4r M,~ . i. ' 4U j .—o "a “I: .. x l [I -0 ~"1 '” ,— my) , . 1 ' 9:...- +1-M “a: ffirr. \.._3 J4“ ..-—~ ’ Z , 37!? 03701:? .,.- £02001 U J0] A .6 :1’ SEP 0 32003 6812 In 0] “-M-d. gur‘ “.“goanéldr-‘u ' THE POLITICAL ECONOMY OF GRAIN MARKETING AND STORAGE IN THE SAHEL By David Carrington Wilcock A DISSERTATION Submitted to ‘Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultural Economics 1977 ABSTRACT THE POLITICAL ECONOMY OF GRAIN MARKETING AND STORAGE IN THE SAHEL By David Carrington Wilcock Weather-induced instability in domestic food grain supplies in the subsistence agricultural economies of Sahelian West Africa is the overarching, immediate problem which motivated this study. The dimen- sions of this problemr-exaccerbated and widely publicized during the 1968-73 regional drought-and potential solutions involving increased grain storage and improved marketing institutions and services are explored within a political economy conceptual framework. This frame- work applies a general theory of structural systems dependency to the particular historical and environmental conditions of West Africa. In this perspective complex coalitions of actors and institutions in the center and periphery of an evolving world economy functionally serve the interests of increasingly concentrated financial capital. In Sahelian west Africa this has led to structural underdevelopment which is characterized by rapid urbanization with high unemployment, formation of new elite classes, increasingly unequal distribution of wealth, rural stagnation and increasing dependence on foreign food as well as capital. In the future these patterns of dependence may be exaccerbated or an alternate path may be followed toward greater David Carrington Wilcock economic and political independence and self-reliance. The latter option stresses rebalancing domestic production and consumption in decentralized patterns of self-generating development whose structures and incentives are consistent with the realistic promotion of stated Sahelian objectives of increasing food self-sufficiency and autonomous economic growth. Following an overview of Sahelian macro-economic indicators the study focuses on trends in regional grain production, consumption and food imports. Regression analysis of 1951-75 aggregate production levels showed long term stagnation in grain production aggravated by severe drought particularly in 1972-73, little evolution in estimated crop yields except for those under irrigation, and substantial varia- tion in production levels within and among Sahelian countries which indicates potential for expanded locally and nationally held grain stocks. A review of traditional Sahelian grain marketing reveals that, within the context of relatively free local grain markets, non-market factors play a substantial role in determining differential ability to engage in spatial and temporal arbitrage. This, coupled with cash and food crop marketing and pricing policies disadvantageous to pro- ducers, tends to limit the expansion of production and self-generated growth in the agricultural sector and leads to the underutilization of cost-efficient traditional grain storage technology. The remainder of the study involves a detailed examination of issues in Sahelian government involvement in grain marketing, an ana- lysis of existing grain boards in the six countries, and proposals for possible incremental change in these institutions and policies to meet David Carrington Wilcock Sahelian deve10pment objectives. In addition to macro-economic issues, the key policy areas examined are those involved with translating traditional emergency and buffer stock options to the Sahelian envir— onment and the specification of appropriate analytical procedures. Existing Sahelian grain boards are hampered in their pursuit of marketing and storage objectives by inappropriate institutional structures-which lead to overcentralization, duplication of effort, disfunctional incentive systems, very high operating costs and dis- tributional effects biased in favor of certain urban groups--by insuf- ficient storage and transportation infrastructure, and by overly rigid and badly timed official grain price structures. The general strategy outlined for institutional reform involves holding greater quantities of grain at all levels of a federated na- tional grain marketing system to be based on strengthening village level storage capacity to provide a risk-minimizing service to all segments of the rural population. This is illustrated by the village stock system proposed in a detailed case study of Upper Volta. Urban supply stabilization, inter-regional transfers and the holding of strategic emergency stocks would be assured by streamlined regional and national institutions, aided by practical programs of technical assistance in grain stock management and the promotion of increased functional contacts among Sahelian grain boards. ACKNOWLEDGMENTS Among the many persons who have provided guidance and inspiration during my graduate studies I would particularly like to express my sincere appreciation to Carl R. Eicher, my major professor and thesis supervisor, for the helpful advice and friendship he has offered me. I would also like to thank the members of my guidance and thesis committees: Derek Byerlee, Lester Manderscheid, Lawrence Officer, James Shaffer, vernon Sorenson and warren Vincent. I am grateful to the United States Agency for International Development, particularly to David Shear, Director of the Office of Sahel and Francophone West Africa Affairs of the Africa Bureau, for providing a number of field work opportunities in the Sahel which led to this study. I would also like to thank M. Jomni of OFNACER in Ouagadougou; M; Dosso of OPVN in Niamey; MM. Golding, Von Teutem and Meuller of the FAO, Rome; and Dr. Shlomo Reutlinger of the World Bank for generous assistance in this research effort. My fellow graduate students in the Department of Agricultural Economics have provided insights and friendships over a four-year period for which I am most appreciative. I would particularly like to thank Mark Newman, Edouard Tapsoba and Tom Zalla who provided needed encouragement and advice during the completion of this study. ii Jan Carlin and Janet Munn typed the drafts and final version of this study with great skill and patience and I thank them sincerely. To my wife, Karen, and son, Daniel, I owe my greatest debt of gratitude and I dedicate this work to them. iii TABLE OF CONTENTS Chapter Page I INTRODUCTION . . . . . . . . . . . . . . . . . . . . . 1 Poverty and Crisis in the Sahel . . . . . 2 Problem Definition and Study Objectives . 8 II UNDERDEVELOPMENT IN THE SAHEL: A CONCEPTUAL FRAMEWORK . . . . . . . . . . . . . . . . 12 Introduction . . . . . . . . . 12 A Political-Economy Model of Center-Periphery Domination and Dependency . . . . . . . . . 15 Model Definitions . . . . . . . . . . . . . 16 Historical Note on the Center- Periphery Mbdel . . . . . . . . . . . . . . 18 Definition of Center-Periphery Domination . . . . . . . . . . . . . . 20 Mechanisms of Domination . . . . . . . . . . . . 23 III ECONOMIC CONDITIONS IN THE SAHELIAN PERIPHERY . . . . 52 IV TRENDS IN SAHELIAN FOOD GRAIN CONSUMPTION AND PRODUCTION . . . . . . . . . . . . . . . . . . . . 73 Food Grain Consumption . . . . . . . . . . . . . 74 Population Growth and Urbanization . . . . . . . . 83 Trends in Food Grain Production, Yields and Imports . . . . . . . . . . . . . . . 87 Production . . . . . . . . . . . . . . . . . . . 89 Yields . . . . . . . . . . . . . . . . . . . 95 Correlation of Yield and Production Levels Variation . . . . . . . . . . . . . . 97 Grain Imports . . . . . . . . . . . . . . . . . 99 V PRIVATE GRAIN TRADE AND STORAGE . . . . . . . . . . . 103 Market Structure and Operation . . . . . . . . . . 107 Marketed Surplus . . . . . . . . . . . . . . . . . 110 iv Chapter Market Efficiency and Competition . Distributional Effects of Market Operation . . . . . . Traditional Private Grain Storage . VI ISSUES IN SAHELIAN GOVERNMENT PARTICIPATION IN GRAIN MARKETING AND STORAGE . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . Grain Board Tasks . . . . . . . . . . . World Food Crisis and the FAO Food Security Assistance Scheme . . . Grain Stock Policy Issues . . . . . . . . . Macro-Level Policy Issues . . Emergency Stock Issues . . . . . . . . . . Buffer Stock Issues . Methods of Analyzing Grain Storage VII SAHELIAN GRAIN BOARDS: STRUCTURE AND PERFORMANCE . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . Country Reviews . . . . . . . . . . . . . . . . . . Mauritania . . . . . . . . . . . . . . . Senegal . . . . . . . . . . . Chad . . . . . . . . . . . . . . . . . . . . . . Mali . . . . . . . . . . . . . . . . . . . . . . Niger . . . . . . . . . . . . . . . . . . . Considerations for Sahel-Wide Grain Storage Efforts . . . . . . . . . . . . . . Cbnseil de Z'Entente Stabilization Project . . . . . . . . . . . . . . . . . . . Proposals for Nigerian Food Security . . . . . . Sahelian Food Grain Marketing and Storage Institutions and Policies: A Summary . . . . Institutional Structure . . . . . . . . . . . . Objectives . . . . . . . . . . . . . . . . . . . Means to Implement Objectives . . . . . . . . . Price Policy . . . . . . . . . . . . . . . . . . Relief Efforts . . . . . . . . . . . . . . VIII UPPER VOLTA: A CASE STUDY IN THE IMPLEMENTATION OF GRAIN MARKETING AND STORAGE PROGRAMS IN THE SARI! C O O C O O O O O O O O O O O O O 0 Introduction . . . . . . . . Elements of Voltaic Rural Development Strategy . . . . . . . . . . . . Page 111 114 116 119 119 119 124 126 129 132 136 138 150 150 151 151 154 160 164 171 176 178 180 183 184 186 187 189 191 194 194 195 Chapter Government Grain Marketing in Upper Volta . . . . . . . . . . Emergency Grain Stocks in Upper Volta . . . . . . Emergency Stock Implementation: The First Two Years . . . . . . . . . . Lessons and Possible Changes . . . . . . . . . Buffer Grain Stocks . . . . . Operations to Date . . . . . . Proposed Village Grain Stock Program . Grain Price/Supply Stabilization for Noanroducers . . . . . . . . Toward a More Integrated National Food Grain Marketing and Storage System . . . . . . Overview of a Possible System . . . . . Practical Measures to Improve the Existing System . . . . IX SUMMARY CONCLUSIONS AND POLICY RECOMMENDATIONS . . . . . . . . . . . . . . Summary Characteristics of Grain Marketing and Storage Institutions . . . . . . Objectives . . . . . . . . . . . . . Institutional Structures . . . . . . Means of Implementing Program Objectives . . . Performance of Government Grain Institutions . . . . . . . . . General Policy Recommendations . . . . . . . . . General Strategy . . . . . . . . . . . . . . . Investment Priorities . . . . . . . . . . . . Research Priorities . . . . . . . . . . . . . APPENDICES, A STATISTICAL TABLES TO ACCOMPANY CHAPTERS III AND IV . . . . . . . . . . . . . . . . B A COMPARATIVE ANALYSIS OF PER TON STORAGE COSTS FOR BULK SILO AND BAG WAREHOUSE GRAIN STORAGE . . . . . . . . . . . . BIBLIOGRAPHY . . . . . . . Page 199 205 206 210 213 213 213 217 227 227 230 240 257 257 258 260 263 263 263 266 268 271 280 290 Table III-1 III-2 III-3 III-4 III-5 III-6 III-7 III-8 IV-Z LIST OF TABLES The Sahel and West Africa: Area, Population, GNP and GNP Per Capita, 1975 O O I O O O O O O O O O O O O 0 Estimated Relative Importance of the Sahel Zone in Each Country . Agricultural Income Per Capita as a Percentage of Nonagricultural Income Per Capita for Selected Countries Sectoral Allocation Gross Domestic Product, Factor Cost Estimation for 1970 (in Percentages) . . . . . Official Commercial Trade, Annual Averages, 1969-71, by Sahel Country Export Value by Selected Categories of Exports, 1963-70 (in Percentages) Concentration of Sahel Trade To and From Major Export and Import Partners with Comparisons to Selected Other Countries (1965) . . . . . . . . . Total Bilateral Development Assistance (From.Market Economies) and Multilateral Assistance (1972-74 Annual Average) in Relation to 1973 Total Government Budget Expenditures (Including Investment) and to Total Estimated 1973 Gross National Product by Sahel Country . . Total Daily Per Capita Calories Consumed by Food Class for Sahel Countries (in Percentages) . . . Estimated Sahelian Consumption Levels of Main Staple Commodities (Average Kg Per Head Per Year, 1967-71) . vii Page 53 56 58 59 61 62 65 71 75 78 Table IV-4 IV-5 IV-6 IV-7 VII-1 VIII-1 Estimated Urban and Rural Staple Grain and Root Consumption in Upper VOlta and Mali, 1972 (in Average Kg Per Head Per Year and Percentages) . . . . . . . Population of the Sahel by Sector (in Millions and Percentages) . . Sahel, Nigeria and Total West Africa: Annual Average Gross Production of Main Staple Foods ('000 Mt and Percen— tages of West African Totals) Mean Annual Grain Production and Coefficients of Variation by Group and by Country, 1961-75 (Thousands Metric Tons) . . . . . . Pearson Correlation Coefficients for Total Cereal Production by Sahelian Country, 1951-75 and 1961-75 (Coeffi- cients with "*" Significant at .05 or Better) . . . . . . . . . . Sahelian Food Grain Imports, . Averaged for 1969-71 and 1974 . . Malian "Bareme" (Marketing Margins) for Millet, 1974-75 . . . . . . . Comparison of Per Ton Storage Construction and Annual Operating Costs . . . . . . . . . . . . . viii Page 80 85 88 91 98 101 167 223 Figure II-l II—2 II-3 II-4 II-S III-l IV—l VI-l VI-Z VII-1 VIII-1 LIST OF FIGURES Page The Basic Structure of Center Periphery Domination . . . . . . . . . . . . . . 22 An International Interaction Budget . . 24 Feudal Center-Periphery Interaction Structures . . . . . . . . . . . . . 26 Center-Periphery Structural Transformation: Evolutionary Paths for the Periphery . . . . . . . . . . . . 45 Alternate Models of Peripheral Structural Transformation . . . . . . . . . . . 46 The Sahel: Major Ecologic Zones and Principal Cities . . . . . . . . . . . . . . 55 Trend in Combined Total Cereal Production for Mali, Niger, Senegal and Upper Volta, 1951-75 . . . . . . . . . . . . 92 Cereal Supply Situations and Cereal Office Actions . O I O O O I O O I I O O I O O O 121 Characteristics of Different Levels of Analysis of Marketing and Storage Activities of Cereals Office . . . . . . . . . . 140 Selected Fiscal and Product Flows for Senegalese Government Actions in Food and Cash Crops . . . . . . . . . . . . . 157 Republic of Upper Volta: Regional Development Organisms, Their Head- quarters and Location of Initial Emergency Stocks . . . . . . . . . . . . . . . 208 ix CHAPTER I INTRODUCTION In recent years there has been increased international concern with natural disaster and related development problems in the Sahel, a contiguous belt of West African countries which is usually defined to include Mauritania, Senegal, Mali, Upper Volta, Niger and Chad.1 In this study, I will concentrate on two general tasks: development of a structural theory of political-economic change to explore under- development in Sahelian West Africa, and, within this context, the specific examination of applied policy issues in the area of food grain storage. This work is based on available secondary data, on reports furnished by national governments and international organi- zations, and on several field assignments in the region. It is hoped that the study will be useful to policy makers in the governments of the Sahelian countries, the United States Agency for International Development, and other bilateral and international donor agencies, as well as be of some interest to academic colleagues. 1Some definitions also include the small riparian country of the Gambia and the Cape Verde Islands off the coast of Senegal. "Sahel" is an Arabic word meaning "shore", referring here to the "southern shore" of the Sahara Desert. 1 Poverty and Crisis in the Sahel By virtually every quantitative measure commonly used the coun- tries of the Sahel rank very low in the provision of government ser- vices and productive infrastructure, in agricultural modernization, and in overall levels of material development. With a few exceptions they are on all lists of the world's least developed countries. Within most of this region only 5 to 20 percent of the population live in urban areas although this percentage has been increasing rapidly in recent years. The bulk of the population is comprised of nomadic herders and sedentary farmers, and both groups are highly self-sufficient and only partially integrated into a cash market economy. Cotton and peanuts have been the major international agri- cultural exports with cattle playing an important role in West African trade. Ironically, for a region at the periphery of the world economy, the Sahelian countries are highly dependent. They are dependent on imported energy, technology, skilled personnel, foreign aid to meet investment and sometimes recurrent budget expenditures, outside markets for primary commodities, and the overall health of the eco- nomic system in more developed areas. Even food-primarily for urban areas-has been increasingly imported in recent years. The agricultural sector is heavily dependent on levels of rainfall occurring in general east-west belts which gradually decrease and taper off to nothing in the sands of the Sahara (see Figure III-l in Chapter III). As in other traditional societies in arid and semi- arid environments, highly specialized nomadic pastoral and sedentary agricultural systems have evolved over time to minimize risk and help insure survival under very harsh conditions. It is important to understand the nature and extent of these physical and economic dependencies because, in large measure, they define the parameters of the current problems of the Sahel and give some clues as to likely and possible evolutionary paths open to the region. Policy makers, no matter what their objectives but particularly those con- cerned with longer-run investment strategies, should attempt to take these factors into consideration to maximize their effectiveness. A description and interpretation of these patterns of Sahelian depen- dency, seen in a political economy framework, will be outlined in the following chapters to provide perspective for the more focused analysis which follows. The vulnerability of Sahelian agricultural systems was dramati- cally demonstrated during the 1969-73 drought. During these years of abnormally insufficient rainfall, the magnitude of production shortfalls and degree of physical suffering varied substantially by sector and subregion of the Sahel. Overall, however, the conse- quences were grim. It has been estimated that up to 100,000 deaths could be directly attributed to these unusually severe conditions, occurring primarily among the most vulnerable, the very young, and old.2 Perhaps one-fourth of the total livestock herd perished and cash crop exports fell drastically. The production of the primary staple food grains, millet and sorghum, was heavily affected. Total output of these grains, which had not increased as rapidly as popu- lation between 1960 and 1968, was off by a third or more 2For a very clear assessment of drought impacts, see Caldwell [1975]. particularly in 1972 and 1973 although there was great variation in these shortfalls among subregions. Even though the 1974 crop season saw a return to more normal aggregate grain production levels across the Sahel, individual countries continue to be hampered by local grain shortages due to natural causes ranging from insufficient or poorly distributed rainfall, to flooding, as well as insect and rodent infestation. These local shortages (and surpluses), in ways typical to all agricultural systems, give rise to sharp fluctuation in market- ed grain surpluses and in attendant food grain prices on both an intra- and inter-annual basis. In ways observed throughout history, these extreme supply and price movements carry with them the risks of consumer-led political instability and producer disincentive effects. The institutional response to extreme and normal fluctuations in food grain supplies has been complex, attracting both praise and criticism. First, the governments of the Sahel states were often quite limited in their capacity to provide immediate emergency assist- ance to their own populations. Politicallyindependent from France only since 1960, these governments are small, highly centralized institutions with very tenuous control over rural areas and often little positive impact on the economic and social life of the peasant population. Attempts of Sahelian governments to directly and indi- rectly exert a stabilizing influence on fragmented food grain markets through price policy and the direct actions of grain marketing boards have been largely unsuccessful to date. Second, the bilateral and international aid community, once mobilized, provided significant amounts of grain and other food aid