boutiques, and thriving other adult male must be available to help with village markets at Dar-es-Salaam, Kie, and Da- heavy farm labor. Two successful merchants- boura suggests that at least part of the high in- one a kola merchant and the other a mechanic- come from crop production is being reinvested were forced to abandon their trade when they lost into the development of commercial enterprises. key family members (a married brother in one Some of the large income disparities we ob- case, an older unmarried son in the other) served between farm families with manual tech- through out-migration. nology and those with animal traction and tractors may be offset by off-farm employment. Linkages among Diversification, Agricultural Four of the six manual male farmers with low Intensification, and Sustainability crop and livestock incomes at Kie/Dar-es-Salaam have secondary occupations (masonry, carpentry, The AVV Planned Settlements trade), and one engages in dry season irrigated gardening. Only two of the six are recent mi- DIVERSIFICATIONAND NEW TECHNOLOGY. Diversi- grants (one in 1982 and one in 1985). None of the fication need not necessarily be at the expense of seven male farmers who farmed without animal additional investment in intensive farming or traction or tractors at Daboura indicated second- higher levels of crop productivity. In 1988/89 the ary occupations, however, except for two FulBe net CFA value of crop production per ALE was 12 households who (not surprisingly) indicated percent higher for the study farmers at Linoghin herding. than at Bombore, and 20 percent higher than for the Mogtedo households. Ten years ago there was Kompienga no visible difference between the blocs in terms of net productivity per ALE (Murphy and Sprey Greater opportunities for dry season irrigated 1980:69). gardening and trade were the principal reasons Linoghin, the planned settlement group with cited for moving by the male and female settlers the greatest actual and future potential for diver- who immigrated from the AVV to Kompienga. To sification into activities other than mining, had date, the state has not yet begun to develop the ir- average per household cash and kind expenses rigated perimeter that is planned for the lake for production that were 24 percent higher than at bank. Nevertheless, both male and female settlers Mogtedo, and over 200 percent higher than at benefitted from high levels of off-farm employ- Mogtedo-Bombore and Mogtedo V3 (Savadogo, ment in trade and services (Table 6.3). In 1988- Sanders and McMillan 1989:43). Seventy-nine 1989, these revenues were still being fueled by a percent of the study farmers used some com- large contingent of wage laborers involved in tree pound fertilizer on their fields at Linoghin, 85 clearance. As these wage earning opportunities percent at Mogtedo, but only 45 percent at Mogt- decreased, so did the profitability of commercial edo-Bombore (Savadogo, Sanders and McMillan enterprises. It is doubtful that the earlier high lev- 1989:34). els of off-farm employment can be regained un- Linoghin also is the planned settlement where less other changes, like the development of there was the greatest interest in the development irrigation, occur. of soil conservation measures like diguettes (small dikes), manure pits, and straw.5 To assist farmer Household Level Factors That Affect the groupements with the initial development of Development of Off-farm Employment diguettes, the AVV-UPl provided trucks. Even after the AVV program ended, some farmers con- In addition to village level factors, a variety of tinued to rent trucks on their own in order to household level factors affect the distribution of expand their diguette networks. One family in- off-farm employment within a given village. vested 37,500 CFA (the equivalent of the mean ex- Some of these are cultural and include local atti- penditure per household on fertilizer, insecticide, tudes toward women's participation in trade and seed, and wage labor at Linoghin for 1988/89) on handicrafts. The size of the family labor force is installing stone diguettes (personal communica- also important. Settlers repeatedly emphasized tion, F. Guira, DEPC, AVV, 19 April 1990). The that in order for someone to engage in a profitable highest level of interest in stone diguettes was in

55 the most isolated Linoghin village, V6, which is Settler "Step" Migration from the AVV 20 to 25 kilometers from the Linoghin market. Al- to Kompienga though too distant from the market for many family members to engage in trade easily, the V6 The importance of diversification in the lower settlers are close enough to Ouagadougou to de- rainfall, northern river basins was brought to velop profitable off-season crops-in this case, light by the immigration of some of the most suc- fresh corn. In 1989 the V6 farmer who sold the cessful AVV settlers to Kompienga. The Kompi- greatest amount sold 20 carts of fresh corn; the enga census identified 55 male heads of lowest amount was 4 carts. household who were former AVV settlers in the The dynamism of the Linoghin settlers and planned settlements at Mogtedo, Mogtedo-Bom- their commitment to diversification is reflected in bore, Rapadama, and Mankaraga. More than 30 other areas as well. Average household expendi- families left the bloc of Mogtedo, cutting all ties tures on education-especially secondary educa- with the AVV.Overnight, the number of children tion-were much higher at Linoghin than in the attending the AVV Mogtedo primary school other settlements. The percent of debt repayment dropped by a third. The actual number of families on short-term credits has generally been the high- that left was even larger, but some did not appear est of all the settlement groups studied (Figure in the official figures since their moves were clas- 6.3). In comparison with the other blocs, Linoghin sified as "dissociations," with only part of the also has had one of the most successful records family leaving. for the development of groupement managed en- Two of the households who left V3 at Mogtedo terprises, the income from which has been used were very wealthy, with large cattle herds (one to maintain basic infrastructure like roads and with 25 animals and another with over 50) as well wells. Especially remarkable was the Linoghin as sizable incomes from off-farm employment. The groupement's development of the first kindergar- male heads of these wealthy households had lob- ten in the region. Settler turnover has been very bied intensively to get the planned settlement low at Linoghin. Although a number of wealthier group to support development of a central market settlers have moved into full time commerce on the road linking the planned settlements to the based in Ouagadougou, they have in almost ev- town of Mogtedo. One of the men had even of- ery case left some member of their family respon- fered to subsidize the price of grain sold at the sible for the farm. market in order to attract outside merchants. An- In contrast to Linoghin, the more isolated, less other settler had been recognized as the province's diversified neighboring blocs of Mogtedo and leading "progressive farmer" at the 1987 agricul- Mogtedo-Bombore have experienced high turn- tural fair and had received over $1,000 worth of over. Of the 255 households who acquired AVV prizes. The six other families who left V3 were all farms at Mogtedo, only 58 percent (148) are still better off than average in terms of the size of their living there. Eighty-one percent of the 345 house- livestock herds and past crop production levels. holds that have had official rights to an AVV farm The settlers' main motivation for leaving was to at Mogtedo-Bombore were still living there in move to an area with greater opportunities for dry 1989. Other factors, like the settlers' lack of un- season irrigated farming and regional trade. 6 derstanding about the project when it first start- "Step" agricultural migration, like that of the ed, undoubtedly contributed to the higher rates AVV farmers to Kompienga, is not new. The dete- of settler turnover in the six Mogtedo villages rioration of opportunities in an area of in-migra- which were created five years before Mogtedo- tion has always encouraged agriculturalists to Bombore. Nevertheless, settlers' perception of the move. This sort of serial migration has been an in- lack of local opportunities for developing nonag- tegral feature of agricultural migration on the ricultural sources of income contributed to the Mossi Plateau for centuries. Thus, it is of little sur- high rates of settler turnover in the more isolated prise that many of the settlers from the Sanmaten- AVV blocs of Mogtedo, Mogtedo-Bombore, Kai- ga (Kaya) region who left the AVV for Kompienga bo South, and Kaibo North, and to lower rates of had lived in other areas before coming to the AVV. turnover in less isolated blocs like Linoghin and As many as 50 percent of the Mossi living in Kom- Bane, which offer greater opportunities to earn pienga today had lived for periods ranging from income from trade and services (see also Annex 1, two to ten years in other areas of the OCP valleys Table A-3). or plateau before migrating to the town.

56 Figure 6.3 Rate of Credit Reimbursement, AVV Planned Settlements

Percent of repayment 100

40~~~~~~~~~~~~~~~~~~~~~

1~~~~~~~~~~~~~~~~ 0

Year

E Linoghin A Mogtedo U Mogtedo-Bombore

Sourc: AV, Credit Agricole, March 1990.

57 The cost of this step migration from one OCP ied were purchased with cash rather than credit, as area to another is not only economic, but institu- is almost always the case in the AVV sponsored tional and environmental as well. Out-migration settlements. of wealthier farmers, as we have seen at Mogtedo Although intersectoral linkages between agri- and certain villages at Rapadama, means the loss culture and the service sector in Solenzo may be of cash resources earned during the early period considered weak in comparison with the AV, the when crop production at a settlement site is usual- settlers have shown a strong tendency to reinvest ly most productive. These cash resources could in social infrastructure. The Solenzo area, which is have been reinvested in the development of more one of the top cotton producing areas in the coun- intensive, sustainable environmental management try, also boasts one of the highest rates of agricul- like the diguettes. Another incalculable cost is the turalist investment of ristournes (returns) from loss of a first generation of settler leadership. The farmer managed cotton markets into village commercial farmers who left Mogtedo for Kompi- schools and health facilities (personal communica- enga included some of the most ambitious leaders, tion, Cellule Suivi-Evaluation, CRPA du Mouhoun, who had worked closest with the AVV administra- March 1990). tion. Many of these leaders had accumulated con- As the population density increases, diversifica- siderable investments. tion is likely to become increasingly important at The cash resources that the former AVV settlers Solenzo. One of the goals of the PNGT village land used to purchase improved housing or rental prop- management program in the western cotton boom erties and to rent tractors and hire labor at Kompi- areas is to stabilize new land clearance by restrict- enga could have been reinvested in the ing settlers and hosts to a fixed land area registered development of the AVV sponsored settlements on a village map. In principle, the system has much and the surrounding regions from which these rein common with the AVV system that gave farmers sources were gleaned. Instead, farmers responded a fixed 10 or 20 hectare farm. How farmers respond to what they perceived as the lack of future oppor- to this circumscription of their land area-by inten- tunities for high levels of income growth by moving sifying production with more manure and fertilizer on. In contrast, the rate of settlers leaving the project (as is happening in the AVV planned settlements at for agricultural production elsewhere has been Linoghin), or by moving to a less densely settled much lower at Linoghin, an area that is basically the area in the Mouhoun or further south to the Leraba same agriculturally, but has greater possibilities for or Comoe Basins (as happened in the planned set- diversified cropping and nonfarm income growth. tlements at Mogtedo)-will be closely related to the settlers' perception of future income earning oppor- Solenzo tunities. We predict that in the absence of local opportunities to develop off-farm employment, The technical package for cotton has demonstrated tractor farmers and ambitious animal traction farm- its steady reliability in the agroclimatic zone suited ers with large family labor forces are likely to move to this crop. As a result, farmers show a strong to areas where land is easier to acquire and/or willingness to invest in fertilizer and labor saving where opportunities for diversification are greater. technology to increase crop productivity per ALE. The success of this migration will depend, however, One reflection of the settlers' perception of the on local attitudes in the receiving area and requires long-term opportunities in rainfed agriculture in further investigation. the region can be seen in their patterns of rural financing for crop production technology. National Policy Implications research shows that the Solenzo zone has one of the highest rates of cash purchases of agricultural The important role of economic diversification has material (personal communication, Cellule Suivi- a number of implications for policy at both the na- Evaluation, CRPA du Mouhoun, March 1990). Al- tional and local levels. though local credit has tended to stagnate, the area has experienced an increase in animal traction National Level Planning Issues equipment, such as plows and carts. This agrees with the results of our economic survey, which Any long-term policy to encourage the dynamic showed that a high percentage of the plows, and development of Burkina's rural sector will neces- almost all of the carts, owned by the settlers stud- sarily be linked to the creation of a certain

58 synergism between agricultural and nonagricul- * EARLY EMPHASISON FACILITATING REGIONAL tural activities. The establishment of small enter- ECONOMIC AND SOCIAL INTEGRATION. Early plan- prises to transform agricultural products, for ning is needed to incorporate the needs and con- example, could encourage the development of cerns of indigenous inhabitants into a project in transportation infrastructure, facilitating the order to minimize host-settler conflict and the movement of agricultural products between vil- negative impact that conflict can have on diversi- lages and regions. fication. As food processing tends to be a female activi- * UPGRADE EXISTING MARKET AND ADMINIS- ty, this activity is especially important for pro- TRATIVE CENTERS AND CREATE NEW CENTERS. In grams aimed at raising women's incomes. A view of the costs of creating new administrative national program to encourage the development centers, attention should be given to upgrading of food processing for major and secondary ur- existing regional centers when they exist. As ban markets could thus be a means of redistribut- more and more immigrants arrive, however, new ing revenues to women and encouraging centers will emerge. The experience of the AVV women's participation. shows that it is not always easy to predict where The next question, then, is how best can na- centers will grow as new lands settlement in- tional, regional, and local governments facilitate creases. For this reason, we recommend that out- the development of such activities? A policy en- side aid to develop basic roads and infrastructure vironment that is conducive to investment is es- should be phased in over a 10-15 year period, sential. Tax concessions and temporary thereby increasing the chances that these roads subsidies might be considered to encourage pri- can be used to reinforce the development of dy- vate businesses to relocate to some of the high namic, stable centers. potential valley areas. Agricultural fairs and "development weeks" can be used to educate Notes farmers and local administrators about the potential of such programs. The products pro- 1. Both male and female settlers from the Kongoussi region duced must also be moved between regions and said that this was changing, however, and that their women had tended to increase the amount of land they were farm- villages. It is here that national and regional pol- ing privately with longer periods of living in the valleys. icies are likely to have their greatest impact-by 2. A comparison of our results with the results of the eco- either encouraging or discouraging local food nomic survey conducted by the CRPA suggested that we processing and trade between regions. A nation- might have underestimated the incidence of non-traction al program to improve infrastructure in the livestock. In view of the importance of livestock in future plans for diversification in the area, this is a topic that west, as was discussed in Chapter 4, would en- needs additional research. courage greater regional specialization of pro- 3. One should note that the woman chosen for the interview duction. Under these circumstances, certain in each household was usually the most "economically ac- areas like the higher rainfall river basins in the tive" woman in the family. Computing statistics based on west would tend to become specialized in com- these women tends to inflate data. The data can be inter- preted, therefore, as expected maximum productivity per mercial cereal and crop production as well as woman. cotton. Other areas like the lower rainfall north- 4. This increased demand was partially due to the revolu- ern river basins that currently produce large tionary government's requirement that officials wear cloth- amounts of cereal for urban consumption are ing made from locally produced cotton for work and likely to specialize in intensive livestock raising, official ceremonies. The progressive relaxation of this mea- likelym handicraft production, andvegetable sure is expected to lead to a decreased demand. nonfarm handicraft production, and vegetable gardening for the cities. This sort of regional 5. Wedo not feel that the high levels of interest of farmers in the new technology are fully reflected in the official project specialization demands an overarching policy records. The extension records list only the number of hect- for augmenting basic transportation. ares covered with diguettes that were constructed as part of the UPI program; our observations consider the diguettes created under project auspices as well as those created Local Level Planning Issues through personal initiative. 6. Although economic factors appeared to have primary im- * SITE SELECTION. In view of the importance portance, the final decision by individual families to go or that settlers attach to developing diversified stay was also influenced by such factors as intravillage so- sources of income, policymakers should concen- cial conflict and strong social ties with other settlers who trate on less isolated areas, where the opportunities for diversification are greater.

59 Farming Systems

The term "farming system" encompasses the FarmingSystems in Areasof Unassisted different management techniques, crop rota- Spontaneous Settlement Niangoloko and tions, technologies, labor systems, land tenure Kompienga norms, and macroeconomic policies that affect local crop and livestock systems. The farming The first two study sites-Niangoloko and Kom- system is a subsystem of the broader house- pienga-are examples of how unassisted settle- hold production systems discussed in Chapter ment tends to reproduce extensive cultivation 6. A specialized study of household farming systems with little attention to sustainability and systems is justified by the fact that production high potential for social conflict at higher popula- of crops and livestock is likely to remain the tion densities. principal source of cash revenue and food for rural households living in the river basins Niangoloko (McMillan, Painter and Scudder 1990). The case studies illustrate the potential of Crop production in the Niangoloko region focuses new lands settlement in the valleys to increase on the extensive production of cereals. While cot- total production and, in so doing, to raise total ton production has undergone a tenfold increase agricultural income levels and settler welfare. since 1985,it is largely confined to a small number This increased production is engendering of villages in the Leraba Basin. Moreover, the chief wider regional effects of increased on- and off- stimulus behind the sudden increase in cotton pro- farm employment, livestock production, and duction (from 50 tons in 1985-86to 500 tons in cereal production. 1987-88)has been the recent return migration of The present chapter provides a brief over- Burkinabe from Cote d'Ivoire rather than any new view of farming systems at the study initiative on the part of the Burkina govemment. sites. Detailed examination of the study results Despite this recent growth in cotton produc- is presented in Annex 5, in K. Savadogo (1989a, tion, the area's extension services have remained b, c, d) and Savadogo, Sanders and McMillan limited; four agents, including the area director, (1990). This overview is followed by a discus- cover 15 villages. Extensive livestock production sion of key factors that contributed to or de- and trade continue to be the major sources of set- tracted from the development of sustainable tler income growth. crop production methods, and some of the Farmers in the region traditionally farmed a implications that this has for future policies. field for only five years before leaving it fallow for an average of 20-25 years (Nana 1989a). An abundance of reserve fields made this lengthy

60 fallow period possible. Traditional cultivation Basin region had not benefitted from any special techniques included a system of mounding that donor funding for crop and livestock develop- follows the water flow from torrential down- ment. Crop production continues to be heavily pours-a technique that reduces the effects of soil focused on subsistence food production. Rates of erosion. While two of the study villages showed adoption of animal traction and fertilizer were visible erosion, no special technique to offset its among the lowest in Gourma province. The sam- effects had been developed, as farmers tend to re- ple failed to reveal any use of fertilizer (Table 7.1). spond by abandoning their fields at the first sign The Burkina government was correct in assum- of erosion. ing that rapid spontaneous settlement would ac- If immigration continues at present rates, clear- company construction of the dam. For this ing new fields (i.e., expansion of extensive farm- reason, it was adamant about preparing a coming) will no longer be an option in some villages, prehensive regional development plan to rein- especially those near Niangoloko, the major ad- force the area's weak infrastructure and extension ministrative and market center, where immigra- services. To date, however, no special extension tion rates have been the highest. At that point, services have been developed except for the relo- farmers will undoubtedly want to spread the vil- cation of four villages from the flood basin and lages' field areas into adjacent classified forests. for another nine villages on the basin's edge that The most immediate threat to the forests, howev- are losing all or part of their fields because of er, is from pastoralists. Cattle, in the area of the flooding. The relocation program, administered Classified Forest of Toumousseni, for example, by the AVV and DGMOK (Direction Generalede la are estimated at 10,000 head. One solution to the Mattrise d'Ouvrage de la Kompienga), includes a va- problem is to introduce experimental programs riety of extension programs whose administra- like the Classified Forest Project at Toumousseni, tion was supposed to revert to the regional offices which helps local pastoralists and agriculturalists of the relevant national ministries after 22 to develop sustainable sources of revenue from months. 1 The town of Kompienga is slated to forest products. Projects like this can create a benefit from a series of special development pro- group of local people with a vested economic in- grams that includes irrigated gardening and terest in forest preservation. Over the long run, flood recession agriculture on the lake banks, as however, this sort of multiple use forest project well as special infrastructure and extension pro- must be combined with the development of other grams to promote intensive farming (Agrotech- sectors, including crop and livestock production. nik 1989). In the meantime, there are no special programs for settlers living in the town, and very Kompienga limited services for villages other than those in the AVV program. Kompienga is another example of extensive agri- Despite significant differences in the household cultural systems that develop in areas of unassist- organization of production, and in the relative im- ed spontaneous settlement. Prior to 1982, when portance of agricultural (crop and livestock) ver- construction started on the road linking Fada to sus nonagricultural (trade) sources of income the proposed dam site, immigration to the Komp- among the three principal ethnic groups (the in- ienga valley was very much like immigration to digenous Gourmantche, and immigrant Mossi Niangoloko. As in Niangoloko, the Kompienga

Table 7.1 Mean Input Costs for the Male Head of Household for Different Ethnic Groups, Kompienga, 1988/89 (CFA)

Ex-AVV Non-AVV Yana Gourmantche All Number 7 6 6 6 25 Fertilizer, urea, phosphate, manure - - - Insecticides - - - Hired labor (cash and kind) 22,171 9,255 38,875 3,250 18,539 Hired equipment (cash and kind) 2,344 1,000 3,125 5,475 2,344

- Not available. Source: Savadogo, Sanders, and McMillan 1989.

61 and Yana),the crop production systems of all three workers. In any case, a large percentage of the groups are uniformly extensive. surrounding area is rapidly approaching the fifth What distinguishes the Kompienga farmers year of continuous cultivation-a time when the from spontaneous settlers at the other study sites indigenous Gourmantche farmers would begin is their high use of purchased labor (Table 7.1). letting the land lie fallow under the system of They spent an average of nearly 20,000CFA (in- crop rotations they have followed for many gen- cluding cash and kind expenditures) on labor as erations. This type of lengthy fallow will not be opposed to 5,000-10,0000 CFA in the AVV and an option for the new migrants, who have occu- less than 1,000CFA at Solenzo. This high level re- pied almost all the good rainfed land for a ten ki- flects the substantial cash resources enjoyed by lometer radius around the settlement. the average Kompienga settler. These settlers are not typical low resource farmers. Attracted by the Effortsto Promote Higher Yielding, Sustainable long-term possibilities of the area for irrigated Farming Systems farming and trade, former AVVsettlers were willing to pay from 100,000-150,000CFA for truck The Solenzo and AVV sites are good examples of transportation to bring their families to Kompien- the two most important models of government ga. At the site, 52 of the 55 former AVV settlers assistance to agriculture in the river basins. purchased their house sites from departing dam workers at an average cost of 100,000CFA. None The SolenzoAssisted vs. AVV SponsoredCrop of this cash reserve was being spent on fertilizer. ExtensionModels The only study farmer who used any fertilizer during the 1988/89 crop year had won the fertil- The crop extension services offered to settlers in izer as a prize in a "progressive farmer" contest at the Solenzo area make up one model for assisted the AVV. spontaneous settlement. The program that is pro- In 1988, the total production in kilograms and moted at Solenzo was developed over several de- net CFA value of production per worker was cades-first by the French colonial government, highest for the indigenous Gourmantche and for and later by the semiautonomous Burkinabe cot- the Yana farmers who had been living at the site ton agency, SOFITEX.Although the basic pack- for several years (131,313CFA/ALE and 125,841 age focuses on commercial cotton production, it CFA/ALE, respectively for net value). Net crop includes a balanced mixture of crop rotations, fer- income and production were lower for the Mossi tilizer, and cultivation techniques that, used to- settlers, both former AVV and non-AVV (64,366 gether, was expected to reduce the long-term CFA/ALE and 72,145 CFA/ALE, respectively), deleterious impact of cotton production on area most of whom had been living at the site for only soils. In recent years the program has also pro- one to two years (Table 7.2). The immediate pros- moted the use of high levels of manure to offset a pects for sustainable income growth in livestock decline in soil organic matter. and crop production look dim however. When the AVV was created in 1974, the pro- The majority of settlers now farming in Komp- gram advocated many of the same agronomic ienga moved there in 1987 and 1988. Some of themes that the cotton programs in Burkina's their fields had been farmed earlier by dam southwest were promoting. Farmers in both re-

Table 7.2 Crop and Livestock Production per Adult LaborEquivalent (ALE),Kompienga, 1988/1989

Mossi Ex-AVV Non-AWV Yana Gourmantche All Average family size (no. residents) 13 (7) 15.5 (6) 7.3 (6) 6.7 (6) 12.9 (25) Average family labor force (ALE) 5.9 6.5 4.8 3.8 5.2 Cereal production (kg.): sorghum-maize per ALE 988 1,129 1,666 1,911 1,357 Net crop income (CFA per ALE) 64,366 72,145 125,841 131,313 91,616 Net livestock income (CFA/ALE high scenario) 19,116 102 a 12,155 2,578 8,522

() indicates number of households a. verified 3/26/92 Source: Savadogo, Sanders, and McMillan 1989, and unpublished analyses of field data.

62 Table 7.3 The AVV Index of Adult Labor project, they signed an agreement pledging to fol- Equivalents (ALE) by Sex and Age low the recommended program for intensive farming. Their participation was closely moni- Age Male Female tored. During the first five years, each planned 0-12 - - settlement included one male extension agent for 12-15 0.5 0.25 every 25 settler households and one female agent 15-55 1.0 0.75 for every 50. Although the project administration 55-65 0.5 0.25 only rarely exercised its right to expel farmers for not adhering to the proposed technology and land use techniques, the threat of expulsion, or of gions were supposed to cultivate a proposed ro- being refused access to short-term credit, carried tation of cotton, cereals, and legumes that would, some weight. In contrast, the adoption of the new in combination with the recommended levels of cotton technology package was voluntary and mineral fertilizer, allow them to attain higher, conformed with the individual household's fi- sustainable yields. Both programs focused heavi- nancial situation and goals in Solenzo. ly on commercial cotton production, which pro- Climate and soils mark a third difference be- vided the means for reimbursing short-term tween the two programs. The AVV planned set- credit for inputs, and longer-term credit for ani- tlements are located on the northernmost fringe mal traction equipment, such as plows and carts. of the area deemed suitable for cotton. Lower av- An important difference between the two sites erage, less evenly spaced, rainfall made cotton was that the AVV restricted the amount of land production a fairly high risk activity even in an that settlers could farm. A system of weights that average year. The Solenzo area benefitted from a assigned a labor value to each registered family better set of natural conditions. member (Table 7.3) was used to determine the amount of land that a family was officially al- Early Response to the AVV Sponsored and Solenzo lowed to clear and farm.2 Total farm size was re- Assisted Crop Extension Models stricted to either one or, in the case of a very large labor force, two ten hectare farms. Each project Avv. Although the 1979 farm monitoring survey farm consisted of a one hectare home site and six conducted by the AVV showed that average bush fields measuring 1.5 hectares each. Under yields for sorghum were below the expectations the prescribed system of crop rotations (Table of the project planners, yields were still two to 7.4), each of the six bush fields was supposed to three times as high as the figures recorded for the lie fallow for two years as part of a six year rota- McMillan (1983) case study of home villages in tion that included sorghum, cotton, and legumes. the Sanmatenga (Kaya) region. The survey also The AVV also specified the type and level of fer- showed, however, that the increase in settlers' in- tilizer, seed, and insecticide to be used on each comes was primarily a result of the expansion of crop. There was no such restriction at Solenzo. the total area cultivated and the natural fertility A second important difference between the of the new soils rather than the successful intro- AVV and Solenzo programs was that participa- duction of the recommended package of intention in the proposed package of intensive cultiva- sive technical innovations (Murphy and Sprey tion methods was not voluntary for the AVV 1980). In particular: (1) cotton was the only crop sponsored settlers. When settlers joined the for which settlers were following the recommended package of intensive cultivation techniques; and (2) even for cotton, settlers tended to Table 7.4 Recommended AVV Rotation diminish their use of the intensive package with longer periods of residence in the project. Based Year Crop on this evidence, Murphy and Sprey concluded I Sorghum that it was unlikely that the settlers could contin- 2 Cotton ue to maintain the same high levels of agricultur- 3 Cotton-peanuts-cowpeas al income growth on a fixed 10 or 20 hectare AVV 4 Red sorghum-millet farm. 5 Fallow What we saw then, after the fifth year, were 6 Fallow two different patterns of coping with perceived

63 limitations on future income growth in crop While the area planted in cotton has decreased production. as a percent of total area planted (from 42 percent The first pattern, adopted by the wealthier in 1979 to 20 percent in 1988) (Table 7.5), the per- AVV farmers and larger households, involved recent of total area planted in cereals increased investing their new earnings from cotton crops from 56 percent in 1979 to 71 percent in 1988 (Ta- into the development of non-crop enterprises. ble 7.5). During the same time period the average This usually included: efforts to launch at least kilogram production per unit labor decreased one family member into a high earning second- from 500-700 kg/ALE to 100-300 for cotton but ary occupation in trade or services; development increased to over 1000 kg/ALE for sorghum, mil- of a cattle herd that was boarded with neighbor- let and maize (Table 7.6). ing FulBe;3 and cultivation of the largest area pos- An analysis of fertilizer trends for the three sible within the allocated cultivation zone, AVV blocs shows a steady decline in purchased sometimes via the use of rented tractors. fertilizer since 1979 (see Figure 4.2). In most cases, A second pattern observed in less successful the first dip coincides with the fifth year of settle- households, which were often smaller, involved a ment.Mineral fertilizer was used on 87 percent of shift toward limited practice of the prescribed cotton fields and 33 percent of white sorghum cultivation methods plus limited diversification fields at Mogtedo V3 during the 1979/80 crop- into livestock. ping year. In 1988 fertilizer use was almost exclu- The 1989 restudy of a subsample of farmers sively on cotton. that were included in the 1979 AVV farm moni- SOLENZO. In contrast, the cotton package at So- toring survey sample indicated a continuation of lenzo has allowed migrant farmers to enjoy a sus- these earlier trends. Field measurements (of the tained rate of income growth from commercial collectively worked household fields) show that farming since the late 1960s through a steady in- settlers are farming an area per unit labor slightly crease in yields and a steady expansion of the to- larger (1.31 ha) than that recorded in 1979 (1.15) tal area farmed. The average yields per hectare (Table 7.5). The 1988 figure, however, is still with- for cotton increased from 544 kg/ha in 1972/73 to in the range of what was recommended by the 870 kg/ha in 1980/81, and to 1,076 kg/ha in project (1.04-1.71), which suggests that the set- 1985/86 (See Chapter 4; Table 4.5). During the tlers' patterns of production are only slightly same time period, total cotton production in- more extensive than they were in 1979. creased fiftyfold (from 801 tons to 41,500 tons).

Table 7.5 A Comparisonof the AVV'sRecommended Crop Area with the Actual Area Planted in 1979 and 1988 (with and without compound fieldsa)

Hectares per unit labor Percent of area planted in different crops Cowpeas with w/o Sorghum! peanuts comp. comp. millet Corn Cotton beans Rice Other Recommended: - 1.04-1.71 33-48 - 38-42 16-31 - Observed: 1979 AVV Mogtedo/Bombore (household) - 1.15 54 2 42 2 - -

AVVLinoghin (household) - - 40 5 30 9 - 16

Mogtedo V3Y (private and household) 1.95 - 50 12 29 7 1 2

Mogtedo V3 (household) 1.74 ------

Observed: 1988/89AVV all (household) Linoghin, Mogtedo, Mogtedo-Bombore,and Mogtedo V3 1.31 - 60 11.3 19.9 9.2

- Not available. a. 1979 AVV Statistical Service Survey did not count the 1-hectare homesite (compound field) in the measure of field area Sources: Murphy and Sprey 1980; Savadogo 1989b.

64 Table 7.6 Cropand Livestock Productionand ProductionExpenses per Adult LaborEquivalent (ALE) in AVV Planned Settlements, 1988/1989

Linoghin Bombore Mogtedo Mogtedo V3 All Mean residents per households 10 (20) 11 (20) 11 (20) 9 (20) 10.25 (80) Mean labor force per household (ALE) 3.8 4.3 4.3 3.5 3.98

Kilograms/ALE Cotton 1979/80 521 575 575 709 1988/89 288 100 309 88 Sorghum, millet, corn 1979/80 - 391a 391a 902 1988/89 1,052 1,086 853 1,397

CFA/ALE Crop inputs (fertilizer, insecticide) 31,417 11,327 32,876 15,211 Hired labor (cash and kind) 5,999 2,385 5,625 0 Rented equipment (cash and kind) 10,564 1,675 277 555 Net crop income 61,314 54,988 51,071 55,899

Annual livestock income (for male heads of household/ALE) 21,897 6,797 14,764 5,415 8,522b A. high income, low cost/loss scenario 15,570 2,022 7,977 -1,480 6,223 B. low income, low cost/loss scenario.

() indicates number of households a. Murphy and Sprey 1980:78. Figure includes households in lst - 5th year installation and only household fields. b. Excludes V3 data. Sources: Savadogo, Sanders and McMillan 1989:30; Murphy and Sprey 1980.

Fertilizer use in the Solenzo area is among the ton fields received NPK fertilizer at 100 kg/ha or highest in the country. The most concentrated ap- more, close to or above the recommended levels plications tend to be on cotton and, to a lesser ex- of 150 kg/ha for NPK and 50 kg/ha for urea; 57 tent,oncorn,whichisgrownasacashcropbythe percent of the maize fields received lower tractor farmers. The CRPA estimated fertilizer amounts of NPK fertilizer. Only 23 percent of the use in 1987/88 to be 143 kg/ha NPK for cotton, red sorghum, and 21 percent of the white sor- and 80-90 kg/ha NPK, with about 95 percent of ghum fields received any fertilizer. In general, the the area planted in cotton and 80 percent of the tractor farmers used the largest amounts of fertil- area planted in corn receiving some fertilizer izer per hectare. Only the tractor farmers, who re- (CRPA 1989b:32-33). Our data show a similar pat- ceive a guaranteed price for their production tern for 1988/89 (Table 7.7): 80 percent of the cot- from the cotton parastatal that finances their

Table 7.7 Number of Fields that Received Different Doses of NPK Fertilizer at Solenzo

Fertilizer (kg/ha) Cotton Red sorghum Maize White sorghum Millet Soy beans 0 2 20 10 11 21 0 1-49 2 0 2 1 0 0 50-99 1 1 5 1 0 0 100-149 13 2 4 1 0 1 150-199 8 2 1 0 0 0 200+ 3 0 6 0 0 0 Urea only 0 0 2 0 0 0 Manure only 1 1 0 0 0 0 Total 30 26 30 14 21 1 Manure and NPK fertilizer (2) Source: Land Settlement Review.

65 tractors, normally grow corn in large quantities productivity per ALE between animal traction (an average of 2,864 kilograms per ALE versus and motorized tractor farmers (Figure 7.1). Even 124 kilograms/ALE for animal traction farmers deducting the substantial costs of mechanized and 88 kilograms/ALE for manual farmers). cultivation (estimated at 650,0000 CFA per family The lower use of fertilizer on cereals can be at- per year) and the high costs of fertilizer (Annex 5, tributed to several factors. When farmers fertil- Table D-2), net crop income for the tractor farm- ize, they tend to think in terms of fertilizing a ers was three times as great as that for animal rotation, with the residual effects of the fertilizer traction (Table 7.8). passing on to the corn and sorghum crops to be SOFITEX (1989) research shows that the yields planted in the same field the next year. The CRPA obtained by farmers with tractor technology tend observed, however, that only 26 percent of the to be slightly higher than those recorded for ani- sorghum and 55 percent of the corn fields includ- mal traction farmers (CRPA 1989b) after an initial ed in its farm monitoring program in 1987/88 learning period. It is difficult to ascertain, howev- were planted on former cotton fields; 48 percent er, what portion of the high yields per hectare en- of the sorghum and 28 percent of the maize fields joyed by the tractor farmers is attributable to their were preceded by other maize or sorghum fields. much greater capacity to clear and farm "new" High levels of fertilizer use on cotton, in combi- land. At Solenzo, 51 percent of the collectively nation with the high quality area soils and timely worked fields at Kie/Dar-es-Salaam (the newer insecticide spraying, have allowed the Solenzo settlement) and 23 percent of the fields at Dabou- area farmers to sustain high yields. While there is ra (the older settlement) had been farmed for six increasing concern about the loss of soil organic years or less, in contrast to the AVV, where all of matter in the soil, this is not yet alarming. the collective project fields had been farmed for The study showed substantial differences be- ten years or more. tween net crop revenues for farmers at different Increasing population densities will probably levels of technology (Figure 7.1). Animal traction constrain the more ambitious farmers in their at- in Solenzo was associated with crop income lev- tempts to exploit the larger areas. Our best evi- els that were 2.5 times as high as manual traction dence for this is the recorded differences in net (Table 7.8). There was another big jump in crop income per ALE for the tractor farmers at

Table 7.8 Crop and Livestock Production per Adult Labor Equivalent (ALE Solenzo 1988/89)

Dar/Kie Daboura All Animal Animal Animal Manual traction Tractor Manual traction Tractor Manual traction Tractor Average family size 8.0 (6) 13.5 (8) 29.5 (5) 8.0 (7) 7.5 (6) 37.0 (4) - - - (no. residents) Average family labor force 2.8 5.3 10.9 2.6 2.5 13.8 2.7 (13) 4.1 (14) 12.2 (9) (ALEa) Collective field area 4.4 (6) 7.0 (9) 37.7 (5) 2.7 (7) 5.8 (6) 29.9 (4) 3.5 (13) 6.5 (14) 34.5 (9) (per farmb) Collective field area (ALE) 1.4 1.6 3.4 1.5 1.9 2.2 1.4 1.7 2.8 Cotton (kg. per ALE) 148 548 2,789 133 513 869 138 537 1,824 Sorghum-millet (kg. per ALE) 422 694 808 411 1,252 737 410 838 771 Maize (kg. per ALE) 42 135 4,292 132 94 1,448 88 124 2,864 Net crop income all fields 24,416 79,821 521,167 39,134 104,988 182,850 31,978 86,398 297,331 (CFA per ALE) Annual Livestock Income 1,000 11,200 71,907 3,737 -4,911 10,921 (CFA per ALE) for male heads of household

-Not available. ()indicates number of households. a. See Table 7.3 for weights used to derive ALE. b. Measurements for the fields of the household head and one wife were estimated by crude calculations of length and width. A subsample of fields was subjected to more rigorous field measurements in order to verify the accuracy of our estimates. Sources: Savadogo, Sanders, and McMillan 1989 and McMillan calculations.

66 Figure 7.1 Average household field area farmed collectively, net crop income per adult labor equivalent (ALE) for the major technology groups at Solenzo, and comparison of net crop income for the AVV and Solenzo

A. Average household field area farmed collectively C. Comparison of net crop income for the AVV and (Solenzo) Solenzo

Hectares Thousands 4 550 500

2 K ~~~~~~~~~~~~~400

' u r n ~~~~~~~~~~~200

Dar/Kie Daboura All 100 0 B. Net ccropincome from both privately and S collectively worked fields (Solenzo)

CFA (Thousands) 600 ____Solenzo _

4009

200 w

0 Dar/Kie Daboura All

Manual Traction _ Tractor

Source:Tables 7.5 and 7.7.