to the drought affected countries. From 1973 through 1975 alone this aid had a value of over $400 million.3 Given normal problems in aid and relief coordination and given inadequate food handling, storage and transportation infrastructure which necessitated the use of costly improvised logistics, the relief operations were judged quite effective. There were, however, instances in which donated grain was of a quality unfit for human consumption, in which supplies rotted on the docks in coastal countries or in mismanaged government warehouses or were diverted to black market channels for private gain. This type of problem and the high cost of using ad hoc distribution systems have reinforced the concern on the part of the Sahel govern- ments and the donor community for increasing food production and stocks, improving food storage methods and implementing related food security strategies. The concern is both to diminish the need for future assistance by increasing the resistance of the Sahel to drought conditions and to improve the efficiency of distribution channels and facilities for relief purposes. During the drought period a number of organizations were estab- lished to help channel relief assistance and begin planning for more substantial longer-run recovery and development projects, among them OSRO (the Office of Sahel Relief Operations) of the Food and Agri- culture Organization (FAO), UNSO (the UN Special Sahel Office) and CILSS (the Permanent Interstate Committee for Drought Control). The latter organization, in response to donor offers of assistance, pre- pared initial lists of development projects in the six countries. This has led to the creation of a quasi-official grouping of 3U.S. Agency for International Development [1976a]. bilateral and international donors and Sahel governments, the CTub des Amie du ShheZ, whose purpose is to coordinate and plan for sub- stantially greater long-term development assistance to the region.4 In the past two years donor agencies have completed at least eight major studies of development strategies in the Sahel,5 as well as numerous special sector studies. In one of the most ambitious studies, FAO assembled a hypothetical list of development investments totaling $7.6 billion over the next twenty years6 and seen as neces- sary to meet the general objectives for the region agreed on by CILSS: Reduce the consequences of emergency situations in the future, insure self-sufficiency in staple foods (cereals and meat), and accelerate economic and social development, particularly in the least developed countries of the region. All of these studies have made reference to the positive role food grain stocks could play in the economic transformation of the region but provide very little systematic, detailed analysis of the subject. Concurrently with the peak of the Sahelian drought, severely depleted grain stocks in North America and rising prices for petroleumrbased inputs into food production technologies added to renewed world concern over the adequacy of food supplies. As one response to these circumstances, the constitution of food grain ASee Shear [1976] for a description of the history and objec- tives of the "Club". SSee IBRD [l975d]; MIT [1974]; OECD [1976]; SCET International, SEDES and ORSTOM [1976]; UN, FAO [1976a]; UN, 80 [1974]; USAID [1975b]; and USAID [1976c]. 6am, FAO [1976a]. 7USAID [1976c, p. 7]. reserves to be held internationally, regionally or nationally re- ceived considerable promotion and debate. In 1973, past FAO Director General Boerma proposed a system of nationally-held emer- gency reserves and this has remained the basic FAO position since then. The United States proposed a more international reserve system to the World Food Conference in 1974 stressing cost and risk shar- ing and Open flows of crop production information. The size, loca- tion and net benefit incidence of alternative reserve systems has also commanded significant attention in the recent literature of agri- cultural economics.8 While little progress has been made on proposed international grain reserves, FAO through its new Food Security Assistance Programme has been quite active on the national level in third world countries with a program of food security evaluation studies9 and within the past year has begun to help institute tightly run emergency food grain stocks in three of the Sahel countries, Mali, Upper Volta and Niger, and has engaged in some preliminary planning for the others. By 1976 all six countries had operational cereals marketing boards engaged in grain storage and various open market operations. All now employ food grain price controls, most involving fixed producer and consumer prices for sorghum, millet, maize and rice. In several cases attempts at grain market intervention have been pushed several steps further with the cereals marketing board or another arm of the central 8See, for example, Eaton and Steele [1976]; Hillman, Johnson and Gray [1975]; Reutlinger [1976]; Rojko [1975]; Trezise [1976]; Van Teutum [1975]; and walker and Sharples [1975]. 9For two examples in Africa, see UN, FAO [1974a] and [1975a]. government being given theoretical monopoly marketing powers although they clearly do not have the means to fully exercise such powers. Many of these policies have yet to receive careful study or evalua- tion partially due to the sensitive political nature of agricultural marketing and price policy and partially because various factions have differing, yet often vaguely formulated, ideological views on the wisdom and propriety of government intervention in food grain markets. Problem Definition and Study Objectives The first order problem facing the Sahel is that food grain supplies are chronically or cyclically inadequate and unreliable to meet development needs. There are clearly numerous potential strate- gies or groups of strategies that could be employed in attempting to find a solution to this overarching condition.10 Investment pro- grams-to increase agricultural productivity through the self- sustained use of new technology, to increase total agricultural out- put through acreage expansion as a result of improved water and land management, or to expand the reach and depth of markets in primary products and consumer goods-all involve substantial periods of time to reach fruition. Investments in food grain stocks under appropriate complementary policies and management conditions offer the potential of contributing to food security and supply stability in the short run lOR.H. Green [1975] provides a broader, yet succinct definition of what he calls the "paradigmal rural African problem". It "combines uncertain rainfall, mediocre soil, low unit value crops, scattered population, meagre input levels for both physical (e.g., seed, fer- tilizer) and knowledge (e.g., extension, new techniques) items, and resultant constant minimum level of cash income and recurrent near- famine conditions". and can provide a strengthened base for further transformation of the agricultural sector. This is not an either/or situation; in- vestments in holding food stocks complement and facilitate other productive investments regardless of the ultimate objective function of the society. Grain stocks are held by most societies at all levels of economic development; the crucial questions involve the functional roles of stocks and how their use is related to the overall political- economic evolution of a given society. Failure to understand the possible functional roles of food stocks in a particular economic setting or attempts to transplant stock systems from one economic environment to another have resulted in costly failures which may, in fact, promote increased supply in- stability and maldistribution. We must make our investment choices carefully in an environment of extreme resource scarcity where human survival and well-being are at stake. It is in this light that the overall purpose of this study is to provide a pragmatic policy ana- lysis of alternative roles for food grain stocks11 as one part of the solution to the problems of aggregate food grain supply insta- bility and more general underdevelopment of the Sahel. 11Terminology in the area of food grain storage tends to be a bit imprecise, with the names of various types of "stocks" and "re- serves" being used interchangeably. In this study stocks will be used instead of the more ambiguous reserves as the generic class with types of stocks named according to their principal functional purpose. Therefore, emergency stocks will refer to those stocks, held primarily by government bodies, to be used under specified rules to relieve "severe" food deficits, particularly where there is a risk of loss of life or severe malnutrition. Buffer stocks are those stocks, collectively held, whose functional role is to mitigate or "stabilize" fluctuations in supply (or price) whether on an intra- annual or inter-annual basis. I will also refer to pipeline stocks, those marketed grain supplies held by the private commercial sector for potentially stabilizing temporal arbitrage or potentially 10 To achieve this overall purpose this study has the following more specific objectives: 1. To outline a political-economy conceptual framework which pro- vide insights into underdevelOpment in the Sahel and which helps place policy alternatives in historical and socio- economic perspective (Chapters II and III) 2. To assess the quantitative dimensions of the Sahel food grain supply problem through the analysis of trends and variation in production, weather-production relationships, and longer- term patterns of food demand, particularly changes in con- sumption in urban areas (Chapter IV) 3. To describe private Sahelian food grain trade with particular attention to patterns in the marketing and storage of millet and sorghum (Chapter V) 4. To analyze, within the framework of Sahelian underdevelopment, government participation in food marketing and storage, parti- cularly its holding emergency and buffer stocks (Chapter VI) 5. To review the structure and initial performance of the six Sahelian grain boards which are the focus of government market intervention (Chapter VII) 6. To explore in detail in a case study on Upper Volta the economic and institutional dimensions and practical problems associated with emergency and buffer stock programs (Chapter VIII) 7. To summarize the findings and implications for technical and material assistance to Sahelian grain stock actions and for further macro and micro level research needs (Chapter IX. It would be helpful to the reader to understand that the analy- ses in this study focus on different levels and aspects of an inter- related socio-economic system over different time periods. The analysis of Sahelian underdevelopment in the first few chapters destabilizing speculation, and to farm-level stocks, those stocks held by rural production units, the "compound" or extended farm family, and used for a variety of purposes but principally home consumption. Clearly these labels are arbitrary because the cate- gories are not functionally distinct (e.g., emergency stocks can play a buffer role and buffer stocks may diminish the need for emergency stocks, etc.) and their use is not meant to conceptually prestructure the analysis which follows. ll adopts a long term perspective in which system goals and basic socio- economic conditions evolve over time. This general political- economic framework plus the specific chapters on Sahelian grain pro- duction, consumption and trade provide background and insights into the last four chapters which focus more narrowly on government parti- cipation in food grain marketing and storage. In these latter chap- ters the analysis is more limited in time and concentrates on the structure and performance of Sahelian grain boards and the practical institutional and economic policy Options open to them in modifying their operations to more fully meet national policy objectives. In summary, the component chapters of this study are complemen- tary. The political-economy framework will be used to describe the current economic state of the Sahel. It thus provides the context for understanding the economic and institutional constraints which define "micro level" policy options in grain storage, a rationale for alternative grain storage policies which might not be fully con- sidered on economic grounds alone, and suggested patterns of insti- tutional and policy structures more consistent with underlying poli- tical and economic conditions and, thus, more likely than other pat- terns to foster the progressive structural transformation of Sahelian economies. CHAPTER II UNDERDEVELOPMENT IN THE SAHEL: A CONCEPTUAL FRAMEWORK Introduction In this chapter, I elaborate a conceptual framework for the ana- lysis of underdevelopment in the Sahel which provides a context for the more restricted analysis of grain stock questions to be presented in subsequent chapters. A number of descriptive and analytical con- cepts, diverging somewhat from standard economic analysis, are intro- duced and developed in their Sahelian context. The focus is on the resource base, international position, and economic-institutional viability of the Sahel countries today, how these factors evolved to their present state and the likely path of their future change. This will be done through the elaboration of a theory of structural systems dependency which incorporates selected observations and generalizations on Sahelian socio-economic life, most of which are widely accepted by those knowledgeable of the re- gion.1 1This elaboration of theory through the use of observable real- ity is in the tradition of historical or dialectical (dynamic) mater- ialism. Materialism stresses that ideas, philosophies, etc., take form.within the influence of (and are determined by) real material conditions. The dialectical (or dynamic) method, according to Gurley [1976, p. 18] makes the following assumptions: 12 13 To be useful this exercise must be capable of pragmatic appli- cations; thus, it must be well grounded in the existing structural and institutional positions of the individual countries and of the region as a whole. This is particularly true given the very narrow range of realistic policy options open to Sahelian leaders. Gerald Helleiner wisely points out that, "the essence of success- ful development policy is knowing what must be assumed constant dur- ing any particular period, what is truely variable."2 In order to follow his prescription we need an analytical base which can provide an adequate description of basic socio-economic conditions and rela- tionships and also give us some clues toward explaining the seemingly ironic situation whereby most foreign assistance in rural development in West Africa has been judged to have failed in achieving desired change while, at the same time, we can point to growing numbers of successful foreign instigated commercial ventures in the region. It must be able to demonstrate the key relationships between a successful food policy-which, it is safe to say, has eluded most West African states-and a successful general economic policy. Finally, we need an analytical system which can practically address specific develop- ment goals. R.H. Green, who has provided many years of pragmatic assistance to the Government of Tanzania, suggests a number of goals, "(1) all things are in constant change; (2) the ultimate source of the change is within the thing or process itself; (3) this source is the struggle of opposites, the contradiction, with- in each thing; (4) this struggle, at nodal points, brings about qualitative changes, or leaps, so that the thing is transformed into something else; and (5) practical-critical activity re- solves the contradictions." ZHelleiner [1976, p. 14]. 14 beginning with an explicit value preference for movement toward greater distributional equity, which I will adopt as a reasonable set with which to evaluate the applied usefulness of the analytical model to follow: 1. Attainment of a radically less unequal income distribution eliminating the most glaring urban/rural, educational level, and class anomalies; 2. Sustained progress toward ending absolute poverty in per- sonal consumption capacity and access to public services; 3. An integrated assault on entrenched rural poverty; 4. Primary attention to identifying and meeting mass needs, including consumption goods and public services; 5. Attainment of a better balance between national and sectoral production and use structures; 6. Greater flexibility in resource use based on greater know- ledge (itself the most flexible of resources) and a great- er coherence between domestic supply (production) and de- mand (use); 7. A more defensible international economic position based on higher national economic integration and flexibility, providing greater power to secure less unequal bargains; 8. A ratio of domestic to foreign investments (in personnel, knowledge and finances) adequate to ensure effective na- tional control is maintained.3 The analysis which follows makes use of the observations and analytical procedures of a number of noneconomic disciplines as well as various schools of thought within economics. They are called upon for the conceptual insight and richness of detail which they bring to the structural theory of domination/dependency. Depicting alternate theories in rigid terms so as to provide easy targets for academic demolition does not meet a reasonable criterion of practical utility. 33.3. Green [1975, pp. 117-118]. 15 Neither a demonstration of the lack of relevance of neoclassical macro-economics for sub-Saharan Africa nor an analysis of Marx's failure to foresee peasant revolutions against world capitalism would have useful significance for the government officials charged with operating a Sahelian grain stock system. A Political-Economy Model of Center-Periphery Domination and Dependency The most striking feature of the international political- economic system is pervasive inequality in the distribution, control and ultimate consumption of human and physical resources. Since these levels of inequality clearly cannot be attributed simply to the distribution of physical wealth we must look to the dynamic in- teraction of various sectors or components of this international sys- tem for some explanation. It is assumed that various shifting alli- ances or coalitions of components in the system over time gain differ- ential access to power and the derived benefits which result from domination over other components. Specifically, nation-states and subnational groups are in center-periphery domination and dependency hierarchies such that some states and subnational groups coexist in relations of harmony of'interest and others in conflict of’interest. It is the operation of these asymmetrical relationships which pro- vides the dynamic (dialectical) drive or motor to the system as a whole. What is proposed is a political-economy model of center- periphery domination which provides specific definition for terms such as "underdevelopment" and "dependency" and which hopefully provides an accurate and useful descriptive framework for the Sahel 16 as a component of a larger West African region, which in turn, is structurally integrated into the international system. Definitions and acknowledgement of historical origins are the first orders of business. Model Definitions A.political economy model is one which includes as endogenous a range of political and social factors which are normally held to be outside, or exogenous to, more narrowly defined economic models. This is in the tradition of the classical political economy of Smith, Ricardo and Marx before the narrowing introduced by Austrian neo-classicists and the term is used largely as a warning that a wider range of variables is being explicitly employed.4 The terms center and periphery apply to groupings within the international economic system, particularly nation-states and sub- national population groups. Center and periphery are somewhat 4There are, of course, many important implications which stem from this difference in approach. While beyond the scope of this study, it is possible to link these theoretical differences to sig- nificant divergence in underlying value theory and to elaborate the theoretical havoc introduced into neo-classical economic theory through modification of assumptions such as the exogenous determin- ation of demand. Dobb [1973, p. 25] pointed up the implications of the clear non-neutrality of economic theory stressing that: "The form and angle of generalization, according to what it selects for emphasis and what it casts into shadow, cannot be held without influence, not only upon human attitudes and beliefs, and hence upon social activity (e.g., whether 'social engineering' is attempted or radical institutional change), but upon the intellectual diagnosis of particular social and economic problems." We will see later on, for example, the difficulty of applying con- ventional theory dominated by "exogenously determined demand" (i.e., "producing for the market") to the traditional Sahelian rural eco- nomy where peasants are producing for use (subsistence) rather than for the market. l7 arbitrary groupings useful for theoretical simplification and model application. As such they may range from discrete bivariate group- ings (i.e., all nations or all population groups within a nation combined into two groups) to a continuous function ranging from most central to most peripheral. The terms are useful because they con- note two variables important to the development of the model, the idea of domination and the idea of this process occurring over space. The normal meaning of dbmination, a disproportionate directed flow of determining or guiding influence among units, is employed here. The model here is a sub-type of dominance systems which can be elabor- ated in personal or international terms. Two of the descriptive terms which stem from the center- periphery analysis are the related ideas of underdevelopment and dependency. Underdevelopment represents two things: (1) the conven- tional description of the less materially advanced, more powerless countries or subnational groups within a country, and (2) the idea that underdevelopment is a product or state in system interaction. This means that integration into the larger system implies change from a traditional, isolated society (or undevelopment as A.G. Hopkins suggests)5 to an underdeveloped position within the larger system and, further, it represents a shift in thinking from the con- ventional view that underdevelopment is a stage on a unidirectional path toward attainable development. As Alain de Janvry puts it: Underdevelopment cannot be treated apart from development if backward areas or countries are related by the market to the advanced areas or countries. In fact, within the world capi- talist system, a theory of underdevelopment and rural poverty 5A.c. Hopkins [1976, p. 32]. 18 needs to be a theory of economic space which can explain how the contradictions of development in certain areas transform, in other areas, traditional societies into underdeveloped ones.6 This does not preclude the movement of individual peripheral groups or states toward more central or developed (dominant) posi- tions but it does imply that the continued stratification of groups and states is a precondition for system maintenance. Dependence refers to that set of subordinate model interaction relationships which apply to the peripheral groups and states within the system. Detailed examples of dependent relationships in West Africa will be presented below as well as several further definitions as part of the specification of the model. Underdevelopment is descriptive of the state of the periphery; dependence refers more to the process of interaction between the periphery and the center. Historical Note on the Center-Periphery Model The use of the center-periphery terminology and some of its specific ideas have their modern origins in the work of Raul Prebisch who employed them in the context of Latin America.7 These ideas were further developed by writers such as Frank and Cardoso and the work of de Janvry8 represents a comprehensive review and updating of the Latin American case. It is important to note that the model must 6Alain de Janvry [1975, p. 490]. 7United Nations, Economic Commission for Latin America (ECLA) [1951]. An interesting statistical updating and elaboration is con- tained in Anibal Pinto and Jan Knakal [1973, pp. 34-89]. 8Frank [1967], Cardoso [1972] and de Janvry [1975]. A review of the Latin American "dependency literature" is contained in Chilcote [1974]. 19 be specified for different parts of the underdeveloped world given the widely differing ecological, demographic and institutional pat- terns which prevail across these regions. For example, the extensive concern in Latin America with land tenure arrangements and the exis- tence of Latifundia is not relevant in West Africa with the excep- tion of the plantation areas of a few coastal countries such as Liberia and the Ivory Coast. There are, however, close similarities between the land tenure problems in many Latin American and other African countires such as Ethiopia. For West Africa the exploration of the center-periphery model has been dominated by Samir Amdn and his followers.9 Much of the work in East Africa has centered on and been conducted in Tanzania 10 All of these works stress or the former East African Community. adaptation of center-periphery ideas to the particular historical, resource, demographic, political, and economic patterns of various parts of the African continent. The general model presented in this chapter draws heavily on the work of Johan Galtung, Norwegian conflict theorist.11 His 9Among his many publications, see particularly Amin [1970, 1972 and 1973], a very useful series of case studies be edited [Amin, 1975] and one of his most recent works [Amin, 1976]. 10See, for example, Green [1975], Brett [1973] and Cliffe and Saul [1973]. 11Galtung [1972]. His neutral use of the term "imperialism" applied to the center periphery domination system will not be followed here due to the impossibility of its dissociation from overwhelming ideological connotations. Galtung, while demon- strating the neutrality of the domination pattern, fully acknow- ledges his indebtedness to Lenin for major ideas incorporated into the model structure. See Lenin [1967] and Gurley [1976] for a very lucid overview of Lenin's work. 20 presentation has been the clearest and most ideologically neutral, stressing that the domination patterns explored apply, with some modifications, to socialist as well as capitalist variants. Definition of Center-Periphery Domination In its most general statement, center-periphery domination is "a system that splits up collectivities and relates some of the parts to each other in harmony of interest, and other parts in re- "12 Conflict of interest can lations of. . . conflict of interest. occur when there is a gap in the "real interests" of system compo- nents. These real interests can be summarized as living conditions (LC) which can be made up of indicators such as income, material standard of living as well as by notions such as quality of life, independence, etc. There is coanict of'interest when system com- ponents are coupled together such that the LC gap is increasing; there is no conflict, or harmony of'interest, when the components are interacting in such a way that the LC gap is decreasing.13 'lzGaltung [1972, p. 94]. 13Galtung immediately adds a number of cautionary notes and qualifications to this very simple schema. First, system components must interact in order for there to be a conflict of interest. Second, the theory, at this level, describes conflict of interest (not goals) among system.components (not individual actors). Third, he notes the logical problem of a constant LC gap and states that this could represent a conflict if wide enough. Fourth, the LC gap has two dimensions, its initial absolute width and whether it is decreasing or increasing. This permits more detailed theoretical presentations to talk about degrees of conflict and harmony of in- terest, static views of absolute gaps at various points in time and the more important dynamic changes in the LC gap over time. Fifth, a value preference for equality is internalized in the model since a system interaction structure which increases the LC gap (creates more inequality) is not in the interest of the weaker party. 21 Using these building blocks the simple case of a two nation system can be defined. A domination system . . . is a relation between a Center and a Periphery nation so that (1) there is harmony of interest between the center (group) in the Center nation and the center (group) in the Periphery nation; (2) there is more conflict of interest within the Periphery nation than within the Center nation; and (3) there is conflict of interest between the periphery (group) in the Centeianation and the periphery (group) in the Periphery na- tion. This simple two nation case can be represented diagramatically which reveals further assumed elements of system composition and in- teraction. Figure II-l represents this case and summarizes a num- ber of important points: 1. The center socio-economic group of the Center is linked to the center socio-economic group of the Periphery through a harmony of interest. There is a conflict of interest within each of the two nations. We assume that the LC gap even in the Center nation is, at best, constant. A key point is that there is more conflict of interest in the Periphery nation (dotted line within circle) represented by a wide and increasing LC gap,15 than in the Center nation where redistribution measures keep the LC gap approximately constant. There is structural conflict of interest between the two Peri- phery socio-economic groups (dotted line). The center of the Center grows faster (larger area on circle) than the center of the Periphery which serves as a conveyor belt for a net transfer of resources. Alliance formation is avoided between the two peripheries as the Center nation becomes more cohesive and the Periphery nation less so. 14Galtung [1972, p. 97]. 15Irma Adelman and Cynthia Morris [1973] provide very convinc- ing evidence for this assertion in their multinational statistical study, Economic Growth and Social Equity in DevelopingyCountries. 22 onHuosafiuom 23 7. There is, therefore, a net conflict of intereSt between the two nations (not represented) which is often referred to, but this should not obscure the basic fact that center-periphery domination is a combination of intra- and inter-national re- lations. The "bridgehead" relationship between the two cen- ters with a commonality of interest is necessary, by defini- tion, for the operation of the model. Mechanisms of Domination Galtung proposes two general mechanisms by which the Center dominates the Periphery and which also apply to intra-national re- lationships. The first is the vertical interaction relationship in which vertical international exchange is considered to be unequal if the interaction is cummulatively asymmetric in terms of what the two parties get out of it. There are two indicators which can be used to examine the symmetry or equality of system vertical exchange; they are inter-component effects and intra-component effects. This is illustrated in the case of economic interaction between two na- tions (components) in an international interaction "budget", shown in Figure II-Z. In this example, "Underdeveloped B" is exporting raw materials in return for manufactured goods from "Developed A". Because we are examining both inter- and intra-national effects we can see that, even if the terms of trade between raw materials and goods were equal and the negative by-products (pollution, depletion,“ internal exploitation) could be considered equivalent for the two nations, the overall relationship is still asymmetric since A gains positive benefits from spin-off or multiplicative industrial or tech- nological development while B gains few equivalent benefits from its importation of manufactured goods. This tendency is visible in many actual trade relationships. First, there is the omnipresent gap 24 Hmunbm ZOHHUfiumwoz\u=o wafinuoz mmoou uaosnoam>oa warehoumz no oauqu monouomwonmx HmuuumoooH 3mm o>fiuamom\aH Amaoamv Amzoahv muoommm muoomwm muoommm muoowwm HgOflumfllmku—HH HooowuooluoucH HoaowumclmuuoH amoOfiumcluouaH maoauoz moaoao>omummas um maofiumz oomoao>on n< oumum\aoauoou«n 25 in levels of'product transformation with more advanced countries generally performing the higher levels of transformation. Secondly, since LDC economies, by definition, are less functionally integrated, they would, in general, benefit from a lower growth multiplier effect from imports. As we will see, these conditions certainly hold for Sahelian and West African trade patterns. The second mechanism of domination is the feudal interaction structure. This is a general principle by which boundaries, bar- riers, institutions, geography, transportation patterns, etc., tend to structurally restrict and channel system interaction into limited hierarchical patterns. Very crudely put, these patterns tend to promote inequality through the structural functioning of the prin- ciple of "divide and rule". This can be illustrated in the diagrams in Figure II—3. Type A represents, for example, the general rela- tionship between three colonial metropoles and their captive peri- pheral countries. This relationship is characterized by vertical relationships between the Centers and their Peripheries, a lack of direct interaction among the Peripheries linked to one Center and among all the Peripheries, and only indirect Periphery interac- tion with other Centers and Peripheries by way of respective Centers. (This, in an overly simplistic way, could represent the historical interaction of French, German and English colonies in West Africa.) Type B could be thought of as representing the classical dendritic marketing and transportation patterns which existed in colonial West Africa. Areas of primary product production in two colonial areas (the pp1 and ppz) are linked by largely independent channels to the colonial metropole (cC) through regional market towns 26 mmmafiobmem ZOHHUm «onH«uoouun :quS: no monsoon .+ owuwauuum +, II+ “momma newcumuu .mlud ouawauuoou pew musoeevooea husuo mamas» concessions“ one vuaoao>ov noeuo ouszou anon» no huuaaou uaosooumnuu no oaonouuafl nnowuouuaouuo floccuuaauouan .n heaped“: .u neuron ueoao>oz oofiuouwuuouo>nn acuuooauaouo>wn saw: noucanmau afloueuax ouoeo««~< .H Honouuqeuouan .n A:o>uoaouwoumev A:auoeouuooo¢:v :HuUd—vflfl= :Gufluuvaz: 006600 Hfluflflg o< aoanuwuo> Haouuuuom .HHH «do: Hosea nonuo Homoeom dozen .auuuu< .o caucuses 0. v chaos .uuooo huo>H dosamucoa .aoquu< .m mourn x 4 Humane .ouvan can. ouuou .aaouuuoueoz .< uoH amxo nuance .HH announce r. onuouooo oaaouoau announce vain“ . oaom . . . 030m «7 canon canon. honoreoaov ea unease ooaoocoamvaa eouuanwuuoqoou deceased nvuusou «to. sum: nowadauouoaauu coauaEHOuueauu nouuosuouucouu aoauaauouaauuu Hawcoaoonooa mme5Ho>ohmwuoam unsecuuoao>om huoaoauaao>u Hafiaoaootouz canon buo> no noduoouwo unuuauu .H such show such luck oaouunm «unavoEuoucH ououooauouau aeouuxm souuououuu< uo o~o>oa ozouuo> u. nouumwuouowuoeo new N dose: a «one! nouewuuu> aouohm oouooaom onHou no museum :ouaoanouluuom uo« coauoooom: mouuuuou omaamcan osoaouuvcu :aaov eon: nouuuunowuo swan: houdunauuo>coo honouuou uuauoua no uuoexu :uooauuo>au now ouuaaau room: aaouaaox [r V ulduu Annuioun .uaoau .uaolhoHeeoen and: peace Saunas Hoodeououown «nanoqoo owaueucx macaqueuoa humaouuo unmanned nmaaoaou on 833 “ ounouooiem xoo~a 0» some ouuowum 0. v unusuao v >~owamu ao~o>oa counceuofiuuoa .quieao cause so now: >~o>uuua .uaouuoeaw uoox layman huuomnoeo Adana-non louoau huuaesuuunxv unwuuoaauen Addaeowuoaoh loan». moonstone :uuoou «mono: “ sauna Hana» .aonu: nu :naooum huocnauoe: nu“: maouw m:«~:¢ v :o>wuouuu«euao<: moons douse .aanuo aw :uiaouu Hogans: saw: eaoum unuuam auounuuo unoleo~o>oo uo auooflonm .4 nodawaua> uaaoeoom alodnoum deacon .~ one“ .H ao«unu«aoau= .u nouuaoopu .N unusuawa .H aouuuuno«uo kuoaoo .n noauaauou Auowuov moon» ouaoeoooto«oom .4 uoanqaus> monounoonouoom mongooou ~u3uom .n nouuuoa Huouunaom .N monouun< .n ouguaaOg Aqauouau .0 .> .>H such oauuuxm such «unavoauouan anon snowmanuouau anon oluuuxu N some: H Home! nouuuuouuu< «o ndo>ua coonuw> an auuuaquououuoeo on. noun-«un> Ina-hm wouooHom zo~sna noloa donuaoo was nonuoeuuuunaa “mood nouoaona .aonnoa Iaaouoo ona>qna nouonncanm aowuaucunno nonan .onanan unann0isu «no: Sonnoz aouuooununno fiasco n w pagan: eouuonumanafloa House a« aoaoaoauuun« .couunaalsouw oua>ana ed hueoqowuuo ounoflonm eouuwueouno coon: .vnwauso spanned pen: nevusnunumuo «nuances: hno> use one! nuns use noaoum Hauueoo amen nonna uno- . can: sows: hmoHoaeoou 4“ coonaon hwnoao Hanan Aaaowuownq> hoary Illlllt handgun nooouonm Aoo>a~ueo an onouunoumuv ono>omv noaan mango was nunoeo .ucoaanavo Houneuo .aauuimu cannons“ OOUH-agflh HIHOUQZ om aunoanm .N Anooou Hangman \nolaueOUV ounoean .H announce moans .u uo>nnaouaw mo uuuouuo can acnuuonna .e cabanaooad dunnouuaueoz .n canon ouoaxhnoawm .N noun Ildnlooua Annunoni encode can guano: .H uo>uuaooeu .o no>ah no>wu confine nenunh .0 eugenoeno uqsoaooo ouo>nne lacuna no oua>unm sod ano> A .+ can: ano> we aownonueooaou .n Inch lnom Inch anon oaunuxu «unwousnuuam ouowvoanoueu uaonuxm coaunuonmu< mo ono>on unanna> no nuwnaunouownonu can N Home: A some: nounanno> Banana oouooqom Zeuhczzommzawux unoaowucma one aneonoaso o>uuoounoo mo one: no osmium anaconda maununxu no cause .ouunfianuaoooo on: caucusem now hun>nuo=vona Hunonanoqnua meanwonocw cu muscun .coquaaoacoa Hmaowume .mumnnouus uaooH on wennouoauoeqa nnoo II on=u~=UIo=oa onunou manna nuoxnoa manor non oouonnunu mneunoazo oua>unn on mosey sononnio o>e~oau .vouqaonunoo knead: nouuoovonn unoixo .noaun 4 names on: .oumunonal nan .naoaanavo Hauuaou uaaunoaflu announce mono .e cowuaowuwono>un .n uaaenao Honauusounw< .N unseen coma .H nouooo dwnauunuwnm< .U ong30u peck nauuou ununuooocn .n anon gmhuxm anon ouunooanouen anon Uufld‘OQ—HUUGH anon oaonunu N Homo: H dove: nounuuonuu< no oHo>oA asouna> no ouwuannouoanaeo one uoaauuno> souuam monocuom ZOHH<=fl0hmz<¢B A<¢=Ho=th Ano>H .ommmfimImoeaou .moenou .mamnu .manflmu .aoonmamu .eueom mooofioeHm .mnan .nmmn ".u.a .oonmeanmmz .monmm eneonw new .uoooonm powwow nom .aofiumaonmm “moand 38mm vanes .nmmH ”womaom II Nam nee NHH uaa New mownmm umoz «0 neoonom mm Honmm m.~ n.~ oeN o~m.e~ coo.mmn new.n eunnn< new: neeon n.n n.~ oem oem.m ooa.em nee.n eeunnL< new: no nomaamamm o.m m.~ OHN omo.ma cam.Hn «Na wenomnz e.ou ~.~ own oee.~ oom.e~ ~e~.m nmeem neeon .ou n.~ on one oo~.n enu muno> name: .n- n.~ own oen.n con.e can nemeeem .HI ~.~ QOH one coc.e mo~.H nowwz .q H.N com omN oom.H Hmo.H maemufinomz .H H.~ em ohm ooe.n oc~.H “an: .NI m.H on w cum m ooa.m ew~.H mono muammo\mzu coaumanaom mnuoeen Amuse neonnnna conev nooov nooo «any ANV manna susonu onwemo\mzu mzu aoqnmaonom mon< knuenoo mean .annmao mam mzo nz< mzo .zonn 5&3.) \ ' \.,::..\ * IM‘,I so r . ....... [.11n ..\....I. 1. . . .1 ago mug! ayes. . 7 ix vamp—O 45:02.": DZ< mmZON QGOJOOM m0fi<2 unmr .menan_ owa .o m «a ma mu me am eonnoomonm e an um um co mm moum>wnaoo omn< “mo neoonom "meono .n n ma mm on mm on eoHanaaoe neoonom .w oaoN Honmm one .N mm me am mm m II ewH:H:UIoammsm ma hm Hm mm on mm emwaonmm II II on an no es emnonmm meow he unmonom "mono mama .H muHo> none: Homoeom mono flaw: nownz maemnfinomz moanuaoou %MHZDCU moHH¢ANM QHH%ZHHmM NIHHH wands 57 76 percent in Senegal, to 85 percent in Mauritania, 89 percent in Upper Volta and a high of 91 percent for Chad, Mali and Niger.2 These percentages are, of course, closely related to rates of urbanization to be discussed in the next chapter. Further, as our model predicts, the estimated relative returns to agricultural labor in contrast to other economic activity are much lower in the Sahel than they are in most other developing countries and certainly less than the relative returns in Center countries as shown in Table III-3. These relationships are also evident in estimates of the sec- toral attribution of Sahelian Gross Domestic Product as shown in Table III-4. In that table, we see that in Senegal, for example, 30 percent of GDP comes from agriculture (which contains 75 percent of the labor force), etc. Further, we note the major influence of mining in the Mauritanian economy, the large proportion of "value added" in the commercial and transportation/communications sectors and the major role of the government sector in all of the economies. It should also be noted that the mining sector percentages are most likely significantly higher in 1977 due to Mauritania's doubling of iron-ore exports, to Senegal's increased export of rock phosphates which enjoyed a brief but sharp price rise in 1974, and to increases in Niger's exports of uranium ore. Finally, these GDP figures give only rough estimates of actual economic activity and the estimates 2U.S. Agency for International Development [1975a, p. 6]. AGRICULTURAL INCOME PER CAPITA AS A PERCENTAGE 58 TABLE III-3 OF NONAGRICULTURAL INCOME PER CAPITA FOR SELECTED COUNTRIES Countries Years 1960/62 1964/66 1969/71 ‘Mali 8 8 8 Mauritania ll 8 9 Niger 10 12 12 Senegal 11 13 11 Tchad 7 8 11 Upper Volta ll 13 10 Average of: Underdeveloped countries 28 27 27 Industrialized countries 54 56 55 With: France 37 37 39 United States 61 66 73 Great Britain 103 103 108 SOURCE: P.L. Iacoacci and E.F. Szczepanik, "Income Inequalities Between Agriculture and Other Economic Activities at Different Levels of Development," UN/FAO/PAD, draft document, 1975; quoted in UN/FAO [1976a, Vol. 1, p. 2]. 