67 Daboura and Dar-es-Salaam/Kie. Net productiv- later joined by the indigenous Bobo. Similarly, the ity per ALE for the tractor farmers at Dar-es-Sa- indigenous Mossi in the Ganzourgou region were laam/Kie was more than two and a half times not interested, at first, in the intensive crop pro- that recorded for Daboura (Table 7.8; Figure 7.1). duction program promoted by the AV, nor were Daboura, located along the main highway linking they particularly interested in living in large Dedougou and Solenzo (Figure 7.2), was one of numbers near the rivers. The early success of the the first villages to experience extensive immigra- AVV settlers, however, attracted large numbers of tion-much of it before 1974. Dar-es-Salaam did spontaneous settlers from the region's indigenous not begin to develop as a cultivation hamlet of villages who moved into areas adjacent to the Kie until after 1974. Dar-es-Salaam possesses a sponsored settlements. From these examples, one vast uninhabited bush between the core village could argue that new settlers are often more open and the Mouhoun (ex-Black Volta) River. Because to innovation than are their hosts, but that, over of this large area of uncleared land, the Dar-es-Sa- time, differences in receptivity to change often laam settlers and hosts can still expand toward taper off. the river. * Family Size. Family size can work both for For farmers to maintain their high levels of and against intensification. If crop production yields, they will need to combine fertilizer use technology and prices make intensification prof- with other intensive cultivation practices. Espe- itable, families with more labor are better able to cially important are: (1) diversifying the existing absorb the higher labor costs of such technology. system of crop rotation to include crops that are Smaller families would be hindered from taking not as depleting of area soils as cotton and maize, advantage of the technology package by their rel- and (2) increasing the degree of integration of on- ative lack of labor. Conversely, if intensification farm livestock with crop production to make use does not pay, the same large families are likely to of organic fertilizer. devote their larger labor force to greater extensification-i.e., in farming the largest area possible Factors Affecting the Development of with lower yields, rather than in farming a small- Sustainable Farming Systems er area with higher yields. Under the existing technological and institutional structure, the size Each of the two dominant development models of the family labor force tends to be positively currently in use in the river basins tends to induce correlated with the size of the total area farmed. greater extensification. Several macro- and micro- Even at the AV, which has official restrictions on level factors encourage this pattern. the amount of land per family, larger families did * Security of Tenure. Secure tenure can contrib- not appear to farm a total area that was demon- ute to the development of intensive farming. Un- strably smaller per ALE than families with small- fortunately, recent efforts to provide settlers with er labor forces, in part because larger families greater security by allocating "land (rights) to tended to cultivate a larger number of "illegal" those who work" seems to have encouraged both fields outside the project limits. A similar rela- migrants and indigenous farmers in the Solenzo tionship between family size and land area was area to expand the total area that they farm. observed for farmers in the Solenzo region. With * Opportunities for Diversification. The exist- existing crop technology and prices, the size of ence (or lack) of alternative opportunities for di- the family labor force tends to promote extensifi- versification can have a positive (or negative) cation and diversification rather than crop inten- effect on intensification. One example of this is Li- sification. noghin, where farmers are reinvesting cash re- * New Technologies. Mechanization is another turns from crop and noncrop enterprises into factor that can contribute to, as well as detract rainfed agriculture at a higher rate than in the from, intensification. New technology, such as other two planned settlement groups. animal traction and tractors tends to be used ex- * Attitudes toward Experimentation. Immi- tensively, even though these technologies were grants can be a force contributing to agricultural originally intended to help farmers overcome the innovation just as they can contribute to greater labor bottlenecks associated with more labor in- extensification. In Solenzo, for example, the Mossi tensive cultivation techniques. Our data showed immigrants were the first to adopt animal traction that in 1988, the average land area per ALE and cotton production in the 1960s. They were farmed by the animal traction farmers at Solenzo

68 FIGURE7.2 IBRD23848

BURKINA FASO LOCATIONOF THE STUDYVILLAGES IN THE SOLENZOSUBSECTOR: DABOURA, KIE, AND DAR-ES-SALAAM(A CULTIVATIONHAMLET OF KIE) LOCALISATIONDES VILLAGESD'ETUDES DANS LE SOUS-SECTEURDE SOLENZO:DABOURA, KIE ET DAR-ES-SALAAM(HAMEAU DE CULTUREDE KIE)

YeraDskoro o .4\ Da \ \ Dissanku "

/ \ / /M1%ontionky-( tY . y/q B~~onza/__ J Yeresssoro2

Ki~~~~~~~~.

U '>* ~Sanakui - / C Ban/W Herdougou

Dar-Salam'

------. ,~.,.. d c', ... go,f- ,A.-hdfl, a mu a I - - - ~ I..,&G ...4 \6.d _ WA- pA

STUDYVILLAGES M A L I i _ W~~~VLLUGESD'ETUDES . MARSH / , NIGER ALLWEATHER IMPROVED UNPAVED HIGHWAY ROUTEPRINCIPALE NON-GOUDRONNEE DE TOUTESAISON f BURKINAFASO ALL WEATHERUNIMPROVED SECONDARY ROAD / OOgadougou ,<. ROUTESECONDARIE NON-AMELIOREE DE TOUTESAISON / / UNIMPROVEDSECONDARY ROAD ; Area of PISTESECONDARIE NON-AMELIOREE BENIN -_---_- UNIMPROVEDSECONDARY ROAD / PATH SENTIERNON-AMELIOREE CTE DIVOIR HAN

JULYi1992 was 21 percent greater than the land area per ALE * Farm-FamilylExtension-Agent Ratio. The So- for the manual farmers; the land area per ALE for lenzo case study demonstrates that if extension the tractor farmers was 64 percent greater than themes are profitable and entail low risk within a for the animal traction farmers (Table 7.8). Net particular agroclimatic, market, and policy envi- crop income for the animal traction farmers was ronment, settlers will likely adopt them with little 170 percent greater than for the manual farmers, intervention from the government. In contrast, if and net crop income for the farmers with mini- farmers think that a particular extension theme is tractors is 244 percent greater than for the animal risky or unprofitable (like the original package of traction farmers (Table 7.8). These figures suggest intensive cultivation methods promoted at the that the more ambitious Solenzo farmers were AVV from 1974 to 1983), no amount of top-down pursuing a strategy in which they are trying to supervision by the extension service (e.g., the maximize their returns to labor, rather than per high ratio of one extension agent to every 25-50 unit land. This is a rational strategy as long as households that characterized the AVV for most land is freely available. of the 1970s) can force settlers to adopt it. * Aversion to Risk. Attitudes toward risk are * Credit. The credit programs linked to com- important in determining the pattern of respons- mercial cotton production at the AVV and at So- es to new economic opportunities at each of the lenzo helped a first generation of farmers adopt study sites. At the AVV, for example, poorer the new crop innovations. The Solenzo program households (with smaller labor forces) were gen- shows, however, that if income levels become erally less willing (or able) to absorb the higher sufficiently high, settlers are capable of financing risks associated with cotton production in the a large part of the next generation of agricultural sponsored settlements. Settlers know that in the technology. Over half the plows owned by the event of a total or partial crop failure-not an un- tractor farmers at Solenzo, and almost all of the usual occurrence in this area where rainfall can donkey carts and draft animals currently used by fall below 800 mm a year-they will still be restudy farmers at the Solenzo and AVV study sponsible for reimbursing all credits for fertilizer sites, were purchased with cash. and insecticide at the time of the cotton sale (see Chapter 4). Macro-level Factors that Contribute to or i Availability of Land. The availability of land Detract from Intensification under easy tenure arrangements encourages extensive agriculture. It is less expensive (in time Many aspects of the national policy environment and labor) to abandon all or part of a field that is have contributed to the development of extensive no longer producing well in favor of a newly cultivation systems. cleared field than to invest time and cash in inten- * National Input Policies. Until 1987, modern sive crop production methods. In certain villages crop inputs (NPK, urea, insecticides) were subsi- in the Mouhoun and outside the AVV sponsored dized by the state. Input subsidies have been settlements, however, there are already signs of slowly reduced since 1983, leading to a steady land shortages. One instance of this is the differ- rise in fertilizer prices that has affected the utili- ence between Daboura and Dar-es-Salaam, al- zation of fertilizer, particularly on cereals. The ready mentioned. impact of higher input prices in the long run will Ease of access is not only determined by the ac- depend on output prices and price policies and tual physical availability of land, but also to what on the efficient commercialization of the cotton we refer to as "social access" to land. The village crop. As long as the cotton market is guaranteed of Keraba in the Mouhoun CRPA is an example of by SOFITEX, and as long as the world cotton an area where outside immigration has been lim- price does not drop below a critical level, we ex- ited, despite the good agricultural potential of the pect that input use on cotton will not drop signif- area. This is in large part because Keraba's indig- icantly. Cotton production is thus likely to remain enous inhabitants were resistant to any and all the most intensive crop production model. outside immigrants reducing "social access," a The Burkina government has recently begun fact that distinguishes the village from the more to emphasize organic fertilizer. Efforts to im- receptive communities and community leaders of prove the quality and efficiency of organic fertil- Daboura and Kie. izer used on crops is likely to increase the demand for organic fertilizer and ultimately

70 help to improve soil quality. Increased use of or- Cereal prices have tended to be higher and less ganic fertilizer from compost pits might help to volatile in the AVV planned settlements largely reduce the bottleneck of exclusive reliance on because of the area's easy access to the grain mar- manure. The use of organic fertilizer, in conjunc- kets in Ouagadougou. Given higher prices and tion with mineral fertilizer, can lower soil acidi- more accessible markets, the AVV settlers have ty and contribute to the development of more gradually been decreasing the amount of land de- stable production systems. voted to more labor intensive cotton and increas- * Producer Prices, Commercialization, and Sec- ing commercial cereal production. Many farmers toral Development Policies. State intervention in in the southwest would be willing to shift from sectoral policies (prices, commercialization) has cotton to commercial grain or legume production notable impacts on producers' responses. For ex- if this was a practical alternative; under present ample, in the current institutional setting, cotton conditions, it is not. is favored at several levels. First, the cotton market is guaranteed to the producer. SOFITEX will Policy Implications buy all the cotton a farm family is capable of producing. In personal conversations with SOFITEX The development of more sustainable farming officials, we were told that Burkina's cotton is systems in Burkina's underpopulated river ba- considered high quality and easily sold on the in- sins requires attention to a number of details that ternational market. The principal problem is that affect the profitability and risk associated with the international cotton market fluctuates widely different cropping enterprises. Some of these fac- depending on the relative demand for cotton ver- tors are at the micro-level of individual farm fam- sus other synthetic fibers and the level of produc- ilies or local extension services; others are more tion in countries such as China. macro, at the level of national food, input, and ex- In contrast to cotton, Burkina's cereal produc- port policies. tion enjoys almost no price support. Although cereal markets are, at least in theory, guaranteed by Reinforce the Existing PNGT Program OFNACER (Office National des Cereales),these prices are not respected because OFNACER's ca- The case studies have discussed several micro- pacity to intervene in the market is very limited. level requisites for developing sustainable farm- Agriculturalists are therefore obliged to sell their ing systems. These include security of tenure, the products in local markets where, as discussed in initiative of settlers, the involvement of hosts and Chapter 4, prices vary significantly, both between pastoralists, availability and condition of roads, and within years. Examples of this can be seen in and access to basic social and economic services. the variations in white sorghum prices in 1988/89 Burkina's Village Land Management Program from 28 to 40 CFA/kg at Solenzo, from 44 to 50 (PNGT) is a promising micro development model CFA/kg at the AVV,and from 50 to 70 CFA/kg at that addresses these issues, as well as the more Kompienga (Annex 3, Table B-3). The especially conventional concerns of crop and livestock ex- low prices at Solenzo are typical of the southwest tension services. Especially important is the fact and do not encourage production. In addition to that the PNGT is not an isolated regional pro- the problem of fluctuating markets in these rich gram. The basic institutional model has been surplus areas, it is sometimes difficult to find any adopted nationally and enjoys a strong base of buyers because of the great distances to markets national and regional support. and the expense of transport. Thus, we found that in the southwest of Burki- ReinforceRegional and NationialCapacity to Support na, agriculturalists tend to cultivate large areas of Settlement Houiseholds cotton, the crop that is most reliable in terms of extension, markets, and prices. Despite its great Efforts to promote sustainable crop production potential, marketing problems and unstable pric- systems are more likely to be successful in con- es make sorghum production less reliable. The junction with a strong institutional base. Care percent of total area devoted to sorghum is grad- should be taken, however, not to apply the PNGT ually decreasing as sorghum is progressively program mechanically. First, the precise structure downgraded to a subsistence crop. of a local program will need to be adapted to the local cultural and economic context. Second,

71 within any local context there are likely to be im- bor, or land resources of those who arrived earli- portant differences in cash and labor resources, er. These newer, "poorer" settlers are either needs and concerns, experience, and entrepre- unwilling or unable to gamble on the high risk, neurial capacity of farmers. No single, monolithic high potential, income producing strategies that extension program is likely to be sufficiently sen- attract wealthier settlers. sitive to these variations within a population. Extension programs must work with evolving ReinforceAppropriate Crop Research Programs patterns of diversity within a population. The AVV extension model, for example, focused on a Our study identified two research areas critical to single set of crop recommendations. It was antic- both the short-term and the long-term develop- ipated that differences would exist or develop be- ment of the valleys. The first area involves rein- tween larger and smaller households, but forcing current efforts to promote water primarily in terms of the total area farmed, rather harvesting technology for agroclimatic zones than in terms of the types of crop and noncrop en- with less than 800 mm average rainfall (McMill- terprises developed. A longitudinal analysis over an, Painter and Scudder 1990).Large sections of fifteen years, however, reveals increasing diversi- the Nakambe and Nazinon fall within this agro- fication in terms of types of technology used, climatic zone. Although the river basins are gen- crops grown, and the relative importance of ani- eraily considered to be of higher potential and to mal husbandry and off-farm employment. have fewer soil problems than adjacent plateau Similarly, substantial differences exist among zones, they share many of the same soil character- the activities, resources, and incomes of the man- istics and experience the same low, irregularly ual, animal-traction, and tractor farmers at Solen- spaced rainfall. zo. Earlier extension models focused on the idea A second priority area involves research to of moving settlers progressively from lower to help increase the diversification of local crop higher technology levels. One of the important and livestock systems in the southwest. The results of recent research by the Mouhoun CRPA present crop extension program focuses almost Monitoring Unit was to show great variation exclusively on cotton, with lower levels of sup- within technology levels and movement between port for maize. Both crops are hard on area soils. technology levels over a five year period. Within In addition, cotton demands almost twice the the category of manual farmers, for example, amount of labor per hectare as other crops. Re- were farmers who preferred to rent animal trac- search can contribute to the development of tion or tractor plowing equipment, rather than to more diversified crop production systems by ex- own and look after their own oxen. This same cat- ploring opportunities for developing rainfed egory, however, could include young farmers grain legumes, on-farm livestock, and dry-sea- without the resources to rent, who were aspiring son irrigated gardening. New technology for to animal traction. The question was then raised, these crops will be important. Equally, if not given high rates of variation within categories, more important at the present stage, however, is how meaningful is "technology type" as the key research to identify some of the key processing, variable with which to identify groups of farmers storage, marketing, and socio-economic con- for a particular extension effort? straints that restrict the dissemination of crop If local development programs are successful and livestock production technology that has al- at a first stage, they create a group of farmers with ready been developed by INERA (Institut higher expectations and resources in a second d'Etudes et des RecherchesAgricoles), SOFITEX, stage. If "captured" by profitable crop and live- ICRISAT (International Crop Research Institute stock technology and local opportunities for di- for the Semi-Arid Tropics), and other national versified investment, the first generation of and international research institutes. successful settlers can become a driving force for In both instances, researchers are likely to pro- income producing investments and activities. If mote more realistic programs and policy sup- not captured, this same energy will be turned to- ports if there is a conscious attempt to link the ward cultivating the largest possible area. results of their studies to past and projected mar- Successful development is also likely to attract ket trends, and to the socio-economic characteris- large amounts of additional immigration. These tics of farmers. later migrants are not likely to have the cash, la-

72 Reinforce Local Projects with Sound National Policies technologies, rather than to choose the less to Support Regional Marketing Systems risky route of extensification. We are not advocating strengthening govern- Total cereal production in Burkina experienced ment marketing boards. Our research on the dy- an upward trend over the last decade. Although namic food distribution systems that have production dropped during the 1983-1984 developed along the paved highway to the AVV drought years, the dip was considered less severe sponsored settlements suggests that a more prac- than those associated with earlier successive tical route will be for government to enact poli- drought years. One reason is the substantial in- cies that reinforce the private sector's ability to crease in the total area farmed in cereals. If the carry out this role. The case studies provide addi- present course continues, the valleys will contin- tional evidence to support national programs that ue to be cleared and Burkina will continue to in- deal with food surpluses through improved stor- crease its annual production of cereals and cotton age, better roads, and the development of on- until the new land resource is exhausted. Many of farm livestock feeding programs. the northern river basins are already becoming saturated, with migration shifting south. Notes One can hypothesize that increased cereal production is a response to increased family 1. The Kompienga Relocation Project(Projet Transfert des Popu- consumption and cash needs. However, one of lationsde laKompienga) was created 13 June 1988.Spedal fund- the unintended results of successfully increasing for the project was slated to end in March 1990 (AVV 1990). ing total cereal production has been to hold 2. This potential for labor was measured by a systemnthat assigned labor values to persons in the family according to age down local market prices. A ten year analysis of and sex (see Table 7.3).Since an adult male was considered to local market prices at Linoghin shows that the have the work capadty most readily transferred to a variety of median per kilogram price of white sorghum at 1.0.tasks,Women, this was children, the standard and older unit workers and was were assigned assigned a value lesser of the AVV is 25 percent less than the price record- values. On the basis of its total index (the sum of all the labor ed in the AVV market surveys in 1979 (Murphy values for family members), a household can be classified by and Sprey 1980:58,143;Annex 3, Tables B-3 and farm "type." Type ia,wforexample, withthesmallestlaborforce, included families with a labor index of 1.75-2.25,which could B-7). Adjusted for inflation, the real price of sor- consistof a man and wife (1.75),a man, wife, and a teenage boy ghum is 54 percent less. Better farm prices (2.25),orothercombinationswithsumssinthatrange. encourage farmers to invest more in intensive 3. See Annex 5, Table D-5.

73 ~~8

Importance of Markets and Service/Administrative Centers

Settlement-Related Development and Markets market and administrative centers. In 1960, the town of Mogtedo was listed as having 60 house- Some of the clearest evidence we have of the wid- holds with 200 people (Table 8.1). Since 1973, the er regional effects of new lands settlement in the town has been the administrative headquarters OCP river basins comes from the substantial for AVV activities in the upper Nakambe, and by growth in existing markets and the creation of 1985 planned and spontaneous settlement to areas new markets since 1974. Only 3 of the 20 markets surrounding the earlier AVV planned settlements frequented today by settlers in the AVV planned was sufficient to justify the creation of a new ad- settlements at Mogtedo, Linoghin and Mogtedo- ministrative department based at Mogtedo. The Bombore existed prior to 1974 (Figure 8.1). town itself has grown to about 6,000 inhabitants. New lands settlement in the area has also had a The Mogtedo market has grown from 100 stalls, 5 major impact on the growth of several regional shops and a single grinding mill in 1975 to more

Table 8.1 Settlement and Development in the Town of Mogtedo, 1960-1989

1960 1968 1975 1980 1985 1989 Households in town Indigenous 60 60 - - - - Immigrant - 200 - - - Inhabitants 200 1,300 3,374 - 4,550 6,050 Stalls in central market - - 100 300 600-650 800 Businesses in separate buildings - - 5 14 - 75 Mills - - 1 - - 10 Etalagistes (semi-detached stores) - - - - - 122

Persons selling in boutiques and market a market day, rainy season (August 1989) 1,914 market day, dry seasonb 4,000 non-market dayc 273

- Not available. a. Persons selling (not attendance) in boutiques, the market, detached businesses, and etalagistes. b. Estimated "normal" attendance. A restudy of the market in March 1990 failed to confirm this figure. Indeed, the number of people selling actually decreased from August 1989. This was attributed to the large number of Mogtedo merchants selling at new markets created at the gold-mining sites. c. Non-market day = 273 Source: E Kabore and J. Guigma, Enquete Marche, August 1990.

74 FIGURE 8.1 IBRD 23850

BURKINA FASO LOCATION AND RELATIVEIMPORTANCE OF EXISTINGAND ABANDONED MARKETSFREQUENTED BY SETTLERSLIVING IN THE AVV SPONSOREDSETTLEMENTS AT MOGTEDO, MOGTEDO-BOMBORE, AND LINOGHIN, 1998/89 LOCALISATIONET TYPESDE MARCHESEXISTANTS OU ABANDONNESPAR LES COLONS DES PERIMETRES AVV DE j* iMOGTEDO,MOGTEDO-BOMBORE ET LINOGHIN, 1988/89

7.6I 210 | MOGTEDO VILLAGE

9. V6 AV MAAAGA6 19 ------. TRDAPADAMAII R -O----NA\

) \MO~~GTEDO BOUDRY /' t ,024 > 122 0GRAND MACE RE_ NAX @ / o U 3 /~~~~1 ZORGHO

XBOMBRy MAKARAGA \

.< AVVNGOGADEGHN9 RAPADAMA \ \ OUAYA~~~~~~~LGUI _,

MAJOR. REGIONALMARKETS: \ { d . -GRANDS MARCHESREGIONAUX: \ st_ .MOGTEDO I.OWN)- o , o 3. WARDOGO rTNM

SUB-REGIONALMARKETS: MARCHES SOUS-REGIONAUX.V 4.B. VDV AVV AVV MANORAPADAMAKARAG TANEMA ,lyo .p~ fTA Ol Bo rop 7. VS AVV MOGTEDO-NOMBOREV R 9 V6 AVV MANKARAGA 2 _ 11. RAPADAMATRADITIONAL A I 5. GADEGHIN/V9 RAPADAMAAVV 2 A 6. GOLDMINE MARKET AT NOBSI N o 12.YAIKA- hen e ro ni h'r noo lt r 13.TENLOUKA-M Vd2 25. AdVhV b .. onthM 1. VOAVV MANKARAGA VG-MGTEO V3/ Oopg-,hMGTED.VW24. mop nyjudp

fiAREA VILLAGEMARKET: MARCHE DE ZONE /NIVEAU VILLAGES): 14. DAMONGOMALI '

0 ACEVILLAGEMAKES 0 ABANDONED MARKETS:o IE

*ESTABLISHED26. INMOGTEDO-BOMBORE PRESENT LOCATION 25. BEFORE AVVMOGTEDTE 1974 DIVOuagaEugo(BENINGG maNA 'MARCHEEN PLACEAVANT 7974 LINOGHIN G

JULY1992 than 800 stalls, 75 shops in separate, permanent the 21 markets have been created or moved since buildings, 122 shops in semi-permanent build- the OCP began in 1974.1 ings, and 10 mills in 1989.We counted an average The town of Solenzo has itself developed into a of 600 men and 900 women selling in the central major market center. Our census counted 980 market area on the two days we conducted our people selling in the central market in August census during August 1989,a period when it was 1989, twenty boutiques, twelve grain mills, two raining almost daily. bars, and two gas stations. Nine boutiques, the The market created by the AVV sponsored set- bars, and the government wholesale store (Faso- tlers at Linoghin in 1974 is today the most impor- yaar) are open even on nonmarket days. Al- tant market center, after Mogtedo, on the paved though we do not have precise figures, the major- highway between Ouagadougou and Koupela. ity of store owners appear to be immigrants. The market has grown steadily (Table 8.2) and Even the young cultivation hamlet-recently now numbers some 300 stalls. Market officialsin- turned village-of Dar-es-Salaam boasts a lively dicated that as many as 5,000 people may attend village market that originated in 1985.The village, the market during the dry season. In 1990, the which serves FulBe and other pastoralists in the market included four businesses in separate, per- immediate area, has several mills (privately manent buildings (one large boutique, two bars, owned) and numerous boutiques. Daboura, the and a coffee shop), in addition to open air restau- other study village located on the main highway, rants and service centers (mechanics, tailors) that has a larger market that included 536sellers in late are open every day. Food concessions and beer August 1989.On nonmarket days, we counted an sellers operated late into the evening, every day average of four boutiques, several people selling of the year. cooked food, two mechanics, and two small res- New lands settlement has had a similar far taurants in the main market area. Except for the reaching impact on the number, size, and volume dolo makers, all of the local merchants seem to be of area markets in Solenzo. Recent research by the immigrants, either FulBe pastoralists or Mossi. CRPAdu Mouhoun (1990a)counted some 21 markets in the area (Figure 8.2). Twelve of these mar- Settlement-Related Development and Markets kets have been created since 1966,when large-scale in the Niangoloko and Kompienga Areas Mossi immigration to the area started; sixteen of The much lower rates of immigration to the Kom- pienga Basin and Niangoloko region since 1974 are reflected in the small number of markets and ser- Table 8.2 Dry Season Attendance and Number vices outside the prefecture's regional administra- of Market Stalls at the Linoghin Marketa, tive headquarters at Pama (for Kompienga) 1975 - 1989 (Figure 8.3)and Niangoloko town. The economic role of these two market and ad- Averagenumber of personsattending ministrative centers is also different. Long distance Personl trade at both Pama and Niangoloko is primarily Linoghin merchants concerned with imported manufactured products. Number of Linoghin spontaneous from Ouaga In contrast, the principal force driving the devel- Year stalls AVVfarmers immigrants and outside' opmentraf the pmo and fore d makts has 1975 32 6018 0 opment of the Mogtedo and Linoghin markets has 1976-78 40 100 40 20 been the rising sales of the area's crop and live- 1979-81 80 120 40 25 stock products and, since 1988,gold. 197981 130 200 12000 25 Niangoloko has long been an important market, 1982-84 130 200 00 60 transportation, military, and customs center on the

1988-89 300 800 2,500 1,700 na.paved Economic highway activities linking at Cote Niangoloko d'lvoire withcontinue Burki- to a. In contrast to other area markets, the largest Linoghin mar- focus more on international trade than on regional kets usually occur during August and September because agriculture and livestock production. large numbers of seasonal migrants from Ouagadougou are resident then. Settlers' relatives are also numerous at this Prior to construction of the Kompienga dam, time. the prefecture administrative center at Pama was Source:Guira, F.1989. the most active regional market. Pama, like Niangoloko, was the garrison headquarters for

76 FIGURE8.2 IBRD23851

BURKINA FASO ZAi Prh. b Th. inP.PrdnbyM .i A. LOCATION RELATIVEIMPORTANCE, &o. 6r AND DATE CREATEDAND RELOCATED .n ,. OF MARKETSIN THE SOLENZO SUBSECTOR,1989 LOCALISATION,TYPES DE MARCHES,ET DATESDE CREATION 'Desse ETDE DEPLACEMENTDES PonV MARCHESDANS LE SOUS-SECTEUR Dira oni DE SOLENZO, 1989

s~~~~~~~

Moussakongo

;~~~~~~i or o;kou Bonk

B ma; BIe _onza ontio .00U0 koro YressoroO 2 " \ Maw/l'* 9 \_/

'&B iolSiguinoguin Sgy Ban Hedgo ' UiQ Biassaga/LanfiercA. Dar-SaIam Gnassoumadougou/ /F Koutou/ ...... _-- <

MAJOR REGIONALMARKETS: -- - MARSH 0 GRANDSMARCHES REGIONAUX: - - MARAIS 1 BSOLENZO(VILLE) 1900-1966 ALLWEATHER IMPROVED UNPAVED HGHWAY 3. BOENA 1968 ROUTEPRINCIPALE N)ON-GOUDRONNEE DE TOUTE SAISON 4. DIRA 1946-1950 _ _ ALLWEATHER UNIMPROVED SECONDARY ROAD 5. BONZA 1985 ROUTESECONDARIE NON-AMELIOREE DE TOUTESAISON SUB-REGIONALMARKETS: UNIMPROVEDSECONDARY ROAD MARCHESSOUS-REGIONAUX: PISTESECONDARIE NON-AMELIOREE 6. GNASSOUMADOUGOU/ KOUTOU 1975 ---- UNIMPROVED SECONDARY ROAD / PATH 7. TOKORO 1987 SENTIERNON-AMELIOREE 8. BIALE 1987 9. HERDOUGOU 1986 10. SIGUINOGUIN 1981 11. DABOURA 1900-1974 12. DAR-SALAM 1985 _j 13. KIE 1890-1984 M A L I , AREA VILLAGEMARKETS: NIGE MARCHEDE ZONE (NIVEAU VILLAGES):WITHOUTSMARKET: MANIGER / 15. BAYE 1976 - VILLAGESSANS MARCHES: 16. DINKEMA 1890-1987 22. MONTIONKUY / BURKINA FASO \, 17. MOUSSAKONGO1957 23. YERESSORO / Ouagadougou 24. BAMA O VILLAGEMARKETS: 25. MAWE MACHESVJLLAGEOIS: 26. DESSE Area of 0 27. MASSO map BEI 1.DISSANKUY 1985 28. PONI fS EI 19. KOMO 1890-1985 29 BaN -C.3N 20. LEKORO 1910-1974 30. KOUAKOUA .4 21. DIMKORO 1950-1980 31. BIASSAGA/LANFIERA C6TE DIVOIREl< GHANA

JULY1992 FIGURE 8.3 IBRD23849

-'--4,o \ NIGER

0 Kiougou Kandago D _c6TE D'IVOIRE HN \

ToFado-HgourmB \KN BURKINA FASO LOCATIONAND TYPEOF t , ) 7s., NaBUMARKETS IN THE SOUTHERN ;Sakong a KOMPIENGA BASIN, 1987-1988 0'O t x, tX LOCALISATIONET TYPESDE 25 \ \ ' MARCHES DANS LE SUD DU BASSIN {Kohogo ,/ \v '\ >/ \ DE LA KOMPIENGA, 1987-1988 I* Ib, Na)ia ouani jKis6ouga Komin-Yana CRPMKE !0 ~ ~ ~~ ~ ~ ~ ~ ~ ~ ~~~~ O CROPMARKET 0 1r IMARCHES AGRICOLES -s6tIgZongo Yolge

r 11 CROP AND LIVESTOCK MARKET {' //Boussira60g ,2< g7 _ hMARCHESAGRICOLES ET MARCHES DE BETAIL

\ '\,q Salam6aore

\\w;Yonds \ 5, g \ Source: Agrotechnik, 1989

Ouargay \\ 90Koadiga

\ Yakti6ao1 } X

to \ Napaoo..... KompiembigaQ Pama

\ Bagmoussa

/0 -. Tind~~~~~~~~~~~~angou -- ang 4 \\ OSoudougui

Sankanse ' Koudiorgou Diablga7Q -- ' ~ _ _Nat'0 _ ntanga"^Kompienga ~~4 QDiem6end6 / Ratjentanga § J 0 10 20 Pognoa KILOMETERS

Ti..i nc,A. A. pr-.o.d AyTA. WuI.., '. neb;"Lf f*

jpl..i~y, - A. p.1 .1 TA.Vdd lo-A G.W, .. yh.dww -6.. Iu .If..ys #-ihe.. - y my r..dr.ww p m , / T O G O 7 I-J JULY1 992 customs and military officials who patrol the sacks of grain would be sold; in the dry season, Togo and Benin borders. Unlike Niangoloko, they cited a higher figure of 800 sacks (96,000 kg) however, Pama never developed into a major sold per major market day (markets are held ev- trade center, in large part because until 1985 ery third day). Eight hundred sacks of grain rep- transportation between Pama and Fada could resents the annual cereal needs of 400 persons take as long as eight hours over poorly main- (using the FAO estimate of food grain needs of tained dirt roads. The same poor roads and isola- 240 kg of cereals [Murphy and Sprey 1980:74]). tion that kept Pama a small garrison outpost On the same day we counted the sale of 150 goats protected the area as Burkina's most plentiful and and sheep. unspoiled game park. Both Linoghin and Mogtedo have emerged as An active market and trade center sprang up important regional trade centers. In contrast to almost overnight when dam construction began other markets in the region, market activity is at Kompienga in 1985. Tradesmen and merchants generally highest at Linoghin in August because from all over Burkina were attracted to the site. many seasonal migrants and extended family The workers' demand for imported manufac- members are in the area to assist with the early tured goods, food, and drink also stimulated re- harvest. Officials estimate that during the period gional trade with Togo, Nigeria, Benin, and when we conducted our study (August 1989), Ghana. The agricultural settlers who moved into fewer than 20 percent of the persons attending homes vacated by departing construction work- the market might be from the AVV sponsored set- ers were attracted primarily by the prospect of tlements (Table 8.2). developing irrigated farming. They were also Mogtedo also boasts a high concentration of drawn by the idea that Kompienga might one day freestanding boutiques, mills, and an active mar- develop into an important regional trade center. ket that serve the entire region. The important Although the number of businesses and the num- role of the market in redistributing agricultural ber of people attending the town market has de- products was reflected in our census of wholesale creased dramatically since the inauguration of grain and livestock merchants. On a single mar- the dam, the town still boasts a sizable number of ket day we counted more than 35 people em- individuals who indicated during the census we ployed as wholesale grain merchants (eight conducted in August 1989 that trade or services based in Ouagadougou and 27 based in Mogte- were their primary or secondary activity. These do), nine wholesale peanut merchants, 22 whole- village entrepreneurs will undoubtedly make a sale goat and sheep merchants, seven wholesale major contribution to the development of inter- chicken merchants (with 65 assistants), and seven national trade of imported products for other ar- wholesale cattle merchants (five from the region eas of Burkina and Niger once the paved road and two based in Ouagadougou). linking Fada, Pama, Kompienga, and the Togo The presence of these active regional markets border is completed. and the opportunities for selling cereals has been a major factor contributing to the AVV settlers' Markets and Settlement Related Development shift toward commercial cereal production.