59 .Hamlqm .mn no moanmu aonw mm>finmo .H .Ho> .moan. onom noauo N.OH m.ma m.m m.oa H.mH m.¢H omeomoo mam nonnmnumH:HEo¢ o.m m.m m.~ o.m N.q o.a monoaaoo new nonumnnommeona m.na n.a~ m.ma m.n H.mH w.nH oonofifioo m.q H.~ o.~ o.n H.m m.m nouuonnumnoo «.ma m.m~ ~.~ c.m m.ma m.w innumooeH II ¢.~ H.o N.Hm II II mafia“: m.e¢ ~.om H.5m e.mm H.mc m.cm mnnnaoownw< muHo> none: Homoeom nowfiz resonanomz was: mono moanuanoo mnonoom $nu .eenani onnme "moeeom ON em a mu cm ma munanH mm co on mm on «a munonxm oomnu neoonon mm muooooni Hanananoanm< om mm om mm on om munoiEH m aw ma on ma «H mnnoexm mad ncoonon mm oomne ma cm «a ma NH om mononeH e we a mm nn e eunoexn Acoaaaan mnao> none: Hmwmnom nomnz wwemufinsmz «Hm: pogo mofinunoou nenzaoo Ammem nm .nnumemn .mmu< q<=zz< .nneen ne4 :Onuoo ouoooone unaooe moo monsoon onHo> noon: Howooom nowfiz onamnnnooz «do: peso ooflnneooo hnowouoo enumemn .mneomnn no Amno Hmomxm cIHHH mum<fi 63 pattern of dependence on one crop or raw material for total export earnings: Chad, 79 percent on cotton; Mauritania, 88 percent on iron-ore; Senegal and Niger, 58 percent and 60 percent, respectively, on peanuts. Mali and Upper Volta are only somewhat more diversi- fied with 77 percent and 82 percent of export earnings respectively coming from similar proportions in each country of peanuts, cotton and livestock on the hoof. Livestock export figures are notoriously unreliable and underestimated, generally due to a large amount of smuggling to avoid paying taxes. Our description of Sahelian exports certainly would not be com- plete if we did not include what, for several countries at least, is their most valuable export, namely, human labor. Labor migration, whether of a voluntary or involuntary nature, has played a major role in shaping the economic history of West Africa. This includes various periods of internal and external slavery, colonial forced labor migration (which was officially abolished in French areas only in 1945), internal rural-urban migration of the classic variety, and seasonal or permanent international migration from the interior coun- tries primarily to the plantations and cities of coastal West Afri- can countries. As one illustration of the magnitude of the latter flows, it is estimated that perhaps one third of Upper Volta's ac- tive male labor force is in permanent or temproary foreign residence, primarily in the Ivory Coast and Ghana.5 This provides a very im- portant component in the center-periphery development of the region. Voltaic workers in the Ivory Coast (in this case, the center of the 5For a thorough exploration of this phenomenon in a center- periphery framework, see Gregory [1974]. 64 Periphery) receive wages, much of which go to Upper Volta as sub- stantial remittances. But these workers, however, do not have easy access to the economic growth of the "Ivorian miracle" due to the difficulty they have in gaining effective ownership rights to land and property. The economic impact of remitted wealth in Upper VOlta, while not negligible in some instances, has not proven to be of much overall importance as witnessed by the secular stagnation of the Voltaic economy. These factors are among those which caused Samir Amin to harshly categorize Mali, Upper Volta and Niger as "labor-reserve economies", partly because this was one of the chief roles they played in the colonial era of forced-labor and partly due to the current, post-independence realities of massive inter- national migration.6 Another trade relationship which is embodied in the feudal interaction structure of our model is the concentration of an al- ready undiversified trade with a limited number of trading part- ners. This pattern is dramatically confirmed in Table III-7 which contains the percentages of Sahelian exports and imports to and from the leading trade partner and the top three. Similar data are pre- sented, for comparison purposes, for the Ivory Coast, Nigeria, France and the United States. Not surprisingly, France was the leading exporter to every Sahelian country and the top importer for three of them. These data are for 1965, five years after indepen- dence, when preferential trade agreements were still in effect with France; there has been a diversification of trading relationships 6Amin [1973]. 65 ofi>oaoownw my monoHnoSHoz Hz manna mu .nom .oom .>om mo none: Mam: oonnnnad mononnonuoz <2 .eom .oom eoanou emu monoum rouge: m: .nom nooowonoz m2 ooaonm mm acumenn among: an nnez n: eneoeapo mu Homoeom Zm comma mm mooeoo <0 oenno .nom o.oanoom 0mm haouH 8H mnooaaonoq\aonwaom Am onnomnz oz nooou hno>H UH onnowH< Hm “oooo onoonnoa wenoonao .onnoz on» mo oonn< onaoeoom onomno ”MUMDOm no .me .«o ne mu «no .en . none: ens .mo .en me em as .mno .en em nm neweeam nz .oem .mn me mm n< .82 .en em on nomnz «no .m: .en em he ens .mn .n: so an mnemonnaez mmma .omm .en we en zm .mo .on nn on nnez m: .eou e .en .e .eenu< seem unto: .nemn "moneomo .nmnnunom .ee .mn> . menses .mnen. ween "mangoes . mum . Q . oON MHQMH .Moonhmmw Hmofiumfiumum _ mGOHumZ GOUHGD "mousomw en nnn m.nm om.on m.oe o.e~ e.em euno> nose: e um e.emn om.nn s.en m.e~ e.me newpepm on man o.ee oe.en n.nm m.om m.on nemnz nn em m.me on.n~ e.e~ m.nn m.nn eneeennaez en es o.mn on.mn m.on o.nm m.em nnez nen nee m.ee ow.nne n.me o.en n.m~ peso nomosm Houoa Hononoanuaoz Hononoanm mzo nmeonuez A.nnnz many Amamv spouse menses neonnnnz wees mo neoonom noeonuoz nom mad oonmuowoom neoEQoHo>on mo ow< Houoa o mnaa Hooae< HouoH Hooae< owono>< anmmmH knueooo MMHZDOU ammZH UZHQDHUZHV mMMSHHazmmxm Huunbm Hzmzzmw>ow A¢HOH mnaa OH ZOHH< A<=zz< «mlmhaav mozHQ A 5 HH e N m «H ouoomonn HHo OH N o m N a ooaoon H c H OH N N nowsm N e e . H m m onooon new ouoom Aeov Anny Amsv Hmev Ammv Ammv Haoemnom\uonnnzv we we on co on no oHoonou onHo> none: Howoaom nomHz oHnoansmz HHmz pogo ooHnneooo mooHo ooom Amneanznoenm znv mnnenzaoo nn=H MHmHo newHos noev no unoHo>H=ao oHonw on oonno>aoo ononon\ouoomn .oHaou no noon oHoonoo nonno HHo moonHoaHo .nnenmnH amen "someom emn nmn non nmn non men neeone>neee enenw neuon eon new nmn emn men men on neeon NH me on m Ne w onon=n\onoon Houoa m o m m Ne m ouoon nonno e mm mm II II n o>oooou m II II II II II meow NmH cHN mmH HNH NMH NmH oHoonoo Houoe mmn mun can we mun men eeaewnom\eonnnz OH mH N N CH N owns: m NN H «H m m noon: e on o NH oN N oon ouHo> nonn: Howoaom nosz oHnoanooz HHoz mono ooHnncaoo nnnooaaoo mMHHHDQZZOU manMH onHNZszco z¢v NI>H NHmo Hoaea< H<:zz< AmHH MHMHo GOHuoH>ov onomaon onn mH noHuoHno> mo unoHonuoouo .mI< enema .xnnnnnn< "mueaom II Aecm. none: mu. ewe mm. «en ce. on nu. nnm newnnpm II II II II II II on. earn «n. can we. en mm. N on. nmnm nomnz em. mm pII nII eII nII on. we ennnnnnnnz en. era «a. men mu. nm on. man nnnz mu. men nII eII nII eII an. see were .>.o coo: .>.o coo: .>.o coo: .>.o coo: nuance nnnon pone annex anemnom\unnnnz non: nnnanoo nOMU Amzon onenmz mnz mo mHZHHUHhhmoo nz< onBo:nomn ZHH mHm Ohm. moo. 00m. 000. 0 .mm. I .2. 835:8 to I new. 8. N. "616...“. sense So. ”no; i as: 91.8». I 8:88... man “225an no I A __ o .No: stotm 285m Amn.ummvo +28: .1183 26:688.. l l l .l l l L m_nN mmwm mnmm mNmN mmmN morn mmnn mOmn mhmn mvmm miov ( .LINOOO ') NOllOnOOHd WVLOJ. 93 in 1972-73 at the height of the drought, and climb back up to about their 1964-70 average in 1974 and 1975. Two regression lines, calculated for 1951-75 and 1961-75, are plotted on the same graph and demonstrate that we must be very care- ful in our choice of base period when analyzing production levels over time. Slopes of opposite sign such as these characterize many of the crop and country-specific time series analyzed over the two time periods. The negative slope is highly sensitive to changes in data points as the low R2, large standard error for the slope coeffi- cient, and wide variance in the data points indicate. It should also be noted that the "fit" of the 1951 series increases when the data are subjected to natural log transformation, although it is not 2 . 's" in these Circumstances. acceptable practice to compare "R A summary of the more statistically significant regression re- sults10 are contained in Appendix A, Table A-6. For each country there are really two production trends which are of interest, millet/ sorghum and rice. Maize is difficult to analyze in these aggregates because the production is relatively small and, in much of the region, it is grown in small fields directly surrounding or adjacent to the family compound which complicates estimation and can lead to great variations in actual yield. Taking millet/sorghum first, in Appendix A, Table A-6, we can see that estimated production in Mauritania and Chad dropped sharply and fairly continuously from 1961 through 1974. Mali also shows a 10That there is no linear or log trend in the production figures is also "significant" in that it generally indicates no major sys- tematic change in aggregate output of that crop, just wide year to year variation. 94 strong decline in production since 1961 but more recent reports in 1975 and 1976 indicate record crops and problems of surplus dis- posal, particularly for government held supplies.11 For Senegal, Niger and Upper Volta, millet and sorghum trends generally follow the aggregate pattern in Figure IV-l, increasing fairly strongly over the entire 1951-75 period but generally stable or slightly declin- ing on the average since 1964. It should be noted that the rate of increase shown in Figure IV-l for 1951-75 is about 1 percent per year and thus is about half the rate of aggregate increase in total Sahel population which is about 2 percent per annum. Increases in aggregate production are thus more nearly consistent with the rate of rural population growth which may then be a reasonable indicator to use when projecting production trends. Rice production increased overall at a rate about twice that of sorghum and millet. Mali, which produces somewhat under half of all Sahelian rice, showed some decline since 1961 but this is largely due to the higher vulnerability of output from major projects, such as the Office de Niger, to rainfall shortages. Niger, Upper Volta and Senegal, which have all significantly increased controlled irri- gation production of rice, showed strong, continuous rates of produc- tion increase, particularly over the longer time period. 11See, for example, Steedman et a1. [1976]. 95 Yields Trends in yields follow patterns not greatly dissimilar to those in production levels but the figures themselves generally have less absolute significance since yields can vary so widely within countries and by type of input "package" employed. Average yields by crop and by country for the two time series are contained in Appendix A, Table A-5 and the most significant regression results are in Appendix A, Table A-6. Again reflecting the impact of the drought years, yield trends for the coarse grains and maize decline sharply over the 1961-75 period. Yields of rice--some of which is under irrigation--showed much more resistance to drought and fairly strong increases in yields were recorded in Niger and Upper Volta. As in other semi-arid lands, weather variation, particularly rainfall variation, is strongly associated with Sahelian cereal yields. Where agricultural data systems are very weak, it may be that certain statistical relationships among more readily measura- ble natural phenomena, such as rainfall and river flows, will pro- vide improved prediction and forecasting capabilities which can be very important to a grain stock system. In an attempt to examine this for the Sahel, average annual rainfall figures for Upper Volta for 1951-75 were regressed on and correlated with both production levels and yields for all cereals and cereal averages. This did not produce too many encouraging re- sults, but several relationships are worth noting. First, a strong negative linear relationship was observed between Bob—01013350 rain- fall levels and total annual Upper Volta rice production, much of 96 which comes from the "Bobo" region. (This result has also been reported for irrigated rice production levels in Niger and is hypo- thetically attributed by other authors to decreased radiant energy caused by greater cloud cover.) Also a strong positive relationship was found over the 1961-75 period between averaged Upper Volta annual rainfall and sorghum yields. Even with only fragmentary results such as these, inquiries along these lines seem to offer potential for re- latively cheap early warning crop forecasting systems. 12This is confirmed in the recently published work by Davy et a1. [1976]. A number of general conclusions can be drawn from this important baseline study: 1. Weather and resulting crop response is extremely localized in West Africa within the longer-run averages of each broad ecolog- ical belt. Since much of the rainfall in the region is convec- tional in nature and highly specific in incidence there can be great variability among close-by rainfall stations within the same climatic zone. Therefore, rainfall averaging should in- clude readings from as many stations as possible stratified by ecological zone. 2. Even with these data problems very high correlation coefficients have been computed in the Sahel between averaged national and sub-national rainfall figures and the yields of some crops, with these results often as precise as those obtainable from experi- ment station data. For example, as a rough rule of thumb for Senegal and Niger, the authors suggest that when comparing yields between succeeding seasons, for every millimeter in- crease in rainfall there may be 1.5 to 2.0 kg per Ha'1 increase in millet/sorghum yield per hectare. 3. In addition, large and small river flows, reflecting rainfall and runoff in their respective catchment basins and a close relationship between maximum and mean annual flows, also pro- vide promising means of assessing future average yields both on national and sub-national bases. 4. These statistical relationships must be worked out for each na- tion or sub-national region based on an adequate number of ob- servations and an in-depth understanding of the agronomic rela- tionships at play in that particular environment. For an example of a national-level weather prediction model which has been tried with encouraging results for Niger, see LaBonne and Garzia [1975]. For further introduction also see Thompson [1975] and Kelly [1975]. 97 Correlation of Yield and Production Levels Variation We are interested in the correlation of production and yield variations to assess the extent to which temporarily and chronically deficit regions or countries could be theoretically supplied from surpluses in neighboring regions or countries. Highly correlated total production by crop could indicate that this "venting of sur- plus" to mutual advantage would be more difficult to achieve. Table IV-7 provides an illustrative summary of the correlation analyses focusing on the relationship in total cereal production by country over both the 1951-75 and 1961-75 periods. (More detailed versions of these results are contained in Appendix A, Tables A-7 and A-8.) What is of particular interest is the correlation of pro- duction levels in adjoining countries in the region where it would be possible to conduct inter-grain board exchanges under the right supply conditions. The data for Chad and Mauritania refer only to 1961-75. During that period both countries showed fairly strong secular declines in production levels and this results in a high "r" value. This also explains the high coefficient with Mali for both countries. Chad production levels are not particularly strongly associated with those in neighboring Niger except with Niger sorghum which also showed a consistent secular decline. Mauritanian production was highly correlated with that of Mali but not of Senegal. For the other four countries, we can examine these correlations over both time series. For Mali, aside from the relationships men- tioned above, only its relationship with Niger production levels is significant over the two time periods. Cereal production in 98 TABLE IV—7 PEARSON CORRELATION COEFFICIENTS FOR TOTAL CEREAL PRODUCTION BY SAHELIAN COUNTRY, 1951-75 AND 1961-75 (COEFFICIENTS WITH "*" SIGNIFICANT AT .05 OR BETTER 1951-75 Mali Niger Senegal Niger .35* Senegal .17 .43* Upper Volta .07 .57* .71* 1961-75 Chad Mauri- Mali Niger Senegal tania Mauritania .62* Mali .73* .75* Niger .42 .64* .63* Senegal -.06 .30 .26 .26 Upper VOlta .13 .23 .00 .41 .41 SOURCE: Appendix, Tables A97, Ar8. 99 Senegal, Upper Volta and Niger is fairly strongly correlated over the 1951-75 period but less so over the drought-dominated 1961-75 period. As Figure IV-l showed, all four countries showed a sharp drop in total production in 1972 and 1973 but the levels were lowest in Niger and Upper VOlta in 1973, Senegal in 1972 and Mali was equally low in both years. Although not explored in detail, patterns of intra-country varia- tion also show similar contradictions in the correlation of total output by region. In summary, it is probably safe to say that in most years severe shortages will probably be fairly localized and it should be possible to shift portions of relative surpluses to ad- joining regions and countries (assuming, of course, that transport costs do not far exceed the spread in relevant prices across the two areas). It should also be clear that in unusually had years it may not be possible to count on much assistance in this respect from neighboring production zones, particularly if two bad years in a row hit the region as a whole, such as occurred in 1972 and 1973. Grain Imports For West Africa as a whole the 19603 produced sharp and con- sistent patterns of increase in food grain imports largely due to the twin phenomena of rapid urbanization and food policies which have provided limited production incentives to farmers. Average total food grain imports in 1969-71 were 75 percent higher than the average import levels for 1961-65. Senegal and Mauritania alone, out of the 21 countries in IBRD's west Africa region, accounted for an average of 27 percent of total regional grain imports over the 100 1969-71 period. The decline in self-sufficiency has been much more rapid in these two countries than the four others in the Sahel. Traditionally wheat and wheat flour have been the major grain imports to the region providing bread and pasta products primarily to a small population in cities and towns. In recent years the amounts of rice and other grains imported (mostly maize and coarse grains) have also grown substantially. The drought period, of course, dramatically increased food imports, much of it food relief which will be examined briefly in the next chapter. Table VI-8 shows some of the principal variables in the Sahel- ian food import situation. First, we can see that in more normal years Senegal and Mauritania dominated food imports, with an aver- age of 81 percent of total regional imports (65 and 16 percent re- spectively). Next, it is clear that wheat and flour have been the dominant import in the four interior countries with rice important to Senegal and Mauritania. Other grains (maize and coarse grains) also made up substantial portions of imports in Mali, Mauritania and Niger. The relative dependency of the two coastal countries is indicated by the fact that average imports for 1969-71 amounted to 78 and 44 percent of estimated domestic production in Mauritania and Senegal respectively. Import figures are also shown for 1974 and reflect the height of massive food relief shipments to the region. Note that the large amounts of donated maize and sorghum push the relative proportion of "other grains" to 54 percent of total imports. It should not be overlooked, however, that rice and wheat imports also increased sub- stantially. SAHELIAN FOOD GRAIN IMPORTS, AVERAGED FOR 1969-71 AND 1974 101 TABLE IV-8 Country Wheat 5 Flour Rice Other Grains Total Percentage of Domestic ('000 m) 2 ('000 MT) 2 ('000 Mr) 2 ('000 Mr) Production Chad 1969-71 10.8 96 - 0 0.4 4 11.2 2 1974 9.0 20 - 0 37.0 80 46.0 8 Mali 1969-71 13.3 44 4.6 15 12.3 41 30.2 3 1974 20.8 9 65.3 28 146.0 63 232.1 27 ' Mauritania 1969-71 16.7 22 19.4 26 39.0 52 75.1 78 1974 19.4 15 32.0 25 75.0 59 126.4 373 Niger 1969-71 5.5 52 0.1 1 5.0 47 10.6 1 1974 10.7 6 0.1 0 176.0 94 186.8 24 Senegal 1969-71 107.8 35 150.9 48 53.6 17 312.3 44 1974 116.7 24 230.5 48 130.0 27 477.2 57 Onggr volta 1969-71 25.5 66 1.8 5 11.5 29 38.8 4 1974 20.8 28 - 0 54.0 72 74.8 6 Sahel Tota1 1969-71 179.6 38 176.8 37 121.8 25 478.2 11 1974 197.4 17 327.9 29 618.0 54 1,143.3 27 SOURCE: various issues of the PAD Trade Yearbook. 