Active markets both feed and are nourished by Impact on Provisioning successful settlement related development. The sale of agricultural products at good prices Impact on Marketing gives farmers money with which to purchase a wider variety of foods and consumer goods. An important role of markets is to facilitate the These include industrially manufactured and im- sale of agricultural commodities from the imme- ported goods like sugar, salt, batteries, matches, diate area for consumption both within and out- soap, coffee, tea, gasoline, kerosene, shoes, side the area. Livestock and poultry are also scarves, secondhand and new clothing, and bicy- traded at these markets. cle and motorcycle parts. Other items widely We counted 375 sacks of grain (45,000 kg) leav- bought and sold at the markets are manufactured ing the Linoghin market on 4 August 1989 (Guira, locally. These include processed or produced con- F. 1989). Market officials at Linoghin estimated diments or spices, cooked foods and alcoholic that during the average rainy season market, 400 beverages, farm tools, pottery, baskets, clothes,

79 and cloth. Finally, the market centers provide ondary occupation after agriculture. This latter area inhabitants with a variety of services such as figure does not count the hundreds of single tailoring, letter writing, photography, ironing, males and married men, who were not recog- milling, barbering, and the repair of bicycles, mo- nized heads of household, who cut trees and torcycles, and farm tools. The importance of these made charcoal from the wooded areas the dam services, and of processed foods and condiments, was flooding during 1989. are reflected in the large number of sellers in these categories at local markets. Our expendi- Impact on SocialIntegration ture survey showed that a high percentage of men's and women's purchases are for special ser- Effective market integration cannot be mea- vices, condiments, personal consumption items sured simply in terms of market size, the per- like tobacco and kola nuts, and "market foods" centage of settlers' income spent on purchased like doughnuts, rice, beer, and sodas. goods and services, or the percent of income derived from off-farm employment. Markets also Impacton Diversification play an important role in promoting social integration. This is especially important in areas of The increase in demand by rural people for im- new lands settlement, where large numbers of ported and locally produced products feeds effec- individuals have no established framework for tive demand in the area for a variety of off-farm interacting with one another. The important so- activities that generate earnings for the men and cial and political role of markets was reflected in women who practice them. Their growing de- the extreme care with which villages would mand adds impetus to the diversification of choose a market site. In addition, both the cre- household and regional economies in the OCP ation and displacement of the markets studied river basins. were ritually sanctioned by one or several well Opportunities for diversification were greatest known area religious and political figures who in villages located close to major markets. At Li- belonged to the host ethnic-if not the host resi- noghin, for example, an average of 52 percent of dential-community. the recorded income for women settlers was from For the average farmer, the social function of nonagricultural sources like trade; at Mogtedo, markets is probably even more important than near the gold sites, the figure was 43-54 percent their economic role. Both men and women go to (Table 6.3). markets to socialize. Pocket money spent on beer, Five percent of the heads of immigrant house- nonalcoholic sweetened beverages, fried dough- holds cited commerce as their primary reason for nuts, breads, yogurt, and gathered fruits is not immigrating to the Solenzo subsector; the figure just a luxury, but a central element of one's social was 27 percent and 16 percent for Daboura and existence. Without markets and the periodic dis- Solenzo, which have active markets. Eighteen traction that they afford, rural life is dull. The lack percent of the immigrants to the Niangoloko sub- of easy access to markets was underscored more sector cited commerce or special services like frequently than any other problem by women liv- butchery, repair services, or wage labor as their ing at Mogtedo and Mogtedo-Bombore; the next primary occupation; another 10 percent listed most frequently cited problem was insufficient services or trade as their secondary occupation. and unreliable water points. Women settlers who Twenty-seven percent of the immigrants to the migrated from the AVVsponsored settlements at Niangoloko subsector cited the area's opportuni- Mogtedo to Kompienga expressed a high level of ties for trade and employment as the principal satisfaction with their move, despite the more re- feature that attracted them. Eleven percent of the stricted opportunities that they had for private 455 heads of household at Kompienga in August crop production. The chief source of their satis- 1989 indicated that trade, fishing, handicrafts (in- faction was the active town market. cluding dolo making and cooking), and services The largest markets occur when the three day were their primary occupation; another 6 percent market cycle coincides with religious holidays, of the household heads in the town listed char- allowing the immigrants to combine the social coal making and woodcutting as their primary role of markets with the very important social occupation; 20 percent of the heads of household role of religion. Sunday markets are always large, listed services, handicrafts, or trade as their sec- as most Christians are conscientious about not

80 working on Sunday, regardless of the period in have initiated construction either in Ouaga- the crop cycle. On these days, Christian migrants dougou or in a smaller urban center. Those who and indigenous inhabitants first mingle at the lo- build in Ouagadougou usually are merchants. cal Catholic and Protestant churches, then ad- Only a few farmers have actually left the Li- journ to the market. When market days coincide noghin area to engage in full-time trade; the more with Friday, Muslims from throughout the region typical pattern is for part of the family to leave pray together, then attend the market, an event and the other part to remain and continue to op- especially noticeable at Linoghin, with its new erate the farm. mosque located just outside the market. The lower rates of permanent out-migration As settlers' incomes grow, and they increase from Linoghin appear to be strongly related to their levels of investment in areas outside of agri- the much greater local opportunities that exist culture, they also expand the range and nature of there for diversification. Those opportunities ex- their social and economic linkages, an occurrence ist because of greater proximity to and settler in- invariably reflected in the scale of outside atten- tegration with national and local markets. dance at local markets. Based on average dry season attendance, we classified the markets PossibleNegative Impact of Marketson Settlement attended by the settlers at Mogtedo, Mogtedo- RelatedDevelopment Bombore and Linoghin, and those in the Solenzo region into four categories (Figures 8.1 and 8.2): Ease of access to active markets and adminis- major regional, subregional, area village, and trative centers can have negative consequences as simple village markets. well. The factors that encourage the development It was common knowledge that eight of the 35 of dynamic market and administrative centers wholesale grain merchants who purchased grain (ease of access to all-weather highways and con- at Mogtedo, and two of the seven wholesale live- centrations of population) are also factors that en- stock merchants, were based in Ouagadougou. courage deforestation for fuelwood. Similarly, most of the settler-owned boutiques lo- The opportunities for diversification created by cated around the Linoghin market are rented by markets can attract spontaneous immigrants to merchants from Ouagadougou. Even on a rainy river basins that would not ordinarily be consid- market day in August 1989 we counted about 70 ered attractive because of their isolation or their outside merchants (three full busloads) from lack of opportunities for cash cropping. Thus we Puetenga at the Mogtedo market. Market officials found that the highest proportions of agricultural at Linoghin estimated that 35 percent of the per- immigrants to the Niangoloko region are in vil- sons attending the Linoghin market during a lages with easy access to the Niangoloko market; "typical" late August market came from outside immigration rates to more isolated villages are the area. The tin roof on the Linoghin market still low. Similarly, the highest rates of immigra- mosque was paid for by one of the wealthiest tion to the Kompienga Basin are to the town of merchants in Burkina-a wise investment on his Kompienga and to the larger villages along the part in establishing social linkages with a group road that also have larger markets. AVV blocs of of farmers and merchants in an area that is obvi- Linoghin and Rapadama with easy access to the ously going to develop into a major market and paved highway to Ouagadougou have attracted administrative center over the next two decades. more spontaneous settlement than distant villages. Because active markets attract greater num- Impact on Reinvestmentof IncomeEarned from bers of spontaneous immigrants, villages near CommercialFarming and LivestockProduction market centers are almost always the first to experience the effects of environmental degrada- The most successful settlers in the AVV planned tion associated with over cropping. settlements tended to invest in real estate and trade in other areas. A total of 21 male heads of FactorsThat Affect Market Dynamics household in just two of the planned settlements at Linoghin have either started or plan construc- An effective market and administrative center en- tion in Ouat adougou or at Ziniare, the regional ables area producers to sell their surpluses in a capital of their adjacent home area. We estimate timely fashion; to obtain the agricultural and that 10 to 15 percent of the Linoghin households manufactured commodities and services they

81 need, but do not produce; to invest in the devel- istic that distinguishes the markets in which we opment of nonfarm services, manufacturing, or see a high degree of settler integration or social trade; and, ideally, to have access to certain basic access is that they generally did not exist prior to social services that are not available in the villag- the start of large-scale immigration. We also es. The effectiveness of markets is influenced by found successful integration in areas where pre- numerous factors. existing markets were small, and it was the immi- * Location on a major highway. By far the most grants, rather than the indigenous inhabitants, important feature that facilitated development of who built the major commercial enterprises. a market was whether it was located on a major AVV settlers created the market at Linoghin. highway. Two of the largest regional markets at Although the market has grown to include over the UPI were on the major highway to Ouaga- 1,000 buyers and sellers on a regular basis, it re- dougou. The active market at Daboura, near So- mains a settler-created and settler-run market. By lenzo, was also located on a road linking the welcoming indigenous inhabitants and sponta- department headquarters for Solenzo with the neous agriculturalist and pastoralist immigrants provincial center at Dedougou. The historic lack to the market, the settlers gained new friends in of development of a major market center south of the region. Fada, near Burkina's borders with Togo and Be- Similarly, increased market activity in the So- nin, can be attributed to lack of a good all-weath- lenzo region could be traced directly to immi- er highway linking the zones. grant Mossi. Many Mossi were active traders and * Location in a frontier zone between different the primary people owning boutiques in the new types of settlement. The AVV villages that devel- villages. oped markets were generally located in the fron- Increased market activity in the Kompienga tier area between sponsored settlements and district was also attributable to Mossi migrants. spontaneous agriculturalist and pastoralist settle- Extension agents emphasized that the Kompiem- ments. Project markets that were created in areas biga market did not even exist until Mossi mi- more centrally placed vis-a-vis a group of grants arrived and started it. planned settlements (usually right in the middle) In contrast, the active regional market at Mogt- were often abandoned (Figure 8.1). A recent edo was not created by immigrants. In 1974, study of markets in the Solenzo subsector shows when the planned settlements started, the market that larger markets tend to be concentrated either was very small, with fewer than 100 open-air along the main highway or in the "frontier" zone straw stalls or hangars. Nevertheless, the town of between the highway and the new settlement Mogtedo did exist, and there was already a core taking place toward the Mouhoun (Figure 8.2). of town merchants. Since that time both the town The Agrotechnik research on markets in the Kom- and the planned settlements have developed, but pienga Basin shows a high concentration of vil- on parallel tracks. Immigration to Mogtedo town lage markets near the main highways with al- has largely been from the more populated areas most no interior or frontier markets (Figure 8.3). of the Ganzourgou region-a pattern similar to * Social access to markets. Location near a mar- that observed for the spontaneous settlements ket is not sufficient to bring about effective mar- around the AVV sponsored settlements. Al- ket integration. The settlers must also have what though the majority of settlers living in the spon- we describe as "social access" to the market. The sored settlements (outside the town) are from the more distant Linoghin sponsored settlements are same ethnic group, fewer than 30 percent are located almost as far from the market at Linoghin from the Ganzourgou region. While the settlers as are the Mogtedo planned settlements from have been welcome to spend their money in the Mogtedo. The key difference between the two is town, efforts to build commercial linkages with social access to the market. At Linoghin, settlers the town have been thwarted. That is to say, al- were actively involved in both buying and trad- though they have physical access to the Mogtedo ing. At Mogtedo, settlers were socially and eco- market, they have not had social access. nomically excluded from the fullest possible * Concentration of civil servants and paid work- participation in the market because of their hos- ers, and presence of services at market locations. The tile relations with the indigenous inhabitants. presence of groups of civil servants in villages The factors that contributed to or detracted facilitates market development. Civil servants from social access were complex. One character- need food and condiments and create a secure

82 base of demand. The concentration of AVVtrain- al markets that offered higher prices for their ing facilities at Mogtedo was definitely a factor in crops and livestock, and lower prices for the con- the early development of the market. Civil ser- sumer products they wanted to buy This did not vant and worker patronage attracts attention to a improve the situation for most women settlers, market and helps it grow. Again, however, al- however, as they were still, by and large, restrict- though the presence of civil servants can facilitate ed by lack of transportation to markets in the im- development, it is insufficient in and of itself to mediate area. support a large, dynamic market. One of the settlers' recommendations for ad- Another factor that contributes to the strength dressing this issue was that planners reinforce the of a market is its location near other sorts of ser- development of small, settler-created markets vices and infrastructure. Location near distribu- from an early date by carefully phasing the devel- tion points for agricultural inputs and near mills opment of basic infrastructure and roads. It is this was particularly important. These facilities draw overlay of services, markets, and infrastructure buyers and sellers to the market and reinforce its that seems to create strong decentralized market many roles. and service centers. Without these inland centers, * Dateof creationof a marketin relationto the set- the basins become undiversified production ar- tlement cycle. Farmers in the older AVV settle- eas, from which benefits are merely drained off to ments suggested that the date of creation of a the paved highway and funneled into the capital. market in the settlement cycle is extremely im- Ceteris paribus, phased infrastructure develop- portant. The AVVvillages were created quickly- ment would be easier to carry out in areas of in a period of two to three years, in most cases. spontaneous settlement, where settlement takes Although the AVV had clearly delineated a mar- place more gradually, than in areas of sponsored ket site in the plans for each village, in most cases settlement. little interest in creating a market emerged until * Gold Mining. The discovery of gold usually the third to fifth year of settlement. Settlers com- leads to the development of a market at the min- plained that they were too busy clearing their ing site. The new market center houses the official fields to worry about creating a market. If they gold purchasers as well as places where miners needed something, they would leave the bloc and can spend their money on food, water, fuelwood, attend the Mogtedo market or one of the smaller imported consumer products, and various servic- indigenous markets in the area. As a result, the es. A very large market has developed at Nobsin, settlers tended to fuel the development of these the major gold site. We counted more than 40 markets, rather than their own. By the time they merchants at the Nobsin market site in August were ready to establish their own markets, their 1989,when mining was supposed to be officially small, new markets had to compete with much closed to encourage the miners to return to farm- stronger area markets that they themselves had ing. During a later census, in early April 1990,we helped create. By this time, local demand for mar- counted 75 merchants at the four principal sites in kets was also reduced because a growing number the Mogtedo region (Table 8.3). Smaller markets of males had been able to purchase bicycles and/ with about 20 to 30 merchants (gold purchasers, or motorcycles. With better transportation, male water carriers, food vendors, merchants selling settlers were able to frequent more distant region-

Table 8.3 Number of Miners and Merchants at the Mogtedo Area Gold Sites, April 1990

Merchants by origin by actitvity Site Miners Total Ouaga Mogtedo Boutiques Petty trade Meat Rice Water A 400 42 7 35 3 17 5 7 10 B 160 14 2 12 0 5 1 3 5 C 100 13 1 12 0 4 2 4 3 D 45 6 0 6 0 1 2 1 2 Total 705 75 3 27 10 15 20 Source: Land Settlement Review.

83 clothes and metal cookware) have sprung up at are not very optimistic that the sudden infusion of some of the less important mining sites. off-farmincome into the AVVplanned settlements As mentioned earlier, the income earned from at Mogtedo will be reinvested in ways that will im- mining and selling products at the mining sites prove village income earning opportunities. has raised the living standards of women and of We expect that the most lasting effects of min- unmarried and married males attached to (but ing will be to solidify Mogtedo's role as an im- not heading) family units in the AVV planned portant regional trade and administrative center. settlements at Mogtedo. It is symptomatic of the In late March we again counted the number of low level of development of services and off- people selling at the Mogtedo market in order to farm employment in the AVV planned settle- see how large the market grew during the dry ments, however, that almost none of the entre- season. We expected the market to be almost preneurs who have moved in to provide services double the size we recorded in August 1989. To for the miners are former or current AVV set- our surprise-and to the surprise of the presi- tlers. Instead, the largest merchants at the sites dent of the local Revolutionary Committee-the were nearly all from the more distant town of total number of people selling in the market cen- Mogtedo (Table 8.3) rather than from the adja- ter and surrounding areas had decreased. The cent planned settlements. principal factor explaining the drop was that Because of the danger that we might be per- farmers and merchants were being attracted to ceived as government spies, we were unable to the very active markets at the gold sites. As the explore the issue of gold revenues in great detail. largest proportion of merchants and tradesmen Settlers were reluctant to tell exactly how much at these sites seem to be from Mogtedo, the in- they had earned when we interviewed them for come earned from these activities will presum- the economic survey. We did find that women ably be reinvested in that town. It would be who reported income from gold earned an aver- wrong, however, to assume that the dynamic de- age of about 14,000-20,000CFA, with some wom- velopment of Mogtedo is the result of gold min- en reportedly earning as much as 60,000CFA in a ing. Mogtedo had emerged as a major market single season (Table6.3). center long before the first rumors of gold were We also conducted more detailed interviews heard and gold fever became rampant in late with 15 male gold miners. The miners, who were 1987 (see Table 8.1). interviewed by another settler, were guaranteed * Roleof infrastructure.Contrary to our expecta- complete anonymity, including age and place of tions, infrastructure appeared to be less important origin. The recorded income levels over the last than location in the growth of the larger village year, along with indications of how this income markets that were away from the main highways. has been spent, show half the miners earning Efforts to create AVV markets along the important 50,000 CFA or more (Table 8.4). The interviews secondary roads linking the AVVblocs to the town confirm our observation that most of the settlers' of Mogtedo never worked (see abandoned mar- money is going toward consumer goods rather kets at Mogtedo V3, Mogtedo VI, and Bombore than reinvestment in agriculture or in developing V3). In contrast, the AVV markets located away trade or service activities in the region. Thus, we from the paved highway that flourished in the short run (V4-5Mogtedo) or for longer periods (V5 Bombore) were located in the frontier areas men- Table 8.4 Reported Net Income Earned from tioned earlier. In these villages, ease of access to the Gold Mining for 15 Miners main regional market did not appear to be an absolute necessity. Ease of access, for example, was definitely not a factor in the development of War- 1-0,000 l l dogo, the third largest regional market in the area. 10,000-20,000 3 For a great part of the rainy season, the only way 20,000-50,000 2 to reach that market was by crossing the Nakambe 50,000-75,000 3 tributary by boat. Wardogo was very favorably po- 75,000-100,000 3 sitioned, however, as a key market in the border 100,000+ 2 area between planned settlements, spontaneous Total 11 4 agriculturalist immigrants, and pastoralists. Thus, Source.' Land Settlement Review, while permanent year round highway access to a

84 market is important, it is possible for large markets kets, planners can discern what services and in- to develop in more isolated areas-if these areas frastructure investments will be needed to are important centers of exchange between differ- enhance the facilities in an area. It is important for ent land use groups. development to be phased in, and sensitive to im- * Level of governmentinvolvement in creationof portant shifts in market dynamics. the market. No direct relationship was observed between government planning for markets and SpecialMarket PlanningProblems in Areas of the patterns of markets that developed in the ar- SponsoredSettlement eas surrounding the AVV planned settlements. Indeed, there appeared to be something of an in- Phased reinforcement of market centers will verse correlation between government planning probably be easier to carry out in areas of gradual and markets that survived for more than a few spontaneous settlement than in areas of spon- years. The AVV had focused on placing market sored settlement or near dams where large areas and service sites centrally with respect to the new are apt to be settled in a very short period of time. infrastructure built for a particular group of self- Given the importance of socioeconomic centers, contained planned settlements. One of the signif- planners dealing with sponsored settlement in icant lessons of the AVV is that in order to suc- more isolated regions should consider the option ceed, markets must incorporate a much wider of a staged approach to market center develop- public, including spontaneous agriculturalists ment. This option might include two stages: first, and pastoralists. The growth of markets on the linking a small sponsored settlement core having periphery of the sponsored settlements high- a center with outlying areas of spontaneous set- lights their usefulness at the interfaces of the tlement; then, in a second stage, reinforcing the different types of settlement that shape develop- development of roads and markets that may de- ment in a particular area. velop in other locations in response to wider area settlement and development after ten or fifteen Policy Implications years.

Reinforcethe Developmentof Regional SpecialPlanning Needs of RegionalTrade and Market Capacity Administrative Centers

Markets in the OCP river basins form a hierarchy Planners must anticipate the impact of successful in terms of periodicity, number of participants, settlement related development on existing mar- volume, nature of commodities and services ex- ket and administrative centers. In the short run, changed, and radius of the market's drawing successful settlement is likely to expand the size power. Price analysis is a simple tool that econo- of area markets. In the long run, the associated in- mists have used for generations to identify ineffi- crease in opportunities to develop services and ciencies in market dynamics. This same tool trade will attract increasing numbers of mer- could be used to alert policymakers to marketing chants and expand the number of permanent bottlenecks that reduce the development poten- boutiques and the demand for housing. Thus tial of new lands settlement in an area. planners can predict growing competition for commercial and housing space. Planners should Phasein DevelopmentPlanning in Orderto Reinforce thus anticipate a growing need to delineate com- the EvolvingMarket Hierarchy mercial and housing areas. Urban zoning, if carried out by local leaders Market hierarchies are not fixed, but change over working with the administration, can protect the time, along with patterns of settlement and settle- rights of indigenous inhabitants while it improves ment related development. The existing hierar- migrants' social and economic access to develop- chy of markets provides planners with a point of ing urban markets. Such planning is probably bet- departure for efforts to assist settlement and to ter carried out in the early stages of urban growth promote a broader base of area development. By rather than later, after problems have begun to examining market growth in relation to settle- emerge. The only example we had of early plan- ment areas, and product prices in these markets ning is at Linoghin, where commercial space near in relation to product prices in less isolated mar- the AVVmarket was zoned from the very begin-

85 ning, and allocated as registered parcels. Spontane- Note ous immigrants and indigenous inhabitants who wanted space in the market were able to apply for 1. In the legend on Figure 8.2, where two years appear next to a and receive one of the registered spaces. This early market, the first date indicates the year the market was created; zoning was one of many factorsthat facilitatedthe the second marks the year the market was moved. peaceful growth of the Linoghin market.

86 ~~9

Integration of Hosts, Settlers, and Pastoralists

The Importance of Integration source of green vegetables in the daily sauces in most settler diets, and medicinal herbs, herb spe- Settlers moving into an OCP river basin are cialists, and healers. Learning from their hosts aware that their long-term success is linked to about the ritual status of specific sites, appropri- their ability to become integrated into regional ate areas for burial, market sites, and the local hi- and local economic and social systems. These sys- erarchy of political and social groups, helps tems engage settlers in a wide range of relation- settlers feel more at home (See McMillan 1983: ships with (McMillan, Painter and Scudder 1990): 205-207). For their part, settlers provide new mar- (1) the indigenous agriculturalists or hosts kets for host farmers' products, contributing to who lend them land; the well demonstrated correlation between new (2) pastoralists who have traditionally grazed lands settlement and expanded area markets. Set- their animals, and had villages or camps tlers also stimulate area development to the ex- in the area; and tent that they bring information or experience (3) other immigrant agriculturalists and pas- with such new technology as plows, tractors, cot- toralists. ton (Niangoloko, Solenzo), or new linkages to toalst.. .. otietaeadamnsrto(Naglk Mutually beneficial linkages are different at outside trade and administration (Niangoloko, each stage of the settlement process. To design Solenzo, Kompienga). The first immigrants in some of the Solenzo, more effective programs, policymakers need to . f . s . be aware of the relationships between the three Niangoloko, and Kompienga villages were often groups, and of the factors that contribute to or de- sponsored by host families. The typical pattern ' . . . ~~~~~~~wasfor a group of kinsmen to request land from tract from the peaceful integration of their differ- a o ru fknmnt eus adfo enactinmthers pan indigenous inhabitant. Early immigrants then typically sponsor the immigration of later settlers from the same home area. Even when this occurs, the sponsor will usually take the newcomer to the A complex interweaving or synergism exists be- same friendly sponsor who helped him or her to tween settlers (agricultural and pastoral) and first acquire land. The settler family thus increases hosts who provide access to land (Figure 9.1). The its allies, while at the same time increasing the al- inhabitants also provide settlers with lies of its own sponsor. By granting land, the in- indigenous dintnsalopoleseteswt information on suitable settlement sites, soil qual- ndgenous sponsor gradually increases his power ity, and water sources. They are also knowledge- in the community. If immigrants and the indige- able about gathered foods, which are the main nous population are from the same or related ethnic groups, these social bonds will often be

87 Figure 9.1 Integration of hosts, settlers, and pastoralists

Political Draft animal Protection Purchases Grain sales Plowing and gifts Services Cash gifts I~~~~~~ I Animal health services Hosts

Protection of animals Information on water and New technology Manure soil quality, indigenous foods, crops, medicines, New political and Milk and rituals institutional linkages Herding Political prestige

Pastoralists -_ Draft animal sales Product purchases and sales Product purchases and sales Livestock commerce Politicalprotection Grain purchases Livestock Market information Settlers

Politicalalliance A

Political Draft animal protection purchases Plowing Grain sales services and gifts

Cash gifts

Source: McMillan, Painter, and Scudder 1990.

88 reinforced by later intermarriages between the lin- acquire cultivation and grazing rights. This ap- eages of indigenous grantor and migrant grantee. proach to acquiring land rights helped create pos- Such sponsorship and mutual aid typically pro- itive relationships between indigenous vide the power base for the first generation of in- inhabitants and area pastoralists, minimizing digenous and migrant political leaders who guide conflict and promoting peaceful collaboration. the new hybrid communities. The value placed by settlers on peaceful, mutu- Agriculturalist-PastoralistRelations over Time ally beneficial integration with the host community is reflected in the tendency of spontaneous At all of the sites, the first immigrants to the re- settlers to continue traditional land tenure prac- gion were pastoralists. Unfortunately, because of tices despite a 1984 change in national land ten- difficulties in identifying the immigrant pastoral- ure law. Settlers know that official land title or ists, the site reports often underestimate their role not, they will have few opportunities for a satis- in area land use patterns and settlement related factory social or economic life if they have antag- development. onistic relationships with the local people. Nana and Kattenberg's research on patterns of spontaneous settlement showed that pastoralists Integration of Hosts and Settlers with were slowly moving into the upper Nakambe re- Local Pastoralists gion near Linoghin, Mogtedo, and Mogtedo-Bom- bore in the late 1960s(Nana and Kattenberg 1979), Successful settlement related development in long before the AVV planned settlements were Burkina is invariably linked to increases in the created. The combination of access to local and re- size of livestock herds of agriculturalists and pas- gional markets, availability of pasture land, isola- toralists and to greater linkages between the agri- tion from agricultural settlements, and access to culturalists and pastoral populations (Figure 9.1). water made the area attractive to pastoralists in Livestock continues to be the principal source of the rainy season. The late sixties showed a grow- investment for rural farmers; it augments their in- ing number of pastoralists shifting their "perma- come as well. For pastoralist populations, live- nent" rainy season camps to the river basins. stock provides their principal source of income By 1979, a second wave of pastoralists, attract- and-to the extent that livestock or livestock ser- ed to the area by the prospect of working as herd- vices are exchanged for grain or used for meat ers for the newly rich AVV settlers, flooded into and milk-food. Livestock manure is an impor- the upper Nakambe basin. In the AVV planned tant part of any long-term strategy to develop settlements, the "herder" pastoralists were gener- sustainable cropping systems. Small livestock are ally poorer families who had lost all or part of also one of the most important sources of income their herds through drought, disease, or some and investment for women. other personal misfortune. Guarding the wealth- High levels of conflict between pastoralists and ier agriculturalists' animals is a way of reconsti- agriculturalists hurt the agriculturalists' chances tuting their herds. In return for working as of negotiating herding and manure contracts. Lo- herders they receive grain and cash. From the cal pastoralists, in turn, can be penalized with larger livestock owners, herders also receive a high fines for the destruction of farmers' crops by percentage of the young animals born under their animals, although determining whose animals care and have rights to drink and sell the animals' are at fault is complicated because such a large milk. number of itinerant pastoralists move in and out Herding out their cattle to pastoralists has four of the settlement areas. major benefits for hosts and settlers (McMillan, Although pastoralist-agriculturalist conflict Painter and Scudder 1990). One is better use of was a factor at all of the sites, it was more pro- family labor, since the demands for herding fre- nounced in some. In general, factors that seemed quently are beyond the capacity of farming house- to promote more peaceful relations between pas- holds (Delgado 1979).Second is the reduced risk toralists and agriculturalists were some of the of theft. High labor demands and risk of theft same factors that promoted peaceful relation- were two important reasons why about one- ships between migrants and indigenous inhabit- fourth of the families in the AVV settlements no ants. Especially important was whether the longerhavedraftoxen,despiteeasyavailabilityof pastoralists approached indigenous leaders to credit with which to purchase animals. A third

89 benefit is better care of livestock, because herders permanent camps around the AVV sponsored have knowledge of and accessto more distant for- settlements. age, as well as a better understanding of local me- Another option for pastoralists without herds dicinal plants and modem preventive measures. of their own is sedentarization, and with sedenta- The fourth is a reduced likelihood of conflicts over rization comes increased cropping. The Solenzo crop destruction, as animals are herded away settlements, for example, include a sizable num- to from fields during the cropping season. ber of pastoralists farming at levels similar In the AVVvillages, pastoralists generally pre- those of agriculturalists. Three of the four FulBe the ferred to herd the animals of one powerful family pastoralist households that were included in or a group of related powerful families. Pastoralist farming systems survey had adopted animal households thereby gained political protection in traction. Savadogo (1989d) observed that crop- addition to the other benefits of the relationship. ping systems of the sedentarized FulBe resem- The wealthier agriculturalists' interest in facilitat- bled those of the agriculturalists. Passing through to ing integration with the herders is reflected in an the customary authorities to obtain permission incident we observed in one of the AVV planned graze and farm may have contributed to the suc- settlements. When the first rumors circulated cessful integration of pastoralists in Solenzo. about creating a village land management pro- Since 1983an increasing number of pastoralists gram, wealthier settler leaders took the initiative have oriented their immigration away from the of encouraging the FulBe pastoralists to organize more saturated northern basins of the Nakambe, to ensure that their needs would be considered. Nazinon, and Mouhoun (former White, Red, and The same success that attracts livestock herders Black Volta Valleys). This trend was reflected in re- tends to attract more immigration by agricultur- the case study research, which shows a fairly alists. This increases pressure on crop and pasture cent increase in pastoralist immigration into the resources over time. The degree of pressure de- Kompienga Basin (Agrotechnik 1989). Large- pends on the rate of immigration, which in turn is scale pastoralist immigration to the Niangoloko a function of other attractive features (e.g. oppor- subsector and the Toumousseni Classified Forest tunities for trade) that may attract settlers to an is also recent. area. As population densities increased, the first In sum, pastoralist immigration to Burkina's generation of pastoralists' access to water and OCP river basins follows a regional trend toward pasture became more circumscribed. the shift of transhumant patterns further south. The emerging "crisis" was accompanied by a Countries such as Ghana, Benin, Togo, and C6te steady escalation in accusations and formal litiga- d'Ivoire received a growing amount of spontane- tion concerning livestock damage to crops. We ous pastoralist immigration from the Sahelian can surmise that a similar pattern of escalating countries in the 1970sbut, by this move south, the conflict developed at some point in the past in the "FulBe problem" was merely exported, not re- Solenzo subdistrict. solved. Burkina's OCP basins are now catching Faced with growing conflict, pastoralists must the "backwash" of this earlier immigration. In choose between leaving or developing new pat- late 1987, several thousand alleged Burkinabe terns of closer, and generally more sedentarized, pastoralists-virtually the entire FulBe popula- interaction with agriculturalists. Pastoralists with tion of Ghana-were expelled from that country significant herds that make them less dependent by government edict. The expelled FulBe were on the activities of neighboring agriculturalists quickly attracted by the prospect of huge water generally chose to move on. When such migra- resources and markets near the Kompienga Dam. tion occurred in the area around the AVV spon- This sudden increase in the number of cattle and sored settlements, a substantial growth in market of FulBe put additional pressure on the Kompien- activity appeared in the frontier zone between the ga Basin's already stressed pasture and water re- sponsored settlements and the river (Markets 3,5, sources and on agriculturalist-pastoralist social 7,16,18, and 22 in Figure 8.1). relations. Emigration away from advancing agricultural- Social disturbances between agriculturalists ists is not a choice for pastoralists with small and pastoralists also caused many pastoralists to herds who are forced to work as herders. It is flee across the C6te d'Ivoire border in late 1988- these pastoralists whom we find still living in some in such a hurry that they abandoned their herds. The repulsed pastoralists fled into the OCP

90 basins creating a sudden, dramatic, almost imme- tion densities, and increase as settlement densities diate increase in the area's pastoralist population, increase. in agriculturalist-pastoralist conflict, and in the * ImmigrationRhythms. Indigenous social and threat of livestock depredations in the area's clas- economic institutions can effectivelyaccommodate sified forests. high rates of immigration if it occurs steadily over a long period of time. Sudden, massive streams of FactorsThat Affect Integration Dynamics immigration-of pastoralists or agriculturalists- tend to overload the capacity of local social and Various factors contributed to or detracted from economic institutions to negotiate mutually benefi- the peaceful integration of hosts, settlers, and cial land use patterns. The fairly sudden break- pastoralists into regional social and economic down of these systems, and parallel increase in systems. social conflict,can be seen at Kompienga. * Authorization of Land TenureRights. One of * Present Role of Village Extension Groups. In the principal factors influencing host-settler and most areas of Burkina, farmers are required to be- host-pastoralist relations was whether or not set- long to a settler extension groupementin order to tlers passed through the customary channels to acquire credit. The groupementis then responsible acquire land. for insuring that individual settlers repay their * Same Area of Origin. Other factors being loans. To acquire credit, settlers are forced to join equal, there are fewer problems between settlers, the same groupementsas the indigenous farmers. and between settlers and hosts, when they come This shared membership has reinforced the social from the same area of origin. The fact that the and economic integration of hosts and settlers. AVVsettlers at Linoghin were from the same Ou- These groupementstructures have been less ef- britenga region as the indigenous inhabitants fective for incorporating pastoralists. While sever- whose land they were granted by the AVV was al of the sedentarized pastoralists living at one of several factors that reduced conflict be- Daboura were equipped with animal traction, our tween the two groups. By 1989, we counted: impression from other interviews was that this (1) five daughters of spontaneous immi- tended to be the exception rather than the norm. grants who were married to AVV settlers In general, the pastoralists have not been readily in the Linoghin Planned Settlements, Vil- incorporated into existing groupements. Instead, lages 1, 2, and 5; most official dealings with the pastoralists have (2) six daughters of AVV sponsored settlers focused on special pastoralist groupements. who were married to spontaneous mi- * Religion.Religion was frequently cited as a grants; and factor that worked both for and against peaceful (3) four daughters of indigenous inhabitants integration. First, the animal sacrifices made of the region who were married to AVV through their sponsors when immigrants re- sponsored settlers, also in Villages 1, 2, ceived land were an important sign of their re- and 5. spect for (not necessarily adoption of) the local customs and authority of the indigenous people. This pattern of intermarriage and increasing Seod prciigtesm.ol eiin sociallackofsocial integrationintegration contrasts sharplysheAply with ahed the ggross s whetherSecond, practicingCatholic or the Muslim-encouraged same world religion- col- lack of social integration in the AVV planned set- laboration. Religious instruction was a third way tlements at Mogtedo and Mogtedo-Bombore. that religion provided a mechanism to promote Shared area of origin and, even more impor- social integration. Agriculturalist (or, in the case tant, long-term acquaintance, also influenced the of the Classified Forest at Toumousseni, pastoral- AVV settlers' cioice of which FulBe families to boarthi anml wit. lst) settlers who were Koranic masters would board theirwith. animals ~~boardchildren of other settlers as well as the chil- * Distancefrom Indigenous Villages,Rate of Im- dren of othersers a thes chi- migration,and Pre-existingPopulation Densities. The dren ofnidigenous farmers i the region. This Avv'AVV plannedplne setmnssettlements thattha werewer createdcreae in lessles wherewas considered they had previouslyto be an asset had forto send hosts their in areaschil- populated areas have had fewer conflicts with the hre they pares h theo ir chil- indigenous inhabitants. Similarly, conflicts over to other parts of the country for Koranic livestock are generally minimal at lower popula- schoolig. Just as religion can facilitate, it can also work against settler integration. In the Solenzo area,

91 for example, the Yatenga immigrants tend to be habitants. After 1983, the indigenous chieftain- elected Muslim; the indigenous Bobo, animist. Religion cies were theoretically replaced by the was usually offered as the reason that the two CDRs (Committees for the Defense of the Revolu- as a groups did not intermarry, despite long years of tion). Since the immigrants tended to vote living together. block, they would dominate the CDRs in the vil- switch Strong pressure was exerted on AVVsettlers to lages where they were predominant. This convert to either Christianity or Islam when they in the political hierarchy almost always brought in immigrated to the AVV.In most cases, the pat- about a fairly immediate and direct increase terns of conversion followed political lines. migrant-host conflicts even in villages where In one AVV village, for example, almost half hosts had always been quite hospitable. the settler households were from the same home Similarly, one of the unintended consequences to village, and over half those settlers from one of of the 1984 Agrarian Reform Act has been seven clans in that same home village. The three make agriculturalists more reluctant to host they clans most closely identified with the chieftaincy transhumant pastoralists on their fields, lest were almost exclusively Muslim. Each of the oth- then try to claim residence as "permanent" inhab- cul- er clans, without any historic claim to the home itants and acquire "permanent" pasture and true in village chieftaincy, tended to be a mixture of ani- tivation rights. We found this especially mist-Christian. When they first immigrated, one the Kompienga region where area villages have ex- of their leaders suggested that the entire group been inundated by "stranger" pastoralists convert to Islam as a means of reinforcing the fact pelled from Ghana. they were from the same home village. The non- Muslim groups were unanimous in their refusal. Policy Implications One of the reasons given for this was the associa- is tion of Islam with the traditional chieftaincy.One The most successful development planning settler said: likely to be that which strives for a balance in pro- groups In the home village we were under the sway of moting the economic interests of all the as the traditional chiefs. Why should we convert affected-host and migrant agriculturalists, and put ourselves back under them? well as pastoralists. Interventions or events that * Markets. If settlers were instrumental in the threaten the economic, social, or political security for creation of new markets that benefitted the indig- of any one of these groups reduce the chances enous people as well as themselves, this eased successful integration. Lack of successful integra- economic and social relations between the differ- tion, in turn, reduces the chances that immigrants from ent groups. We observed this at Linoghin, Komp- will invest the first generation of returns iembiga (an immigrant community near new lands settlement in the development of more Kompienga) and Kompienga town, and through- sustainable land use practices. Successful integra- groups out the Solenzo and Niangoloko region. Several tion increases the chances that the three pastoralists invested in large boutiques in the will be willing to collaborate to resolve future ag- Mogtedo market. There were also a number of problems with soil degradation, reforestation, re- prominent pastoralist merchants at the Kompien- riculturalist-pastoralist conflicts, and natural be ga market. The active social and economic partic- source zoning. Most likely to succeed will rules and ipation of the pastoralists in these markets programs that combine land tenure culti- provided one of the most important means for zoning (to protect the rights to pasture and immigrant pastoralists and agriculturalists to be- vation areas of both immigrant and indigenous to come better acquainted with one another. The agriculturalists and pastoralists) with efforts markets-especially the Kompienga market- reinforce the capacity of local institutions. also provide an important center for the nonfor- mal adjudication of disputes between pastoralist Need to PromoteVillage Zoning for Agriculture, and immigrant leaders. Pastoralism,and CommonResource Management * Political Allegiances and Political Change. Ter- Government reforms to promote more democrat- Burkina's ProgrammeNational de Gestiondes land ic village governance have sometimes had the un- roirs(PNGT) is a promising model for village the intended consequence of creating new sources of use management that is designed to facilitate The conflict between immigrants and indigenous in- integration of hosts, settlers, and pastoralists.