102 Food policies dealing with grain imports will be crucial in the years to come, particularly in considering stock policy. There is an increasing awareness by the Sahelian governments of the strong implications for self-sufficiency of trends in urbanization, income- elastic substitution of higher prestige grains for sorghum and millet, and the effects price policies can have on stimulating food produc- tion. Efforts are being made to expand domestic rice and wheat pro- duction and to develop substitute flours with a substantial coarse grain content.13 In addition, several of the Sahelian countries have imposed tight import quotas on items such as rice. So far, however, these efforts have been fragmented and do not address the main issues of international dependency and domestic stagnation which underlie these trends in food production, consumption and imports. These areas will be examined again in later chapters. 13The Technical Institute in Dakar, Senegal, is working on a composit bread made of 30 percent millet and 70 percent wheat. The Government of Senegal has decreed that by 1978 all bread in the country will be of the composite form. Shifting from all wheat to the composite bread will not result in a decline in protein content since millet contains as much protein as wheat flour. However, the composite is not as white in color as bread made from wheat flour. CHAPTER V PRIVATE GRAIN TRADE AND STORAGE This chapter will serve two purposes. First, it will provide further specification of the center-periphery model in the area of food grain marketing and storage. Second, this description of current knowledge of private Sahelian grain trade will help build a transi- tion to the remaining chapters of this study which deal with Sahelian government participation in food grain markets. In doing this, a number of generalizations concerning private grain trade and tradi- tional grain storage practices will be drawn from the literature on West African agricultural marketing. Rural agricultural markets--often misunderstood, romanticized or manipulated on the theoretical level for ideological purposes-- play a vital role in the process which incorporates the Sahel into West African and larger international economic systems. Paralleling the tendency for agronomic research to focus heavily on the develop- ment of export cash crops, much of the research in West African mar- keting has focused on the evacuation of cash crops. The literature on food crop marketing is much more limited and quite diverse, with sharp differences in methodology and approach between studies in the 103 104 French and English traditions. We can, however, draw conclusions which are appropriate to the focus of this study.1 West African food marketing studies can be roughly forced into three groups, two of which follow food marketing convention and, not illogically, focus on urban food-sheds, and a third is a set of anthropological studies which are largely confined to an examination of the marketing process in a village environment. The first group involves food marketing studies conducted in English speaking west Africa. As in other domains, this group is dominated by studies undertaken in Nigeria with those centered on the sorghum/millet areas of Northern Nigeria of particular interest. William.0. Jones' investigation of staple food crop marketing is by far the most well known. He and his colleagues attempted to analyze lTo fully explore the role of agricultural markets in the nature and evolution of west African center-periphery relationships one should include a review of: (1) economic histories of traditional long distance trade and the nature of surplus extraction in precolon- ial Savannah kingdoms, (2) sociological investigations of the impor- tance of ethnic diasporas (such as Lebanese, Hausa, WOlof and Dioula trading clans) and specialized market channel participants (such as traditional market place "landlords and brokers") who facilitate exchange over the "feudal interaction structures" of Chapter 2, (3) the literature in economic geography on the spatial organization and evolution of rural market networks--much of which focuses on the function of rural periodic market systems in west Africa, (4) the area of rural agricultural factor markets which includes land allocation and use studies, labor migration studies, etc., and (5) the role of modernizing institutions such as cooperatives in village-level agricultural marketing, particularly the extensive cooperative movements in Senegal and Niger. For an introduction to these topics, see my previous paper, "The Political Economy of Agricultural Marketing in West Africa," Department of Agricultural Economics, Michigan State University, East Lansing, September 1976 (unpublished working paper). In addition, many of the most important works are carefully annotated in Arditi [1975]. Other bibliographic sources are Hopkins [1973]. Bohannan and Dalton [1962] and Miracle [1970]. 105 the structure and performance of key staple food systems, including flows, exchange levels, personnel and facilities, in the supply areas of four African cities: Enugu and Ibadan, Nigeria; Freetown, Sierra Leone; and Nairobi, Kenya. The principal focus of this work was to examine market price formation through an analysis of variables such as: . . . market information, channels of grade, bulking, trans- porting, preservation and storage, contracts, financing, risk bearing, traders' margins, restraints on trade, weights and measures, quality standards, fees and taxes, and arbitrage among markets.2 Greater detail is available in component studies by Whitney [1968] on Eastern Nigeria and Matti and Atere-Roberts [1968] on Sierra Leone and the related study by Gilbert [1968] on Northern Nigeria. At about the same time as the Jones study was being undertaken, Rolf Gusten [1968] conducted a very interesting study of foodstuff flows in western Nigeria. His work differs from.those mentioned above in that he made no attempt to describe causality or efficiency in the system he was "tracking". Lawson [1967] undertook a more descriptive study of Ghanian food markets at about the same time in the 19608. More recently, Northern Nigeria has been the focus of renewed exploration of food crop marketing with Ejiga's [1972] study of the cowpea trade and 34M. Hays' [1975] examination of the marketing and storage of food grains, a study which is somewhat similar in approach to Jones' earlier work. A second group of studies, those in the francophone countries, are similar in their focus on the provision of urban food supplies 2Jones [1973]. 106 although they tend to be more descriptive and lack some of the theo- retical and analytical sophistication of the anglophone studies. This may be partially compensated for by a greater refinement in descriptions of product flows and in cartographical representation. Couty [1965] examined the production and marketing of millet around the regional capital of Maroura in the northern Camerouns Diamara Department. He devoted substantial attention to the impacts of market prices and their fluctuations on the competition between the cultivation of millet largely for subsistence, and cotton for the export market. SEDES, the large French para-statal research organization has conducted a large number of market research efforts ranging from studies of commercial possibilities in specific commod- ity markets to global assessments of marketing strategies, such as their widely used study in Niger [SEDES, 1963]. Recently it com— pleted a study of the problems of "provisioning" five large cities with food in 1985, given current trends in urban growth and consump— tion (see SEAE [1973]). Le Chou [1966] conducted a descriptive study of commercial trade in the Bouake region of the Ivory Coast and Bollinger [1974] has done the same type of study for all of Upper Volta. The United Nations study on food grains in Chad [Auffet, 1974] which was cited in the previous chapter, contains fairly extensive portions dealing with private marketing as does a very recent study on cereal price structures and marketing mechanisms in Mali [Haik, 1976]. The third group of marketing studies which are of use in con- sidering Sahelian staple foods are anthropological studies focus- ing on the role of markets in the context of the traditional economy. 107 A useful beginning is provided by Bohannen and Dalton [1962] who edited a collection of studies on pre-independence local African markets. Claude Meillassaux [1964] conducted important micro-level research among the Gourou in the Ivory Coast and later edited one of the most useful collections of West African market studies which incorporates many elements of the center-periphery framework used in this study [Meillassoux, 1971]. Finally, and also focusing on the impact of outside capitalism on the traditional economy, is an important study by Claude Raynaut [1973] on millet exchange at the village level in Niger. The studies mentioned here and others have been reviewed to draw summary conclusions on the following topics of direct relevance to the establishment and operation of government grain stocks: grain market structure and Operation, the amount and nature of marketed surplus, grain market efficiency and degree of competition, and the distributional impacts of the operation of grain and other staple food markets. Market Structure and Operation Here we are interested in the disposition over space and time of the total production of food grain. A farmer's production of millet and sorghum can be disposed of in the following ways: 1. Consumption by the farmer and his family which generally involves the majority of the annual crop. This means that the bulk of the crop must be stored at the local level over at least part of one year; 2. Gifts and other forms of traditional non-monetary exchange (such as Zakka, Islamic grain tithes in Northern Nigeria) which gen- erally take place within the village or within ethnic clan and family structures; 108 3. Sales to other village residents (for storage, resale or trans- fer to local markets) and/or other market participants in var— ious types of area markets. Relatively isolated villages are linked both to isolated periodic markets and less isolated ones which are on roads permitting at least some regular vehicular traffic. These periodic markets are then linked to larger bulking markets and to urban markets, both within and outside a given region. Traditional marketing nomenclature can be applied to the wide range and large numbers of participants in west African grain mar- kets. Producers, part-time village traders and transporters often may use donkeys or bicycles to convey grain (and, importantly, a variety of other products) to and from local markets. Rural assemblers bulk quantities of grain collected in villages or rural markets and generally make arrangements with transporters to move these supplies to wholesalers, urban assemblers (who have the capi- tal and ability to specialize in inter-regional shipments), and retailers in larger urban and regional markets. It is important to note that such patterns constantly overlap and are often "short- circuited" reducing the number of participants when, for example, farmers can take their grain directly to a regional market or when travelers may profit on a long trip by carrying a few sacks of grain directly to an urban market. In a large market many of these types of transactions may take place concurrently in the same part of the market. Market periodicity permits both "thin" supply and demand to be concentrated, and allows traveling buyers of farm products and sellers of consumer goods to increase the effective size of their 109 markets. Communication and transportation to and among markets is a major barrier to fuller economic integration. In effect, "islands of competition" (again, in overlapping patterns) are imperfectly linked together by means of communication and transportation fre- quently subject to various types of local oligopolistic control. Neither the government nor private parties regularly make actual market prices available either to farmers or traders. In more iso— lated rural areas transportation is either very costly or is avail- able on a discontinuous basis when, for example, a trader can get a truck to a remote village and obtain surplus production at extremely low prices. The degree to which "cartelization" dominates grain wholesaling in urban areas, as it does other types of longer- distance agricultural trade such as cattle, kola nuts and export crops (when they are not handled by government monopolies) is not fully known and may be difficult to ascertain, particularly when government official prices and other market interventions drive much of the trade from public market places and into private homes and courtyards as has happened in Niger, for example, in recent years. Large wholesalers who operate in major cities do seem to spatially allocate the urban food supply shed among themselves although price competition probably places some rough bounds on such practices. 110 Marketed Surplus The percentage of food grain production that farmers take to market can vary greatly across the Sahel. It varies among regions according to the overall food balance, it varies within regions by farm family composition and size, by ability to store grain (and meet current cash needs by other means), by access to markets, by farm enterprise combination which, in turn, is influenced by relative food and cash crop prices available "at the farm gate", at the near- est market, etc. The rule-of-thumb figures often used that the "average farmer" may sell 10 to 20 percent of his grain production may coincide with national aggregate consumption and production esti- mates but mean little in any particular village or to any strata of farm families. Hays reports, for example, that up to 40 percent of his sample of farm families sold no millet or sorghum at all although they grew both. Given volatile seasonal grain prices the amounts and timing of grain sales by family type can be highly important as we will see below. Hays also found that, in the aggregate, about 80 percent of grain harvested in Northern Nigeria went into village level storage for later consumption, gifts and sales. Gifts in general amounted to about the same proportion of aggregate production as sales but it is difficult to estimate what part of gifts is eventually marketed sur- plus due to complex patterns of purchase and resale, or sale of pre- vious gifts. One result that Hays' data clearly reveals is that larger farmers generally sell disproportionately more grain than smaller farmers and they sell higher percentages of that grain later in the season when they can take advantage of seasonal price 111 increases. As we see below, Raynaut's data for one village in neighboring Niger go far beyond this level of generality in assess- ing the dynamic importance of differential family ability to produce, store and trade grain. Market Efficiengy and Competition This aspect of food grain marketing has received the greatest attention in studies in English-speaking areas reflecting a greater concern to demonstrate and validate certain types of assumed market performance. Market efficiency is the key phrase in these studies. Ignoring the warnings of such critics as A.A. Schmid [1974] who warn that "efficiencies do not exist in nature" and, thus, must be pro- vided detailed, multi-faceted definition, market competition and efficiency are measured through an unidimensional analysis of com- parative market prices over space and time with price differences compared with transportation and storage costs. Hays puts this very directly by stating that he wants to test four hypotheses, that: l. The traditional market structure is basically competitive and fulfills the requirements generally specified as necessary for competitiveness. 2. The margins at each stage of the marketing chain are not greater than marketing costs. 3. The markets for millet and sorghum.are closely interrelated and the intermarket price differentials are not greater than transport and handling costs. 4. Seasonal price increases for millet and sorghum are not greater than the cost of storage.3 3Hays [1975, p. 2]. 112 He, like Jones et al., before him, basically confirms the first two hypotheses in Northern Nigeria and has mixed results on the other two. Spatial arbitrage is seen as inadequate in grain mar- kets when measured by theory's dictate that intermarket price differ- entials should differ only by transport costs. He finds that while different sub-systems are internally competitive they are not effec- tively integrated due to poor information, weak physical and institu- tional infrastructure, risk bearing behavior and an effective seg- mentation of the market. Similarly, temporal price differences are greater than they "should be" but this seems to provide the possi- bility of some distribution of excess profit to market channel par- ticipants at all levels. Thus, he, like Jones, goes on to make re- commendations that the role of government should be restricted to making "facilitating arrangements" to improve market intelligence and roads, to provide improved standard measures and market facilities, to establish a national marketing service, and to concentrate more attention on the farm input delivery systems. Jones also suggested methods of fostering further integration of urban food systems through strengthening the system of bulking markets in supply hin- terlands, and even advocating the creation of commodity exchanges to help facilitate these transactions. Another recommendation was that government should only directly intervene in the market when rapid shifts in demand present unusual opportunities for monopolistic exploitation by private traders. Few would argue with these conclusions and recommendations as far as they go. They simply do not go very far if one is interested in a broader definition of rural development. There is a strong 113 tendency to concentrate attention sequentially on the horizontal pat- terns of market competition at various levels in the market channel and to look at vertical organization and integration in fairly sim- ple terms. This can miss important patterns of concentration and trade restrictions along the dendritic interaction structures which link freer market "islands" of different sizes.4 These spatial and temporal structures are dotted with ad hoc and permanent, traditional and modern, institutions, rules, borders, boundaries and property rights which only the specialized few can master and which are not completely captured in comparisons of inter-market price and cost differentials. The relationship between institutional organization and the arrangement of economic activity over space is of particular importance. A working hypothesis, inspired particularly by Edgar Johnson [1970] is that, for much of West Africa, the organization of space for agricultural production and marketing is not optimal for a modernizing yet independent, more self-sufficient and locally- controlled system. Finally, the most important objection one can make to this approach is that it does not consider, except by assumption, the relative welfare positions over time of different classes of market participants. 4Marvin Miracle [1968], on the other hand, focuses on the very obvious patterns of cartel formation in the vertical structure and operation of west African food systems and may, in fact, overempha- size the 1ocal-level marginal quantity and quality adjustments in food purchases and sales which he sees as further demonstration that the standard assumptions of African market competitiveness are great- ly overstated. He does, however, begin to focus on the key role mar- kets play in the process of capital accumulation (largely through violations of competitive assumptions) and speculates as to whether the process of accumulation might not be accelerated if certain types of reforms were implemented. For further details see this insight- ful article. 