92 PNGT model, which is described in greater detail pally because the zones created areas where they in Chapter 10, combines zoning (to delineate land could live without competition and conflict with to be used for crop cultivation, livestock,and man- agriculturalists. Once pastoralists throughout the aged and unmanaged forests) with the creation of valleys had heard of the existence of such (and outside institutional recognition of) a village projects, they kept asking us if a program like the land management committee. The land manage- one at Nouhao would be created near Kompien- ment committee is supposed to include elected ga. In the pastoralists' opinion, the zones repre- representatives from all of the major social groups. sent their best chance for living peacefully and In theory, only village pastoralists (that is, pasto- protecting their lifestyle. ralists who have settled permanently near a vil- An intensive case study, focusing on the evolu- lage and who farm) and village agriculturalists are tion of a small sample of family groups within entitled to keep their animals in the village area these agropastoral zones, would make an excel- zoned for pasture. Under the model, any iimi- lent contribution to the literature on settlement grant family-pastoralist or agriculturalist-must and development in the OCP river basins. Unfor- pass through the committee to acquire land rights tunately, we were able to conduct only a brief sur- in the village. Once constituted and officially rec- vey of ten pastoralist families who had been ognized, the village land management committee living in the Gadeghin zone for two agricultural can place certain restrictions on the acquisition of seasons and in the general area for periods of five land rights, such as specifications concerning land to ten years. We were impressed by the pastoral- use practices and the amount of land that can be ists' enthusiasm for the project and the efficiency legally cleared. with which the pastoralists had organized their If implemented early in the settlement process Revolutionary Committees (CR) to patrol the rather than later as part of an effort to rehabili- zone boundaries in an effort to keep out non- tate land that has already been damaged by ex- member pastoralists and their herds. The CR and tensive cultivation (such as our study sites at village land management committee also have Daboura near Solenzo and the area surrounding successfully met to discuss various problems the AVV sponsored settlements), the PNGT with animal depredations, manure contracts, and model offers an unusual opportunity to protect herding arrangements with settlers living in the interests of indigenous inhabitants by allow- sponsored settlements as far away as Mogtedo- ing them a certain amount of control over immi- Bombore. Especially impressive was the fact that gration onto their land. Second, it gives settlers the herders appear to be organizing these activi- some legal recognition of their claims. Third, it ties on their own with only minimum directives protects village pastoralists and settlers who from the AVV-UPI1staff. We were also impressed own cattle by assigning them a clearly delineat- by the fact that agriculturalists in the surround- ed area for pasture. ing area seemed to support the concept. In the agriculturalists' view, if the Gadeghin Needfor LargeAgropastoral Zones scheme is successful, it will provide a more stable base for agropastoralism in the Nakambe basin. Zoning village lands for agriculture, pastoralism, Although the Gadeghin settlers are expressly for- and common resource management is not a via- bidden to board agriculturalists' animals within ble alternative for settlers with large herds. For the agropastoral zone, this is difficult to enforce. this reason, most area planning models need to Nor is this necessarily desirable since area farm- include the delineation of one large contiguous ers would naturally prefer to board their animals region reserved exclusively for agropastoralists with a nearby family that they know will remain with large herds. in the region. In turn, pastoralists who plan to re- Burkina Faso has launched several of these ex- main in the zone have a vested political and eco- perimental agropastoralism zones. Again, as in nomic interest in helping their agriculturalist other types of assisted settlement, the goal is to fa- neighbors to control theft and to develop peace- cilitate sedentarization of the pastoralists' produc- ful herding and manure contracts. tion systems by providing roads, wells, schools, This strong base of settler initiative and rela- and health, veterinary and extension services. tively peaceful relations with the neighboring Most pastoralists we worked with strongly fa- agriculturalists bodes well. By far the most press- vored the creation of such pastoral zones, princi- ing problem the agropastoral zones face today is

93 overstocking. When we conducted the survey, the Classified Forest at Toumousseni is one exam- the President of the Village Land Management ple of this type of innovative program but one Committee estimated that there were about 5,000 that is primarily attractive to agriculturalists. animals living in a zone area planned for 2,000. A strong priority needs to be to develop experi- Zone leaders are struggling with the best means mental programs to associate the pastoralists liv- of getting settlers to regulate the size of the herds ing in villages near classified forests. Transhumant they will be authorized to keep in the zone. Crop pastoralists are less likelyto know or to respect for- production programs as well as efforts to grow est boundaries. In view of the difficulty of distin- improved fodder for the animals have been liter- guishing pastoralists who are eligible for forest ally trampled in the dust by overstocking. grazing from those who are ineligible, it is probably better to forbid any sort of grazing in the Need to PromoteIntensification of LivestockSystems forests. For rules against grazing to be enforced, the lo- For the concept of area zoning for pastoralists to cal sedentarized and semi-sedentarized FulBe be effective, it must be combined with pro- must be associated. To accomplish this requires grams to intensify livestock production sys- the development of complementary zoning for tems. To date, local livestock production agropastoralists in adjacent areas, along with pro- programs have emphasized animal vaccination. grams to help pastoralists intensify their own Far less attention has been focused on questions livestock production systems. For the pastoralists of animal nutrition and breeding, programs that to respect area zoning, they must feel a strong could presumably make it possible for pastoral- economic and social investment in remaining in ists to reduce livestock densities without losing the agro-pastoral zone. To date, we are unaware income. of any programs that have been in existence long enough to provide demonstrated models. Indeed Need to DevelopPrograms for ControlledPastoralism past experience with the development of more nearClassified Forests sedentary pastoralist production systems in the Niangoloko region has been quite negative. The Our research supports other studies that show question of how best to combine zoned areas for that the most immediate threat to Burkina's clas- livestock with programs to promote more inten- sified forests is grazing, not cultivation. Present sive agropastoralist systems is still, in our opin- programs to regulate land use in the classified for- ion, one of the weakest areas in development ests focus on agriculturalists, however. The FAO planning for all of Burkina's OCP river basins. project to develop renewable income sources in

94 Knowledge and Management of the Natural Resource Base

The Dilemma of Spontaneous New tlers' fields to lie fallow for the length of time nec- Lands Settlement essary to restore soil fertility. Increased pressure on agricultural land is generally associated with The extensive crop and livestock systems being rising levels of social conflict. Predictable sources established in Burkina's river basins make sense of conflict include livestock damage to agricultur- under a given set of environmental conditions alists' fields, growing problems with potable wa- when "new" land, with good potential for rain- ter because of higher demand and pump fed agriculture, is still easily acquired and maintenance requirements, pastoralist resent- cleared, and when there are low social and eco- ment at being denied access to former grazing ar- nomic costs to abandoning worn-out land. eas and water sources, and a diminished supply If immigration rates are low, as in the Niangolo- of easily accessible fuelwood. ko area and the Kompienga Basin (before con- If agriculturalists and pastoralists have an op- struction of the dam), extensive agricultural tion, when these problems start to arise, they will cultivation is generally ecologically and socially move. This is a process that is as old as African ag- sustainable; population densities are low, and agriculture-a process that has become romanti- ricultural lands are still plentiful. In addition, cized and played out in the origin myths of there is usually little in the way of social or eco- Burkina's different villages and ethnic groups nomic infrastructure other than access roads and (Izard 1965). markets that makes one village setting more at- The traditional Mossi patterns of marriage, lin- tractive than another for outside migrants. Under eage, and politics have for centuries facilitated the these conditions, the primary factor determining reinstatement of settler crop production, live- household production levels is the size of the fam- stock, trade, and market networks in areas of new ily labor force. The rational strategy for settlers lands settlement (Skinner 1957,1960, 1962, 1964a, under these circumstances is to clear and plant the 1964b, 1970;Izard 1965, 1970, 1971;Izard-Hertier largest area possible and to abandon the older and Izard 1959).Moreover, the Mossi people ex- fields at the first sign of erosion. Farmers in the pected that a certain amount of lineage fragmen- Niangoloko region, for example, have traditional- tation to colonize new lands would be a natural ly farmed a field only five years before leaving result of their extensive cultivation practices. land fallow for an average of 20-25 years. In the Once established in a new region, the settlers past, the abundance of reserve fields made this would request a representative of their former long fallow possible. chief. This new chief would then be linked to their As population densities increase, it becomes former chief as well as to the former chief's alli- more difficult to find new land to allow the set- ances with other village, regional, and provincial

95 chiefs in the Mossi kingdoms. By requesting and plows and tractors to cultivate the largest area receiving an official "chief," the colonists reinforce possible. At the Solenzo site, a few successful their linkages with their former home areas and tractor farmers were able to cultivate as much as the wider kingdom. Historically, these social and 40-50 hectares per family with mini-tractors. This political linkages played a vital role in validating accounts for the sizeable difference in total land the territorial rights of different groups and in fa- area farmed per ALE between the AVV sites, cilitating peaceful trade relations between areas where controls were exercised, and the Solenzo with different resource endowments. region, where they were not (Figure 10.2). This traditional system for colonizing "new" Thus, while the land use restrictions in the AVV land worked as long as land was plentiful; it be- sponsored settlements contributed to the devel- gan to break down when land became scarcer. As opment of more "stable" production systems (i.e., population densities increased on the Mossi pla- less prone to extensive land clearance than were teau, it became increasingly difficult for agricul- the unrestricted cotton systems), overall, the pro- turalists to let the land lie fallow for a sufficient duction practices were not more sustainable. period of time to renew soil fertility. Moreover, neither settlers nor extension staff were able to control unauthorized immigration of Early Plans for and Response to the AVV cultivators and pastoralists onto village lands de- Model for Sustainable Agriculture signed for other purposes. By 1979, the area around most of the AVV sponsored settlements The early AVV program was a deliberate attempt was being occupied by spontaneous settlers. The to offset some of the deleterious land use practic- production methods used by these farmers were es commonly associated with spontaneous new even more extensive than those of the sponsored lands settlement. The first step of the AVV land settlers. By 1987, when most of the spontaneous use model involved a series of aerial photos and settlers at Rapadama had been living there for ten surveys of soil and water. On the basis of this re- years, they were already facing a crisis. Almost search, all the land in a given river basin was clas- the entire region had been occupied and the sified according to its best and highest use for settlers were experiencing increasing problems forestry, livestock, crops, or tourism. These sur- with declining yields, a proliferation of Striga veys were then used to develop a tentative plan hernonthica,and other weeds (AVV 1988). for a series of settlement, managed forestry, live- The AVV was also unable to control illegal stock, and dam construction projects (Figure woodcutting in areas around the sponsored set- 10.1). tlements. Thus, by the time more effective nation- Planned settlements were created in groups or al controls were put in place in 1984, the thick blocs of six to seven villages, in areas selected for savanna bush forests that once encircled the AVV the quality of their soils. The settlers' cultivation planned settlement in the upper Nakambe (ex- was restricted to one or two ten hectare farms, White Volta) were almost entirely cleared. which was considered to be the maximum amount of land that they could farm with the pro- Early Plans for and Response to Zoning for posed program for intensive cultivation. In addi- Agropastoral Areas tion, each planned settlement included areas set aside for forestry and livestock. In return for the Land use planning for pastoralists in the OCP riv- right to cultivate a farm in one of the "improved" er basins has focused on the development of project villages-with access to roads, schools, large, separate agropastoral zones at Gadeghin, health facilities and wells-settlers were required Sondre-Est, Nouhao, and Leo. Both pastoralists to adopt a recommended package of intensive and agriculturalists who can demonstrate previ- cultivation techniques and to respect the project ous residence are eligible to participate. Each zoning. zone is supposed to be equipped with basic vet- A comparison of the project plan for restricted, erinary services, dispensaries, schools and roads, registered land use with actual land use patterns as well as water points for humans and animals. ten years later shows that the settlers did respect In return for the right to herd its animals in the the broad outlines of the crop cultivation bands. area, a herder family was required to follow a rec- In contrast, in areas such as Solenzo where no ommended program to promote intensive crop such controls were imposed, the settlers used and livestock production.

96 FIGURE10.1 IBRD23853 BURKINA FASO PROJECTEDINSTALLATION OF AVV PLANNED %% .... SETTLEMENTSAND OTHER TYPESOF LAND USE PLANNING IN THE NAKAMBE AND NAZINON (EX- ,..... tWHITE AND RED VOLTA) RIVER BASINS, 1974 OurJsaoFgou-F S C VES Ll PZONES DE DEAVELOPPEMENT ENIVISAGEES POUR ZIS A 1.ASE, PHAS2 , D ALA REGION DES BASSINS DU NAKAMBE ET DU COMMUNAGRAZING 3RDPHASE INSTNAZINON 9EX- VOLTA BLANCHE ET ROUGE4 ZONESDi ARCURSDi PASE3 ALCONVETEPA- PROGRAMME L OCP, 1974

1 LINOGHIN ® NATIONALCAPITL N A ;- I

2 MOGTEDO CAPITA..L NIGE

%% k~~~,uoodougoum I

... N~~~~~o6ersIg.so n..E 4-,oyno.n.iorco ,n.o

CLASIFEDORETGHAAPHASEE EN, 19776 m COMMUNALGRAZING 3RDPHASEJINSTALLATION, 1981-84 F g ZONES DEPARC'OURS DE BErAIt F PHASE3D'INSTAUAIION, 1981-84 RANCHING 4TH PHASEINSTALLATION, 1 985-88 3 3wCLSSMIDBORESTZONE DE RANCHING KB57PHAERONTIERHAE 40DINSEXPERNAIMENT, TALUA CON, I 985197-7 -88 'i 4 ZBNESAtSE 2N\HS.NSALTO, 978 BURINAAS

LINOGHIN f NATIONAL CAPITAL M A L I - CAPITALE I

2 MOGED INTERNATIONAL BOUNDARY NIGER 3 BOMBORE FRONTIERE INTERNATIONALE . 4 BANE i BURKINAFASO \_

5 KAIBO NORTH Source: AVV | |0' dloA,. of | "IN 6 KAIBO SOUTH M.bimopbo fprnp-doby Th.hrfBW-llo clufi *L uaqoadou nap 7TIEBELEO oi G-op.Th. omloxwI ENI| oow bwovsbono

.uerfiodorl ..... | | COTE Dg{VO;RA ~~~GHANA0°.

JUNE 1992 Figure 10.2 Averageland area per unit labor for different technology groups at the AVVand Solenzo, 1988

Hectares 3.5

2.5~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~! '.'VSS;SS,.'.

Prvael Solenzos ;d-.-- G,G

Source:Tables0 7.1 and 7.8.f8i , :7 S G .: ' .G AVV Solenzo::::... SR:..9y . 2S .. iiiiii; i i:'i:':;...... "'' 98: 0: ;0Privately 0 X:aT ,:!id,, EE.EEEE', ,Di;tSiEE;E!iEESS,dtEi: EEEEE..T X iti: :t:0:; i: lli:l t:iG:t.:00000:00;:::.,...... : i _ S:g,::::,,, .....WS.S'.:.:.... ',,.S- , ...... 5.g.. co5 lec:t!i i:lyll:tvei :: :; ^ E: :: : t ::: :S:::020444'':;'"'"''""E:

SouceTale 7olnzan78

98 Burkina's agropastoral projects have been fices and rituals associated with initiations. Forest plagued by many of the same planning problems products (fruits, leaves, wood) have always as the AVV planned settlements. Chief among played a central role in the local rural economies. these is the determination of zone boundaries In contrast to crops, forests and forest products and lack of clear land tenure. Each of these agro- are considered "a gift from God," and a collective pastoral zones was selected in the early 1970s as good. For this reason farmers have generally not part of the initial AVVplanning for the OCP river been involved in active forest management, out- basins. In general, however, the actual projects side selective weeding to allow high value trees did not get under way until after the first decade such as the karit, nere,baobab, prunier, and Aca- of sponsored settlements, when mounting levels cia albida to survive in cleared fields. Even this of social conflict between cultivators and pasto- right was denied to outside migrants, who were ralists forced planners to do something with and, restricted from planting or harvesting trees on therefore, something for the pastoralists. By then "borrowed" land. many of the zones designated for pastoralist de- In an effort to preserve some of the country's velopment, like the present Nouhao and Gade- remaining forest areas, the colonial government ghin Projects, had already been occupied by exercised its legal right to reserve certain areas as spontaneous settlers. classified forests. The majority of these classified This delay in planning and forced relocation of areas were created in the country's low density spontaneous settlers can disrupt the evolving river basins (Map 23844).Local populations were patterns of economic and social integration in a pressed into assisting the administrators with the region. At Gadeghin, for example, more than 200 delineation of the protected forest zone. Forced families had to be removed, when the AVVbegan labor was also used to carry out reforestation to enforce zoning and to create the proposed base projects, especially along the major roads. This of infrastructure in 1980. Almost all the "illegal," history of forced labor and imposed limits con- spontaneous, agriculturalist settlers had immi- tributed to the development of hostile attitudes grated to the site since 1974,when the zoning had toward forest agents-attitudes that still remain. first been recommended. They blamed their Since independence Burkina has found it in- forced removal on the sponsored settlers-a fact creasingly difficult to satisfy the growing de- that soured the unusually hospitable relations mand for fuelwood. By 1985, only the extreme that had existed between the two groups before southeast and extreme southwest were consid- this time. These ill feelings and the forced dis- ered to have sufficient fuelwood supplies (Map placement of the spontaneous settlers contribut- 23852). This combination of rising demand for ed to the virtual collapse of the Gadeghin/V9 fuel and construction wood, plus growing demo- Rapadama market (number 5 on Figure 8.1) graphic pressure on land in the northern river ba- which had become the most prosperous interior sins has made it more and more difficult to market on the "frontier" between the sponsored control illegal cutting, cultivation, and grazing in settlements, the pastoralists, and the spontaneous the country's classified forests. Although precise agriculturalist settlement in the river basin. The figures are lacking, one survey identified 31 clas- hostility associated with the forced relocations, sified forests with varying degrees of forest dep- coupled with the collapse of commercial activi- redation (Minist2rede la Question Paysanne1987). ties, contributed to the departure of seventeen Twenty-six of the 31 forests that were identified households from the closest AVV sponsored set- as having problems are located in the river basins tlements in 1988. of the "cotton boom" provinces of Houet, Mou- houn, Bourgouriba, and Sourou. The classified Early Plans for and Response to Zoning for forests near roads and those within a one hun- Classified Areas dred mile radius of major cities like Kaya, Ouaga- dougou, Banfora, and Bobo-Dioulasso are The national government has faced similar sorts especially vulnerable. of problems in its attempts to protect classified Until recently, the low population densities and forest. Traditionally forests were considered dan- relative isolation of the classified forests in the gerous places where wild animals, serpents, and most southern provinces of Tapao, Gourma, Sis- spirits (both good and bad) reside. The spiritual sili, Comoe, and Poni have protected them. For qualities of the forests were reflected in the sacri- example, conflicts between settlers and the ad-

99 ministration were minimal at our case study site agrarian reform. The themes and approach in the Classified Forest of Toumousseni until adopted were heavily influenced by the extensive 1983. From 1954 to 1982, only 33 verbal repri- experience of the AVV in land use and coopera- mands were issued by the administration: 21 for tive development (Guyon 1986:12).The commit- hunting, five for illegal wood cutting, one for ille- tee presented a four-step village program. gal grazing, and six for illegal bush fires. STEPONE: Information and Election. A series of Since 1983, however, the rate of pastoralist im- information sessions to inform villagers of the migration to the area has increased dramatically goals and the necessity for this type of collective (SAED1988). About 60 percent of the households program precedes the village's election of a "land identified as immigrant in our interviews with management" committee responsible for deter- leaders in the four villages around the Toumous- mining land allocation and dealing with outside seni Forest had immigrated since 1981 (Figure authorities. The land management committees 10.3). Since we focused only on immigrants, this include representatives of the major social groups does not count FulBe whom the village leaders living in the village-recent immigrants, the in- classified as non-permanent residents. A recent digenous population, and pastoralists. survey by the local prefecture estimates the ratio STEPTwo: Delineation of the Village Frontiers. of pastoralist immigrants to agriculturalists to be The rights of each group-pastoralists, agricul- as high as seventeen to one. turalists, recent immigrants, and long-time residents-must be represented in the delineation of Alternative Models for Village Land village frontiers. These boundaries are based on Management: PNGT soil and topographic maps that are usually based on aerial photos. The delineation is organized by By 1983 it was clear that top-down projects like the local land-use committee working in cooper- the AVV and the system of classified forests ation with regional authorities and neighboring which attempted to legislate zoned land use were villages. almost certainly doomed to failure. Without vol- At Solenzo this step was subdivided. First, the untary collaboration from the local inhabitants, it villagers proposed their own land management was too expensive to enforce either the bound- plan. Then the extension service presented a sec- aries or the proposed land use practices. One of ond "counter-proposal" to discuss with the local the primary goals of Agrarian and Land Tenure people. Reorganization was to develop a more holistic STEPTHREE: The Village Contract. This step in- approach to natural resource management. This volves the negotiation of a signed contract be- approach would combine comprehensive land tween the community (represented by the land use planning for agriculture, livestock, and for- management committee) and the State. The com- estry based on the natural potentials of the areas, munity agrees to respect certain themes for soil with the voluntary implementation of these pro- and forestry conservation, improved pasturage grams through the village and regional Revolu- and suppression of bush fires. In return, the state tionary Committees (CRs). agrees to help the villages with the realization of Growing out of the August 1985decree on the basic infrastructure and land improvements. In implementation of the ReformeAgraire et Fonciere, theory, the village contract and land survey guar- the Departementdu Monde Ruralde la Presidencedu antee official recognition of the villagers' rights to Faso (DMR) elaborated a proposition for a pro- the land and any future improvements that they gram of village land management, "gestion des make on it. terroirsvillageois." At the initiative of the DMR, STEPFOUR: Realization of the Terms of the Con- two joint meetings of the World Bank, the GTZ tract. This step refers to the realization of the (Gesellschaftfiir TechnischeZusammenarbeit [Ger- terms of the contract by the villagers and the rel- man Development Agency]), CILSS (Comite'Per- evant state agencies. A national Coordination manent Inter-e'tatsde Lutte contrela Secheressedans Unit (Cellulede Coordination)was created in 1986 le Sahel)and the CaisseCentrale were organized in to help ministries implement a series of pilot 1986.The goal of the meetings was to define a na- projects following the village land management tional program that would help villages and re- model. The Unit, which provided some limited gional development authorities with the assistance with conceptualization of the pilot implementation of the themes advocated by projects, was attached to the Ministry of Plan,

100 Figure 10.3 Year of immigration for settlers living in villages surrounding the classified forest at Toumousseni, 1989

Number of families

1973 1977 1981 1985 1989 and before Yearof immigration

Source:Nana 1989

101 with special aid from the World Bank, Caisse Cen- most active participants in different project activi- trale, and other donors. During the first three ties (cutting and selling firewood, charcoal mak- years, the Cellule's activities focused on monitoring, beekeeping) have been the youth. ing the pilot projects in an attempt to refine the The genius of this sort of project which increas- program guidelines and recommendations es farmers' and pastoralists' income from renew- (PNGTV 1989a). able forest products is that it creates a group of A number of experimental programs to work people with a vested economic interest in control- with spontaneous agriculturalist and pastoralist ling illegal cutting, grazing, and farming in pro- settlers in the OCP basins have grown out of the tected forests. The country's fledgling Ministry of agrarian reform movement. Some of these have Environment and Tourism does not have the been directly incorporated into the national manpower to do the job alone. To be effective, PNGTV program, which is now referred to a na- however, such projects need to be combined with tional village land management program, which the development of other sectors, including crop is now referred to as the Programme National de and livestock production. Gestiondes Terroirsor PNGT. One of the first PNGT test sites integrates spon- Development of Industrial Forestry Projects taneous settlers into the extension, credit, and land management programs of the older AVV Any long-term solution to protecting Burkina's planned settlements at Rapadama. Another type forest resources must also include the develop- of experimental assisted settlement program ment of managed forestry projects to offset urban aims at the incorporation of agriculturalist and demand. One example of this approach is the pastoralists hosts and settlers into joint land use managed, industrial forestry project at Wayen committees. The program is being tried in three created by the AVV. villages in the CRPA du Mouhoun that have experienced different rates of in-migration. The So- Policy Implications lenzo program basically follows the PNGT model of creating village land-use committees. Special Importance of National and Regional Support for emphasis is being placed on attempts to promote Village Land Management Programs more sustainable production practices and to stabilize rampant forest clearance. The settlers in the PNGT pilot program at Rapad- Although it is still in the experimental stages, ama emphasized that a program like theirs was in several instances the PNGT has been quite only possible if there was a strong base of national successful in introducing more sustainable prac- and regional level administrative support. This tices. One of the most interesting is the Projet point was brought home to them by the early lack Amenagement des Terroirs et Conservation des Res- of support from regional administrators for a sim- sources dans le Plateau Central (PATECORE) in the ilar program in the adjacent block of Linoghin. Bam province of north central Burkina. Al- It was originally planned that Linoghin, which though it is not an OCP river basin, the project had equivalent if not higher rates of spontaneous provides an example of the development poten- immigration, also would be included in the pilot tial of the PNGT concept. assisted settlement program created by the AVV at the UP1. This did not occur, however, because Land-Use Planning in Connection with Forest one village of spontaneous settlers-many of and Wildlife Preserves them civil servants and wage laborers from Oua- gadougou and the FulBe (who aligned with the Small-Scale Projects to Develop Renewable Income dissident spontaneous settlers)-refused to coop- Sources from Classified Forests erate with a preliminary census. The dissident settlers then carried their case to the newly creat- Projects like the FAO multiple use forestry project ed Ministry of the Peasant Question and got the in the Classified Forest at Toumousseni are a project officially terminated, but not before promising model for development in areas near threats of physical violence were made against classified forests. Although still very new, the Tou- AVV and UPI staff, who were conducting the mousseni project was enthusiastically received in studies, and against settlers, who allowed them- the three villages where it was implemented. The selves to be interviewed. The civil servants' fear

102 of reprisals for being absentee landlords appears Clarificationof Adjudication Procedures to have motivated their lack of cooperation. Clarification of the procedures for adjudicating Clarification of the Role of the Village Land land disputes and individual tenure rights would Management Committees give "teeth" to current legislation. To date, the process for negotiating disputed land claims or A striking feature of all the village land manage- certification of family tenure rights is still unclear. ment committees was the speed with which hosts and settlers were able to organize village land Coordination of Participating Institutions management committees. After less than one month of discussion, groups were organized at The concept of interdisciplinary collaboration be- both Daboura and Rapadama. We were im- tween different ministries (Agriculture and Live- pressed by the high level of enthusiasm and the stock, Water, Environment and Tourism, Finance, efficiency of the committees. Caution must be ex- Peasant Cooperatives) is considered vital to the ercised, however, to ensure that the organization design, implementation, and evaluation of an in- of the committees does not become an end unto tegrated, holistic approach. To achieve this sort of itself. These committees must be given the means interdisciplinary coordination on the ground, to carry out development programs on their own. however, takes a great deal of time. The institu- Care must also be given to work in collaboration tional mechanisms for achieving collaboration at with traditional authorities in the villages. the regional or national level need to be more effectively determined (CRPA du Mouhoun 1990b). ClearDelineation of VillageBoundaries Extension Planning for future projects should include greater attention to and a longer time period for the de- Extension programs continue to focus on either lineation of settlement boundaries, a task far crop or livestock production, with relatively little more complex than it would first appear. It was attention given to how they are interrelated. cited as one of the principal problems at Daboura Questions of dry season irrigated gardening and (CRPA du Mouhoun 1990b). Neighboring villag- nonfarm employment have been addressed in es often have difficulty agreeing on village limits. only a few isolated cases. The extension programs Other problems are created by house and field need to take a much broader perspective. sites that appear in frontiers. Boundary delineation must therefore be viewed as a long-term, Research evolving process rather than something that is carried out at a single point in time. The PNGT program foresees a vital role of research in connection with every stage of a village Delimitation of Areas for Pastoralists land management program (Figure 10.4). To date, however, these research programs tend to follow Settlers and hosts realize that long-term develop- the lead of extension and continue to focus most ment of their own diversified production systems heavily on crop production. Research projects is linked to good relations with the pastoralists. need to take a broader perspective on the land Facilitating development of large agropastoral use patterns, economic returns, and linkages be- zones as well as village areas dedicated to pasto- tween different types of agricultural and nonagri- ralists enhances the probability that the two cultural employment. groups will be able to regulate their separate and Farmers and extension agents alike are tired of a interrelated land use patterns with less conflict. proliferation of descriptive, baseline studies. When- Programs are more likely to succeed if there is a ever possible, village research should be linked to clear, early attempt to indicate the boundaries for the evaluation of particular types of innovations or agropastoral zones with painted or inscribed programs in order to make recommendations for markings that are easy to identify. future action. Caution should be exercised so that the energies of the extension staff and the inhabitants are not drained by an endless stream of dispar- ate student theses. Thesis research is important, but

103 Figure 10.4 Interrelated research and interventions: the PNGTV approach

Interventions Researchand studies

Organizationof a multi-disciplinaryteam to Identificationand choiceof test sites coordinateprogram(mobile team) v Human resourcestudy . V lIaining of mobileteam and other CRPAagents Analysisof topographicand aerialmaps of involvedin program villageland resources v V Informationsessions with the villagecommunitLies Additionalresearch and choiceof villagesfor projects V Creationof villageland management < < a

Delineationof villageboundaries | D>> Elaborationof villagecontracts YrV Zoning of villageland areas < < | Naturalresource inventory V V VUllagecontracts: negotiation and action plan DD> Assessmentof needs and resources

Source:CRPA du Mouhoun1990

104 it should be linked to ongoing research efforts.Fail- VitalRole of NationalFuel wood Policies ure to do this is an imposition on the farmers, especially in popular, highly visible development National policies that determine the urban price programs like the PNGT and the former AVV.With of fuelwood and of alternative energy sources rareexception, the final stage of any researchproject will play a crucial role in reinforcing local land should include the translation of research findings management programs. For example, although into graphic illustrations that can be discussed with the industrial wood plantation at Wayen was suc- village land management groups. cessful in organizing regular harvests and mar- Researchersneed to emphasize the incorporation kets, the project did not become self-sustaining of settlers in the design and implementation of re- and profitable until national legislation was intro- search studies. Old "top down" research models, duced after 1984that restricted illegal cutting and that many of us still use, date from a period when raised the price of imported fuelwood. Important primary education levels were low and it was diffi- complementary policies included: culttofindindividualswithinthevillagew/hocould (1) The "3 Luttes" (the Three Battles), help carry out surveys. This is no longer the case in launched 22 April 1985,against bush fires, most villages. If the inhabitants see the practicality animal depredations to forest, and abu- of a research project and they trust the individuals sive wood cutting; conducting and supervising the survey to keep (2) The strict regulation of fuelwood and their responses confidential,the quality and policy charcoal organized at three levels, (the significanceof the research can increasedramatical- small transporter, the wholesale trans- ly. Research with high levels of farmer direction is porter, and urban merchants) that began also more likely to reflect the evolving needs and in 1985; concernsof settlers and hosts as their economiccon- (3ThdelomnofaNtnlCo is ditions improve. Failure to report findings to the sion to Fight Against the Effects of the sponsoring village groups can jeopardize their will- Drought (CommissionNationale de LutkeCon- ingness to cooperate in future research efforts. tre lesEffefs de la Secheresse)(CNLCES); and PlanningSurveys ~~~~~~(4)The adoption of a National Plan for Fight- Plannm

105 11 ~~~~~~~~~~~~~~~~~~~~~~~~~~.

National Management Institutions

The type of assisted spontaneous settlement advo- Creation of the Autorite des Amenagements des cated by the OCP Land Settlement Review and Vallees des Volta (AVV), 1974 Burkina's National Program for Village Land Man- agement (PNGT) requires planning and manage- The official mandate of the Autorite des Ame'nagement at multiple levels (McMillan, Painter and ments des Vallees des Volta or AVV was to study, Scudder 1990). While negotiating the "village con- promote, coordinate, and execute (or control the tract," villages request roads, wells, and health and execution of) the operations necessary for the educational facilities, as well as special surveys of economic and social development of the land of- water, soil, societies,and vegetation. Under the ex- ficially placed under its administration (AVV isting administrative structure of most African 1985d). The agency's goal was the optimal devel- countries-indeed most countries in the world- opment of natural resources (soil, forests, pas- each set of tasks is coordinated by a separate min- ture, water, and wildlife) in the underpopulated istry or ministry branch. At the same time, phasing areas of the Volta Valleysand their major tributar- projects over a long period of time as recommend- ies. Special emphasis was put on the conservation ed by the Land Settlement Review and the PNGT, of soil resources in connection with the area's best requires that this assistance be coordinated. Not land use (ibid.). This official mandate was de- only must the administrative tasks be coordinated, clared for a twenty-year period. The AVV's activ- they must be coordinated in such a way that they ities then would be placed under the established respond to and complement a strong base of local structure of regional and national development institutions and initiative. The sheer volume and agencies (Ministry of Education, Ministry of Ru- complexityof the tasks involved has led many gov- ral Development, Ministry of Health, Regional ernments throughout the world to advocate spe- Development Organizations, for example). cial, autonomous management structures to A complex planning process in which the coordinate development activities in areas of new project would conduct comprehensive topo- land settlement or resettlement. graphic, hydraulic, and sociological surveys was Burkina Faso was the only one of the original central to the AVV planning model. On the basis seven OCP countries to establish a management in- of these surveys, each valley was divided into stitution to supervise development planning and five broad categories of land use (AVV1985c:17): interventions in its OCP river basins from the start. * areas suitable for dryland agriculture with a Thus, the history of that institution-the AVV- high proportion of good soils suitable for provides other countries with an interesting van- planned settlements; tage on the potential benefits, as well as costs, of * areas suitable for irrigated agriculture and adopting this strategy. dam construction;

106 * areas suitable for intensive livestock or feed of installation, settlers had to be encouraged to programs; immigrate; therefore, special agents recruited set- * areas most suitable for natural wildlife tlers from the overpopulated areas most severely parks and tourism development; and affected by the 1968-1973drought. * forest reserve areas,including areas for com- To encourage settlers to move, they were pro- mercial wood production near major population vided transportation to the site, assisted with centers. field clearance, and provided a monthly ration of grain, oil, and fish until they harvested their first EarlyEmphasis on Planned Settlement as a crops. Each settler household was entitled to one, Strategyfor Development of the Valleys or in the case of an extremely large labor force two, ten hectare farms that consisted of six 1.5 The PlannedSettlement Model hectare bush fields and a one hectare plot on which to construct a house. The project was also The early AVV program focused on the design responsible for the design, testing, extension, and and implementation of groups of model settle- evaluation of a new intensive dryland cropping ments in areas with good agricultural potential package-one that would presumably allow and a few more specialized industrial or irriga- farmers to increase yields while retaining soil fer- tion projects. The first settlements were created in tility (Chapter 7). the Nakambe and Nazinon river basins (ex-White In addition to the planned settlements, the and Red Volta) (Table 11.1and Map 23854). AVV undertook a number of sectoral develop- Once an area was determined suitable for agri- ment projects as well as the planning of a series of cultural settlement, the project was responsible dams and irrigation schemes (Table 11.1). for the selection of village, field, and house sites; installation of basic infrastructure (wells, roads, Administration,Costs, and AnticipatedBenefits bridges, extension worker housing); and coordination of economic services. In the very first years The original AVV project plan foresaw the evolution of the AVV planned settlement program in Table 11.1 Activities of the AVV,1973-1981 terms of: an annual increase in planned villages, the type and level of infrastructure, and special YBear Number of livestock and forestry projects. Although the pro-

Planned Settlements posed agricultural program required a greater Linoghin Nakambe 1973 8 outlay of cashand laborthan the settlerswere ac- LMnoghed Nakambe 1973 8 customed to, it was anticipated that these costs Mogtedo Nakambe 1974 6 would be offset by higher yields. Moreover, it Kaibo Nord Nakambe 1975 4 was assumedthat the settlersper capita food pro- KaiboeNord Nakame 1975 4duction and cash income would increase every Tiebele Nazinon 1975 4 year. During the first three years, the annual ad- Djipologo Bougouriba 1977 6 dition of a new field would make this increase Bombore Nakambe 1979 7 possible. After the third agricultural season, any Manga Est Nakambe 1979 4 subsequent increase would come from the greater Rapadama Nakambe 1979 8 use of fertilizer and labor on the existing crop

Other Projects Started area and the expansion of the settlers' livestock Industrial Wood Project Wayen 1976 activities, rather than the addition of new fields. Pilot Irrigation Project Bagre 1980 To coordinate these activities, the AVV operated its own heavy equipment for land preparation Planning Studies Undertaken and road building, hydraulic development, and Livestock Project Nouhao extension services. The development program Hydro-electric Dam Bagre was coordinated through an elaborate central- Hydro-electric Dam Kompienga ized organization divided into five directorates: Hydro-agricole Development Vallee du Sourou and General, Finance and Administration, Planning, Haute Vallees de la Volta Noire Construction, and Agricultural Development Project for Irrigation Development Along the Black Volta (Figure 11.1).Each of these directorates was bro- Project for the Tourism Development of Po ken into divisions and subdivisions. The agency

107 Figure 11.1 Organigram of the AVV administrative structure, 1974-1982

Ministere du D6veloppement Rural

Secretariat General

Direction Gen&ale de UA.VV Cellule coordination des aides

|Direction des Etudes, Dieto de Travau Direton de Etds Direction Admindstrative Direction de la Mise Programmeset et Financi're Directiondes avau, en Valeur

* ServiceEtudes de base * Servicedu budget * Serviceroutes * ServiceTransfert installation * ServiceEtudes de Suivi- * Servicefinancier * Serviceforages * ServiceDEveloppement *valuation * Servicede Ia comptabilit6 0 Servicebatiments agricole * ServiceEtudes Topographiques* Serviceapprovisiorinement * Servicetravaux agricoles * ServiceAspecs-woo * ServiceEtudes ExpiEnmentales* Servicedu personnel Motorises humains(Eco. famn.) Service cr&Iit *IService garage S ServiceSemences et 0 Servioecredit agricole protectiondes vegetal * Elevage

SECTEUR

BLOC < BLOC

VILLAGES

Source:AW 1985c:37.

108 was presided over by a board of directors that the project, estimated to have increased to included several high level government adminis- $12,000-$15,000per settler household installed trators and was chaired by the Minister of Plan- (Table 11.2). ning. The General Director of the AVV was Other areas of concern included: appointed by cabinet decree. (see AVV1985c for a * Limited Adoption of the ProposedExtension more detailed explanation of the activities of the Package.Settlers were not practicing the proposed different directorates). package of intensive cultivation techniques. Spe- Some of the projected benefits of the AVV cifically,Murphy and Sprey (1980)found: (a) set- planned settlement program over a 20-year peri- tlers who had been at the project for shorter od were: (a) to settle an estimated 650,000 per- periods of time followed the extension package sons in planned villages-290,000 persons, or more closely than those who had been there long- 38,000families, in rainfed agriculture and another 360,000 in irrigated agriculture associated with project plans for dam construction; (b) to Table 1 cost to the BEVAofSet one reduce population pressure in some of the more Family accordig to the BEI-AgrerReport densely settled areas of the plateau; (c) to control (February 1978) settlement and development in the valleys in or- PercentSubtotals der to minimize the negative consequences of CFAF of total % these higher population densities through l.Preliminarystudies project attempts to link access to project lands Pedology 1/20,000 30,800 1.18 with the adoption of intensive farming practic- (4,000ha a year) es; (d) to provide a regional grain surplus, that Geohydrology 1/20,000 would allow the country to offset a large part of (40,000ha a year) 45,000 1.72 its projected food deficit; (e) to triple Burkina's 2.90 production of the export crop, cotton; and (f) to 2. Settling the farmers improve the standard of living for the 15-20 per- Recruiting and moving 50,000 1.91 cent of the country's population who would be Transportation WPF food 19,900 0.76 living in project areas (AVV 1974, Ouedraogo Marking location of villages 1976, Nikyema 1977). and fields 61,425 2.34 The 20-year cost of the AVV development pro Preparing master plan for bloc 2,000 0.08 5.09 gram (1973-1993) was estimated at 60 billion CFA 3. Infrastructure (US $240 million) in the 1971 planning paper Roads 557,000 21.31 (AVV 1974). This amounted to a project cost of Equipment and personnel 139,900 5.35 828,571CFA (US$3,300) per family installed in an Technicalassistance 45,700 1.75 AVVfarm, 35,024CFA (US $139)per hectare and 28.41 56,000,000CFA (US $221,306)per village for the Housing 187,600 7.18 Nakambe and Nazinon valleys, which were for- Equipment and personnel 23,000 0.88 seen as the major area of intervention. In 1978, ap- Wells 81,60 3.12 proximately 3 percent of the budget came from 4 Landl 3.12 the Burkina government, 12 percent from funds 4. Land preparation that the AVV generated through its own commer- Deepplowing (212,100) 368,70014.10 cial activities, and the remaining 85 percent from Harrowing (103,600) foreign donors-mostly France, the Netherlands, 5. Equipment and supplies 40,700 1.56 and the European Economic Community (see 6. Field personnel for 3 years 138,000 5.28 AVV 1985c for a more detailed discussion of actu- 7. Update cars and furniture 34,200 1.31 al project costs, funding, and achievements from 8. General Expenes AW 491,700 18.81 1973-1984). 9. Technical assistance 297,000 11.36 (10 experts for 3 years) Early Results of the AVV, 1973-1979 Total 2,614,225 100.00 (about By 1979, criticism of the AW at the national and $12,500) international level was increasing. In particular, it Source:BEI-Agrer, Report on the AW Program, 1978-1982, was becoming difficult to justify the high costs of cited in Murphy and Sprey 1980:86.