114 Distributional Effects of Market Operation Micro-level studies being undertaken in west Africa by agricul- tural economists and practitioners of other disciplines are helping to broaden our knowledge of the functioning of rural systems. Howe ever, the illumination provided is sometimes more constrained than need be because partial equilibrium analyses may hold constant key variables whose movement should be followed over time. In the tradi- tional field of micro-marketing studies this is particularly true. Great imporvements can be made simply by broadening the scope of these studies to give traditional conclusions about the nature of competitive markets some contextual meaning. The main point is not to dispute the empirically testable degree of competitiveness of mar- kets but to examine what happens to the participants who deal in these markets and to the societies which contain these transfer mechanisms. A good example of a more integrated, systematic approach which can realistically comment on distributional impacts is the important study of peasant grain trading in a Niger village by French anthro- pologist, Claude Raynaut [1973]. He supplemented an original study of production patterns and the socio-political organization of the village and its links to the outside world, by recording all internal and external grain transactions by village adults, their timing, and the purposes for which they occurred (supplement to subsistence food stocks, storage for seasonal resale and profit, repayment of tradi- tional credit in kind, purchase for preparation of food for sale to wage laborers and single adults, etc.). He found that competitive markets, responding to traditional seasonal and biologic rhythms, have a very important role in the process of rural class formation and 115 wealth accumulation in a traditional society linked, if only tenuous- 1y, to the modern state and to more advanced capital-dominated factor and product markets. Moreover he found that relatively better-off farm units (those with more advantageous traditional kin obligations and relationships, a larger number of active adults, or access to better land, etc.) are more able to buy and store grain at harvest time, engage in supplementary activities (prepared meals for sale) and sell more grain later in the year. Poorer farmers in order to meet cash needs--head taxes, small consumer items, etc.--have to sell more of their grain at harvest time and earlier in the crop consump- tion year. This can lead to a depletion or exhaustion of farm unit grain stocks and the need to sell labor services for wages (in the local area or in other regions) or for payment in kind in order to meet food needs at higher "hungry season" prices. These effects can be cumulative with more prosperous farmers (or, increasingly outside traders or salaried urban residents) better able to take advantage of modern credit and technical assistance packages to produce cash and food crops with improved technology and a greater use of wage labor. (This process is most advanced in parts of Senegal.) These patterns are simply the rural portion of the center-periphery sys- tem which is transforming West Africa. It should be stressed that these patterns do vary substantially from place to place and that they are, in many cases, only partially visible and may seem to be contradictory. However, we can put efficient markets in some mean- ingful context and stress the need to see agricultural trade, includ- ing food trade particularly as urban food markets "deepen" and expand, 116 as an important element in a larger system of asymetric structural economic relationships with strong distributional consequences. Traditional Private Grain Storage The bulk of West African grain is stored in private facilities either at the farm level or at various levels in food marketing channels. Grain storage is another complex area because of the large variety of ecologically-adapted local storage techniques employed and because of the lack of reliable and detailed evidence concern- ing storage costs, efficiency and losses, particularly at the village level. Hewever, a few generalizations can be drawn from the litera- ture and used as background in this study.5 First, at the village level, farmers store millet or sorghum on the head or threshed, in a variety of home made containers ranging from mud-brick silos to woven straw graineries. As it moves into marketing channels grain is usually bulked into 100 kilogram (220 pound) sacks and is stored in these sacks in a variety of buildings as it moves to larger markets. Private houses and warehouses are used and almost no grain, even in inter-regional shipments, is handled in bulk due to the lack of appropriate facilities and infrastructure and the need to resack the broken bulk for distribution and resale. Grain losses in traditional storage from excess moisture, insects and rodents can vary considerably by region, type of grain, type of storage and stochastic events such as unusually heavy rains, or 5For an introduction to grain storage in West Africa see Abdel- Aziz [1975], Ennis et a1. [1975], Giles [1965], Hall [1970], Hays [1975], Lindblad, Newman and Vinita [1975], Nyanteng [1972], Volun- teers in International Technical Assistance [1976] and various issues of the Bulletin of Grain Technology. 117 attacks of insects and rodents. Storage losses do, however, tend to be higher in wetter, more humid areas and certainly are higher when grain is stored over one or more rainy seasons rather than just with- in the consumption year. The selective use of insecticides on farm granaries and gas fumigation on warehouse grain stocks can greatly reduce insect losses. Moisture losses are cut when grain is stored in relatively small farm granaries and when more sophisticated dry- ing techniques are used on larger bulks of grain. It is not completely meaningful to attempt to compare costs of grain storage at different levels of a highly segmented economy with large subsistence sectors. However, rough estimates do illustrate the cheapness of farm level grain storage and the rapid increase in cost associated with the use of more technologically advanced, large quantity storage systems. Hays estimated the average yearly storage cost in traditional storage (accounting for granary construction by special craftsmen, depreciation and grain losses) per ton of grain to be about $5.30 at 1973 prices.6 This is compared, for example, to modern "low-cost" portable storage facilities at about $30 per ton per year and Wbrld Bank estimates which for the Sahel put modern warehouse operating costs at about $44 per ton per year (or $58 per ton per year if warehouse construction costs are amortized over a 20 year period).7 These figures are somewhat misleading since we are moving from a semi-monetized portion of the economy to a more 6Hays [1975, pp. 33-38]. ZAlternative grain storage cost calculations are explored in Appendix B which accompanies the analysis of the costs of government emergency stock holding in Upper Volta in Chapter VIII. 118 fully monetized one but they do point to the relative ecological effi- ciency and stability of traditional, localized grain production and S torage systems . CHAPTER VI ISSUES IN SAHELIAN GOVERNMENT PARTICIPATION IN GRAIN MARKETING AND STORAGE Introduction Grain marketing and storage in the Sahel are undergoing rapid transformation as government food marketing and storage institutions attempt to replace or supplement private grain trade ostensibly to promote the goals of increased food self-sufficiency, distributional equity and mass mobilization for economic development. However, Sahelian economies are being incorporated into a dominant world eco- nomy which, as we saw in earlier chapters, is producing predictable consequences in a center-periphery structural system, such as very rapid urbanization. This in turn has led to more centralized, capital-intensive distribution mechanisms and structural conflicts which make the attainment of all of the above goals increasingly difficult. Grain Board Tasks When we consider food grain marketing and storage specifically, this part of the political economy can be divided into three major task areas which are a function of the consequences of widely fluc- tuating grain supplies. Assuming a demand for grain which only changes over the long run, short term variation in supply and prices 119 120 produces consequences for various market participants and provides a definition, albeit somewhat arbitrary, for distinct cereals office activities. When conditions are "normal" over much of the country, the cereals office can concentrate on commercial supply activities and improving facilitating infrastructure (in addition to activities necessary to maintain emergency stocks, etc.). When supplies are very short, there are the risks of starvation, severe malnutrition and windfall gains through private liquidation of grain stocks at very high prices. These call for the use of emergency stocks. The oppo— site extreme situation of massive excess supply risks waste and re— sulting producer disincentive effects and should call into action sur- plus disposal actions which can be the bane of stock programs, but to this point have not had to be given much attention in the Sahel. In intermediate cases, the cereals office will probably act most effectively in buying and selling stabilization stocks and attempt- ing to regularize supply or price. These relationships are summar- ized in Figure VI-l which represents in a sense the standard "Mars- hallian cross". This schema, of course, employs a major simplification since these processes take place over time and space, interact with and complement each other, and are linked to larger systems of politi- cal, economic and social relationships. Following the logic of Figure VI—l, we can see that the dividing line between emergency stocks and buffer stocks is necessarily arbitrary. In the Sahel, as we will see, however, a very strong structural and operational line has been drawn between the two functions, particularly in the operation of FAD-backed emergency stock projects. This is due to 121 Grain Grain Negative Important Cereals Supply Prices Consequences Office Actions Starvation, Malnutrition, Emergency Low’ High Stock Release Speculative 1‘ Profiteering Stabilization Sales 4 Improvements in Average Average Market Infrastructure Stabilization Purchases waste, " Surplus High Low’ Producer Disposal Disincentives FIGURE VI-l CEREAL SUPPLY SITUATIONS AND OFFICE ACTIONS CEREAL 122 three factors: the desire of the FAO group to succeed in the effi- cient operation of at least one component of a total food grain marketing and storage system, the lack of an overall integrated food grain policy and the infrastructural and institutional means to carry one out. Three caveats need to be stated at this point. First, changes exogenous to the domain of the cereals office can affect the scope and degree of activities possible for the office to undertake. For example, new physical and biological technologies (e.g., irrigation, improved seed varieties and cultural practices) can reduce aggregate supply variation just as improving commercial pipeline stock holding procedures and infrastructure can make better use of existing supplies and reduce the need for stabilization and emergency stock programs. Second, grain production and exchange relationships are a sub- set of a larger group of production and exchange relationships which, as we have seen, have often structurally and functionally worked to the disadvantage of the rural, surplus-producing population. Thus, to be viable, new grain trade relationships must be consistent with and supported by modifications of the larger set of inter-sectoral, inter-socio-economic group trade and non-trade relationships. If not, seeming reforms may simply fail or may backfire leading to declines in efficiency and an exacerbation of negative distributional conse- quences. Therefore, the larger set of center-periphery relation— ships will help order choices among alternative policy and implemen- tation strategies. As we saw in the concluding portion of Chapter II the basic dichotomy in policy options is between those which cluster around a trend of continued systemic maintenance of increases in 123 center-periphery dependence and those which promote movement towards a more self-reliant, independent system. Stemming from the second, a third caveat is that policy makers do not generally face a simple choice between two distinct packages of opposing policy options. They work in the grey areas of "second best" solutions where methods of judgement decision and trial and error prevail. But there is a pragmatic necessity to choose. As we will see in the next chapter, all the Sahelian countries have cereals offices, all are committed to holding both emergency and buffer stocks of coarse grains (and most have already begun these efforts), all are attempting to directly intervene in cereals mar- kets, all are leery of private traders, all use price controls in cereals marketing, and, most importantly, all are enmeshed in an interrelated system of international and internal relationships which are vitiating their ability to promote their self-proclaimed goals of "food self-sufficiency and autonomous economic growth". All of this is given. The challenge now is to elaborate some of the major issues involved in government participation in grain market- ing and storage and to discuss the likely implications of pursuing alternative courses of action over time. In doing this we must prag- matically synthesize the insights derived from the more theoretical macro-level political-economy model with the insights gained from an analysis of Sahelian government grain marketing and storage experi- ments. However, in order to understand the preliminary responses to the grain crisis in the Sahel in the early 19703 it is necessary to examine the international grain crisis during the same period. 124 World Food Crisis and the FAO Food Security Assistance Scheme The holding of special grain stocks for emergency needs has a long historical tradition in both high and low income countries. The recent catalyst for international attention to grain storage pro- blems was the world grain crisis of 1972-74. World grain stocks reached historical lows, world prices jumped more than threefold with- in a year, and a concerned public in high income nations saw widely publicized reports of drought conditions in Asia and Africa. This contributed to the sense of urgency which surrounded the World Food Conference in 1974. A number of alternative grain stock programs were discussed at the conference but no agreement was reached on the creation of internationally held grain reserves. This led the FAQ Council to adopt a resolution proposing "An International Undertak- ing on World Food Security".1 In 1975, the Council then created a standing Committee on World Food Security which has held two sessions, one in 1976 and one in 1977.2 The FAO Director-General, in order to more fully implement the Committee's recommendation that FAO provide special assistance to developing countries to enable them to achieve national food secur- ity objectives, formally created the Food Security Assistance Scheme in April 1976. While drawing most of its personnel from and centered in FAO's Commodity and Trade Division, FSAS, as an FAO inter- disciplinary activity, draws expertise from other technical units within that organization. The major stated objectives of FSAS are to: lResolution 1/64 of the FAO Council, Rome, November 18-29, 1974. 2See UN/FAO [1976c, 1977a, 1977b]. 125 (i) Formulate national food security policies and action plans in line with the International Undertaking on Werld Food Security and identify projects and pro- grammes in support; (ii) Mobilize external resources for these programmes and projects; (iii) Undertake a technical review of the progress of implemen- tation in conformity with the guidelines of the Undertak- ing and of the Committee on WOrld Food Security; (iv) Evolve national programmes from an emergency phase to a mediumrterm approach; Activities under the Scheme, to be financed from extra- budgetary resources, cover such fields as: building of storage facilities and protection of stored grains; supply of grain for initial reserve stock building; strengthening of marketing sys- tem and produce handling infrastructure; commodity policy ad- vice; technical review of national food production programmes; improvement in national early warning systems and related data collection and analysis; and training (storage management per- sonnel, cooperatives, marketing operations, commodity intelli- gence). Financing of $16.2 million had been pledged to FSAS by February 1977, with the Netherlands, the Federal Republic of Germany and Swit- zerland the principle donors up to this point; other nations are cooper- ating by coordinating their existing bilateral programs with FSAS activities. Country food security policy reports have been completed for thirteen countries (including Niger and Nigeria in west Africa) and FSAS projects were slated for 1977 implementation in eight coun- tries, all in west Africa except for Tanzania.4 The four Sahelian projects, as we have seen, were initiated as emergency drought re- lief activities and have been taken over by FSAS which is attempting to consolidate its current projects primarily in West and East Africa. BUN/FAD [1977a, p. 1]. 4The countries are Cape Verde, Chad, Guinea Bissau, Mali, Niger, Nigeria, Tanzania and Upper Volta. 126 FSAS stresses the importance of basic cereals stocks in these coun- tries and is willing to help provide a complementary "package" of necessary supporting projects. Stress is placed on building stocks from domestic supplies to encourage local production, and on the immediate provision of management support and technical and training assistance. A fairly coherent set of pragmatic operating rules have been developed to maximize the efficiency of emergency stock holding and these will be discussed below. The remainder of this chapter will be devoted to a brief explor- ation of three interdependent topics affecting government participa- tion in grain marketing and storage in the Sahel: 1. Major policy issues in Sahelian food security strategy 2. More specific operational objectives and program implementation issues in holding emergency and buffer stocks in the Sahel 3. Alternative methods of analyzing grain stock holding activities and a discussion of quantitative techniques to model grain mar- keting and storage decisions. In the following chapters, these more general issues will be explored through an analysis of five grain boards in the Sahel and a more extended case study exploration of emergency and buffer stock programs in Upper Volta. Grain Stock Policy Issues As we have seen, there are a large number of socio-economic and political variables involved in considering Sahelian food secur- ity. These variables can be subjected to different levels of ana- lysis which produce different types of issues-ideological, theore- tical, methodological, operational, etc. We have also seen that in a system of interdependent relationships these variables "move 127 together" over time in both harmonious/integrating and conflicting/ contradictory patterns. This means we must be concerned with the evolution of the whole because it provides the rough boundaries with- in which policy makers and implementors can operate in the shorter run. we will introduce several assumptions to reduce this complexity and allow us to focus on a number of practical policy issues which are of immediate concern to Sahelian grain storage programs. Two additional points must also be stressed. First, while these issues are generally applicable to all six Sahelian countries, each, in practice, must be precisely defined within the environmental limits and resources of each country. Second, we have seen (particularly at the end of Chapter II) that it is possible to trace implications of macro-micro linkages within the general model of center-periphery dependency. This can be done at various levels: in the abstract, for West Africa as a whole and for each Sahelian country. The two extreme illustrative evolutionary models—-those of increased depen- dency and of increased independence-thus have significance for grain stock policy. Given the Sahelian environment, the following issues do not turn so much on whether or not to stock grain but rather on physical and institutional problems and the economics and-politics of alternative storage systems. 0f particularly importance are alternative incentive structures appropriate for different objectives. In addition to these structural and institutional questions, the dependency issue is re- flected most clearly in complementary policies linking grain storage to the process of rural transformation. 128 The policy issues are grouped into macro and micro policy issues for emergency and buffer stock schemes. Before discussing these policy issues we will make the following assumptions: 1. National governments in the Sahel are going to engage in grain marketing activities, particularly: a. Supplying at least a part of urban grain needs b. Making physical infrastructural improvements, particularly in storage and transportation c. Holding emergency stocks d. Employing buffer stocks, along with other measures, to act on supply and/or price variation e. Generally intervene quite extensively in food markets. Many of the conditions required for competitive markets are met, particularly in localized areas. Less Competitive conditions, however, often exist in longer distance trade and at higher levels of product aggregation in vertical channels. Short-term policy choices are quite limited but in the longer run, the Sahelian countries can engage in development policies which will promote greater structural and functional dependence or independence- Higher levels of outside funding (with some implicit constraints) will be available for infrastructural improvements, technical assistance to grain boards, improved data collection and perhaps to cover some grain board operating expenses. Objective functions will most likely stress "autonomous economic development", increasing or maintaining "food self-sufficiency", mobilizing mass participation in rural development and increas- ing distributional equity. While it is possible to observe selected trends which are moving toward the attainment of some of these goals, the major contradiction facing Sahelian countries is that structural and functional interaction patterns associated with economic development and development assistance will make the full attainment of all of these goals very difficult. 129 Macro-Level Policy Issues Self-Sufficiency All Sahelian countries ascribe to the general goal of achieving or maintaining national self-sufficiency in basic food stuffs under normal weather and production patterns. At issue is the identification of the costs and benefits Of pursuing such an objective particularly if this embraces the extreme position of self-sufficiency under all rainfall conditions. While self-sufficiency seems like a more real- istic objective for the four interior countries of the Sahel, this must be examined in the light of the specific conditions of each coun- try. It is necessary to determine the crops to be included (under current trends rice, and particularly wheat, are unlikely candidates for self—sufficiency status as compared to sorghum and millet) and what levels of production shortfall to attempt covering. Under a strategy of greater independence, the "non-economic" benefits of self- sufficiency gain more weight. However, we must still take account of the concept of relative comparative advantage and precisely specify the geographical units for which self-sufficiency would be defined by the local areas, nations, the Sahel countries, West Africa, etc. Trade Versus Storagg A related issue and one of the main currents in the international debate over buffer stocks has revolved around whether supply varia- tion can be more efficiently combatted through domestic storage or by relying on trade in regional and world grain markets or some come bination of both. While generally discussed in terms of total world grain production or in terms of wheat, the argument can be applied 130 to millet and sorghum by noting that the coefficient of variation in world production (for a recent world mean production of about 100 million MT) is substantially lower than that for West African produc- tion (for a mean of about 10 million MT), Sahelian production (about 4 million MT), and that for any particular country (e.g., Upper Volta with mean of about .9 million MT). This is a complex issue that can- not be viewed in zero-sum.terms. It must take into account West African regional and international trade both in terms of the time and cost involved in obtaining external food (which varies greatly by food type and country location) and the potential political and economic costs of having to reply on the outside for food. There may be explicit or implicit political constraints attached to depen- dence on outside commercial or donated food which are not directly obvious when looking at relative prices. Population at Risk Formal and informal procedures must be used to identify grain stock program target groups or the "population at risk" (of starva- tion, severe income reduction due to high food prices, etc.). The nature of the dominant political coalition may influence the definition of target groups, and in turn, affect factors such as information needs, optimal storage location, etc. General Intervention Strategz_ Resolution of the issues above will go a long way to defining the parameters of a general grain marketing and storage strategy. In addition and to some extent independent of the degree of depen- dency careful attention will have to be given to: 131 1. Specification of an integrated food policy to help avoid struc- tural and objective function conflicts; 2. Decentralization, including objectives and patterns of parti- cipation and incentives; 3. The size and timing of food stock operations. Training The issue involves determination of staff training procedures consistent with the achievement of program objectives. For example, in a decentralized grain storage program, participation and control, recruitment and training may be most appropriate if carried out and focused at the local level. Realistic on-the-job training will be facilitated if program infrastructure and operational growth is in- cremental so that the learning process does not foster a dependence on outside support. Infrastructure (Physical and Institutional) As soon as the objectives of grain stock programs have been defined various technical issues such as the cost of transportation, storage and handling can be analyzed using optimization or operations research methodologies. For example, a linear programming transpor- tation model can be used to determine the "optimal location" of new storage facilities given specified existing storage, transportation and storage costs, budget constraints, and the geographical distribu- tion of probable shortfalls. Cost functions associated with infra- structural investment are generally non-continuous. This means, for example, that while different bulk storage systems have technical advantages in long term, large-scale grain stocking, cost savings in bulk handling are only available when bulk equipment is used over 132 most of the system between producer-first handler and retailer- consumer transactions thus avoiding expensive bagging operations.5 Similarly, a standardized grading system may be prohibitively costly until it is used on a wide basis. International and west African Stocks The outlook for both international and West African grain stock cooperation is not particularly encouraging. For international stocks the main issues are how alternative proposed systems could reduce the reaction time and cost of providing food relief assistance. For West African stocks the main issue involves the extent to which reduc- tion in the cost of holding stocks through regional reserves is possible, assuming that satisfactory decisions on location, ownership, manage- ment and trigger mechanisms can be reached. In many instances, the issue will be reduced to what types of functional inter-cereals office contacts, such as information exchanges and reciprocal sales of sur- plus stocks, are possible to initiate. Emergency Stock Issues Objectives Both logically and in actual fact the primary immediate objec- tives for the constitution and operation of emergency stocks in the Sahel are twofold: 1. To have a known quantity of grain available to government auth- orities for immediate distribution under predetermined rules in response to unforeseen events leading to a rupture in normal supply channels, or in response to disasters of natural or human origin which necessitate free or subsidized food aid 5See Appendix B for greater detail in the specification of storage costs. 133 2. To develop within the context of a coordinated set of food grain policies the institutional and infrastructural means to imple- ment such a stock holding and distribution system and minimize the economic cost of doing so. These objectives can be intuitively justified from the point of view of Sahelian governments and international donors on interrelated humanitarian, economic and political grounds. The precise mixture and expression of these rationales will vary as we move along a con- tinuum between a high degree of center-periphery dependence and a more active commitment to and movement toward greater independence and self- reliance. In a more dependent environment the pragmatic emphasis will be on the role of food stocks in minimizing the economic costs of future emergency food assistance (with an implied assumption that such assistance will probably be necessary), on increasing the poli- tical stability of both international and internal dependency struc- tures through selective food aid to both urban and rural populations, and on minimizing "negative humanitarian publicity". In the more independent context, the emphasis will be on minimizing dependence on foreign food aid, on facilitating local economic development by at least satisfying basic food needs and on maximizing political stability through providing a valuable service to the regime's major ally, the rural population. These are clearly exaggerations of polar types; the actual Sahelian situations lie somewhere between the two. Related to and stemming from these primary objectives are vari- ous possible secondary emergency stock program objectives which are derived from FSAS documents dealing with Sahelian countries: 1. Contribution to the rehabilitation and increase in local cereal production 2. Promotion of grain price stabilization through the careful tim- ing of normal stock emergency rotation acquisition and disposal 134 3. Improvement of long-term grain storage techniques (or reduction of storage losses) 4. Assistance in the transformation of the subsistence economy to a market economy 5. Providing occasional stock assistance (grain loans) to other national food aid programs 6. Promotion of multi-national regional marketing storage coopera— tion including short term stock assistance, surplus disposal and more far-reaching actions such as price coordination, etc. Clearly many of these objectives could only be fully met through coordination with price stabilization and marketing institution and infrastructure improvement programs under the umbrella of an integrated set of rural development and food policy strategies. Level of Stocks As we have stated above there are no a priori optimal stock levels since levels must be determined through a balancing of objec- tives with resources under available technological alternatives. This question will be discussed in detail in the following section on methods of analysis. Location of Stocks A decision on stock location must take into consideration the population at risk and their location, the availability of passable roads, the costs of transportation, storage and handling, and oppor- tunities for stock rotation with minimum market destabilization or maximum stabilization. Constitution of Stocks The constitution of stocks involves the use of either domestic or foreign grain supplies, and can be effected under alternative food 135 aid, financial assistance and auto—financing arrangements. Rules for stock reconstitution must also be carefully specified. Trigger Mechanisms In considering trigger mechanisms, the issue is deciding what combination of price and non-price signals to use in declaring a for- mal emergency situation (meeting predetermined criteria) which permit stocks to be distributed. Practical Management In terms of practical management a major issue involves the de~ gree to which emergency stocks should be physically maintained and operationally controlled in isolation from buffer stock actions. Stock Disposal Loss-minimizing excess stock disposal plans must be available. These may range from subsidized exports to developing feeding pro- grams for commercial livestock finishing. Distribution of Emergency Stocks There are several operational issues related to the distribution of emergency stocks. First, the organization responsible for rapid distribution (army, gendarmerie, etc.) must be predesignated and capable of execution. Second, the question of whether distribution is to be free or at subsidized prices must be determined along with the maintenance of an adequate stock rolling fund to rebuild stock levels within a specified time period. Third, a pricing policy for turned-over emergency stocks must be fixed. 136 Regional Stock Coordination and Cost Sharing Regional stock coordination and cost sharing makes greatest sense with respect to emergency stocks. Rules and prearranged mechanisms for facilitating regional emergency stock transfers must be prepared and agreed upon. This may be possible using the offices of regional bodies such as ECOWAS or donor agencies. Buffer Stock Issues Objectives The major issue here is the determination of the beneficiary of stabilization efforts, whether it be rural producer/consumers or various classes of non-producer consumers or both. In addition we must consider other secondary or sub-objectives many of which were discussed above under emergency stocks, which differ only in degree from buffer stocks. The specification of buffer stock objectives will, of course, be largely determined by the fixing of prior goals relating to regional and class distributional equity, general eco- nomic strategy and so forth. Ownership and Control of Buffer Stocks Farmers and traders both own and control their own buffer stocks although their behavior and objectives may differ. In a decentralized, segmented economy, a principle hypothesis is that stabilization can be operated most effectively at the local level. Related to this is the general observation that centralized, hierarchical bureaucracies have a difficult time efficiently operating decentralized activities such as geographically-extensive grain buying, selling and storage. The major issue involved in longer term program planning is how to 137 promote decentralization, increased local participation and control over stock holding which is seen as a public service rather than a speculative activity connected to larger patterns of differential capital extraction. Clearly a national stabilization program needs to also include centralized activities involved in cereal transfers to deficit areas (both urban and rural) and coordinating mechanisms among sub-national regions and with cereals offices in adjoining coun- tries. A federated system built on a strong locally managed base seems, at first glance, to be a logical alternative but it should be remembered that none of the six countries has attempted to use such a system. Pricing Policy This is probably the most troublesome area facing cereals office personnel who are trapped by rigid price and cost barémes, prices which favor urban consumers for political reasons, and pricing deci- sions which are made far too late in the buying campaigns. Uniform prices with differential transportation margins are about as flexible as the Sahelian systems have been to date and this makes sense in terms of the demands of a centralized bureaucracy. With evolution to a more complex system (under either the greater dependence or the greater independence models), increased price flexibility will probably help increase operating efficiency. Uniform prices for surplus pro- duction (beyond the consumption of local areas) with transportation covered at cost would also be possible under a decentralized, locally controlled stabilization system. 138 Structural Reform In addition to the changes inherent in alternate solutions to the issues above, the current systems in several Sahelian countries could be made more efficient through structural reform. The major object would be to streamline the system to reduce the number of phy- sical ownership transactions required in the marketing channel. This will be illustrated for Upper Volta in Chapter VIII. Intervention Strategy As we have seen, price stabilization has been an elusive goal for Sahelian cereals offices. To promote learning, intervention should probably begin with modest goals in a number of market places (deficit, surplus and "balanced"), combined with a simplified data collection system so the general magnitude of intervention effects and the depth of markets under alternative supply conditions can be assessed. Alter- native price trigger mechanisms could be used experimentally, parti- cularly if greater pricing flexibility were allowed a more decentral- ized system. Also success in storage and market operations within one year's market campaign will help build the experience required to begin the much more difficult task of inter-annual stabilization. Methods of AnalyzigggGrain Storagg In United States and foreign literature on grain storage theory and practice, there is a great deal of conceptual and terminological confusion, much of which stems from an imprecise mixing of different levels of analysis and from the complexity of analytical methods avail- able with which to investigate different levels of an integrated grain marketing and storage system. This is a general problem which applies 139 to consideration of programs oriented toward emergency, buffer or commercial pipeline storage operations. First, we should cover the problem of level of analysis. When- ever we prOpose to intervene in a complex, interrelated system by man- ipulating a limited number of variables we want (or should want) to understand the micro and macro level impacts of those actions on the structure, operations and normatively evaluated performance of that system over time. However, in the operation of actual systems there is a division of labor with different actors having different focuses and time horizons, concentrating their attention on different levels of the system and their labor is appropriately evaluated using differ- ent sets of criteria. For Sahelian cereals offices charged with grain marketing and storage, there are basically three levels of analysis which are performed and which are accompanied by characteristics repre- sented in Figure VI-Z. Clearly there are other important dimensions to the different levels of analysis and action in an integrated grain stock system which could be added to Figure VI-2. Also, the categories used are meant to be illustrative and, in actual practice, may involve more overlap and "empty cells" (unfilled tasks) than the chart would in- dicate. Finally, to the extent that such a system is operationally successful, it would involve communications and feedback mechanisms which are not represented. For example, the successful policy maker or cereals office administrator should be concerned with more abstract grain system modeling and system evaluation analysis, with the day-to- day operational analysis of problems faced by office field buying NUHhho m4¢fimmo ho mmHBH>HHU< mwNA 82mmmhhHG ho moHHmHmmHU muauHm hoooauammo munwfimafi one coaum>osou mo>wuuofioo uwumamumoun manounoouo oaaooooo Hoouuuaoa Momma one «woman cues manna monmouaxo HMS soumhm anemone Hmaowuouoeo mo mouuuaom uuouuoo mom womanhood: mfiuoufiuo unoaaaamaom we ooauoofiom ou hufiaanoeou coauosam>m Anmfiaaaoooo mo~3§ so mumowNom Seaman owououm mom on maneuv mNmmuN wsauoxuna :«ouw mowuoxuma sash» msuom secum cause no sowuouonodm mo souusaoom oaawooaum umouuonm ououvosuousH unease; sesame: mafia Aooamuo Auououumwaaaoo Aoowmmo mumouou owouOum Hooowwomv ooummo manouoov cu Hausdomooov HousoaoHeEH nexus hoaaom unhaodo uuuausowum nouo< vousoauo auaaom Honouuouomo \ouuwooauouoH uaaoumumlouumz mummams< mo Ho>mu 141 agents, truckers and warehousemen, and facilitating communication and interaction among all levels of the organization. The literature reviewed in this study covers all three levels of analysis. Much of the more academic literature6 focuses on macro level, systemic analysis of stock operations as they interact with fluctuating supply and with demand conditions evolving more slowly over time. Much of the political economy literature cited in Chapter II--to the extent that it deals specifically with food problems at all--is at an even more global level of systemic analysis. MOst of the donor agency and Sahelian government documentation on staple food marketing and cereals office programs falls into the category of inter- mediate policy analysis and, to a lesser extent, of analysis of short-term micro-level operational problems. Within the area of macro systems modeling of grain stock holding, a wide variety of approaches has been employed. It is crucial for our purposes to note that most of these modeling efforts have been conceived to analyze grain reserves which may be thought of as inte- grated emergency and buffer stocks or basically as buffer stocks which can be put to specific uses under emergency conditions. While this is consistent with the logical structure of grain board activities (see Figure VI—l again) and with all reasonable attempts to use general equilibrium economic modeling techniques, we will see that this more integrated approach is largely abandoned in much of the policy and operational level analysis of Sahelian grain storage. 6See, particularly, footnote 5 in Chapter I for a beginning list of references. 