109 er; and (b) cotton was the only crop on which the sis," as well as the emphasis on increasing food recommended package of cultivation techniques, production, were maintained, but the priorities including monocropping, pesticides, planting in were changed to: (1) interacting harmoniously rows, fertilizing, thinning, timely weeding, and with the indigenous people; (2) increasing the animal traction, was consistently applied. project's integration with regional develop- * Slow Rate of Planned Settlements. By 1979, ment organizations; (3) shifting the earlier em- only 1,834 of the anticipated 9,700-13,700house- phasis on agriculture to the more balanced holds were actually settled (Table 11.3).After the development of all available resources in a re- first year, the delay had nothing to do with the gion; (4) collaborating with the local people and lack of suitable candidates. It was, instead, due to settlers to develop infrastructure that would the administrative and financial difficulty of cre- serve a much wider zone than the original ating sponsored settlements in large enough planned settlement area; (5) reducing the numbers to satisfy the growing demand. amount of infrastructure to be implemented by * IncreasingConflict with theIndigenous Inhabit- the project; (6) de-emphasizing long distance ants in someProject Areas. By 1979 the project was recruitment and emphasizing instead the re- experiencing increasing pressure from various in- cruitment of indigenous inhabitants; and (7) digenous and immigrant groups who, under the shifting the responsibility for infrastructure national land laws (see Chapter 5), could claim and development programs to settler groupe- preexisting or present rights to cultivation and ments (AVV 1981a, 1981b; Vayssie 1982; Kabore, settlement sites. The problem was especially Brilleau and Badolo 1985; Baris, Bonnal and acute in the Nakambe Basin, near the paved high- Pescay 1983; Yanogo 1988). way to the east of Ouagadougou. * IncreasingRates of Illegal,Unassisted Immigra- New AVV Administrative Structure, 1983 tion onto Project Lands. Almost immediately the planned settlements attracted massive amounts In the reorganized model, the administrative of spontaneous settlers that quickly equaled and structure changed to reflect a separation of the then surpassed the number of sponsored settlers functions of infrastructure development and in many areas of the Nakambe and Nazinon planning (ambtagiste)from that of project coordi- (Nana and Kattenburg 1979;AVV 1988; Savadogo nation (realization of particular projects within a 1988b).It was soon clear that large areas of river short, defined period of time) (Figure 11.2). basin would be settled spontaneously before the THEAVV HEADQUARTERS-CENTRAL ADMIN- sponsored settlements that were planned for ISTRATION. The AVV headquarters in Ouaga- thoseareascouldeverbegin. dougou continued to coordinate central administration, planning, coordination, and Reorientation of National Plnning for the OCP support services for all the AVV projects. These River Basins, 1983-1990 activities included: (1) general planning and coordination for the valleys and surrounding ar- From 1981-1983 the AVV donors and adminis- eas; (2) the identification, conception, and tration reorganized the agency's administrative programming of activities for geographically structure and programs. The overall goals of defined planning units (Unites de Plannification "coordinated action and planning for the devel- or UPs) with the relevant ministries and other opment of the zones liberated of onchocercia- administrative structures; (3) financial account-

Table 11.3 A Comparison of the Actual and Projected Rates of Planned Settlement at the AW, 1973-1979 1973 2974 1975 1976 1977 1978 1979 Total Scheduleda 200 300 600 900 1,700 3,000- 3,000 9,700- 5,000 5,000 13,700 Actually Settled 9 195 173 287 302 302 556 1,824 a. According to Upper Volta Five Year Plan, 1972-1976,and 1977-1981. Source: Murphy and Sprey 1980:76.

110 Figure 11.2 Organigram of the AW administrative structure, 1982-1989

Conseil Ld'adrmnistration |

cellule suivi Directeur esvaluation Gen&ral _ cr6dit ~ agricole c

Direction Direction Direction > Administrative Etudes, Pro&rammes Operations Direction et Finandere Controle Techniques Infrastructures

PROJETS INTEGRES PROJETS SPECIFIQUES

------r-----

II liiiil X W

prosjeatet Wayen j jo

Liaison Hierarchique (Administrative Linkage) ------Conception-Coordination-Appui (conceptualization, coordination of projects, and aid in implementation)

Source:AVV 1985c:36

111 ing and relaying financial information to the toral zone at Gadeghin also was part of the UPs; (4) assisting the UPs with the construction decentralized AVV-UP1. of basic infrastructure, installation of migrants, and coordination of credit; (5) and monitoring Assisted Settlement ProgramsIncorporating (suivi-evaluation)the interventions carried out by Spontaneous Settlers in Sponsored Settlements the UPs and special projects. Project administration also was responsible for Early Responsesto SpontaneousSettlement the administration of special projects. Two types of projects were envisioned: (1) integrated region- Early responses to the "problem" of uncon- al development programs; and (2) special projects trolled spontaneous settlement into areas desig- such as irrigation schemes, pastoral zones, or in- nated for sponsored settlements included dustrial forests. various attempts to induce the spontaneous set- THE UPS-ADMINISTRATIVE UNITS FOR REGION- tlers (or migrants sauvagesas they were referred AL DEVELOPMENTPROGRAMS. The administrative to at the time) to relocate to the newly created unit for the regional development programs was sponsored settlements (Nana and Kattenberg the UP or Planning Unit. A UP was a geographi- 1979:19-26).Other programs tried to incorporate cal unit with a certain homogeneity of planning preexisting indigenous and spontaneous settle- problems and opportunities. Each UP was comments into blocs of sponsored villages. Other in- posed of several UDs (Unite de Developpementor itiatives tried to get non-sponsored settlers and Development Unit). Each of the older AVV blocs hosts living in the river basins to adopt the AVV of planned villages was converted into a UD; new program without moving. Not one of these early UDs were created to cover the neighboring tradi- programs was successful. tional regions. A single UP would thus include as The first comprehensive study by the AVV of many as 10 UDs that would regroup areas of spontaneous settlements (Nana and Kattenberg sponsored and non-sponsored settlement. Al- 1979)made it clear that the problem of spontane- though the Director of a UP answered directly to ous settlement onto land designated for develop- the Director of the AVV headquarters, each ment by the AVV had become so large that it project UP was independently responsible for the overwhelmed the capacity of any one institution technical execution and financial administration to deal with it. Moreover, they argued, the AVV of its activities. The original goal was for each UP had never been granted the power to attain the to be funded by one or more donors. ambitious objectives outlined for it in 1974, and In 1983, for example, the planned settlements that the state was unwilling to grapple with ur- along the Nakambe (ex-White Volta) were re- gent but unpopular issues like land tenure and grouped into Planning Unit 1 (UP1). Except for woodcutting. the older AVV bloc of Linoghin, the frontiers of A resolution of the Administrative Council of the new UPI correspond to the administrative the AVV on 26 May 1982 recommended that the boundaries of the Province of Ganzourgou (Fig- Authority undertake the necessary studies to de- ure 11.3). Each of the planned settlement blocs fine a new policy for settlement and soil conser- was then administered as a separate Develop- vation. At the same time, a new Commission on ment Unit (UD) within the planning unit (UP). Spontaneous Migrants was created. The Com- The UP administration was headquartered at mission's goal was to define a global method of Zorgho, the provincial capital. working with spontaneous settlers to allow them In 1989, the AVV-UP 1 grouped together 11 to participate more fully in AVV programs. In UDs. Seven of the UDs are former groups of June 1984 a report was issued outlining the Com- planned settlements or blocs (Linoghin, AVV- mission's research findings and proposals for a Mogtedo, Mogtedo-Bombore, Rapadama, Ouay- new approach. algui, Tanema and Mankaraga); the other four The announcement of the Act of Land Tenure UDs (Meguet, Mogtedo [town and surrounding and Agrarian Reform in August 1984 provided region], Boudry, and Zorgho) coordinated exten- the national support that this new approach to as- sion and development for groups of indigenous sisted spontaneous settlement required. The Re- villages in the non-basin plateau region (Figure form Act created strong national level recognition 11.3). A twelfth planned settlement group at Da- of the rights of AVV settlers as well as the rights kongo is still in the planning stages. The agropas- of the spontaneous settlers-something that had

112 FIGURE11.3 IBRD23855

BURKINA FASO THE AVV PLANNING UNIT 1 (AVV-UP 1) UNITE DE PLANIFICATIONI (AVV-UP 1J

~~~MOGTEDO VILLAGE/-_\ |RAPADAA_) t l \ \ / ~~~ZORGHO

BOBORE| AKN \ / MANKARAGA y \~~~~~OUAYALGUI

r TANEMA J

4 ~~~~~~~~~~~~~~~~~~~~~M A L I,_,

' BURKINAFASO 1 O*suagadougou v

,h.*mpbobrn pedo. Sdyr f.oreh. -- e1 -s of Th. W-vld Areo of _ BonLG-op Th. d.nmosn ... d ondh. boudor-r A._ -n shsJ map f- mpdo -f -mply, fh. p.,l of Th. WorldSonL G-op -ny j.dq-,,l { BNINaE hn,hbg9ol dfFvof y ft-szevy or-nY nrrmn oocprn f ..... caTEDvolREN GHANA \ O0

JUNE1 992 been missing in earlier AVV programs. Subse- resentatives of the UPI held information sessions quently, the PNGT model of village land use com- to explain the new program to sponsored and mittees, created in 1986, provided a nationally nonsponsored settlers in each zone. recognized local institution for integrating set- While the village sociological and boundary tlers with indigenous people in a framework that surveys were being completed, the AVVcollabo- would recognize the legitimacy of land rights of rated with the UPI on hydraulic and soil surveys both spontaneous and sponsored settlers. This to determine the best and highest use of the adja- model was compatible with-and indeed heavily cent soils for agricultural, livestock, forestry, or influenced by-the recommendations of the 1984 protected lands. Based on the soil surveys, the AVV commission. number of hectares of land suitable for agriculture in the area of each village were indicated on New Modelsfor Assisted SpontaneousSettlement the map. This calculation, minus the land in a given AVV village that was already allocated, was The first large scale experiment to incorporate considered to be the amount of land available for spontaneous settlers under the new PNGT vil- distribution (Table 11.4). Other land uses-bas lage land management model began at Rapad- fonds (low-lying areas with potential for irriga- ama in January 1988.This program was one of the tion), agropastoral, and protected lands-were test sites for the national PNGT program (AVV calculated to indicate the total land area for each 1988). Both Rapadama and Linoghin, the two village (Table 11.4).This total amount of available blocs closest to Ouagadougou, experienced high agricultural land was divided by six hectares (the rates of immigration between 1974 and 1987.Pre- recommended area of bush fields per family) to liminary research in 1987 showed 697 agricultur- estimate the maximum agricultural carrying ca- alist and 55 pastoralist families within the UD of pacity of each village (Table 11.4). Rapadama alone (AVV 1988). Only pastoralists To illustrate this system of calculation, Vl was who cultivated in the area (i.e., were permanent considered to have 1,766hectares of land suitable as opposed to transhumant) were considered to for cultivation in the immediate area surrounding have settlement as well as pasture rights, in the it; 450 hectares of this land were already allocated newly defined UD. to AVV farmers, leaving 1,316 hectares (Table In theory, the basic methodology for the assisted 11.4).This 1,316hectares was divided by six hect- settlement program was simple. Instead of requir- ares (the area designated for a project farm) to es- ing the settlers to relocate to an AVV village, the timate the village's maximum carrying spontaneous settlers (both agriculturalist and pas- capacity-219 families if only the areas suitable toralist) were invited to drop their separate status for dryland cultivation are included; 366 families and to join the program on an equal footing with if the lower basfonds as well as higher dryland ar- the sponsored settlers. In return for subscribing to eas were included in the calculation. The census the program, the spontaneous settlers received ac- showed that 43 AVV families and 170 spontane- cess to the same extension program-including ous settler families were already living in the area credit, training for animal traction, and inputs (Table 11.4). The village could thus presumably from the AVVwarehouses. They were also able to accommodate only an additional six families us- have an equal voice in the UD administration. ing the lower estimate of carrying capacity (219), The Rapadama program proceeded in stages. or 153 families using the higher (366). During the first stage (January-February 1987), Once the carrying capacity of each village was the residents of each of the Rapadama planned estimated, plans were made for expanding the settlements worked with extension staff, topogra- earlier base of AVV infrastructure. Boundaries of phers, and representatives of neighboring villag- the now expanded village unit were marked and es, to delimit the village boundaries. During the indicated on the map. second stage, these boundaries were recorded on Responsibility for land management and exten- a topographic map (January-February 1987). In sion services in the newly expanded unit was vest- an overlapping third stage, complementary to the ed in a single land management committee, made second, enumerators working with a sociologist up of elected members from each of the major so- conducted a census of all the migrants living cial groups (settlers whose immigration was spon- within the designated village zones. During a sored by the AVV, spontaneous settlers, and fourth overlapping stage, the sociologist and rep- pastoralists farming in the region). Coordination

114 Table 11.4 Method of Calculating Carrying Capacity in the AVV-Assisted Settlement Program at Rapadama Estimated Land (hectares) Spontaneoussettlers carryingcapacity Already Total attributed Available Available Available Number Village land to rainfed basfond pastoral Reserve Total sponsored territory area AVV settlers land land lands area territory settlers Total Permanent Temporary Rainfed Basfond Vi 1766.30 450 1316.30 887.66 69.60 913 3,636.56 43 170 143 27 219 366 V2 830.35 450 380.35 576.80 - 664 2,071.15 48 54 38 16 63 159 V3 815.75 466 349.75 55.50 6.19 147.61 1,025.05 45 12 10 2 58 67 V4 2347.70 675 672.70 360.85 271.20 709.60 4,689.35 63 221 91 130 278 338 V5 1338.50 450 888.50 584.01 321.50 181.85 2,425.86 44 90 49 41 148 245 V7 2035.75 675 360.75 634.93 223.15 723.67 3,617.50 65 92 52 40 226 331 V8 845.90 450 395.90 362.30 11.15 88.80 1,A08.15 45 15 15 0 65 125 V9 906.56 454.50 452.06 326.06 - 54.78 1,287.40 49 43 43 0 75 129 Total UD 10,886.81 4,070.50 6,816.31 3,788.11 902.79 4,483.31 20,061.02 402 697 441 256 132 1760 Source:AVV 1988. of the different village committees was carried out such basic infrastructure as roads, wells, and ex- by an elected regional land management commit- tension housing. tee. Once incorporated into the new program, set- The newly reorganized AVV is known as tler households (pastoralist as well as agri- ONAT (OfficeNational d'Amenagement des Terroirs, culturalist) received one or more blocs of land that or National Officefor Land Development). In the were registered on a topographical map. In some administrative category of an office under the cases they were allowed to keep their former Ministry of Agriculture and Livestock, ONAT can fields; when fields were not located in the village consult within its own and other ministries in the area designated for crops, they were asked to design and implementation of special projects. move. Spontaneous settlers were usually allowed Many of these projects will undoubtedly contin- to keep their former house sites. New pumps were ue to be in the OCP basins. installed to accommodate the additional settlers- Special regional development projects to plan more than double the official number of settlers and facilitate in areas of new lands settlement will being served by the original AVV-createdinfra- still exist. The new regional projects will probably structure and extension services. be a cross between the old UP structure established In 1989 and 1990,similar programs to incorpo- in 1983 and the PNGT structure. The key difference rate spontaneous agriculturalists and "settled" is that ONAT will cease to have any direct relation- pastoralists were undertaken in all of the older ship to these special planning units, except as a AVV sponsored settlements of the AVV-UP1. consultant. In regard to the latter, ONAT can be contracted to conduct the necessary surveys for the CurrentReorganization of the AVV, 1990 development of an area land use plan. If ONAT re- tains the AVV'sconstruction equipment, the office Reorganizationof the AVV into ONAT (Office might also be asked to intervene in the realization Nationald'Amenagement des Terroirs) of particular types of infrastructure.

On 9 August 1989, the Burkina Faso government ProjectedRelationship of ONAT (ex-AVV) announced another restructuring of the AVV.The to the PNGT new plan transforms the AVV from a structure with a special mandate for integrated develop- Responsibility for land use planning in the basins ment planning in the OCP river valleys to a struc- will presumably be taken over by the National ture specialized in land use planning and the Cellule of the PNGT, under the Ministry of Agri- realization of development projects in rainfed ag- culture. The precise organization of the National ricultural areas (Djigma 1989;AVV 1989a).Under Cellule and PNGT program, and its relationship the new administrative model, many of the with ONAT are still under discussion. A tentative project responsibilities once assumed by the plan shows the AVV as one of several govern- AVV-including crop and livestock extension, ment, university, or NGO institutions that could agricultural credit, reforestation projects, train- work as a consultant to the DirectionAdministra- ing, and rural cooperative development-are be- tive et Financiere(DAF) of the GTV (Gestiondes Ter- ing returned to the appropriate line ministries. roirs Villageois)planning cellule in the design of Under the new decentralized model, the AVV planning studies (Figure 11.4) ceases to have any special mandate for planning Under the PNGT model (Chapter 10), the pro- for the OCP river basins. The new mandate in- vincial PNGT land use committees work in concert volves shifting the extensive expertise of the AVV with village land use committees. Funding for spe- in forestry, livestock, and dryland crop develop- cific interventions is to come from a variety of ment throughout the OCP valleys, to the design sources, including NGOs and local projects,village and early implementation of similar projects contributions, loans from the CaisseNationale de throughout the country. This function includes Credit Agricole(CNCA), as well as special grants the conceptualization of regional development from the national program. The contract signed by plans (schemadirecteur d'amenagement) for differ- the village land management committee represents ent areas and the implementation of land man- the contractual obligations of the state to assist with agement programs under the new PNGT model. the realization of these interventions. In all of this, Currently under discussion is to what extent the the role of the national planning cellule is seen as AVV will retain its former capacity to construct one of facilitating the sorts of interministerial and

116 Figure 11.4 Relationship of ONAT (ex-AVV)to the functions and financial linkages of the PNGT/GTV

Central Celiule ~NACentral financial office GIV (through CNCA- or another ministry)

Planning studies Formulation of areas Provincial land Landuse plan management committee ONAT(ex-AVV) Uruversityof Ouagadougou / NGOs 1 Pfivateconsfting firms Supplementary financing Local finandal office Agenciesof inemgdzie / from the L(through CNCAor National GTV program ] another ministry)

Implementationof NGOs proposed activities / Local Agendes of line ministries Project induding ONAT CN

\ MLUagepoole Implementaton of fundind proposed activites NGOs

* CNCANational Officesfor AgriculturalCredit

117 interagency collaboration that will be necessary to (4) the implementation of a reliable system of realize the village contracts. rural warehouses providing farmers with fertilizer, animal traction equipment, spare Lessons Learnedand Potential Applications parts, and supplementary feed for draft animals which can be purchased on credit. Advantagesof an AutonomousNational Agency (5) the implementation of a reliable system of rural short- and long-term credit during The history of centralized, national planning for the first five years, the repayment sched- Burkina's OCP river basins highlights the fact ule for which was much higher than for that a separate autonomous agency has definite similar types of credit programs in neigh- advantages in early planning for settlement relat- boring regional development authorities, ed development. In particular, if properly fund- (Barrett etal. 1981); ed, it allows the government to organize a (6) the design and implementation of the first complex assistance package much more quickly. program for assisted spontaneous settle- A further advantage is the ability to attract high ment (Rapadama) in 1988, which provid- levels of outside donor funding. ed the model for many of the later PNGT Within two years of its creation in 1974, the pilot projects; AVV had already begun working on five groups (7) the expansion of this initial assisted settle- of sponsored settlements, which would ultimate- ment program to all of the other areas of ly include 24 sponsored villages. Other achieve- sponsored settlement in the AVV-UP1in ments over the seventeen year life of the project 1989-1990; include (Table 11.5): (8) the successful reorientation of many A9 (1) the completion of more than 17,237,000 ha2 of aerial, hydrogeologic, soil, and development programs (notably at the social planning surveys; UPI at Zorgho, UP3 at Manga, and UP10 (2) the design of area development planning at Dibougou) in order to promote the de- models (schemnasdirecteurs d'mngmn)(Tablemodelsschemsdir samnagent) velopment 11.5); of the adjacent plateau areas for 1,700,000ha 2; ( (9) the design and implementation of some of (3) a successful record in trainig of mnaleand the new experimental programs for set- female extension workers and other staff; the nag ed tagrpaoramzoneth and tler managed agropastoral zones (the Sondre-est and Nouhao Projects are shown on Table 11.5;the Gadeghin agropastoral zone is not); Tableof AVV11.5 Areas Census of (10) a successfulrecord for reforestation projects Table 11.5 Census 1987 198 o such as the forestry project at Wayen; ntretoduig18-98(11) the design and implementation of experi- extension Numberof Tol A Numlteraf mental programs in functional literacy that areas tillages po populatpopaion housws became a model for the later development UPI - Zorgho 154 193,735 131,201 22,104 of national programs by the new govern- UP3 -Manga 148 140,072 70,027 18,146 ment that came into power after 1983; UP4- Tiebele 4 1,372 742 177 Viewed in tabular form, over a seventeen year UPS- Bane 4 1,844 988 164 period, the achievements are major (Table 11.5). UPIO - Diebougou 11 4,788 2,647 542 Sondre- Est 1 1,169 530 76 Negative Aspects of an Autonomous Agency Fara - poura 22 19,305 11,041 1,933 Lanfiera 1 4,707 2,396 465 HIGH COSTS. The chief disadvantage of a sepa- Bagre l 1,019 350 109 rate national agency is cost. In 1979 the cost of Kompienga 13 5,370 2,869 1,398 settling one family at the AVV was estimated to Nouhao 53 47,554 25,720 7,878 be between $12,500 and $15,000 (Table 11.6). TOTALAVV 412 420,935 248,511 52,992 This figure includes only the direct costs of set- Source:AVV. tlement in 1978 terms. How does one calculate

118 the wider impact of this funding on an area been adaptive in the early stages, but as planned over a seventeen year period? Total costs of the and spontaneous settlement increased, and as the AVV in 1988 (Table 11.6) were estimated at AVV expanded its programs to respond to this 23,554,857,000CFA francs. The project was then increase, the administrative structure became in- active in 412 villages (figure includes spon- creasingly cumbersome and less able to deal with sored and unsponsored settlers as well as in- the growing complexity of planning issues (AVV digneous hosts) (Table 11.5), and one industrial 1985c:21).The high concentration of administra- wood project, with 420,935 inhabitants. If these tive and fiscal services in the capital created prob- figures are used, then the total cost per benefi- lems for the fiscal management of projects as well ciary village was 57,171,000 CFA ($191,000, at as for obtaining authorization for necessary pro- 297.9 CFA=$1) if averaged out over the entire gram changes. The same concentrations of per- region. This includes the costs of industrial sonnel in the capital made it difficult for the wood projects and other reforestation schemes central administration to be fully appreciative of that would not directly effect individual villag- the need to adjust certain programs (e.g. the pro- es. The cost per individual, using the same set posed package of intensive cultivation methods) of figures, is 55,958 CFA ($188.00) or 447,664 to the evolving land use patterns, needs, and con- CFA ($1,505) for a hypothetical "average" cerns of both settlers and hosts. household of eight family members. Again, this does not include isolated costs of individual Lessonsfor Future ManagementStructures figures but is averaged out over the entire area of activity. Due to the high cost and long-term administra- FISCAL AND ADMINISTRATIVE DEVOLUTION. tive problems associated with a highly central- Another disadvantage of a separate national ized administrative structure, no one, including agency is fiscal devolution. In the original the ONAT (ex-AVV)administration, would rec- project plan, the AVV administrative structure ommend the recreation of the early AVVadmin- was to end after a fifteen to twenty year period, istrative structures in the other OCP countries. after which time the preexisting network of ad- Nevertheless, the history of the AVV does high- ministrative and extension services would as- light: (1) the solid achievements that can result sume responsibility for its various projects. from sustained commitment to special planning This has proven to be far more complicated for settlement related development; and (2) the than originally expected. The very success of the vital importance of having key administrative AVV in attracting donor financing, and the ex- personnel involved in actual interactions with tremely high level of services in the planned set- beneficiaries at the village level. The insights tlements versus adjacent "indigenous" zones gained by the AVV were earned by designing, created jealousies with other government agen- testing, and then evaluating what did and did not cies. These jealousies were exacerbated by the work. These are valuable lessons for the upcom- fact that the national ministries were expected, ing PNGT program. in most cases, to pick up the long-term recurrent While the actual administrative structure and costs of maintaining the extension, education, focus on planned settlement may have been and health services as well as basic infrastruc- flawed, the basic concept (i.e., the need for in- ture (wells, roads) that the AVVcreated and over formed area planning) was a good one. Based which they'd had little original input. on the AVV experience, we can predict that These difficulties were exacerbated by the fact some similar sort of sustained donor commit- that the local people and leaders often perceived ment over a fifteen year period will be neces- the AVV as having forcibly siezed their land sary in order to realize the full potential of while paying little if any attention to their needs Burkina's national PNGT program. Second, we or concerns. Although the AVV expanded its can predict that the PNGT program will en- program to embrace the indigenous villages in counter many unforeseen problems in achieving the basins after 1983, these early resentments the sort of decentralized administration and have remained. funding that it espouses. ADMINISTRATIVEDIFFICULTY IN ADJUSTING SET- Whatever problems were encountered under TLEMENTAND EXTENSIONPROGRAMS. The central- the early AVV, the institution has an immense ized AVV administrative structure may have experience with land use planning, zoning,

119 Table 11.6 Financing of the AV, 1972-1988(FCFA 000's)

0 Sounteoffinanang 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1988* TOTAL % BurkinaFaso State support - - - 50,000 111,958 58,500 58,000 46,350 78,000 30,000 - 10,000 5,000 - - 447,808 2 Treasury advance - - - - - 200,00,0 00 20,000 ------220,000 0.9 Loan fmm state/CCCE - - - - - 700,000 - - 300,0001,100,0001,840,000 A,050,000 - - 860,000 - 5,850,00 24.8 Rural development fund ------30,000 - - - - - 30,00 0.1 P.T.P. - Kompienga ------986,547 986,547 4.2 Subtotal - - - 50,000 111,958 958500 58,000 46,350 398,0001,130,0001,870,0001,060,A00 5,000 - 860,000 986,547 7,534,355 31.9

Outside sources FAC 110,100 202,000 271,500 535,000 711,500 520,000 530,000 618,000 530,000 510,000 390,000 775,000 475,000 165,000 6,343,100 27 FED - - - - 84,136 5,544 519,960 - 70,723 - 1,350,000 - - - - - 2,030,363 9 RFA ------66,000 214,500 253,096 - - - - -1,285,000 - 1,818,596 8 USAID - - - - - 126,125 84,460 - 200,640 ------411,225 2 CEAO ------30,960 - 44,685 ------75,645 0.3 Netherlands - - - - 239,345 217,791 544,174 756,000 672,000 - - 50,000 - - - - 2,479,310 11 UNDP - - 87,000 174,810 115,230 99,725 ------476,765 2 World Bank ------130,000 - - - - 130,000 0.6 Italy ------2,244,000 - - - - 2,244,000 10 Projet UPV/80 ------51,438 ------51,438 0.2 Subtotal 110,100 202,000 358,500 709,810 1,150,211 969,1851,775,5541,639,938 1,771,144 510,0o 1,740,0003,199,000 475,000 165,0001,285,000 - 16,060,442 68 Total 110,100 202,000 358,500 759,8101,262,1291,927,6851,833554 1,686,288 2,169,1441,640,000 3,610,000 4,259,000 480,000 165,0002,145,00 986,547 23,594,797 100 Percentage 0.5 0.9 1.5 3.2 5.3 8.2 7.8 9.0 9.2 7.0 15.3 18.0 0.2 0.1 9.1 4.2 100

*No new financing in 1987. Source: AW, DEPC, 1990. infrastructure development, extension agent the AVV reorganization, that steps be made to training, industrial wood plantations, and agro- document this rich legacy, and to formalize pastoral zones, which is useful to Burkina and what relation the new office will have to the indeed to the entire eleven country OCP zone. PNGT planning cellule and national planning It is especially important at this juncture, after for the OCP areas.

121 Local Participation and Institutions

Burkina's new program for village land manage- To implement this sort of innovative "bottom- ment supports the development of strong, demo- up" development planning requires a radical re- cratically elected village land management thinking of the way development projects are car- committees that represent each of the major social ried out. In the past, most projects were groups-immigrant and indigenous agricultural- conceptualized by planners at the regional and na- ists as well as pastoralists. Whether these innova- tional levels. They were then taken to the local tive conumittee structures succeed will be groupementswho were responsible for implemen- determined by the villagers' ability to see any real tation. The role of the extension agent and techni- benefits from their efforts. These benefits must cian was to ensure that local perspectives were necessarily be in the form of government pro- considered in the design as a means of facilitating grams to channel resources to villages. This sort of implementation. One goal of the PNGT,and a goal defacto decentralization of resources reinforces the that we consider to be absolutely vital to its suc- de jure authority of the village land management cess, is for the local land management committees committees to provide leadership in working with to see strong evidence that they can actively deter- extension agents and local administrators. There- mine which sorts of projects will be carried out in for donors, NGOs and the national government their village. Under this model, the role of the local need to route as much aid as possible through the government agent should be one of education and local land management committees. arbitrage as opposed to one of enforcing or com- Local leaders are well aware that earned author- municating top-down government edicts. ity, rather than their power to enforce, will be the In this chapter we provide a brief review of the secret to success. At the same time, it is unrealistic factors that contributed to or detracted from the to expect the village land management committees development of effective local institutions at the to earn the needed authority and respect over- study sites. In view of their centrality to the cur- night. Too rapid implementation and literal inter- rent PNGT program, we focus mostly on the farm pretation of the reform texts can lead to village extension groups or groupements. fragmentation, such as we now see in the Solenzo region, where large immigrant cultivation hamlets Centralityof Village Extension Groups in Crop have seceded from their mother villages. More- Extension Programs over, the experience of the AVV sponsored settlements shows that the hostile feelings created by Early in Burkina Faso, as throughout francophone recourse to legal edicts to enforce settlers' land Africa, it was felt that extension programs should rights can block settler economic integration and be tied to the development of cooperatives. actually work against development. Membership in a village groupementwas required

122 for an individual to get credit for a plow, fertilizer, how much, and what cultivation methods to use. and pesticides. Membership was accorded to in- Food aid was also distributed primarily through dividuals who were classified as permanent resi- the groupement. dents of the village. Since groupement members One of the first illustrations that settler institu- were jointly liable for the debts incurred by any tions were taking on a more independent stance member of the group, there was pressure not to took place at V3 Mogtedo during the 1979 cotton admit unreliable persons who might not repay sales (in January 1980),when settlers refused to sell their debts. The groupement was also the focal their cotton on the day scheduled by the project.At point for extension programs. this point they had been living in the project village from three to five years. The strike illustrated the Evolutionof the AVV Groupement Model settlers' growing desire to take an active role in con- trolling village relations with the AW. When the AVV was created in 1974, the planned The institution of farmer managed markets settlement scheme attempted to build on this ear- (marchesauto-geres) for cotton in 1980 was a major lier model of groupementorganization for the cre- turning point in the development of local institu- ation of village institutions. Each planned tions. Farmers' representatives would weigh the settlement was to elect officers. Village groupe- cotton and calculate the money owed to each ments were presided over by bloc committees. farmer. Under the new system, the groupementre- The village groupement was the principal organ ceived a percentage of the total cotton sale as ris- through which AVV directives about new exten- tournes. The total returned to the groupement sion themes were issued. could be over one million CFA ($3,000)in good A second groupement was established for the years or as low as 200,000 CFA ($670) in bad women. The women's groupementled by an ani- years. The level of returns as well as the total val- matrice,or female extension agent. Regular meet- ue of cotton sales varied widely among planned ings with the women focused on maternal health, settlement groups (Figure 12.1). Creation of the child care, nutrition, and handicrafts. farmer managed markets empowered local insti- When the settlers arrived in the planned settle- tutions, providing them with funds under their ments, their first task was to clear trees for their own control. Each year the farmers decide, to- home sites and bush fields. As the new settlers gether with the extension agent, how the returns typically knew only a few families from the same were to be used. home region, the extension agents usually ap- After the 1982 reorganization of the AVV,the pointed the first village groupementpresident and village groupementremained the most important other officers. The wife of the groupement presi- village institution for regulating disputes, orches- dent or another older woman was usually made trating extension activities, and creating and leader of the women's group. Although the new- maintaining basic infrastructure. Each collection ly appointed president was often the first settler of planned settlements (UD) has three types of to arrive, extension agents also used a combina- village groupements:a village committee (groupe- tion of age and first impressions about leadership ment villageois), a women's groupement, and a ability to make the appointment. One of the prin- youth groupement. Representatives of all three cipal concerns of the project was to convince set- groups form the village committee of the UD. tlers that the presidency was not a village Wide variation among the three UDs studied chieftaincy to be held permanently by one family (Linoghin, Mogtedo, and Mogtedo-Bombore) has or person. Despite this, a majority of the first characterized the success of the village groupe- presidents were subsequently reelected, some for ments in carrying out group projects. The Mogte- as long as a decade. do groupement, for example, has managed to During the first two years of planned settle- realize only three economic units-an ambu- ment, the settlers received almost weekly instruc- lance, a cereal bank, and a mill. The mill has never tion from the extension agents on new been adequately maintained and runs for only a agricultural themes. The sound of an extension few months each year. The cereal bank has been agent beating an empty metal drum to call male virtually abandoned. The head extension agent heads of households to village meetings was a has had to take a major role in supervising the ac- regular part of life in the early days of AVV. tivities of all three enterprises. Settlers were told where to plant, what to plant,

123 Figure 12.1 Returns from farmer-managed cotton markets returning to AVV settlement groupements (CFA)

Millions 4

Bombore

1980 1982 1984 1986 1988

Sounre:AW UD extensionrecords.

124 Groupementstructures have been more success- tained the resources to construct schools and ful at Mogtedo-Bombore. There they have real- wells. ized a village pharmacy, a library, a popular bank, * Factors That Contributed to or Detracted a cereal bank, and a small water barrage. In the from the Success of Settler Organizations at the past few years the village groupement has ex- AVV Planned Settlements and at Solenzo plored new avenues for earning cash. Most re- * The existenceof cooperativelyrun, income-pro- cently the committee used one million CFA of its ducing economicunits. A recently completed study funds from cotton returns to purchase cereals, of farm extension groups in the Haut Bassin and which were then sold to two other village groups. Mouhoun CRPAs measured the socioeconomic If success is measured in terms of the realiza- impact of extension groups in terms of their con- tion of particular projects, the most successful of tribution to increasing agricultural production the AVVvillage groupementis at Linoghin (Guira, and creating village infrastructure (Nana 1989f). F. 1989).At least part of this success can be attrib- Two factors proved to facilitate the effectiveness uted to the early involvement of the group, usu- of village groups: access to semiautonomous ally led by the extension agent, in the creation of funds for group projects; and the ability to man- a successful market. The success of the Linoghin age these funds. market in attracting large numbers of people and * Roleof leadershipturnover and education.Ade- in generating revenues gave the settlers an early quate management of funds was largely influ- sense of success, which they have transferred to enced by whether a groupement experienced a other activities. turnover of elected officials like the president or The unites economiques(economic units) run by treasurer (Nana 1989f).Funds tended to be mis- the Linoghin groupements were plagued by cor- managed where groups had little leadership ruption in the early years. As at Mogtedo and turnover. Mogtedo-Bombore, the collectively run mill was The educational background of village leaders the worst run of the different enterprises (Guira, was another key factor that influenced the effec- F 1989).The principal reason for this was the in- tiveness of village groups. When education was ability of a groupementto keep accuiraterecords of concentrated in only a few leaders, it was much funds. In recent years, however, the Linoghin more difficult to manage funds correctly (Nana groupementhas been more successful in the real- 1989f). Groups were more effective when there ization of specific projects including a village was a wider base of education that allowed for drugstore, cereal bank, popular bank, library, vil- more accountability of the elected leaders. lage store, and grinding mill (Guira, F. 1989).The * Complexityof Administration. The most suc- groupement collaborated with the AVV to con- cessful group projects at all of the sites were those struct the drugstore building and a small dam. in which the groupementfunds could be used to The base funding for these enterprises came from develop and repair basic infrastructure. The most the cotton sales ristournes. unsuccessful projects were those that required high levels of day-to-day management. Evolutionof the SolenzoGroupement Model The village mills were notorious examples of this problem. Accurate accounting was made dif- The Solenzo groupement model is a direct out- ficult by the rapid turnover of individuals run- growth of the southwest cotton program. To get ning the mills. The person in charge was usually access to credit, immigrants and hosts in the a literate son or occasionally an older literate set- southwest cotton areas have always been re- tler. The position was poorly paid and poorly su- quired to belong to a groupement.National policy pervised. Even without deliberate intent, it was discouraged the multiplication of groupementsin easy for the mill manager to misuse funds, and, areas of new lands settlement. Consequently, new even if the funds were not mismanaged, the man- settlers generally had to join a village group in ager was constantly under suspicion. As a result, one of the older settlements. the position was unattractive and generally taken As at the AVV, the new national program of only as a last resort. farmer managed markets marked an important One result of the financial mismanagement and tuming point in the functioning of these groups. rapid turnover in staff was that little if any provi- With the creation of these markets, the groups ob- sion was made for repairing the equipment. Throughout all the AVV planned settlements, the

125 cooperatively run mills tend to be broken down. areas assigned to settlers who have resigned One possible solution to the problem that was be- from the project. In retrospect, had the AVV giving considered in 1990 was to give responsibility en them more responsibility for these and simi- for the mills to women's groups, because "they lar activities earlier in the project, it might well would presumably be more honest." have increased the settlers' sense of proprietor- * Socialproblems. Social problems among set- ship in the new area. tlers, in particular the animosity between groups The AVVsettlements created at Mogtedo-Bom- of settlers from different regions, contributed to bore benefitted from lessons learned earlier in the the lack of effectiveness of local institutions at project. The system of farmer managed markets Mogtedo. These problems appear to have been was instituted only one year after AVV-Bombore exacerbated in one village by the high concentra- was created. The income from the farmer man- tion of settlers from a particular home village. aged markets was used for various group The initial result was to give this group an advan- projects. There was, then, from an earlier date, a tage over others in taking on village leadership much stronger tradition of farmer initiative in the positions. Later, growing income inequalities cre- implementation of projects than we saw at Mogt- ated jealousies within the group, as the success of edo. This stronger tradition of farmer organiza- certain families from the same home village out- tions may be partly responsible for the lower rate stripped that of others. In general, however, these of out-migration from the Bombore planned set- problems were secondary factors related to the tlement. A comparison between farmer organiza- more general economic malaise that character- tions at Mogtedo and Bombore needs additional ized economic development in the AVV planned investigation. settlements. Although the UDs of Linoghin and Mogtedo * Settlerfeelings of dependency.A key problem were created at about the same time (1973-1974) associated with planned settlements like the AVV and with the same pattern of early institutional is the sense of settler dependence on the project development, the village groupementat Linoghin administration for solving day-to-day social and became more strongly involved with the admin- economic problems. The early days at the AVV istration of the market after the second year. The were characterized by a high level of dependence market generated revenues and created a wide on the project administration. After the fifth year, range of opportunities and problems that re- however, the settlers began to take a more inde- quired group decisions. pendent stance in dealing with the administra- * Nationalpolicies. National policies that con- tion and in dealing with one another. tributed to the development of effective local insti- * Socialand economiclinkages betuween settler in- tutions were: (a) extension policies that created stitutions and externalparallel structures. High set- only one extension group per village with joint tler dependency on project administration for migrant-host membership; (b) policies to decen- handling specific problems is a natural conse- tralize responsibility for cotton markets, thus giv- quence of extension and aid programs that are ing the local village groupementsoperating funds "top-down" or directive.When there is a problem for village projects; (c) programs that created a lo- with the program, settlers look to the agency to cal, regional, and national administrative position resolve it. for the village revolutionary committees (CRs). The AVV has been extremely critical of the di- A key policy that worked against the devel- rective nature of its early extension programs. opment of local institutions involved land ten- These rigid lines of communication, in which the ure changes introduced in 1984 that provided AVVtold the farmers exactly how to plow, when an administrative mechanism for cultivation to plant, how much fertilizer to apply, when to hamlets to break with the villages that original- weed, and when to sell, continued long after ly loaned them land. Under the new laws, once they might have been considered necessary for recognized as an independent village, the set- training purposes. Settlers were given little role tlers had the right to form their own groupement, in the design or administration of village and no longer had to collaborate with the hosts. projects until 1980 when the AVV switched to These secessions have in many cases created farmer managed markets. It is only in the last conflict with the indigenous inhabitants and few years that these same settler organizations done away with the former integrative role of have been given the power to reallocate the field the village groupement.

126 Early Development of Settler Organizations Despite the out-migration, the Kompienga CR at Kompienga continued to meet during 1990. Their activities included the regulation of disputes between the The significance of these different local and na- FulBe and local agriculturalists. The CR also tional factors variables is especially striking at continued to meet with the Prefet and other offi- Kompienga where effective, local community in- cials about such issues as proposals to zone the stitutions were fully operational only three years town, efforts to improve transportation linkag- after large-scale settlement began, with very little es, and the need to improve crop and livestock direct intervention from the outside. This legacy extension services. of cooperation dates back to dam construction, In 1991 and 1992, after the CR's were abol- when the dam authority relied very heavily on ished, the group transformed itself into a com- the CRs in its dealings with the workers. These mittee that promotes small-scale, irrigated linkages facilitated the resolution of concrete vegetable production. In contrast to farmer problems like housing, water, health, and educa- groups at the other sites, a high proportion of the tion. The linkages were reinforced by resource leaders at Kompienga are literate. This strong flows. National and donor financing was used to tradition of village institutions at Kompienga create wells; to delineate house sites; and to create represents a considerable development re- a market, primary school, hospital and comnmnuni- source. At the present time, however, the groupe- ty center. The administration also worked with ment's activities are hindered by the lack of the CRs to organize social events. effective linkages to the regional extension office One could argue that the dam administration in Fada, illustrating the critical importance of ef- had a strong economic motivation for keeping fective national institutions. the workers happy The net result, however, was to create-very early in the settlement experi- Policy Implications ence-a tradition of active settler representation in decisions made by the local administrators in Both national policies and local administrative the town and department headquarters at near- practices influence the development of local ad- by Pama. This tradition continued after the dam ministrative structures. Neither is sufficient in authority turned over its administration of the and of itself. Local projects that espouse a doc- town to local representatives of the different trine of high levels of settler participation may ministries. In early 1990, the leadership of the succeed in the short run but be unsuccessful Kompienga CR continued to have a large num- when these institutions must answer to wider ad- ber of former dam workers complemented by lo- ministrative and extension structures. Similarly, a cal merchants who were not dam workers. The national program (such as the early AVV or CR leaders were vocal in expressing their aspira- PNGT) may espouse high levels of local partici- tions and concerns for the new community. One pation, but be unwilling to relinquish any real ad- of their first activities, after inauguration of the ministrative responsibility or resources. Other dam, was to reorganize the market so that it national programs (like land tenure reform) may would be held on two fixed days each week in have many unintended consequences for creating the hope that this would attract greater trade settler-host conflict. These conflicts may, in turn, from Togo. discourage the settlers' willingness to participate During the 1990 dry season, Kompienga lost in jointly run local institutions. about 100 males (not all of them household Policies that do not build on and incorporate a heads) who immigrated to the proposed site for strong base of settler and host initiative are the new Bagre Dam. One of those who left was bound to fail in the long run. Moreover, failure the charismatic CR head delegate. As of April, to build on the expressed wishes and capacity of only twenty had allegedly found work. The ru- local institutions can lead national institutions mor was that because of the much greater diffi- and donors to advocate inappropriate interven- culty of obtaining land in the Bagre region (due tions that in worst case scenarios may actually to the high population densities of the indige- endanger settlers' social, economic and physical nous Bissa of the area), most of the unemployed welfare. In the final analysis, the impact of any migrants would return to Kompienga for the sort of national policy to affect crop or input rainy season. prices policy (Chapter 4), land tenure (Chapter

127 5), agriculture or off-farm employment (Chap- lation of sound development policies for the ters 6 and 7), pastoralists and indigenous host affected areas. land rights (Chapter 9), or the development of These areas represent one of West Africa's great- markets and administrative centers (Chapter 8), est natural resources. It is unlikely, however, that sustainable land use practices (Chapter 10), na- the full economic returns to river blindness control tional management structures (Chapter 11), or will be realized without sound management and local institutions (Chapter 12) is ultimately de- parallel investment in development programs. termined by the decisions of millions of actors The Burkina case study shows ample evidence living in hundreds of thousands of household for success in areas where new lands settlement units in small communities was assisted. The study also shows the high costs It is therefore important that planners under- of doing nothing. This is a rich history that needs stand the needs, concerns and aspirations of the to be thoroughly analyzed before planners embark farm families living in the OCP river basins. This on a new generation of development planning in understanding is the vital first step to the formu- the valleys.

128 Bibliography

Agrotechnik. 1988.Etude de developpementregional .1974. La Mise en valeur des Valleesdes Vol- dans le bassinversant de la Kompienga.Rapport diag- ta: Principes d'am6nagement et perspectives. nostic.Texte. Frankfurt: Agrotechnik. Ouagadougou: AVV.novembre. . 1989.Etude de developpementregional dans le . 1981a.Nouvelles methodes d'intervention bassin versant de la Kompienga, Refe'rence. de l'AVV. Tome 1. Principes generaux. Ouaga- N':84.70.049.Rapport final. Frankfurt: Agrotech- dougou: AVV.septembre. nik. mars. _. 1981b.Nouvelles methodes d'intervention Ancey, G. 1974. "Facteurs et systemes de produc- de l'AVV.Tome 1, (suite) annexes. Ouagadougou: tion dans la societe mossi d'aujourd'hui: migra- AVV.septembre. tions, travail, terre et capital." In Enquete sur les _ . 1983. Programme de d6veloppement de mouvements de populationa partir du pays Mossi, I'AVV1983-1986. Ouagadougou: AVV.mars. Tome II. Ouagadougou: ORSTOM. .1984a. Etude de factibilite. Unite de Plani- Angel, S. 1985. "Spontaneous Land Settlement on 1. Zorgho. uaaduui AVun. (3 Rural Frontiers: An Agenda for a Global Ap- tom) Z proach." Paper presented at the International tomes.) Seminar on Planning for Settlements in Rural Re- . 1984b.Etude agropastorale. Volet systeme gions: The Case of Spontaneous Settlements. d'elevage. UP1. Zorgho. decembre. Nairobi, Kenya, 11-20November. - .1984c. La production vivriere des AVV (Autorit6 des Amenagements des Vallees perimetres amenages de l'AVV au cours de la des Volta). 1973a. Programme d'infrastructures, campagne agricole 1983-1984. Ouagadougou: 1974-1975.Ouagadougou: AVV.novembre. AVV,Cellule de Suivi-Evaluation. mars. . 1973b.Programme de mise en valeur 1974. . 1985a. Plan quinquennal AVV 1986-1990. Ouagadougou: AVV.novembre. Synthese. Document Provisoire. Ouagadougou: -. 1973c. Propositions d'actions en vue de AVV. I'amelioration de l'elevage. Ouagadougou: AVV. . 1985b. Rapport d'evaluation-Unite de . 1973d. Demande de financement au fonds Planification 1. Zorgho-Burkina Faso. Ouaga- d'aide et de cooperation. Ouagadougou: AVV. dougou/Zorgho: AW. fevrier. .1973e. Bilan des activit6s, annee 1973-Pro- . 1985c.L'Impact socio-economique du programme 1974. Ouagadougou: AVV. gramme de lutte contre l'onchocercose au Burki-

129 na (1974-1984). Ouagadougou: AVV. juin. (10- - . 1990. Rapport sur la gestion des terroirs year report.) villageois de l'UD de Linoghin. Zorgho: AVV- . 1985d. Annexe I. Textes fondamentaux UP1. COMRF. relatifs a la reorganisation fonciere et agraire au Bakyono, A-M. 1989. Etude des migrations: Burkina Faso. Programme de reforme agraire et Burkina Faso. (Mimeo.) gestion de 1'espace UP1. Zorgho. Ouagadougou: Baris, P., P. Bonnal, and M. Pescay. 1983. Am6nage- Direction des Etudes, Programmes et du Con- ment des Vallees des Volta. Mission d'Evaluation. tr6le. avril. Ouagadougou. Novembre. (Mimeo.) . 1988. Programme reforme agraire et ges- Barrett V. G. Lassiter D. Wilcock D. Baker and tion de 1'espace UP1-Zorgho (U.D. de Linoghin Baret V. G.Lsitr'. Wicok D.Bkr n tion de RepademaP. R-porg (U.D. de Linghn E. Crawford. 1981. Animal Traction in Eastern Up- UgD. oe Rapadama). Rapport general. uuaga- per Volta: A Technical, Economic and Institutional dougou:DEPC, Service Etudes et Programmes. Analysis. East Lansing, MI: Department of Agri- . 1989a. Aide memoire. Mission: Evaluation cultural Economics, Michigan State University. de I'AVV. Mise en place d'une nouvelle structure. Batabe, C. M. 1989. Mise en oeuvre d'un projet de Ouagadougou. 3-20 d6cernbre. mise en valeur des terres neuves au Burkina: cas -. 1989b. Document de transfert de la zone de l'autorite d'amenagement des vallees des Vol- pastorale de Gadeghin/UP1 au CRPA du Centre. ta. Ouagadougou: IPD/AOS. 89/9/IPD.AOS. Ouagadougou: AVV.septembre. Becker, B. K. 1985. "Spontaneous/Induced Rural . 1989c. Document de transfert de la zone Settlements in Brazilian Amazonia. Paper pre- pastorale de la Nouhao au CRPA du Centre Est. sented at the International Seminar on Planning Ouagadougou: AVV.septembre. for Settlements in Rural Regions: The Case of 1989d. Rapport: Transfert du projet reboi- Spontaneous Settlements." Nairobi, Kenya, 11-20 sement industrie de Wayen au Ministere de l'En- November. vironnement et du Tourisme. Ouagadougou: Bellot, J. M., and H. Capral. n.d. Etude institution- AVV.9 aout. nelle. Programme PNGTV. Analyse et proposi- 1989e. Rapport des activit6s de I'AVV. tions a l'issue de l'enquete terrain. (Mimeo.) Campagne 1988/89. Ouagadougou: AVV. Benoit, M. 1973a. Espaces agraires mossi en pays . 1990. Rapport annuel: Projet transfert des Bwa.2 tomes. Ouagadougou: ORSTOM. populations de la Kompienga. fevrier. - . 1973b. "Le champ spatial mossi dans les AVV-UP1 (Autorit6 des Am.nagements des pays du Voun-Hou et de la Volta Noire (Cercle de .. dePlanificatNouna,des Volta.Unit Haute-Volta)." Cahiers des Sciences Hu- V allees des Volta-Umite de Planification N°. 1). mansX2)151 * mines X(2):115-137. n.d. Programme zones-tests. Province du Gan- B zourgou. Zorgho: UP1, Cellule Organisation du Berg, E. J. 1975. The Recent EconomicEvolution of MondgouRuralo eo IonCellul. the Sahel. East Lansing, MI: University of Michi- gan, Center for Research on Economic Develop- . 1986. Rapport annuel 1986. Zorgho: UP1, ment. Cellule Production Animale et V6g6tale. Berg, E. J., J. Bisilliat, M. Burer, H. Graetz, R. . 1987. Rapport d'activites. 30 juin. Zorgho: Melville, V. Volyvan, J. Park, R. C. Sawadogo, H. UP1, Cellule Production Animale et V6getale. Sederlof, and K. van der Meer. 1978. Onchocercia- -. 1988. Rapport d'activites. Premier semestre sis Control Programme: OCP Economic Review 1988. Zorgho: UPI, Cellule Production Animale Mission. et Vegetale. BF Proj. Engr. (Burkina Faso Projet Engrais). 1988- .. 1989a. Rapport dac1989 Projet Engrais Vivriers, Rapport annuel 1989.1989a.Rapportons pcivrreux Zorgho .im d'activites no. 4, 1988/89, Ministere de l'agricul- de Planification N. I.o Cellule Production Ani- ture et de l'elevage, Direction des intrants et de la male et Vegetale. Autorit6 des Amenagements m6canisation agricole. des Vallees des Volta, Unite de Planification N 1.

130 Bharin, T. S. 1981. "Review and Evaluation of At- Coulibaly, S., J. Gregory, and V. Piche. 1980.Impor- tempts to Direct Migrants to Frontier Areas tanceet ambivalencede la migrationvoltaique. Les mi- through Land Colonization Schemes." In Popula- grationsvoltaiques. Tome I. Ouagadougou: Institut tion Distribution Policiesin DevelopmentPlanning. national de la statistique et de la demographie. Population Studies No. 75. New York,NY: United Couty, P., J. Y. Marchal, P. Pelissier, M. Poussi, G. Nations, Department of Intemational Economic Savonnet, and A. Schwartz, eds. 1979. Mattrise de and Social Affairs. 1'espaceagraire et Developpernenten Afrique Tropi- Boutillier, J. L. 1964. "Les structures foncieres en cale:Logique Paysanne et RationaliteTechnique. Par- Haute-Volta." Etudes Voltaiques,nouvelle serie 5:84- is: Office de la Recherche Scientifique et Tech- 94. nique Outre-Mer. Broekhuyse,J. T. 1974.Developpement du norddu pla- CRPA (Centre Regional de Promotion Agro-pas- teau mossi.4 tomes. Amsterdarm Departement des torale de la Boucle du Mouhoun). 1988. Rapport recherchessociales, istitut royal des tropiques. d'achevement de projet. Projet de Developpe- 1982a. Production et productiviti agricoles ment Agricole de la Boucle du Mouhoun (Ex-Vol- dans la savanneseche. Amsterdam:* r DepartementO ~~~~~~tat Noire). Dedougou:CRPA du Mouhoun. des recherches sociales, Institut royal des regions - .1989a.Rapport d'activites: Campagne agri- tropicales. cole 88/89. Dedougou: CRPA du Mouhoun. . 1982b. Restructurationdes institutions ru- Avril. rales. Amsterdam: Departement des recherches - . 1989b. Rapport d'achevement de projet. sociales, Institut royal des regions tropicales. Projet de developpement agricole de la Boucle du Caldwell, J. C. 1975. The SahelianDrought and Its Mouhoun (Ex-VoltaNoire). Dedougou: CRPA du DemographicImplications. Overseas Liaison Com- Mouhoun. decembre. mittee, Paper No. 8. Washington, DC: American - .1989c.Statistiques agricoles. R6sultats d'en- Council on Education. quete production. Campagne 1988/89. De- Capron, J., and J. M. Kohler. 1975. "Environne- dougou: CRPA du Mouhoun. ment sociologique des migrations agricoles." In - . 1989d. Le Programme gestion des terroirs Enquetesur les mouvementsde populationA partirdu villageois-Mouhoun programme d'6tudes et de pays mossi. Tome I. Ouagadougou: ORSTOM. rechel2he. Janvier 1988. Dedougou: CRPA du Chambers, R. 1969. Settlement Schemesin Tropical Mouhoun. Africa:A Study of Organizationsand Development. - .1989e.Statistiques agricoles. Resultats en- New York: Praeger. quete production. Campagne 1988/89. De- Clanet, J. C. 1983.Rapport sur le repeuplement des dougou: CRPA du Mouhoun. vallees liberees de l'onchocercose dans l'ouest - .1990a. Enquete sur les marches de Sous- voltaique (secteurs de la Comoe, de la Leraba, de secteur de Solenzo. Dedougou: Service Suivi- Samandeni et de St. Pierre.Ouagadougou: Univer- evaluation, CRPA du Mouhoun. sit6 de Ouagadougou. 20 Novembre. - . 1990b.Le ProgrammeTest de Gestiondes Ter- Commission Technique Interministerielle de la roirs du Mouhoun: Etat d'avancement et premiers Reforme Fonciere et Agraire en Haute-Volta. n.d. enseignements et difficultes.Dedougou: CRPA du Ouagadougou. (Mimeo.) Mouhoun. Conde, J. 1978.Migration in UpperVolta. Washing- DAC (Direction de l'Action Cooperative). n.d. ton, DC: World Bank, Development Economics Ministere de l'Agriculture et de l'Elevage. Secre- Department. tariat General. Communication du Burkina Faso Conti, A. 1979. "Capitalist Organization of Pro- A la Table Ronde sur la Dynamique des regies duction through Non-capitalist Relations: Wom- fonciees et des systemes agraires en Afrique. en's Role in a Pilot Resettlement in Upper Volta." Ouagadougou. (Mimeo.) Review of African PoliticalEconomy 15/16:75-92. Dejou, C. 1987. La Dynamiquede la commercialisa- tiondes ce'realesau BurkinaFaso: Rapports re'gionaux.

131 Ann Arbor, MI: Center for Research on Economic bre 1988. Banfora: Ministere de l'Environnement Development. 2 volumes. et du Tourisme, Direction Provinciale de la Com- Delgado, C. L. 1979. "Livestock versus Foodgrain pie. Projet GCP/BKF/OC6/SWE. D&embre. Production in Southeast Upper Volta: A Resource Gladwin, C. and D. McMillan. 1988. "Is a Turn- Allocation Analysis." In LivestockProduction and around in African Agriculture Possible Without Marketingin theEntente Statesof WestAfrica. Vol. 1. Helping Women Farmers?" Economic Develop- East Lansing, MI: University of Michigan, Center ment and CulturalChange 37 (2):345-69. for Research on Economic Development. Gregory,J. 1974. "Development and In-Migration Deniel, R. 1967. De la savannea la ville:Essai sur la in Upper Volta." In Modern Migrations in West Af- migration des Mossi vers Abidjan et sa region. Aix- rica. Samir Amin, ed. pp. 304-320. London: Ox- en-Provence: Centre Africain des Sciences Hu- ford University Press. maines Appliqu6es. Guira, A. 1989. Compte rendu de reunion de la Deuson, Robert R., and J. H. Sanders, 1990. "Ce- Commission Interministerielle de Concertation real Technology Development in the Sahel: Burki- sur le Schema directeur du bassin versant de la na Faso and Niger," (Viewpoint), Land UsePolicy Kompienga. 14 juin. 7(3):195-197. Guira, F. 1989. Groupement Structures and Vil- de Wilde, J. C., P. F. M. McLoughlin, A. Guinard, lage Markets: Linoghin, Mogtedo and Mogtedo T. Scudder, and R. Maubouche. 1967. Experiences Bombore. Field Report. Draft. with Agricultural Development in TropicalAfrica. Guissou, J. 1977. Etude sur les besoinsdes femmes Baltimore: The Johns Hopkins University Press dans les villagesde 'AVV et propositiond'un pro- for the International Bank for Reconstruction and damme vinterve OuAgadopou: d' Development. Development.(2 volumes.) grammed'intervention. Ouagadougou: SAED. Dione, J. 1989. "Food Security Policy Reform in Guyon, G. 1986. Rapport d'6valuation de la phase Mali and the Sahel." International Economic As- preparatoire d'un programme de gestion des ter- sociation. IXth World Congress, Athens, Greece, roirs villageois Burkina Faso. Ouagadougou: 28Augustio SepthW rle C Caisse Centrale de Cooperation 28 August-I September. D6cembre. Economique. Djigma, A. 1989.Memorandum: Participation a la Hammond, P. 1959. "Economic Change and Mos- definition et a la mise en place d'une nouvelle struc- si Acculturation." In Continuity and Changein Af- ture suite Ala restructuration de I'AVV.5 Octobre. ricanCultures. W. R. Bascom and M. J. Herskovits, Dollfus, 0. 1981. "Ph6nomenes pionniers et eds. Pp. 238-256. Chicago, IL: University of Chi- problemes de frontieres: Quelques remarques en cago Press.

guise de conclusion." In Les Phe'nomMnesde "Fron- -. 1962. TechnologicalChange and Mossi Accul- ti~re"dans lespays tropicaux.Travaux et M6moires turation. Ph.D. dissertation, Northwestern Uni- de l'Institut des Hautes Etudes de l'Amerique versity. Latine 32:445-448. Dr.-ing. Walter International.1987. Etuded'elabora- 1964. "Mossi Joking." Ethnology3:259-267. tion d'un programmede d6placementdes populations .1966.Yatenga:TechnologyintheCultureofa West de la zoned'inondation du barragehydro-electrique de AfricanKingdom. New York,NY. The Free Press. la Kompienga. Wittenbergstrasse, Republique Harris, J., and M. P Todaro. 1970."Migration, Un- Fed&raled'Allemagne: DIWI. employment, and Development: A Two-Sector FAO (Food and Agriculture Organization, United Analysis." AmericanEconomic Review 60:126-142. Nations). 1987. Project Progress Report: Hervouet, J. P. 1977. Peuplementet mouvementsde Am.nagement de la for.t class6e de Toumousse- populationdans les valleesdes Volta blancheet rouge. ni, Province de la Comoe. Ouagadougou. Oagadougou: ORSTOM. .1988. Amenagement de la foret classee de 9____. 1978. "La mise en valeur des vallees des Toumousseni, Province de la Comoe. Mission d'E- Volta blanche et rouge: Un Accident his- valuation Tripartite. Gouvernement Burkinabe/ Gouvernement Suedois/FAO du 5 au 15 decem-

132 torique." Cahiers ORSTOM: Se'rie sciences hu- - . 1971. "Les Yarse et le commerce dans le mainesXV(I):81-87. Yatenga pre-colonial." In The Developmentof In-

____ 1980. Du Faidherbia la Brousse: Modifica digenous Tradeand Markets in West Africa. London: tions culturales1980.Dculuraleset etFaidherada degradation samttaire.saniroe: MOdiaga-Ouaga- rcnIsiueOxford University Press for the International Af- dougou: ORSTOM. rican Institute. 1983_Bilan_de_l'occupation des ten-es des Izard, M., and F. Izard-Hertier. 1958. Bouna: ,,,1~es.1983.,.,,r~es Bilan de ,'onchocercosede l'occupathon ,pr~ des terres10 ansde des MonographieM d'un villagePana de la valleedu Sourou lutteantisi . delnn haute V a-Mali.,.uaga (Haute-Volta). Ouagadougou: Service de l'hy- lutte a idraulique de Haute Volta pour l'Institut des Sci- dougou. (Mimeo.) ences Humaines Appliquees (ISHA) de l'Univer- Hervouet, J. P.,J. C. Clanet, F Paris, and H. Some. site de Bordeaux. 1984.Settlement of the ValleysProtected from Oni Izard-Hertier, F., and M. Izard. 1958. Aspects hu- chocerciasis after Ten Years of Vector Control in mains de l'ame'nagementhydro-agricole de la Valle Burkina. OCP/GVA/84.5. Ouagadougou: OCR du Sourou. Ouagadougou: Service de l'hy- Horowitz, M., D. Koenig, C. Grimm, and Y draulique de Haute Volta pour l'Institut des Sci- Konate. forthcoming. "Involuntary Resettlement ences Humaines Appliquees (ISHA) de l'Univer- in Manantali." In Anthropologyand InvoluntaryRe- site de Bordeaux. settlement:settlementPoiy.rciead7er.MPolicy, Practiceand Theory. M. Cerneaee 1959. Les Mossi du Yatenga: Etude de la vie and S. Guggenheim, eds. Boulder, CO: Westview. . 1 u economiqueet sociale.Bordeaux: Institut des Sci- Hunting Technical Services, Ltd. 1988a. Final Re- ences Humaines Appliquees (ISHA). port:Socioeconomic Development Studies in the On- choceciass CotrolProgrmmerea.Volue 1 Jaeger, W. K. 1983.Agricultural Mechanization: The chocerciasis ControlProgrammeArEconomics of Animal Draft Powerin Africa.Boulder, Main Report. September. CO: Westview Press. .1988b. SocioeconomicDevelopment Studies in the OnchocerciasisControl Programme Area. Volume Kabore, B. R., A. Brilleau, and G. Badolo. d985. 2: National Oncho Zone DevelopmentStudies and Ptea d uagadougu UVV. Ao. DevelopmentProposals. September. P .1988&.Socioeconomic Development Studies in Kabore, F, and J. Guigma. 1990. Enquete Marche. the OnchocerciasisControl Programme Area. Volume August. 3: National Oncho Zone DevelopmentStudies and Kafondo, T. 1989. Rapport de stage: La Gestion DevelopmentProposals. September. des terroirs villageois: Cadre d'application de la *.1988d. Socioeconomic Development Studies in reorganisation agraire et fonciere en milieu rural the OnchocerciasisControl Programme Area. Volume au Burkina Faso. Cas de Rapadama UP1/AVV. 4: Bibliography.September. Octobre. Maitre de Stage: Some Jules Marie. Oua- gadougou: AW. IFDC. 1987. The Fertilizer Market Situation and Kohler,J. M. 1968.Activitis agricoleset transforma- Outlookfor Eleven West African Countries.Muscle tionsocio-conomique de l'ouest du Mossi.Paris: OR- Shoals, AL: International Fertilzer Development STOM. Center. August.

IMF (International Monetary Fund). 1989. Burki- -. 1972.Les migrations des Mosside l'ouest.Par- na Faso. International Financial Statistics Year- is: ORSTOM. book, 1989.Washington, DC: IMF.P. 256. Lahuec, J-P. 1970. "Une communaute evolutive Izard, M. 1965. "Traditions historiques des vi.ag- Mossi Zaonghe (Haute Volta)." EtudesRurales: 37- es du Yatenga. Cercle de Gourcy." t Recherches Voltaiques1. Lesselingue, P. 1975. "Migrations internes-As- . 1970. "Introduction a l'histoire des pects psycho-sociologiques." In Enquete sur les mossi." RecherchesVoltaiques 12-13. mouvementsde population 4 partir du pays Mossi, Dossier I, Fascicule 2. Ouagadougou: ORSTOM.

133 Liese, B., J. Wilson, B. Benton, and D. Marr. 1991. . 1987b. "The Social Impacts of Planned Set- The OnchocerciasisControl Programme in WestAfri- tlements in Burkina Faso." In Droughtand Hunger ca. PRE Working Paper. No. 740. Washington, in Africa: Denying Famine a Future. Cambridge: D.C.: World Bank. Cambridge University Press.

Mabogunje, A. L. 1981. "Objectives and Ration- -. 1989a. Draft Country Case Study: Burkina ales for Regional Population Redistribution in Faso. December. (Analysis of Material from Site Developing Countries." In PopulationDistribution Reports and Other Research at AVV-UP1,Komp- Policiesin DevelopmentPlanning. Population Stud- ienga, Solenzo, and Niangoloko.) Binghamton, ies, No. 75. New York, NY: United Nations, De- NY: IDA. partment of International Economic and Social - . 1989b. "Seasonality, Planned Settlement Affairs. and River Blindness Control." MASCA Research Maitrise d'Ouvrage de la Kompienga (MOK). Papersin Scienceand Archaeology.Vol. 96-120. 1985. "Apercu et orientation strategique de l&-. forthcoming. Adapting Rapid Assessment conomie du Burkina Faso." 8me Conference des Procedures (RAP) to Development Planning in Bailleurs de Fonds du Projet Hydro-electrique de Areas Covered by Disease Control Programs: On- Kompienga. Ouagadougou. chocerciasis." In Rapid AssessmentProcedures. N. Marchal, J-Y.1975. "Geographie des aires d'emi- Scrimshaw and S. Scrimshaw, Eds. Cambridge, gration en pays mossi." In Enquetesur les mouve- Mass.: UN University Press. ments de population4 partir du pays mossi. Dossier McMilan, D., T. Painter, and T. Scudder. 1990. II, Fascicule 3. Ouagadougou: ORSTOM. OnchocerciasisControl Programme,Final Report of . 1978. "L'Onchocercose et les faits de peu- the Land Settlement Review: The Experiencewith plement dans le bassin des Volta." Journaldes afri- Land Settlement in the OCP River Basinsand Strate- canistes48,2: 9-30. gies for Their Development.Binghamton, NY: Pre- Matlon, P. J., and D. Spencer. 1984. "Increasing pared for the Committee of Sponsoring Agencies Food Production in Sub-Saharan Africa: Environ- of the Onchocerciasis Control Programme by the mental Problems and Inadequate Technological Institute for Development Anthropology. May Solutions." American Journalof Agricultural Eco- Ministere de la Question Paysanne. 1987. Opera- nomics56:671-76. tion speciale de demenagement des populations McMilan, D. 1980. Land Rights and Resettlement. illvgalement installees dans les formations Purdue University: Department of Agricultural foresti6res class6es du Burkina Faso. AoCit. Economics. Murphy, J., and L. Sprey. 1980. The Volta Valley _.1983. A ResettlementProject in Upper Volta. Authority: Socio-economicEvaluation of a Resettle- Ph.D19.dissesetatlern, nthProjetoinUpperrVolta. ment Project in Upper Volta. West Lafayette, IN: Ph.D. dissertation, Anthropology Department, Department of Agricultural Economics, Purdue Northwestern Unliversity. University. 1984. Changing Patterns of Grain Production in Nagy, J., J. Sanders, and 0. Ohm. 1988. "Cereal a ResettlementScheme in UpperVolta. Washington, Technology Interventions for the West African D.C.: Center for Women in Development, South- Semi-Arid troi,Aicultur economics7 East Consortium for International Development. 208. 1986. "Distribution of Resources and Prod- Nana, J. B.1989a. Rapport sur le site de a zone de ucts in Mossi Households." In Foodin Sub-Sahar- sous-secteur de Niangoloko et de la foret classde an Africa. A. Hansen and Della E. McMillan, Eds. eoumousseNi(nglish version19 Report Boulder, CO: Westview Press. de Toumoussem. (English version: 1989 Report on the Niangoloko Site and the Toumousseni For- 1987a. "Monitoring the Evolution of est Reserve.) Binghamton, NY: IDA.

Household Economic Systems over lime in _ . 1989b. Rapport sur le site. Programme Farming Systems Research." Development and PNGTV Rapadama. (English version: 1989Site Re- Change18:295-314. port PNGTV Rapadama.) Binghamton, NY: IDA.

134 -. 1989c. Rapport sur le site de sous-secteur programme et propositions d'implantation du- de Solenzo. (English version: 1989 Solenzo Sub- rant sa 1 re phase. Ouagadougou: Ministere du sector Site Report.) Binghamton, NY: IDA. Plan et de la Cooperation, PNGTV.Novembre. 1989d.Enquete Ressources Naturelles de la Queant, T., and C. de Rouville. 1969. "Etudes hu- Zone de Niangoloko. Binghamton, NY. IDA. maines sur la r6gion du Bondo Sourou. Agricult- -. 1989e.Enquete Ressources Naturelles dans eurs et deveurs de la r6gion du Gondo-Sourou." * Ouagadougou: Travauxde CVRS, n°. 1. le sous-secteur de Solenzo. Binghamton, NY: IDA. -1989f. Enqu&e sur les groupements. Oua- Reardon, T., P. Matlon, and C. Delgado. 1988. 1989f.CaisseCentrale. "Coping with Household-level Food Insecurity gadougou: Caisse Centralein Drought-Affected Areas of Burkina Faso," Nana, J. P.,and D. Kattenberg. 1979.Etude prelim- WorldDevelopment 16 (9): 1065-1074. inaire de la question des migrants spontanes. Oua- Reardon, T., C. Delgado, and P. M. Matlon. gadougou: DEPEC, Section Sociologie, AVV. 1992. "Determinants and Effects of Income Di- Nelson, M. 1973. The Development of Tropical versification Amongst Farm Households in Lands: Policy Issues in Latin America. Baltimore: Burkina Faso." Journal of Development Studies. Johns Hopkins University Press for Resources for January. the Future. Remme, J., and J. B. Zongo. 1989. "Demographic Nikyema, J. J. 1977. Memoire de fin d'etudes: Mi- Aspects of the Epidemiology and Control of On- gration organisee de population (AVV). Ouaga- chocerciasis in West Africa" in Demographyand dougou: AVV. VectorBorne Diseases. M. Service, ed., Boca Raton, OCP (Onchocerciasis Control Programme). 1986. FL.: CRC Press, pp.363-386. Report on the Evaluation of the Socioeconomic Remy,G.1973. Les migrationsde travailet lesmouve- Impact of the Onchocerciasis Control Pro- ments de colonisationMossi. Paris: ORSTOM. gramme. Report No. JPC7.3 (OCP/86.7), On Rey, C. 1980. Analyse de la situation agro-pasto- chocerciasis Control Programme in West Africa, rale dans l'ORD du Centre-Nord, Kaya (janvier Joint Programme Committee, Seventh Session, 1980). RHV: Service Departemental de Planifica- Accra, Ghana, 9-12 December. tion du D6partement du Centre-Nord. Ouedraogo, F. 1976.L'amenagement du blocde Mogt- SAED (Soci&t Africaine d'Etudes et de D6vel- edodans le cadrede la miseen valeurdes vallc'es des Vol- ta. Bordeaux: Universit' de Bordeaux Il, Intitut de la Kompienta. Etude socio-seconomique de la de Geographie Tropicaleet d'Etudes Regionales. de daiKompieng.Euadeougou:nSAED depl zone d'inondation. Ouagadougou: SAED. Sep- Paris, F. 1980. Etude geographique d'une zone tembre. d'end6mie onchocerquienne: Bourgouriba et de .1988. Environnement socio-eonomique de la Volta Noire. Dynamique des cultures et d'hab- le foret classee de Toumousseni. Rapport de Syn- itat depuis 1974. Ouagadougou: OCP and ORS- these. Ouagadougou: SAED.janvier. TOM. 13 csdela Sanders, J. 1990. Resource Depletion and Policy 1983. LeOtcpan VoltaNoire.desla Dynamiquedes Reform in the Sahel. (Mimeo.) West Lafayette, IN: cultures et de l'habitat depuis 1974. Ouaga- Department of Agricultural Economics, Purdue dougou: OCP et ORSTOM. University. May 29. _.PNGTV (Programme National de Gestion Sander, J. H., J. Nagy, and S. Ramaswamy. 1990. desTrroirVillgeois.1989aappor"Developing New Agricultural Technology for des Terroirs Villagreols).1989a. Rapport de svn the Sahelian Countries: The Burkina Paso Case," these et d'analyse des experiences pilotes de ges- .eonomicD opnt and Cural Cane,3 tion des terroirs villageois. Ouagadougou: Min- Economic Development and Cultural Change 39 istere du Plan et de la Cooperation, Secretariat (1):1-22. General, PNGTV.Mai. Savadogo, K. 1989a. Analysis of Off-Farm In- -.1989b. Le Programme National de Gestion come. Land Settlement Review Draft Site Report. des Terroirs Villageois: Orientations, apercu du ' nghamton, NY IDA. September.

135 . 1989b. Land TenureSystems. Land Settle- Skinner, E. P. 1957. "An Analysis of the Political ment Review Draft Site Report. Binghamton, NY: System of the Mossi." Transactionsof the New York IDA. September. Academyof Science.Pp. 740-750.June.

-. 1989c. Livestock in the Farming System. .1960. "The Mossi Pogsioure." Man 60:20-23. LandSettlement Review Draft Site Report. Bing- 1962. "Trade and Markets among the Mossi hamnton,NY: IDA. September. People." In Marketsin Africa.P. Bohannon and G. - .1989d. Factors Explaining Household Food Dalton, eds. Pp. 237-278. Evanston, IL: North- Production Systems. Land Settlement Review western University Press. Draft SiteReport. Binghamton, NY:IDA. October. _ . 1964a. "The Effect of Co-residence of Sis- Savadogo, K., J. Sanders, and D. McMillan. 1989. ter's Son on African Corporate Patrilineal Farmand Female Incomes and Productivities in Descent Groups." Cahiers d'Etudes Africaines the River Blindness Settlement Programs of 4(16):467-478.

Burkina Faso. Land Settlement Review Draft Site __. 1964. TheMossi of UpperVolta: The Political Report. Binghamton, NY: IDA. Developmentof a SudanesePeople. Stanford, CA: Savonnet-Guyot,C. 1986.Etat et soci6t6auBurkina. Stanford University Press. Essai sur le politique africain. Paris: Karthala. _. 1970 "Processes of Political Incorporation Sawadogo, S. 1986. Du pays Mossi aux zones in Mossi Society." In FromTribe to Nation in Africa. d'amenagementdes valleesdu Burkina:migration et R. Cohen and J. Middleton, eds. Scranton, PA: mutationsociale au BurkinaFaso. These de doctorat Chandler Publishing Co. de 3me cycle. SOFITEX(Societe Burkinabe des Fibres Textiles).

1988a.Depart des migrants vers la Kom0pi- 1989. Rapport annuel 1987-1988.Culture motor- enga. Zorgho: Unite de Planification (UP) N 1, isee dans l'ouest Burkina. Bobo-Dioulasso:SOFI- Cellule Organisation du Monde Rural et Forma- TEX. tion. Mars. Songre, A.1973. "Mossi Emigration from Upper

- . 1988b. Etude de l'occupation de l'espace Volta: The Facts and Implications." In Employ- dansl'UDdela Bombore.Zorgho: UP1.D&embre. ment in Africa: Some Critical Issues. Pp. 199-255. -.1989. Rapport Pr6lirninaire-136parts de Geneva: International Labour Office. Mogtedo et Mogtedo-Bombore.Binghamton, NY: Terrible, M. 1979. Occupation du sol en Haute Institute for Development Anthropology, Land Volta: Son 6volution entre 1952-56et 1975. Oua- Settlement Review Draft Site Report. gadougou: Centre Regional de Teledetection de Scudder, T. 1981.The DevelopmentPotential of New Ouagadoguou. Lands Settlement in the Tropicsand Subtropics:A Van Raay, G. T., and J. G. M. Hilhorst. 1981. Land GlobalState-of-the-Art Evaluation with SpecificEm- Settlement and Regional Development in the phasison PolicyImplications. Binghamton, NY: In- Tropics: Results, Prospects and Options. Draft stitute for Development Anthropology. Discussion Paper. The Hague: Institute for Social -. 1985. The Experienceof the WorldBank with Studies Advisory Board. February. Government-SponsoredLand Settlement. Report No. Vayssie, J. 1982. L'Evaluation du bloc de Mogte- 5625, Operations Evaluation Department. Wash- do. Ouagadougou: AVV.Mai. ington, DC: The World Bank. Vieira, A. P., J. H. Sanders, and M. K. Habash. Sherman,J. R., K. H. Shapiro, and E. Gilbert. 1987. 1988. "Regional Decentralization of New Agri- The Dynamics of Grain Marketing in Burkina Faso. cultural Technologies: Soybean Diffusion in the Vol.1. An Economic Analysis of Grain Marketing. United States and Brazil." Quarterly Journal of Ann Arbor, MI: Center for Research on Economic InternationalAgriculture 27(2).June. Development. Weitz,R., D. Pelley,and L. Applebaum. 1978."Em- ployment and Income Generation in New Settle-

136 ment Projects." Geneva: International Labour Yanogo, A-F. 1988. Contribution de l'AVV-Ex- Office10/WP 3. WorldEmployment Paper 10, Work- perience Mogtedo-Bobore. Zorgho: AVV-UP2 ing Paper 3. (prepared for CILSS Conference, 7 janvier 1988).

137 Annex 1. Methodology of the Land Settlement Review in Burkina Faso

The Burkina case study research was designed Table A.1 Site and Field Reports Completed and implemented by three senior team mem- as Partof the Land Settlement Review bers. Kimse Savadogo, Chair, Department of in BurkinaFaso Economics, University of Ouagadougou, was responsible for the design, testing, and analysis Savadogo, Kimseyinga of the economic questionnaires at all four study 1989a Analysis of Off-Farm Income. September. sites (See Savadogo 1989a,b,c,d), as well as for 1989b Land Tenure Systems. September. writing draft sections on farming systems and 1989c Livestockin the Farming System. September. diversification for the final report. Jean Baptiste 1989d Factors Explaining Household Food Production Nana, sociologist, was responsible for coordi- Systems.October. nating research at the Niangoloko, Solenzo, and AVV Rapadama Study sites (see Nana Savadogo, Kimseyinga, John Sanders, and Della McMillan 1989a,b,c,d,e,f), as well as for contributing draft 1989 Farm and Female Incomesand Productivities in the sections for the final report on land tenure, re- River Blindness Settlement Programs of Burkina cent political history, immigration in connec- Faso. tion with the classified forests, and immigration types. Della McMillan, anthropol- Nana, Jean Baptiste ogist, Institute for Development Anthropology, 1989a Rapport sur le Site de la Zone de Sous-secteur de was responsible for the overall coordination of Niangoloko et de la Forit Class&ede Toumousseni. research and analysis as well as for research in 1989b Rapport sur le Site. Programme PGTVRapadama. the AVV planned settlements at Mogtedo, 1989c Rapport sur le Site de Sous-secteur de Solenzo. Mogtedo-Bombore, and Linoghin, and on 1989d Enquete Ressources Naturelles de la Zone de spontaneous new land settlement at Kompien- Niangoloko. ga. McMillan was also responsible for writing 1989e Enquete Ressources Naturelles dans le Sous-secteur the first and second drafts of the national case enzo. study based on the site reports, sector analyses, case studies that were contributed by other re- and other research conducted by the team (Mc- searchers (Table A.1). Millan 1989). In addition to the main case study The Burkina case study is based on a mixture research, we benefitted from several special of quantitative and qualitative research at the

138 four study sites and a series of complementary Core Methodology at the Study Sites special studies on market systems, gold mining, the agropastoralist zone at Gadeghin, Socio-economic Surveys settler turnover in the AVV sponsored settlements, and local and national institutions. We Information on the socio-economic effects of set- made a deliberate decision to focus most of the tlement-related development at each site was research at each study site on a small group of drawn from four interrelated surveys. sample farmers whom we attempted to visit FARMING SYSTEMSSURVEY. A total of 141 house- consistently over a three-to-eight month period holds were included in a farming systems survey (the length of fieldwork varied at the different at three of the study sites (Table A.2). The survey study sites). We believed that the trust instilled included a stratified random sample of 27 animal over time improved the quality of the answers tractionand manual farmersin two study villag- we received. The sample sizes used in different es at Solenzo and all 9 of the tractor-owning surveys are described in Table A.2.

Table A.2 Sample Size of the Economic Surveys Conducted in Burkina Faso, 1988-1989: Number of Households

Farmingsystems survey Revenuesand expenditures Site Daboura Dar/Kie All F.S.! Neuw' Pastoralists Solenzo Manual 7 6 13 4 Animal 6 8 14 5 Tractor 4 5 9 3 ALL 17 19 36 12 AVV-UPI Restudy of the 1979 Statistical Service Survey AVV-Linoghin 20 IOC AVV-Mogtedo 20 5 AVV-Mogtedo-Bombore 20 4 4 ALL 60 19 4 Restudy of the Longitudinal Case Study at V3 Mogtedo 20 14 3 Assisted Settlement Program Rapadama 6 Kompienga (town) Former AVV Mossi Settlers 7 5 Mossi Settlers, non-AVV 6 2 Yana 6 2 Gourmantche 6 2 Fulani (pastoralists) 3 ALL 25 11 Special Studies of Pastoralists Kompienga 10 Mogtedo V3 10 Gadeghin 10 a. Subsample of the Sample Households included in the Farming Systems Survey. b. New Households, not included in the Farming Systems Survey but included for other reasons, including long-term familiarity with the research. c. At Linoghin, 10 households were included in the revenue study, but only 4 in the expenditure survey. The decision to augment the size of the revenue sample was made because of evidence of a high level of off-farm employment at the site.

139 households in the two villages for a subtotal of 36 July 1989 at Kompienga, between February and households; a random sample of 60 of the 144 July 1989 at the AVV,and between March and households that were included in the 1979 farm May 1989 at Solenzo. monitoring program of the AVV Statistical Ser- HOUSEHOLD EXPENDITURE AND REVENUE SUR- vice (Murphy and Sprey 1980) in the AVV VEY.A subsample of 56 farm families in the Farm- planned settlements at Mogtedo, Mogtedo-Bom- ing Systems Survey were selected for a more bore, and Linoghin (114of the original 144are still detailed analysis of household expenditures and living there; see Table A.3); 20 households that revenues. In contrast to the Farming Systems Sur- have been included in an intensive longitudinal vey (which interviewed only one male and one case study of 25 households in a single planned female household head in each family), each settlement at Mogtedo (McMillan 1983, 1986, member of the family was interviewed concern- 1987a, 1987b);and a stratified sample from the ing her or his crop, livestock, and off-farm pro- major ethnic groups in the town of Kompienga. duction. Each person was also interviewed about Interviews were conducted with the male house- expenditures for different categories of objects hold head concerning crop production tech- and services for the preceding month and for cer- niques, labor utilization, and inputs on the fields tain large expenditures (house repairs, bicycles) worked collectivelyby all members of the house- for the preceding year. The expenditure and rev- hold. Each family member was also interviewed enue survey included four additional households about crop production on his or her private fields. (from Mogtedo-Bombore) that were not included The male household head and one wife-usually in the main farming systems survey but with the wife considered to have the largest private whom McMillan had worked since 1978. We also crop area or most active involvement in off-farm included three additional households, with large employment-were interviewed concerning pro- herds, at Mogtedo V3 in order to offset the bias duction techniques, input use, and harvests for that resulted from the wealthiest households in their private fields; livestock ownership, expens- the longitudinal case study leaving the project. es, and sales; crop sales; and income from differ- Six farm families that had been newly incorpo- ent nonagricultural sources. We conducted rough rated into the assisted settlement program at Ra- estimates of field size (measuring length and padama were also interviewed. In addition, we width), and verified them with more precise mea- were able to interview three pastoralist families at surements of a subsample of fields at Mogtedo Kompienga. V3, Solenzo, and Kompienga. A smaller number As part of this survey, we conducted a survey of fields were measured for the 60 farmers in the of price changes since 1979 for the major products restudy of the 1979 Statistical Service Survey. that-according to our interviews-were most Each family was visited at least three to five times frequently purchased by the farm families. over several months between December 1988and

Table A.3 Follow-up of FarmersIncluded in the 1979 Statistical Service Survey

Changessince 1979 Totalno. of Replacedby familieswho AVV plannedsettlement Families Left family remainedin Bloc/UD in 1979 project Deceased Expelled members 1989a Linoghin 48 3 4 - 4 45 Mogtedo 72 24 3 - 1 46 Bombore 24 1 - - - 23 Bane 48 2 4 2 1 41 Kaibo-Sud 72 20 1 2 - 49 Kaibo-Nord 48 7 - - - 41 Total 312b 57 12 4 6 245 a. If the male household head was replaced by a member of his extended family, the family is counted as still being there. b. We could locate only 96 of the original 97 families in the original sarnple at Mogtedo. Source:Guira, DEPC,AVV, Septemnber 1989.

140 SURVEYSOF PASTORALISTPRODUCTION SYSTEMS. Natural ResourceIssues Thirty pastoralist families (defined as the unit re- siding together at a permanent camp under the Village migrant and host leaders, as well as a few leadership of one individual) were interviewed at of our key informants from the other surveys, three sites. These include pastoralists living in the were chosen for open-ended interviews on natu- area immediately outside of the town of Kompi- ral resource issues, following a questionnaire for- enga, pastoralists working as herders in the area mat designed by John Buursink. Informants were around the AVVplanned settlements at Mogtedo, asked questions about land-use patterns before and ten pastoralist families in the agropastoral large-scale immigration, current land-use issues, zone at Gadeghin. Each individual in the family and their projections for the future. Results of was interviewed about the number of animals he these interviews were synthesized into a report or she owned, animal losses due to theft or death on natural resource issues at each of the study during the preceding five years, sales of animals sites. and animal products during the preceding years, cash expenses for livestock production, and other ImmigrationPatterns income-earning sources. As at the other sites, the interviews took place over an extended period in EXTENSIVEIMMIGRATION SURVEY.Three of the order for us to adjust to fluctuations in residential study sites (Kompienga Town, Niangoloko, and patterns and for them to get better acquainted Solenzo) lacked basic information on immigra- with us. The major exception was at Gadeghin, tion trends. At Niangoloko and Solenzo we con- where the interviews were conducted over a two- ducted interviews with village migrant and host week period but where several of the pastoralists leaders in order to discover the number of mi- had known the individual conducting the inter- grant households, their approximate size, their view for 10 to 15 years. At Kompienga and Gade- prefecture of origin, and the period when they ghin we attempted to select those interviewed immigrated. These interviews were often verified from groups that the settlers themselves had by personal interviews with the identified male identified as wealthy, average, and poor. Three household heads. The town census at Kompienga pastoralist families-one from each of these gathered similar data on each of the identified groups-were included in the revenue and ex- household heads (male and female) living in the penditure survey at Kompienga. town in August 1989. SURVEYOF LOCAL MARKETS AND SERVICES. A IMMIGRATIONHISTORIES AND SOCIAL INTEGRA- fourth set of surveys focused on the impact of TION.We conducted a series of open-ended inter- new lands settlement on the development of local views with household heads in the main study and regional markets and services. These inter- sample (Table A.2) about their past immigration views were conducted at Kompienga town; in histories and their relationships with other markets frequented by settlers living in the AVV groups in the new immigration areas. To ensure planned settlements at Mogtedo, Linoghin, and comparability, the interviews were reported on a Mogtedo-Bombore; for Niangoloko town; and for standard form. all markets in the Solenzo subsector. Organization of Researchat Individual Special Issues Related to Women Study Sites

To complement the information on women, we The AVV-UP1 interviewed women about their perceptions of their current situation, as compared to their situ- ASSISTED SETTLEMENT PROJECT AT RAPADAMA. ation before immigration to the project. These in- The case study of the UP1 assisted settlement terviews were conducted at each site with the program at Rapadama examines the early eco- woman who was included in the farming systems nomic and social results of this experimental as- survey. Like the immigration survey, they were sisted settlement project at the end of its first year. generally open-ended interviews, but reported Research focused on discussions with the village on a standard form to improve the likelihood of land management committees, extension agents, getting comparable data. and participating settlers. The initial 1987 census of spontaneous settlers who were surveyed for

141 inclusion in the program provided the basis for a settlers who migrated from the same home vil- follow-up study to determine how many of these lage and were in the same project villages in 1979, settlers actually remained at the new site and with related households that did not migrate but what was their level of participation in extension remained in the settlers' home village. This re- programs. Six settlers considered to be represen- search was conducted over a two year period tative of new settlers participating in the program from 1978 to 1980.The results of this comparative were included in the study of household expens- case study were compared with the results ob- es and revenues. tained from the AVV farm monitoring survey PLANNEDSETTLEMENTS AT MOGTEDO,Li- (McMillan 1983). Short restudies of the settlers NOGHIN,AND MOGTEDO-BOMBORE. Research in were conducted during 1983 and 1986(McMillan the AVV planned settlements focused on a re- 1986, 1987a, 1987b,1989a, 1989b). study of sample farmers that were included in an Many of the hypothesized relationships be- AVV-wide farm monitoring survey during the tween length of settlement and settler patterns of 1978 and 1979 agricultural seasons. The baseline socioeconomic development were formulated on survey covered 313 households that were ran- the basis of observations made during restudies domly chosen in all the major AVVvillage groups of the longitudinal case study. in existence at the time (97 of the 313 households were in the contiguous blocs of Mogtedo and Solenzo Mogtedo-Bombore; 48 of the study households were from Linoghin). At the time of the study, the Our research in the Solenzo region focused on sample settlers had been living at the project for two levels. At the regional level, we were con- periods of one to five years. The farm monitoring cerned with gathering accurate information on survey gathered basic information on crop tech- the rate and patterns of immigration to different nology, production, and income for the four 1.5- villages, since the earlier assessment of settle- hectare bush fields (or, in the case of a very large ment-related development in the area by AVV family, four 3.0-hectare bush fields) and 1- or 2- (1985c) had lacked certain vital information on hectare house sites. The results of the farm moni- the imnmigrationprocess. At the village level, we toring survey for 1978 and 1979, are analyzed for gathered information that would be comparable farmers living at the project for different lengths with our economic research at the other sites and of time as well as by size of family labor force with the ongoing farm monitoring efforts by the (Murphy and Sprey 1980). CRPA du Mouhoun. Twenty-seven of the 96 sample families that An extensive migration survey was conducted were included in the 1979 Statistical Service Sur- in 19 villages in the Solenzo subsector. Five of vey of the blocs of Mogtedo and Bombore, and 3 these villages were chosen for more intensive in- of the 48 sample families at Linoghin, had left the terviews on inmigration histories and social inte- project and were not replaced by family members gration with village host and migrant leaders by (Table A.3). A random sample of 60 of the 114re- the senior team member working there. maining households was chosen for inclusion in Thirty-six households representing the three the farming systems survey (Table A.2). The goal main crop technology groups-manual, animal of the farming systems survey in the AVV traction, mini-tractors-were included in the planned settlements was to compare the settlers' farming systems survey. In the manual and ani- patterns of social and economic organization to- mal traction groups, 27 farmers were chosen ran- day (10 to 15 years after coming to the site) with domly on the basis of census data provided by what had been observed in 1979 after they had the CRPA. The 9 mini-tractor households repre- been living there for periods of 1 to 5 years. The sent the entire population of tractor farmers liv- 1979 farm monitoring survey and the extensive ing in the two villages. These 36 farm families records kept by the AVV extension agents were were divided between two villages-Kie/Dar-es- an unusual and valuable resource that enabled us Salaam and Daboura. Daboura is one of three test to assess specific patterns of longitudinal change. sites for the Village Land Management Pro- A second source of baseline information in- gramme (PNGT) in the CRPA du Mouhoun. Dar- cluded a more-intensive longitudinal case study es-Salaam, originally a migrant cultivation ham- of settlers from one of the main recruitment let of Kie, has recently requested and received zones. That case study compares a single group of

142 recognition as an independent village with its income families. Although limited to the personal own extension groupement. network of the research assistant, the smaller A survey was conducted of the Solenzo and sample provided a much higher quality of infor- Daboura market. In addition, we benefitted from mation on farrn and off-farm income than could a survey of village and cultivation hamlet mar- have been acquired by outsider researchers. In kets in the Solenzo subsector conducted by the addition to the 25 farm families, efforts were CelluleSuivi-Evaluation of the CRPA du Mouhoun made to follow, in a less intensive fashion, the in March-April 1990. crop and livestock activities of 10 FulBe pastoralist families. Our increasing familiarity with the Kompienga different immigrant groups allowed us to develop complementary studies on the development Kompienga has been the subject of several excel- of off-farm employment and commercial activi- lent regional surveys on migration and resource ties in the town. issues that were conducted in connection with planning for the dam (SAED 1980;Dr.-ing. Walter Niangoloko International 1987). Our analysis of regional trends is drawn largely from the documents pro- Our methodology at Niangoloko focused on: (1) duced by Agrotechnik, the consulting firm that a migration survey of the 15 villages in the conducted the final regional development plan Niangoloko subsector and the 4 villages around (Agrotechnik 1989). Information on settlement the classified forest; and (2) interviews with in- trends in the immediate river basin is drawn from digenous and migrant leaders in both zones AVVplanning documents. Our separate research about past and present patterns of immigration, has focused on the study of the new town, which income growth, and natural resource use. grew from three large compounds with fewer than 100 persons in 1985 to an estimated 5,000 ComplementaryStudies persons at the peak of dam construction, and numbered 3,239 persons in August 1989, not Guira Frederic, Controleur, DEPC, ONAT (ex- counting civil servants. AVV),prepared a special report on the history of Because the departure of the dam workers is so groupementactivities in the AVV planned settle- recent, no information on the size, ethnic compo- ments at Linoghin, Mogtedo, and Mogtedo-Bom- sition, or activities of the town was available bore, and a second special study on the Linoghin when we started the research in early 1989.More- market. Sommaila Sawadogo, sociologist, AVV- over, the settlers were highly suspicious of out- UP1, prepared a study of settler turnover in the side researchers. We decided, therefore, to focus AVVplanned settlements at Mogtedo and Mogt- the economic survey on a sample of 25 agricultur- edo-Bombore. Moustapha Ouedraogo prepared alists and 10 pastoralists chosen by our research two special studies on gold mining and the Gade- assistant. The research assistant had worked with ghin agropastoralist zone. McMillan in AVV Mogtedo in 1983 and 1986. He The Cellule Suivi-Evaluation, CRPA du Mou- and his family were among the 55 families (452 houn (Chef, Traore Adama) reacted to our initial persons, or 14 percent of the total population ex- discussions of results by implementing a separate cluding civil servants) living at Kompienga who study to examine the history and size of markets left the AVV planned settlements between 1987 in the subsector and goods included in them. Fa- and 1989.The research assistant worked with im- tou Kabore, Animatrice,Bureau Ministere de l'Envi- migrant and indigenous leaders with whom he ronnementet du Tourisme,Mogtedo, and Mogtedo was acquainted to choose "representative" fami- CR DLe'gueChef, Joseph Guigma, worked with us lies from each of the major ethnic groups repre- to prepare a brief history of the town and market sented in the village. Special efforts were made to of Mogtedo, in addition to following market ac- include large, small, wealthy, and average tivity during several seasons.

143 Annex 2. Political History of Burkina Faso

Under colonial rule, Upper Volta was defined as bly installed two new successive national gov- a "military territory" from 1896to 1904, and then ernments referred to as the Gouvernement as the colony Haut-Senegal Niger from 1904 to Renouveau National (GRN), the GRN 1 from 8 1919, before becoming the Colony of Upper Volta February 1974 to 9 February 1976, and the GRN from 1919 to 1932. In 1932 the colony was dis- 2 from 1976 to 1978. The latter encountered membered and divided between the neighboring strong opposition from the labor movements in colonies of C6te d'Ivoire and Sudan-Niger; it was October 1977, which resulted in the dismissal of reunified in 1947. the Ministers of Finance, Plan, Rural Develop- Since attaining independence in 1960, the ment, and Environment. Many political activities country has known several successive political that had been suspended were reinstated in 1977. regimes (Savonnet-Guyot 1986). A constitution was adopted in November 1977 1960-1966.The first national government, re- and elections planned. Legislative elections were ferred to as the First Republic, was the Presiden- held in May 1978 and presidential elections in tial Regime of Maurice Yameogo. He governed June, giving birth to what is referred to as the with the Rassemblement Democratique Africain Third Republic. (RDA), which became the official State Party. 1978-1980. During the Third Republic there 1966-1970.He was succeeded by the military was a return to the multiparty activities of the regime of General Sangoule Lamizana, who ruled three parties that were constitutionally recog- with the Conseil Superieur des Forces Arm&es nized. The Third Republic was brought down by (CSFA),which came to power during street man- powerful opposition from the labor movement. ifestations of 3 January 1966. 1980-1982. In 1980 the military took power 1970-1974.The military government was suc- again under Colonel Saye Zerbo, who presided ceeded by the Second Republic with un regime over the Comite'Militaire de Redressementpour le parlernentairerationalise', based on a constitution ProgresNational (CMRPN). adopted June 1970. General Lamizana remained 7 November 1982-4August 1983. A 7 Novem- in power during this time, but the government ber coup d'etat directed by the Conseilde Salut du was directed by the Prime Minister, whose activ- Peuple (CSP) suppressed the earlier activities of ities were controlled by a National Assembly. A the CMRPN. The CSP was presided over by Com- conflict between the Prime Minister and the Na- mandant Jean Baptiste Ouedraogo with Captain tional Assembly led to social unrest, which re- Thomas Sankara as First Minister. Sankara was turned the military to power on 8 February 1974. dismissed from the government 17 May 1983. 1974-1978. General Lamizana suspended the 4 August 1983 - 15 October 1987. An armed national constitution, and the National Assem- insurrection proclaiming the creation of the

144 Rivolution Democratiqueet Populaireby the Conseil 15 October 1987 - 2 June 1991. In October 1987, Nationalde la Revolution(CNR) presided over by the Popular Front, directed by Captain Blaise Captain Thomas Sankara represents a decisive Compaore deposed the CNR but with the idea of break with the country's earlier regimes. The ob- "preserving the noble ideals of the Democratic and jective of the takeover was l'edificationd'une so- Popular Revolution." Important activities includ- ci6t6 de d6mocratie populaire through the ed the restructuring of the various CRs that were organization of people into Comit6sde Defensede considered to have often abused their power into a la R6volution.The committees were organized to new structure of local,regional, provincial, and so- represent all the different geographic units of the ciopoliticalCRs or Cornite'sRe'volutionnaires. country (village, neighborhood, commune, de- 2 June 1991. A constitutional referendum was partment, province) and such professional and held. The newly adopted constitution reinstates social sectors as government and international of- civilian rule and elections were expected to be fices, military and paramilitary units, and educa- held before the end of the year. tional facilities. On 4 August 1984 the CNR December 1991. Only one candidate, Blaise renamed the Republique de Haute Volta as Burki- Campaor6, ran in the the presidential elections. na Faso, and made parallel changes in the nation- He was elected and sworn in as President of the al flag, emblem, and anthem. 4th Republic on 9 December 1991.

145 Annex 3. Supplementary Tables on Consumer Prices and Price Indexing Table B-1 IMF Consumer Price Index, Burkina Faso

Year 1979= 100 1988= 100 1979 100.0 61.8 1980 112.2 69.3 1981 120.1 74.6 1982 135.1 83.4 1983 146.5 90.5 1984 153.5 94.8 1985 164.2 101.4 1986 159.9 98.8 1987 155.3 95.9 1988 161.9 100.0 Source: IMF 1989; World Bank 1989:63

Table B-2 Price Changes in Important ProductsPurchased by Farmers,1979-1989, and PercentChange (1979=100)

Product 1979 1988-89 Percent change Plowv Donkey cart 40,000 95,000 137 Bicycle 29,000 50,000 72 Yamaha Dame motorcycle 400,000 na Mobylette CT 45,000 165,000 267 Cement (1 sack) 1,100 2,400 118 One roofing tin 1,050 2,300 119 I pack of sugar 125 355 184 I small soap 45 95 111 I medium soap 75 150 100 I large soap 150 280 87 I small pail 600 1,800 200 I large pail 1,250 2,600 108 1 box matches 10 15 50 1 liter gas (reg.) 80 275 244 1 liter kerosene 40 160 300 I large metal plate 600 1,500 150 I bicycle tire 750 1,850 147 1 "jente" velo 750 1,850 147 Cube maggi (2) 5 10 100 I soda 75 150 100 I beer (non-local) - 225 na 1 plate cooked rice 15 50 233 I calebasse dolo (local beer) 05 50 900 1 liter oil peanut 200 400 100 cotton 175 350 100 (5) Bonbons 5 25 400 Rice (I plate of white rice) 250 500 100 n.a. Not applicable. - Not available. Source:Guira, F. DEPC,AVV, March 1990.

147 Table B-3 Price Ranges of the Major Crops at the Study Sites, 1988-1989' White Red Earth Cow sorghum sorghum Millet Maize Cotton Peanut pea pea Sesame Rice Tubers ScenarioI Low prices AVV 44 40 45 33 95 60 60 68 60 80 50 Scenario1 Medium prices AVV 47 45 45 33 95 60 60 115 60 80 50 ScenarioIII High prices 50 50 45 33 95 60 60 167 60 80 50 Note:Gumbo and roselle were given a price of 90 FCFA/kg. Source:Savadogo 1989a. a. There are substantial variations of prices between the high levels in the "hungry period" before harvest and the customary post harvest price collapse. Price ranges are more relevant than either the pre or post-harvest prices.

Table B-4 Retrospective Prices for Major Products, the Linoghin Market 1988-1989a (August, December, March) Whitesorghum Red sorghum Millet Year Aug. Dec. Mar. Aug. Dec. Mar. Aug. Dec. Mar. 1988/89 83 26 41 90 20 40 120 40 60 1989/90 67 25 50 60 26 40 100 40 70

Maize Peanuts Beans 1988/89 60 20 50 83 36 71 240 50 100 1989/90 80 30 60 95 36 71 240 60 130 a. Retrospectiveprices for the three time periods were collected from farmers at Linoghin. In view of large differences between the recorded prices and data recorded by the ORD for Mogtedo and Zorgho during 1984-1985,we are not reporting those results. In- formation was elicited for assiettes(plates). Estimated weights per plate were 3.0 kg for white sorghum, 2.5 kg. for red sorghum, 2.5 kilograms for millet, 2.5 kilograms for maize, 2.1. kilograms for peanuts, and 2.5 kilograms for beans.

148 Table B-5 RetrospectiveStudy of Prices for Livestock, 1979-1989 Guinea Year Chicken fowl Goat Sheep Donkey Pig 1979 250 300 2,250 2,500 13,000 4,000 1980 350 350 2,500 3,000 16,000 4,750 1981 400 450 2,750 3,500 20,000 5,000 1982 375 250 3,000 3,750 20,000 5,750 1983 400 300 3,250 3,700 22,000 6,000 1984 400 275 3,275 3,800 22,500 7,500 1985 375 300 3,350 4,000 23,000 8,000 1986 375 275 3,400 4,250 23,700 9,000 1987 350 300 3,500 4,500 23,700 10,000 1988 175 225 4,000 5,000 24,000 11,000 1989 500 400 5,000 6,500 25,000 15,000 1990 750 600 6,500 7,500 27,500 16,000 Methodology: Retrospective prices based on interviews with sample farmers at Linoghin. Recorded prices cross referenced with prices recorded in other surveys. Source:Guira, E 1989 and AVV-DEPC.

Table B-6 Cereals Food Balance (in 000 tons) Year Domesticproduction Imports 1980 1,028 66 1981 1,2249 59 1982 1,179 74 1983 1,099 77 1984 1,318 173 1985 1,791 190 1986 1,709 67 1987 1,759 60 Source:World Bank 1989:57.

149 Table B-7 Official Producer Prices and Prices Used to Calculate the Value of Production in the 1979 AVV Statistical Service Survey Oficial producer Crop Market surwyf price' Cotton 54 55 (first choice) 45 (second choice) White sorghum 62 40 Red sorghum 55 32 Millet - 40 Maize 57 40 Cowpeas (niebe) 79 45 Peanuts (in shell) 81 37 Rice 72 Millet 65 Cereals (house 50 compound) a. Murphyand Sprey1980:58; 22. b. OfficeNational des C&rIales,Arrette n° 001458/MCODIM/MDR,portant fixation des prix d'achat au producteurdes c6r&aIes localespour la campagne1983-84. Tableau 11. Evolution des Prixdes Productions.

150 Annex 4. Supplementary Tables on Livestock Production Table C-1 Number of Cattle and Sheep/Goats Lost to Disease and Theft, Five-Year Period AVV Solenzo Kompienga Death Theft Death Theft Death Theft Cattle Small 15 0 12 0 25 5 Large 42 27 109 4 9 1 All 57 27 121 4 34 6 Sheep/Goats Small 298 21 74 17 139 4 Large 428 290 178 45 81 4 All 726 311 252 62 220 8 Source:Savadogo 1989c.

Table C-2 Rates of Losses of Cattle and Sheep/Goats by Death and Theft, on a Yearly Basisa (by percent) AVV Solenzo Kompienga Death Theft All Death Theft All Death Theft All Cattle Small 3.0 0 3.0 1.7 0 1.7 7.4 1.5 8.9 Large 3.5 2.2 5.7 7.9 .3 8.2 1.6 .2 1.8 All 3.3 1.6 4.9 5.8 .02 5.8 3.8 .7 4.5 Sheep/Goats Small 13.8 1.0 14.8 15.1 3.5 18.6 24.8 .7 25.5 Large 13.2 8.9 22.1 15.6 3.9 19.5 14.3 .7 15.0 All 13.4 5.8 19.2 15.5 3.8 19.3 19.6 .7 20.3 a. Rates are defined as losses per year due to the relevant cause divided by stock at interview time. This gives a roughapproxima- tion of the probability of losing an animal during a particular year. Source:Savadogo 1989c.

152 Table C-3 Expected Annual Livestock Income for the Heads of Household, by Region Total PerALE (CFAper farm per year) (CFA/ALE) High scenario' Lowscenario" High scenarios Low scenario" Linoghin 80,898 57,599 21,897 15,590 Bombore 29,312 8,723 6,797 2,022 Mogtedo 66,438 35,898 14,764 7,977 Mogtedo V3 18,683 -5,107 5,415 -1,480 AVVC 57,164 32,958 7,934 13,762 Solenzo 143,541 104,335 24,779 18,011 Kompienga 48,257 35,239 8,522 6,223 a. High Scenario:includes lowest cost situation (nonpurchased feed not valued,low loss rates). b. LowScenario: includes highest cost situation (imputed cost for nonpurchasedfeed, high ratesof loss). c. Doesnot includeV3. Data seemunderestimated. Source:Savadogo 1989c.

153 Annex 5. Supplementary Tables on Farming Systems

Despite the constraints on livestock ownership, by labor units (Annex 5, TableD-5) Nevertheless, livestock remains the AVV settlers' most impor- the poorest group-the 20 to 25 percent of farmers tant form of investment. When settlers and exten- who had sold or lost their cattle-ranked substan- sion agents were asked to classify the settlers tially lower on net income as well as on net live- living in a village according to wealth, they invar- stock income (high scenario), and on cereal and iably grouped the families in different categories cotton production per unit labor. by patterns of livestock ownership (Annex 5, Ta- The average family size of the settlers in the cat- ble D-5). The point of reference was the basic egory "AVV Simple" and "Simple/Manual" AVV model of two animal-traction animals (Cat- tended to be smaller than those in the top three egory AVV Simple). Below that were farmers wealth categories. Small families, with limited la- who, for one reason or another had lost and re- bor, were forced to farm small areas and were less fused to repurchase their animal-traction animals able to engage in on-farm livestock activities or in (Category Simple). A third group was identified nonfarm employment than larger families. Other as farmers who had accumulated a few cattle (2 factors, besides size of the family labor force, that or 3) beyond the animals they used for plowing, influenced economic status included entrepre- as well as small livestock. These categories be- neurship, health, age, and losses or gains of key came the "Improved AVV Model." Beyond this workers through accidents of nature (death, dis- were farmers who had fairly extensive livestock ability), ambition (out-migration), and/or school- investments-from 5 to 14 large livestock and ing. As the structure and size of the labor force of small animals (Category Superior). These were individual households changed over the life cy- settlers who often also had one or more family cle, their production patterns and income levels members involved in a specialized trade or in also changed. general commerce. The wealthiest farmers were For example, one family in the longitudinal those with livestock herds of over 15 animals case study that would have been classified in the (Category Exceptional). Like those in the second wealth category "AVV Improved" in 1979 and category to the top, Category Superior, they were 1983, was classified in the category "Simple/ often involved in commercial activities. Manual" in 1988.In 1979 and 1983 the family con- When the 80 families in the farming system sur- sisted of two brothers, their wives and children, vey were grouped according to these wealth cat- and their mother. The older brother earned a high egories, the data did show that the settlers in the annual income from a secondary occupation as a top three categories had a higher average of total mobylette (moped) and bicycle mechanic. The crop income (Annex 5, Table D-5).The differences younger brother supervised the family's crop and were much less remarkable when broken down livestock production, while the older brother

154 traveled to market every third day. When the had recently lost a father. These farmers generally younger brother moved away to establish his own grow cotton and plan to move up to greater levels farm in a neighboring AW planned settlement, of investment in livestock and cash cropping. A the older brother was forced to abandon his trade. second subgroup within the two bottom catego- A combination of illness and lack of labor finally ries includes small families whose activities are forced the older brother to sell the family's oxen. hampered either by the age or health of the male Households can move up in wealth categories head of household and/or one of the wives. Many as well as down. This was especially noticeable of these farmers had lost their older children for one of the case-study farmers who had a large through emigration; a few never had any male family but only young sons in 1979. In 1979 he children who lived beyond infancy. would have been classified in the category "AW For the first subgroup-the young households Simple." By 1988/89, this same head of household undergoing temporary labor shortages-possi- had two sons in their early twenties who took an bilities exist for moving up to more diversified active role in the family's cooperative crop pro- patterns of investment. Opportunities for a duction. The father helped them establish a small change in fortune are more limited for families in local and regional trade in petty consumer items. the second subgroup. Some of the families in this The family's level of crop production, livestock group were "eating" their animals-selling off ownership, and off-farm income placed them in animals to purchase grain or pay medical bills. the AW Superior category in 1988/89. [It is inter- Although differences exist, they are not strik- esting to note that this family left the project in ing, except for the poorest group. While total fam- 1992to settle in the southwest. Since this was after ily crop production, net income, and income from the study ended, they were not counted in our livestock are much higher for the top three than out-migration figures.] for the bottom two groups, the differences are less Farmers in the poorest category tend to be those remarkable when broken down by labor units. In with the smallest family sizes. Within this group contrast, the poorest group-the 20 to 25 percent were two relatively distinct subgroups. The first of farmers who have sold or lost their cattle-is al- subgroup's households were usually headed by ways poorer, ranking substantially lower on al- men who had broken off from a brother or who most every measure of production or income.

155 A. Solenzo Table D-1 Division of Cereal Production between Privately and Cooperatively Worked Fields, Solenzo 1988/1989(in percent) Privatefields Villageand technology type Cooperative All Men's Women's Dar-es-Salaam/Kie Manual 100 0 0 0 Animal traction 79 21 10.5 10.5 Tractor 100 0 0 0 Daboura Manual 95 5 0 5 Animal traction 98 2 0 2 Tractor 100 0 0 0 Source:Savadogo 1989d.

Table D-2 Mean Input Costs for the Male Heads of Household with Different Levels of Farm Technology, Solenzo 1988/1989(per family in CFA) Hired Costsof equipment Fertilizer,urea, rock phosphate, labor mini-tractors manure,and insecticides (cashand kind) (cashand kind) Dar/Kie Daboura Manual 36,018 15,117 2,925(Kie) 1,000(Kie) Animal traction 84,167 36,251 Tractor 920,172 416,645 650,000a a. Expensesfor purchase,maintenance, and deprecationof tractorswere estimated at 650,000CFA per year. Source:Savadogo, Analysis 1989.

156 Table D-3 Average Livestock Holdings for Male Heads of Household at Different Levels of Technology, and Expected Annual Livestock Income per Adult Labor Equivalent (ALE), Solenzo 1988/1989

Average Family Labor Force (ALE) Dar/Kie Daboura All Solenzo Manual 2.8 (6) 2.6 (7) 2.7 (13) Animal traction 5.3 (8) 2.5 (6) 4.1 (14) Tractor 10.9 (5) 13.8 (4) 12.2 (9) Cattle ownership (numbers) Manual - - 8.3 Animal traction 6.2 Tractor - - 25.8 Expected annual livestock income for male household heads (high scenario) (CFA per ALE) Manual 938 3,737 Animal traction 11,182 -4,911 Tractor 71,907 10,921 Expected annual livestock income for male household heads (low scenario) CFA per ALE Manual 200 -14,456 Animal traction 7,561 -16,516 Tractor 66,922 3,168

- Not available. Source:Savadogo 1989c.

Table D-4 Distribution of Cattle Ownership, Solenzo 1988/1989

Number of households Number of cattle Dar/Kie Daboura All 0 11 6 17 1 1 0 1 2 0 1 1 3-4 0 2 2 5-10 2 1 3 11-20 3 2 5 21-40 1 3 4 41 and over 1 2 3 Source: Savadogo 1989c.

157 B. The AVV Planned Settlements Table D-5 Characteristics of Settler Households with Diffetent Levels of Investment in Livestock, the AVV Planned Settlements 19U/1909 Linoghin Mogtedo MogtedoV3Y Bombore Residents (number) 1. Exceptional 20 13 0 23 2. Superior 11 13 11 15 3. AVV improved 0 15 9 0 4. AVV simple 9 9 7 10 5. Simple/manual 8 5 9 9 6. Unclassed 0 0 12 0 Labor force (in ALE) 1. Exceptional 6.2 (3) 5.2 (3) 0 11.0 (1) 2. Superior 3.5 (2) 5.2 (3) 5.1 (2) 4.9 (3) 3. AVV improved 0 6.6 (2) 3.5 (6) 0 4. AVV simple 3.2 (7) 3.8 (10) 2.4 (7) 4.2 (11) 5. Simple/manual 3.5 (8) 1.8 (2) 2.9 (3) 2.9 (5) 6. Unclassed 0 0 6.1 (2) 0

Average Net Crop Income (Private and Cooperative) Average net crop income (CFA) 1. Exceptional 552,508 327,968 0 694,409 2. Superior 393,100 519,532 457,784 345,295 3. AVV improved 0 393,045 273,911 0 4. AVV simple 231,402 218,887 208,302 282,822 5. Simple/manual 223,739 76,399 171,641 141,407 6. Unclassed 0 0 238,251 0 Mean 292,673 283,513 250,429 277,419 Average net crop income (CFA per ALE) 1. Exceptional 89,595 62,470 0 63,128 2. Superior 112,314 100,554 89,323 70,229 3. AVV improved 0 59,327 78,260 0 4. AVV simple 71,991 57,225 87,051 68,001 5. Simple/manual 63,359 43,656 58,848 47,934 6. Unclassed 0 0 38,896 0 Mean 76,766 65,742 72,588 64,329 Net livestock income (high scenario) (CFA per ALE) 1. Exceptional 19,866 -6,914 0 0 2. Superior 21,515 30,825 -2,233 3,984 3. AVV improved 0 16,414 15,046 11,906 4. AVV simple 26,693 13,094 5,005 2,700 5. Simple/manual 18,238 646 4,873 16,492 6. Unclassed 0 0 -3,747 0 Numbers in parentheses are numbers of households: 20 in each settlement. Net crop incone is for total of all crops-cotton plus cereals. a. Two families at V3 defied classification because of their mixed family patterns and production strategies. One is a family of dis- placed pastoralists that still farms only a small amount; the second is a family where half the family emigrated to Kompienga. Pro- duction in both families is somewhat disorganized at the moment because of consklerable mobility of various family members. Source:Savadogo, Analysis 1989.

158 Table D-6 Mean Number of Cattle, Goats, and Sheep, and Mean Anticipated Income, the AVV Planned Settlements 1988/1989 Perfamily Per ALE Linoghin 80,898 21,897 Bombore 29,312 6,797 Mogtedo 66,438 14,764 Mogtedo V3 18,683 5,415 All AVV with V3 48,832 12,208 without V3 57,164 14,486 Source:Savadogo 1989c.

Table D-7 Livestock Ownership for the Study Households in Mogtedo, Mogtedo-Bombore, and Linoghin, the AVV Planned Settlements 1988/1989 Numberof livestock Cattle Sheep/Goats 0 24 20 1 0 0 2 28 13 (1-5 animals) 3-4 9 0 5-10 15 17 11-20 0 16 21-40 4 12 41+ 0 2 Total households 80 80 Source:Savadogo 1989c.

Table D-8 Subdivision of Cereal Production between Privately and Cooperatively Worked Fields, Kompienga 1988/1989 Ex-AVV Non-AVV Yana Gourmnntche All Number 7 6 6 6 25 Percent Cooperative 85 89 99 86 89.5 Private 15 11 1 14 10.5 Women's 6 4.5 0 0 2.8 Men's 9 6.5 1 14 7.7 Source:Savadogo, Analysis 1989.

159 Table D-9 Distribution of Livestock for Male Heads of Household at Kompienga, 1988/1989 Number of households Numberof animals Cattle Sheep/Goats 0 13 9 1 0 0 2 2 6 (1-5) 3-4 0 0 5-10 5 1 11-20 3 7 21-40 0 2 41+ 2 0 Total households 25 25 Source:Savadogo 1989c.

160 IBRD23841I

''t _4_/ k ."'MALI INIE '' v NIGER I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~A 15' 3 i- io X- *-

BURKINAFASO .- \ ~~~~~~~~~ j ~ ~ ~ ~~~~~~~~~~~~~~~~OUDALAN".-BURKINA , ' - POPULATIONDENSITY / > ,/It FAS >qGEI DEN51TEDE POPULATION r- ooOrm >AGUINE'tL~;8 g 1975 .. so U ./ .) , sS CT o0I

Dori@ 9 14 - - tr~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~'Z 4. / wJ @Dj;co <, ~ \t A ' * | S EN . X > OLF.4 N Gui o fDivir

+ 2#Q.mjsK'nK'TmNMTENGA'%>-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~PtV9 D.a 4Rz/GNGA'is 1-4 ' t rraD W-;f iP s )*,-;PA\SORErf-dl\.<_y~~~~~~~~~~~~~~~~~~~~;_rstv}>(yI I\ A N \ \ \ & G\ lyfer@yv v _ > ' 5 ' ' '' ' \'' \\\\\\>~~~~~~~~~~~~~~~~~~~~~~~~~~~\\\6 :;S."-';"---"' > ~~~~A TENGAi-

I v r' 1i s .; , ~~~~~~~~~~~~~~~~~~~~~Z

| ., ., ., ..... ,,, . , wC90URII~B AA H,tM _ [a)XX)_,

~~~~~~~~~~~~~~~~~Toug4[,-, .j . A < 0-10RVRLUE - 10- k? 4 , h s\, \ 8 ;, t zS0-c50SPROVINCECAPITALS/CHESUEU%DESPROViNCESN G N A 9. ;,@ d, NTOACAtA/AIANrOA 13XJ 13'< __POINEBUDAIS/LIIEDoPOIC t~~~~~~~~~''

2~~~~~~ f~~~~~~~tJRITANIA~~~~~~~~~~~~~~~~~IBD23842

.0~~~~~~~~~* J,r~~~~~~~~~~~OL I 2 NIGER BURKINA FASO MAJOR RIVER BASINS AND .- OUDALAN / ,- ~---RURKINA\>-' AREAS COVERED BY AVV PROGRAMS GoonAS~,o >Y_,NIGERIA PRINCIPAUXBASSINS VERSANTS ET o GN6,.zi ZONES D'INTERVENTIONSDE L'AVV .-. c'SOM-

/ ~~~~~~~~~rLANric ~ ~~~~~~~~~~~~~~~~~A3ojf of Goni- I ~~~~~~~~~~~~~~~~~~~~~~~~~OC8AN YATENGA '7

K..~ ~ ~ ~ ~ -1

-13' sou _4KDIG " z "~~~~~~~~~~~~~~I&

A P A 1~~~~~~ Wqd ~4~ OU\~BRITENGA.OUIENAj

.oedg >~ MOUHOUNugou ~ ~ ~ ~ ~ ~ ~ ~ ~ / 0 VALE rG K NAKMOURGOY \ ., KMIEG KENEOOUGOU I SS I L~~~~~~~~~~~~~ ~~~~IGBCUKADQUOG (EX-BLACKVOL .N,gh. GOURM HOVET . 4 .- <:.::.~z od N'o... iao. N 2 Bobo-DioIo,.e VALEY ,rmOUHOURN A KK:. .BE N

P0 .- '~~~~~'. S1SS1UVALLEY NA~~~~~~~~~~~~~~~~~:.ZION KTOMP G .~~~~~~~~~ . V~~~~~~~~~~~~ABALLY Y ~LERABA ~VALLEY ~~ ~ ~ ~ ~ ~ MOUHOL'N VALLEY CLIMATICZONES:X-.HIT VOTA) VALE

H 0 U E2 PT E-LC OTA OE LMTOJS

PONI ® NATIONALCAPITAL / CAPITALTNATIONA

LERABAVALLEY '6 MO~~~~~VLLHEYVA1 YPOIC TDE B H W-- RVNEBUNAIS/UIT

- BOUNDARIES/ FRONTIERESCLIATIC 2* ONES -. INTERNATIONAL comOEP 0 N (EX-BLACKVOLTA)~~~KI:ZONES DINATERVENTOS:O 'V NENTOAE I SAHELIAN/ SAHEUENNE~~~~~~~~~~~~~~~~~~~~~~~~~UN IP

a~~~~~~~~~~~~ 2 -

BURKINAFASO - ER AGRICULTURALAND TRANSHUMANT t - OJAA / 'BURIINKI r_ o.asajauaatFASOV MIGRATIONPATTERNS ( ,' 72 oo-om ' NIGER ->. - FLUXMIGRATOIRE ET TRANSHUMANCE

/ > ' - ' - 06<> ' S * / ~~~~~~~~~~~~~~~~-COTED!VC4fE 'GHANA I i

t 140 3*3 ., 14q- .r t l Gu,a;a

r- : : : : :: SENO3Tb,. . a n p,epa'ed by TheWad Banksstaff 27 f-~~~~~~~~~~aad-r andis far ih 'Ko ousMstNGA 27. it-aa -s at ThaWad Bank Grap. Thea.a-iaatia..s ... d jr. GP ~~~~D ad.. th. banadari., sh-w 22 ~ ~ ~~~~~~~~~~-this ap de eat iply,-anpha

lBogan th~~~~~~~~~~~aya eisa anteea

MALI OB/EI uhbadr s ,62. Ai~~~~~are

. s/ / S , !/ I- . TAPOA (-8

tEdouC R-o . LIMITESDE PROVINCES. (wPEOPLE / Kombissirl / NORQ oup IFd GoraOapaga

1<~~~~~~~~~~~~~~~~~~~~~~1 - Pa B21BAZEGA/,J .\- NG/ 'KK Mna-

4-~~~~~~~~' ..vfcAarlr FOrI / /# CPIAENA Bogand - .=. sttu a?aaNE aayRTIN@

/~ BOULOU<7

Jt'/ / \ ~~~sEo'bo,I uaitsb' 'OUGOLRB I Q N\ -NHUI

(/,nrodara< ~~~~~~~~~~~~~~~~~~~~~GHANATOGO z~~ ~ ~ ~ ~ ~ /'< <'s \'P - udUgiM .> /NENTOA f 80NARE a m COTED IVOIRE40

t ~~~FLUXMIGRATOIRES

/ / ~~~~' . - N>. ~~~~~~~~~~~HEAVYIMMIGRATIONAREAS NTOA AIA

WEAK IMMIGRATION AREAS PROVINCECAPITALS IOEESO 210 40 610 80 -1 oc, ZONE A FAISLEIMMIGPATION ~~~~~~~~~~CAPiTALESPRO ViNCIALES KIOMTES Ii 0

z HOMME ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~INTERNATIONALBOUNDARIES COTED'IVOIRE 40 LIVESTOCK FRNIRSI.ENTOAE 20

1 T5-MAURITANIA IBRD23844

MALI NIGER /~~~~~~~~ BURKINAFASO _-S CLASSIFICATIONOF LANDACCORDING TO ,- / ,'BURKINA" AGRICULTURALPOTENTIAL, AND AREAS *' - < PlS&Y'NIGRRIA DESIGNATEDAS NATIONAL FORESTSAND PARKS Gorom GoronEs O ' CLASSIFICATIONDES SOLS SELON LEUR VALEUR | AGRONOMIQUE,ZONES DE FORETSCLASSEES ET )IGIRE )*/GHIiANAi PARCSNATIONAUX Dp6Do

rlA <~~~~~~~~~~~4 -14- \ g r'S ls ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~rlC I ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ATLANrICGolf oP G-r.o -120 OCEANO …~~~~~~~~~~~~~~~~~~~

120 -~fRRS ' ~1

* / * ) aS d Iz A- _ _ g4 / 56 t \ {~~~~~~~~<59Ou1 6 -$\ 13' 3 ~~~~~~~~~~~~~~T~KoPOe 5 N 6 oo~ 13'

6 6

6 _

,. t<_y_/ t /4/~i_ 1501 RIV4RS/FLADOUGOU ONNEPC,ENTIALIrEPOURLEDEVELOPPMNTROOy3 \ t V ES 150~ -6 ~ ~ ~ ~ ~~ooo~o

I) \f 5 / 6 f_1 /QI EGHFRWNPRIDIzNsERpsrosEEEopM54 0 Rgod ^ \, -- IT6 UUAERILRDEMT'IETN8DI/FRONTIERES

_ MTFOERAL zNsEEusoRNuuEEEopMN -\~~~~~. / /\~ 53

S-- 4'_ - 5- hbEo;. / I6 / / A-_ > \-f f>_ F O R 2~ OA EAESIA1TGRCLUADELOMN -107.-Zt) 4/4) /~~~I I* \ POErA-IEOREEEOPMNARCLL 42I LEG6O- RWN .4 EIDZNSDPRVNEAIASCE-tEXEMVNE5 RPSTOSG EEOPMN ,Y,4^ X~~ S ~ ~ F ,

5 E AZN IIILEI.FOET R AINA AKB E N I N~~~~~~IMT G H AD EL

B. 5 G H A~~~~~~~~~~~~~~~~~~~~3OEAT UTAIT FONGICLUALDVLOMA PRVEOVI ENTAPITALSAL AREA-XDSPRVCE 5 L~~~~~~~~~~~~~~~~~~~~~~~~OENGTHAIPOUROWN LEIDZNSDPOOIINE ARCL DEVELOPPEMENT 6 ~~~-270 LOWDUREE_____ SUIABPEIITY AGRCUTURALCRISNE FOR DEVELOPMENT®NAIALCPTL/AIALAIOLEErE RA FAIMITLOCOTENTALREA POUR LEENESOUEVELUTUEPDVELPEEET II )~~~~~~~~~~~~~~~~J ~ ~ ~ ~ ~ ~AGINAL SUITABILITYFOR AGRICULTURAL DEVELOPMENT POROVINCBOUNDPARIES NATONUXIEGPOIC 6 5 2 BONN~~~~~~~~~~~~~~~~~~~~~~~~~~RSABEPOTENTIALITE POUR LE GEVEL OPPEMENT AGRICOLE RVR LUE

UNOWSUITABLETFORAGRICULTURAL DEVELOPMENT -j - NENATIONAL BOUNDL/ARIESL NATIONTALE

TRASFAGE POTENTIALITE POUR LE DEVELOPPEMENT AGRICOLEINEATNAS

c6TE D'IVOIRE ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~S-oce:H-Rti,Ig T-oI,i-l SRrVic., L-rt.d, I FRRbF38-F39 3 J~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~T JUNE P 92

IBRD238A5

* 0. 1. L.iAURITANIA 41 -~~~~~~MAI ~~~~~~~~2 NJGER

BURKINA FASO OUDALAN ) (URKNA' I AGRICULTURALMIGRATION FROM .* IEI THEMOSSI PLATEAU,1 975 'Gorom-Go,omn FAS Z MIGRATIONRURALE A PARTIR DU I s _ GI ½ PLATEAUMOSSI, 1975 @oGCE D,UVOI R. /GHi ' ' S 4'~~~~~~~Di6 Do,i@94

ATLANTIC Gulf o O... '','..,', - - SEN OCEAN

BAM

SANMATENGA\~ 5,

TIOVU-19 K.y4KovG ( GNAGN F~~~~~ . w 1 f/ a F$i . ~~~~~~~~~~~~~~~~~~~~.a. > (sr *

Eo 5 E0. S U O 0 ANZOURGOU;- 0 I%A'~~~~~ ~~~;Y"'No.n~~~~~ ~~BRAITENGAKORTEG

Wo -~~~~d~ A. . b..d.A.,~'A. A...~ ' X ~. ~> / ~ ,/ ~ / ~;> ~At ~ ~~~~~~~~~~~~~~~~~~~~~Zorgho~ ~ ~ ~ ~ ~ ~ ~ ~~~~T-.dgo @K.uV G OU RM A(\

_/L Cfba 0E-of o. 0, Thd.oMo {; - ' / *- 12 N..B;'' T A P O A

p~oIKESEDOUGOU Th. i~k,+, _OSSII OULGOU

~~A~~~~ 7 / KOR~~~~~~~~~U KAGGIu.STONGO(% LV 2gI% o MINORKPPULATINOUHMF B; ~~~~~~~u.. ,6 L>^ N B E I N / / obo-Dioulcssos . _,G~~~~~~~~OUGORIBA Lb.O '< @-N' . EN

I,OrodomoS,y / . .. ..

I@ 70nsoro s H A - wGN AGANTORGO J }~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~_..... MAJORPOPULATION SHIFT J~~~~~~~~O O P ATEOF MIGRATION: -4.:cMOUVEMENr MIGRATOIREIMPORTANT _s < O M O E j ~~~~~~~~~~~~~p0 N I TAUXD'IMMIGRA710N: MNRPPLTO HF MOUHOUN...-+~~ Z 0 K sGon,Esi Foo _NGoo,nJMOUVEMENFMoGgArOIREDEfAlEIMPORTANCE

s / Jy \ tiz g ~~~~~~~~~~~~~~~~~~~~~~~~~~~NOTABLE(I - 2%j / MODERE(I 2%) RIVERS/ FLEUVES G -0 ME - 06 CD PROVINCECAPITALS CHEF-UEUX DES PROVINCES

i_ ) jw ~~~~~~'-' ' t 63~A(j NATIONAL CAPITAL /CAPITAL NATIONAL 2 t7 S/'/ > ) E3~~~~~~~WEAK (0.2 -0.3%) / FA1iSE(0.2 - 0.3XJ - PROVINCEBOUNDARIES / UANTIEDE PROVINCE

\ i ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ - ~~~~~~~~~~~~~~INTERNATIONAL30UNDARIES / FRON77ERES

COT6 D'IVOIRE ' 2!-Source: F.O.dogo, 1976 IN JUTINAl992 O U E T- ?-

IBRD23846 13~~~~~~~~~~~~ S~~~~~~~~~~~ ~ ~~ ...... ~~ ~~ M~~~~~~~~~~~AURITANIA~~~~~~~~~~~~"OAUIAI

--4 4' 30,/ ig ,° /._. ,

BURKINAFASO OUDALAN - ,-URKINA ) CASESTUDY RESEARCH SITES FAS SITESD'ETUDES \Aorom-Ooyom GU-I F z j |l. \ /'- , ~~~~~~~~SoUM z9.),!., L

1D 1b6 5 (

> T >1 e°U^ 9~~~~~~~~~~~~~~~~~~~~~~~~~~~~o>if~~~~~~~~~~~~~~~~~~~~~~~~~D, ) t 9\<

-~~~~~~~~~ ~~~SENO 14 OCEAN 9 - %a 0 < K~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- 5 GANENA YATENGA| S,

TOAROn0 ~~~BAM -.

&.~~~~~~p.*~~ Th* ~~~~~~~~~~~~~~ A(. 90 jR KOS SIZnr

b- 2b. Oi d8 B~.., AZE G A/ KOUBRI TENGA > A T \ IDoc .. *N7dw.< -,' hh.I.~I,k IR,Id . .O. t/H~~~~~~~~~~~~~~> OU 0i /\> ' Yoko N ' +-. Kmbsir-)O D T * OFG Fi N'Gourmo s. @ - Jt ^ 40 mOUHOUNMOUH~~~~SOOUU R rUX\oZ1 0" 0 9Korebkir,i F NOon P9~

IS. B < " : AEGA ,- TTEAGA KOPEG ,B T A P0 NNA M.>0oiuoro-i ORB U OURTNGAH U I

KENEDOUGOU* A S S S L0B

H O U E T ( RoboDiOLO,,O 'OPNGA uK BENIN

OMOE P 0( N i SIPROVINCEPOCATCESCAPITALS / CHEF UEUX DES ® NATIONAL CAPITAL/ CAPITAL NATIONAL

/ LIMITEDE PROVINCE - PROVINCEBOUNDARIES INTERNATIONALBOUNDARIES / FRONTIERES INTERNATIONALES COTE 9 D'IVOIRE ' U JUNE1992~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~JUE19 L - | n ' i

IBRD23852

IS' 3' T~~~~~~~~~~~~~~~~~~~~~~~~~~ - 0' IS ~~~~~~~~~~~~~~~~~~~~~~~~MAURITANIA BURKINA FASO ------NGE

CLASSIFICATIONOF 'O > QD ALA,N / z 5URKNA I

PROVINCESIN TERMSOF THEIR -

FORESTRYSITUATION Grmom 0 GUINEA ~ ¼Z

CLASSiFICATIONDES PROVINCES N'oDI bo-DIOE EN FONCTIONDE LA SITUATION Do,, 14' ~~FORESTIERE N TI ll f K

0 Osohigooo3 AM

K 0 3~~~~~~~~~A.-%ATE'~GA

0 ~ ~ , Toko -, K.y*9G-,b, -~~T. 15

Too0 8on -~~P J

I 3' G Bogondh~~~~~~N-C)tiRIE d...... -,.d ZW-6~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~' . pdh. d-. , PAO-

POUAGA~DOIGOIJ-oR * @Dhdougoo K-iug. MI - - -r~~E)

'9 - I Z 0 om~~~~~~~~~~~~~~~~Kb-si, Fda N'Goormo

f 0F6~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~E 2

Bopforoo G H A ~ ~~N ~~~~~~~~~~~~~~~~~~~~A LARGSTAOUNTOF ROTETED ORETS kdg

-K E ~~~~~~~~~~~~~~~~~~~~4E E-PRTSEITATSMASE DNE

'I IX'~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

FORGESTSDGADEDSO POETD OET PEMIERTSDERANDEE ZOPROINCEBOUNARIEEREUTEROTPRVINC

FORESTSVTIRULLY INONEXISTENCEBT I AGRRVR UE EGRETSERSANTESUEMENT ENEXDSANTGER NENAINLBUDRIS/POTEE

FORETSPRATIQUEMENT INEX15TANTES ~~~~~~~~~~~INTERNATIONALBONAIS/FNTEES C T E D'IVOIRE ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Soo,oG.i 0 k., S ISP1985, PN-GTV198Fb IS

JUNE 1992

p 1 5 i 29 i / ~~~~~~~~~~~~~~~~~~~~r' ~~~~~4 MAURlrANIA\ > \ I~~~~~~~~~~~~~AUITNI

BURKINAFASO ,l _ , \ s, LOCATIONOF AVV PLANNED - X ODLN )k/'itRIA ~

SETTLEMENTSCREATED 1973-1 984 oooomi--2 NGEI LOCALISATIONDES PERIMETRES AVV r C^UINEA _ IoII zj CREESENTRE 1973 ET1984 w ss lou DVIEc6T~ aGHA i8-t

Dpbo~~~~~~~~ D., 14' rt^ ,G,,r ,cwi".,

IA' ~~~ ~ ~ ~ ~ ~ ~ ~ - Orioyl )n , B~~~~~~~~~~~~~~~~~~~~~~~~~~~~ A M <) \ z I \ w.'c v~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~SN\* X .A x

( J TougrnS - X AKrzyr S~~~ou.it G NAG 4G,,b

d :ndinfgrlh; ¢__ Nounr @\ >3? _@ OUBRI~~~~~~~~~0 YiANMTENGA >sJsolo

psd oS KE P A So~

g~~~~~u7 BZEGA tTAPOAKOUBRITTENG

t -, ,r-.. _i 4 / DJIPOLOGO m f b ) p T O~DIOGGOURMA

-12 B A Z T AW PLNE0ETET RIES/ALUE

~~~~~~~~~~~f% t >5POINCBOUDRIES)MT EPOIC

) \ j I I --- INTERNATIONAL~~~~~~BONARE FRNIEE COT D'VORE T s 5 >> 4- 3- \'< 2t !- oo ! 2- IUSE 199

Distributors of World Bank Publications

ARGENTINA The Middle EastObsurver KENYA SOUTH AFRICA, BOTSWANA Carlos Hirsch, SRL 41. Shenf Street Afrca Book Service(LA.) Ltd. Firrsing tiets: Cajena Guemes Cairo Quaran House. Mftngno Street Oxford Univeruty Prn Florida 165, 4th Floor-Oc. 453/465 P.O. Box45245 Southern Africa 1333Buenos ALres FINLAND NAirobi P.O. Box 1141 Akateeminen Klrakauppa Cape Town 8000 AUSTRALIA, PAPUA NEW GUINEA, P.O. Box 128 KOREA. REPUBLIC OF FIJI, SOLOMON ISLANDS, SF-00101 HeLsinki 10 Pan Korea Book Corporation For slscption arder: VANUATU, AND WESTERN SAMOA P.O. Box 10t Kwangwhamun IntermatLonalSubecipttonSe-etce D.A. Books& jouriAls FRANCE Seoul P.C. Box41095 646Wiutehorse Road Worid Bank Publications Craiglhll Mitcham 3132 66. avenue dl6rna MALAYSIA Johannesburg2024 Victona 75116Pars Universtvof MalayaCooperative Booshop. Limuted SPAIN AUSTRIA GERMANY P.O. Boit1127. JaLan Pantai Baru Mundi-Prensa Libros, SA. Gerold and Co. UNO-Verlag 59700 Kuala Lumpur Castello 37 Graben31 Poppelsdorfer Alle 55 2S001Madrid A-l011 Wien D-5300bBonl MEX5CO INPOT'EC Llbreula Itenleacional AEDOS BANCLADESH HONG KONC MACAO Apartado Postl 22-eW Consell de Cent. 39 Micro Industries Development Asia 2000Ltd. 14060Talpan Mexico DF 00009 Barcelona Assstance Society MIDAS) 46-48WyndŽam Stneet H ouseS, Road 16 Winning Centne NETHERLANDS SRI LANKA AND THE MALDrVES Dhanmondi R/Area 2nd Floor De Lindebooin/InOr-Publikaties Lake House Bookshop Dhaa 1209 Cental Hong Kong P.O Box 202 P.O. Box 244 7480AE H"aakberge 102. Sirf CuttEspipam A. Bnmrschoie INDIA GardinerMawatha 156, Nur AhmuedSank Allied Publishers Privte Ltd. NEW ZEALAND Colombo 2 Chittagong 4000 751Mount Road EbSCONZ Ltd. Madras - 600 002 Pnvate Mal BSag99914 SWEDEN 76, KD.A. Avenue New Market For singLe1tiw Kulna 9100 Bmnchoffic: Auckland Fritres Fackbokforetaget 15 fYN.Heredia Marg Regeringpgtmn 12. Box 16356 BELCrUM Baulard Estate NIGERIA S-10327StockhoLn JeanD[ Lannoy Bombay - 40003 University Pm Limnited Av. d, Roi 202 Three Crowns Building Jericho Fo,rjubscrii ar: 10602r,uels 13/14 Asaf Ali Road PnrvaicMail Bag5095 Wennergren-WilliamsAB New Delhi- 110002 rbadan P. C. Box 1305 CANADA 5-17125 SoLnA Le Dufftseur 17 Clittaranjan Avenue NORWAY CA.?85, 15I0Bn eAmpere Caicutta-700 072 Narvesen Informabon Center SWITZERLAND Bouchervlle. Quebec Book Department Fr singleilits: J4BSE6 layadev Hostl Building P.O.Box 612i EtterbUd Lbraiie Payot 5th Main Road,Gandhinar N40602Oslo 6 1 rue de Bourg CHriLE Bangalore-560009 Ci 1002Lausanne Invertec IGTSA. PAKISTAN Arnenco Vespucto Norte 1165 3-5-1129Kachiguda Mirza Book Agency For swbscpton order: Santiago CrG Road 65, Shahrah-e-Qua2dxAzam Libraiie Payot Hyderabad -500 027 P.O. Box No. 729 Servicedes Abonnementb CHINA Lalhore3400 Casepostale 3312 China fniul & Economic Prarthara Plate. 2nd Floor CH 1002LAuann Publishing House Near Thakor Baug, Na^vngpurt PERU B Da FoSi DongJie Ahmedabad - 380009 Editonal Desarrollo SA TANZANIA SeiJ)ng AparMdo 3824 Oxford University Pres PatualaHouse Laia I P.O.Box 5299 COLOMBIA 16-A Ashok Marg Makaba Road lnfoeiiace Ltda. Lucknow -226001 PHILIPPINES Dares Salaamn Apartado Aereo 34270 Internatonal Book Center BogotaDlE Central BazaarRoad Suite 17D3.Ctylnd 10 THAILAND 60 BayaNagar Condominium Tower I Central Departmient Store COTE D'IVOIRE Nagpur 440 010 Ayala Avenue. Corner H?V.deia 306 SdlomRoad Cartre dEdition et de Diffusion CostaExtension Bangkok Afncaines(CEDA) INDONESIA Makti. MetroManil 04 B.P.541 Pt Indira Limited TRLNIDAD & TOBAGO. ANTIGUA Abidjan 04 Plateau JaLanBorobudur 20 POLAND BARBUDA. BARBADOS, P.O. Box 18l IntemnatonalPublishing Service DOMINICA. GRENADA. GUYANA, CYPRUS Jakarta C320 Ut. Piekna 31/37 JAMAICA. MONTSERRAT, ST. Canter of Applied Research 00-677Warzawa KITS & NEVIS, ST LUCIA. CypnusCollege IRELAND ST. VINCENT & GRENADINES 6. DiogenesStreet Engoui GCovernmentSuppliesAgeny Forsscrpt,anorders: Svstemacs Studies Unit P.O. Box2D06 4-5 Harcourt Road INSJournals #9 Watts Sheet Nicosa Dubtin 2 U. Okrezna 3 Curepe 02-916Warizawa Trinidad, West Indles DENMARK ISRAEL Samnund.Litteratur Yozmot Uterature Ltd. PORTUGAL TURKEY RostenoernsAlI 11 P.O. Box 56055 Livaria Portugal Infotel DK-1970 Fredeksberg C Tei Aviv 615o0 Rua Do Grmo 70-74 NarlabaluqeSok. No. IS 1200Lisbon Cagaloglu DOMINICAN REPUBLIC ITALY Istanbul Editors Taller, C. por A. Licosa Comnissionaria SansoniSPA SAUDI ARABIA. QATAR Restauraci6,nelabeltaCat6lica3D9 ViaDucaDiCalabria 1/1 Jar BookStore UNITEDKINGOOM Apartado de Correos 2190Z-1 CasellaPtale 552 P.O. Box 3196 Micromfo LtdL Santo Domingo 50125Firenze Riyadh 11471 P.O. Box 3 Al ton. Hampshire CU34 2PG EGYPT, ARAB REPUBLIC OF JAPAN SINGAPORE, TAIWAN, England Al Ahram EAsten Book Service MYANMAR,BRUNEI A! Cala Street Hongo 3-ClhomerBunkyo-ku 113 InfornmationPublications VENEZUELA Cai-o Tokyo Pnvate, Ltd. Lbr-na del Este Golden Whmel Building Aptdo. 60-337 41. Kallang Pudding, 00443 Caracas1060-A Singapore1334 i

I RECENT WORLD BANK TECHNICAL PAPERS (continued)

No. 166 Ahmad and Kutcher, Irrigation Plannriingwith Environmental Considerations:A Case Study of Pakistan's Indus Basin No. 167 Liese, Sachdeva, and Cochrane, Organizing and Managinig Tropical Disease Control Programs: Case Studies No. 168 Barlow, McNelis, and Derrick, Solar Pumping: An Introductionland Update on the Technology, Perfornance, Costs and Economics No. 169 Westoff, Age at Marriage, Age at First Birth, and Fertility in Africa No. 170 Sung and Troia, Developments in Debt Conversion Programs and Conversion Activities No. 171 Brown and Nooter, Successful Small-Scale Irrigationiin the Sahel No. 172 Thomas and Shaw, Issues in tiheDevelopment of Multigrade Schools No. 173 Byrnes, Water Users Association in World Bank-Assisted Irrigation Projects in Pakistan No. 174 Constant and Sheldrick, World Nitrogen Survey No. 175 Le Moigne and others, editors, Country Experiences with Water Resources Management: Economic, Institutional, Technologicaland EnvironmnentalIssues No. 176 The World Bank/FAO/UNIDO/Industry Fertilizer Working Group, World and Regional Supply and Demand Balances for Nitrogen, Phosphate, and Potash, 1990/91-1996/97 No. 177 Adams, The World Bank's Treatment of Employene7ztand Labor Market Issues No. 178 Le Moigne, Barghouti, and Garbus, editors, Developing and ImnprovingIrrigation and Drainage Systems: Selected Papersfrom World Bank Seminars No. 179 Speirs and Olsen, Indigenous InitegratedFarming Systems in the Sahel No. 180 Barghouti, Garbus, and Umali, editors, Trends in Agricultural Diversification: Regional Perspectives No. 181 Mining Unit, Industry and Energy Division, Strategy for African Mining No. 182 Land Resources Unit, Asia Technical Department, Strategy for Forest Sector Developmenit in Asia No. 183 Najera, Liese, and Hammer, Malaria: New Patterns and Perspectives No. 184 Crosson and Anderson, Resources and Global Food Prospects: Supply and Demand for Cereals to 2030 No. 185 Frederiksen, Drought Planning and Water Efficiency Iinplicationisin Water Resources Management No. 186 Guislain, Divestiture of State Enterprises: An Overview of the Legal Framework No. 187 Geyndt, Zhao, and Liu, From Barefoot Doctor to Village Doctor in Rural China No. 188 Silverman, Public Sector Decentralization: Economic Policy and Sector Investmnoit Programs No. 189 Frederick, Balancing Water Demands with Supplies: The Role of Managemienitin a World of Increasing Scarcity No. 190 MackIin, Agricultural Extension in India No. 191 Frederiksen, Water Resources Institutions: Some Principles and Practices No. 192 McMillan, Painter, and Scudder, Settlement and Development in the River Blindness ConitrolZone No. 193 Braatz, Conserving Biological Diversity: A Strategy for Protected Areas in the Asia-Pacific Region No. 194 Saint, Universities in Africa: Strategies for Stabilization and Revitalization No. 195 Ochs and Bishay, Drainage Guidelines No. 196 Mabogunje, Perspective on1Urban Land and Land Management Policies in Sub-Saharan Africa No. 197 Zymelman, editor, Asseessing Engineering Education in Sub-Saharan Africa No. 198 Teerink and Nakashima, Water Allocation, Rights, and Pricing: Examples from Japan and the United States No. 199 Hussi, Murphy, Lindberg, and Brenneman, ThleDevelopment of Cooperatives and Other Rural Organizations: The Role of the World Bank The World Bank Headquarters European Office Tokyo Office A 1818 H Street, N.W. 66, avenue d'16na Kokusai Building ' Washington, D.C. 20433, U.S.A. 75116 Paris, France 1 -1 Marunouchi 3-chome Chiyoda-ku, Tokyo 100, Japan Telephone: (202) 477-1234 Telephone: (1) 40.69.30.00 Facsimile: (202) 477-6391 Facsimile: (1) 40.69.30.66 Telephone: (3) 3214-5001 Telex: wui 64145woRLDBANK Telex: 640651 Facsinmile:(3)3214-3657 RCA 248423 WORLDBK Telex: 26838 Cable Address: INTBAFRAD WAS-ENGTONDC

0-8213-2381.4123 81 AGR 100

SETTLE & DEV RIVERBLIN B

400000008523 $10.95

ISBN 0-8213-23814