142 Gustafson, in his classic article on inter-annual grain carry- over levels (i.e., inter-annual buffer stocks), proposed four cate- gories of analytical approaches used to arrive at storage decisions:7 1. The standard method was described as the ZeveZ-ofLstorage approach which attempts to determine an absolute amount of stored grain which would be available in the case of poor harvests, war, etc. This approach was felt to be inadequate because it did not involve a more flexible storage rule which would account for a number of variables concurrently and because no specific level "can be shown to be optimal, that is, no objective way exists for showh ing that one level is better than another". 2. Another method is to specify a plausible functional fbrm fbr the storage rule to determine annual "carry-out" as a function of differing quantities of available supply ("carry-in" and harvest). Then one calculates "expected cost and benefits under the rule, such expected values being functions of the coefficients or para- meters in the rule, and finds store values of the parameters which minimize net expected losses or maximize net expected gains". 3. Storage rules can also be based on deviations in size from the "hormal crop" or from.a trend line in production levels. Gustaf- son attacked these because they generally used unrealistic or simplistic storage rules; there were no obvious probability cri- teria for determining needed starting stock levels and, there- fore, the rule involved was necessarily unstable over time. 4. Gustafson opted instead for a fourth approach, one based on an idealized free market (i.e., "a market in which all stocks are held by private firms, operating under perfect competition, and maximizing expected profits"). With certain simplifying assumptions the stock levels the free market system would hold under alternative supply and demand conditions can then be used to produce an optimum.storage rule under the assumption that what happens under the idealized conditions is desirable and is in fact achievable with public sponsorship or instigation under less than ideal conditions.8 Alternatively, the stock 7Gustafson [1958]. 8Practical difficulties with these assumptions relate partially to the difficulties of defining optimality under conditions of "the theory of the second-best". Gustafson also used some very demanding simplifying assumptions, including zero transportation cost, zero storage cost, all production and consumption at a single point, etc., in order to reduce computational difficulties. Presumably, most of these could be relaxed through the use of more complex models and more powerful approximate computational techniques such as those em— ployed in stochastic simulation. 143 levels produced by this method can simply be compared with the amounts that have actually been stored historically. Of the four methods, this one is most congruent with the ideal approach to economic modeling stemming from nee-classical theory. There are, of course, other ways of categorizing macro-level model- ing of grain stock procedures, particularly as new analytical method- ologies, such as refinements in econometric and systems simulation techniques, permit a wider range of system variables to be analyzed concurrently. A recent article by walker and Sharples9 reviews and compares several of the most prominent recent simulation models while the conference proceedings edited by Eaton and Steele includes several new modeling efforts and a number of insightful pieces of policy ana- lysis.10 Most of the literature is focused on two broad subject areas: (1) United States grain stock holding (alternate policies and rules to-minimize risk and share costs on a unilateral or international basis, reviews of past experiences, etc.); and (2) international stock proposals themselves (with most making obligatory reference to the 1974 werld Food Conference and its predecessors, to a lack of cooPer- tion on data matters from major grain buyers in the socialist camp, to a stress on stock rules rather than levels, and to various insur- ance type schemes for holding international stocks with the developed world to pay all or most of the "premium" for LDCs). This latter pro- posal has been most vigorously promoted by D. Gale Johnson.11 9Walker and Sharples [1975] . 10Eaton and Steele [1976]. 11See, for example, Johnson's article in Eaton and Steele [1976]. 144 To this date very little of this more formal macro modeling of grain stock systems has focused uniquely on LDCs. The Indian subcon- tinent provides perhaps the most notable exception. India has main- tained very extensive food grain stock programs for many years and these programs have been analyzed at various levels. At the macro level, excellent descriptive work was sponsored by the Indian Society of Agricultural Economics and has also been done by Uma Lele and A.M. Khusro.12 The latter is particularly useful to policy makers since it also delves into some of the micro-economic questions of storage cost. Several formal simulation models have recently been developed. Reutlinger, using India as a base, develops a general model which focuses on the question of cost/benefit trade-offs between supply stabilization through international trade versus domestic 13 Another sime storage specified at various levels for illustration. ulation model has been developed by Manetsch using Bangladesh as a base which incorporates a very different set of behavioral relation- ships and even attempts to define benefit functions measured inversely by the number of deaths per time period.14 This type of macro-level economic modeling could be of consider- able use in the Sahelian region since almost nothing of this kind has 12The Indian Society of Agricultural Economics [1969], Lele [1971] and Khusro [1973]. 13Reutlinger [1976]. There is another somewhat different version of this research in Eaton and Steele [1976]. This model might be used as a possible base for a detailed West African stock simulation pro- ject if substantial modification were undertaken to permit accomoda- tion of important structural characteristics of West African grain markets and putting substantial emphasis on the effects of high trans— portation and modern storage costs. l4Manetsch [1976]. 145 yet been attempted, particularly focused solely on food security15 or food grain marketing. To be useful, this type of modeling effort for a particular country or group of countries should include geographic variables (e.g., extensive differences in regional supply balances and trade patterns both within and outside the particular country) and the effects of transportation, storage and management bottlenecks and costs. There are, also, some very important characteristics, such as Senegal's coastal urban area, which might not fit into such a more generalized model. As we move from the macro-theoretical level toward the policy and operational levels, we see that one model characteristic, the pro- portionality between usefulness and the detail of specification, pro- vides one of the biggest drawbacks in terms of continued utility, that of the need for constant updating and modification in response to changing environmental and policy considerations. This is a very expensive proposition for a Sahelian budget unless this recurrent ex- pense is largely met from the outside. It should be clear that most of the international stock models will be of limited direct applica- tion except that they provide further precision on possible changes in transportation time and the real costs of outside relief supplies. As a final critique from the perspective of the paradigm employed in this study, most of these modeling efforts do not include key political- economy variables, although this omission, largely a product of lsThe beginnings of such a modeling effort focused on a Sahelian agricultural sector is contained in Maddox [1976]. 146 political convention and disciplinary tradition, would not be con- ceptually difficult to rectify. As we will see in Chapter VII, most of the studies of Sahelian government grain marketing have been descriptive in nature and have been forced to use secondary data. Stemming from this, most methods for determining either buffer or emergency stock levels have been rather pragmatic. Most of these methods would be grouped under Gus- tafson's "level-of-storage approach" which are fairly uniformly crit- icized as naive by most macro-level policy theorists. In the Sahel, these methods have been used most extensively by FAD/FSAS, and its "multidonor mission" predecessors, in the rough determination of emer- gency stock levels. In the more sophisticated versions the following clusters of variables have been taken into consideration at least on a nominal level in arriving at these calculations: 1. Rough probability of total shortfall is determined in terms of deficits from production trends (or even values). 2. The gross shortfall is allocated among the population at risk (of starvation, malnutrition, etc.) broken down by geographic region, socio-economic groupings such as farmers, herders and urban population by means of: a. Food balance sheets (FAO estimates generally) which divide the total harvest into categories of consumption, seed re- quirements, losses and waste, etc., and b. Average adult and child grain consumption levels. 3. From the gross shortfall (or gross shortfall per capita) is sub- tracted the amount of outside relief grain which can be expected to be delivered to the population at risk given the time required for its arrival in the country and the capacity of the cereals office to distribute grain with its infrastructural and manage- ment capacity. This produces a net aggregate (or per capita) shortfall, some percentage of which (usually fairly high, e.g., 90 percent) is considered to be a reasonable minimum stor- age level. 4. These steps are generally repeated several times with sequential modification of essential components in order to produce 147 "reasonable" or acceptable results in terms of other political or budgetary constraints. As can be surmised from the number of points in the above meth- odology where data may be unavailable or unreliable and at which aggre- gation bias (particularly in a segmented economy) may produce meaning- less results; the act of choosing a given storage level represents an "act of faith" in the intuitive plausibility of our guesswork or in our confidence in the availability of adequate financial and political resources to meet secondary or unanticipated needs produced by errors of judgment or changes in broadly defined environmental conditions. Often the policy and operational methodology for buffer stock holding at the beginning of a particular crop marketing season is equally or even more straight forward. The cereals office is usually heavily constrained by its levels of funding, by officially determined grain price margins, by government purchasing and accounting proce- dures, and by its existing infrastructural capacity (available trucks, sacks, fuel, empty warehouses, personnel, etc.). This means that under optimal conditons it may be able to buy a limited proportion of avail- able supplies and put it into storage for sales later in the year or in the next year. Assessment of operational effectiveness can take place at two levels: (1) the efficiency of operational performance in cost/benefit terms--given a precise definition of objectives and available resources- can be calculated and areas of needed improvement (e.g., minimizing the number of empty "backhauls", cutting warehouse storage losses through correct spraying procedures, etc.) can be specified; and (2) alternatively, the economic impact of its actions can be assessed, primarily through estimates of the stabilizing effects on 148 producer and consumer prices over time and space, and the micro-level impacts on various categories of market channel participants. In the absence of available data and analytical models (which, as we have seen, have their own cost and application limitations) this level of analysis will be done impressionistically using "rule of thumb" calculations and isolated, somewhat random measurement of performance variables. Cereals office officials will say, "compared to the normal fall in harvest time producer prices under these supply conditions, our intervention in Market 'X' in November and December had a substantial impact and kept prices 5 to 8 CFA/kg higher than they otherwise would have been". This is then followed by a similar assessment of price stabilization in consumer markets and estimates of total producer and consumer savings which are then used to "justi- fy" total program.expenditures. Often this process is supplemented by a large volume of time-consuming pro fbrma documentation and finan- cial accounting. The obvious inadequacy of the type of "economic evaluation" out- lined above can be confronted in two ways. First, analytical and evaluation models of food grain supply and demand-—with particular emphasis on private and public storage operations--can be developed and employed if requisite reliable data, and financial and personnel resources can be obtained. Second, the problem can be partially "solved" through a redefinition of the unit of analysis and account- ing which would, of course, have to be preceded by reorganization of collective grain storage and marketing activities to that level. we have seen the functional difficulties inherent in trying to operate a centralized grain market operation in a decentralized economy. At 149 least a partial solution could involve a progressive structural and functional decentralization and redefinition of the basic operating and accounting unit so that they are more congruent with the under- lying base structure of rural production and consumption, the avail- able physical and human resources of the country and stated rural development and distributional objectives. The two solutions, while somewhat contradictory at the theoretical level, are not incompatible in actual practice as an integrated grain marketing and storage system must operate at different levels to meet multiple objectives. Thus, while the basic food security operational and analytical evaluation unit can be decentralized to the village or village cluster level, other transhipment, commercial sales and storage activities can be more narrowly, but more appropriately, analyzed on a more aggregate basis. This strategy will be detailed and further explored in the following two chapters. CHAPTER VII SAHELIAN GRAIN BOARDS: STRUCTURE AND PERFORMANCE Introduction In West Africa, government participation in agricultural market- ing, both during the colonial era and after independence, has focused more heavily on exported cash crops than on food crops, a fact one would anticipate given the importance of primary exports in the lay- ered process of periphery-center capital accumulation and economic development. However, with increasing urbanization, the development of more extensive food grain markets, relative declines in per capita food production and drought induced food calamities, West African gov- ernments have attempted to exert greater direct control over staple food production and marketing. Sahelian grain marketing interven- tions have included producer and consumer price policies, competitive and monopoly buying and selling associated with grain storage opera- tions, and, more recently, the large-scale distribution of purchased and donated outside food grain relief supplies. This chapter will review, for each of five Sahelian countries, the structure, objectives, means and initial performance of national grain marketing and storage institutions and policies. These topics will be covered for Upper Volta in the more detailed case study of that country in the following chapter. In addition, attempts to 150 151 implement regional grain storage and marketing programs, proposals for Nigerian sorghum and millet storage and the major lessons from recent drought relief efforts will be briefly reviewed in this chap- ter. Within the larger conceptual framework which has been elaborated so far in this study, emphasis here will be placed on an analysis of these mechanisms as they affect the practical constitution and opera- tion of sorghum and millet stocks. Country Reviews Mauritania In Mauritania, government coarse grain marketing and storage powers are vested in l'CfTice Mauritanien des Céréales (O.M.C.) created by Presidential decree in August 1975.1 This embryonic or- ganization is not yet fully operational. In late 1976 arrangements were still under negotiation with the German Federal Republic for the provision of initial technical assistance. Crucial rice imports and distribution are handled by the state trading corporation, SONIMEX (Société Nationale dimeortation et d'Exportation), which had also previously handled some other official grain imports. Re— ' lief distribution efforts were coordinated by a national committee headed by the Minister of Health and Social Services and aided in vehicle logistics by the "Service d'Assistance au Profit de la Popu- lation Rurale" of the Army General Staff.2 The structure and objectives of the OMC are similar to those of the other national cereal boards in the Sahel and will therefore be 1Islamic Republic of Mauritania [1975]. 23ee UN/FAO [l973c, 1974c, 1975c]. 152 presented in some detail for illustration. The Director and other administrative personnel of the Office report to an Administrative Council made up of representatives of virtually all national minis- tries dealing with rural problems, the national assembly, the trade unions, the central bank and the chamber of commerce. OMC is opera- tionally controlled by the Ministry of Rural Development but the concurrence of the Ministry of Finance is required for most major operational decisions which involve substantial financial disburse- ments. As spelled out in the Presidential Decree, OMC's overall objec- tive is "to contribute to putting into effect national policies of supplying interior markets with basic cereals". To this end it will constitute (1) buffer stocks to stabilize prices to producers and consumers while assuring a transfer of cereals from surplus to defi- cit areas and (2) emergency stocks to meet supply breakdowns induced by national production shortfalls. To implement these activities OMC is required to: 1. Determine cereal needs in deficit regions 2. Determine levels of marketable supplies in surplus regions 3. Obtain needed supplies by paying producers guaranteed minimum prices which will serve to stimulate greater production 4. Manage transport from supply to storage points 5. Manage necessary grain storage operations 6. Assure a market supply of sufficient quantity to keep grain prices pegged to periodically determined base reference prices 7. To receive and distribute all food grain aid from the inter— national community 8. To acquire in foreign markets all other grain to be imported, except rice 153 9. To aid producer cooperatives in the collection and treatment of harvested grain to be sold to the Office 10. To train the cadres and technicians needed to meet their objec- tives. This is an awesome task in a country which has the least trained personnel and weakest infrastructure in the Sahel. Mauritania's domestic supply areas stretch along the Senegal River and the south- ern border of the country where there are almost no usable roads or other means of modern transport, and whose deficit regions are very isolated and reached over very long, poor roads. Maintaining minimum producer prices and controlling consumer prices is only a dream under current conditions and an extensive stock system with major inter- regional transfers will undoubtedly be quite costly. As a recent AID report concludes, the initial policy decision to decentralize grain stocks and marketing to regional capitals seems a logical one under these conditions.3 It would also make sense that any buffer stock and price support actions be regionalized particularly since measures to promote national integration of the highly fragmented rural economy are just beginning. In addition to a very poor internal road system, severe bottle- necks have existed in import channels and in the availability of ade- quate storage facilities. Most imported grain comes by sea or by road from Senegal. Even with attempts to increase efficiency, the wharf at Nouakchott and river ferries at Rosso (on the main road to Senegal) remain bottlenecks for greater than normal levels of imports. Ware- house storage is considered insufficient for projected regional storage plans although recent construction has raised the total to 3U.S. Agency for International Development [l975b, pp. 014-18]. 154 around 17,000 MT out of an estimated 30,000 MT needed for a minimum storage system. Senegal In many respects food grain policy in Senegal is more complex and varied than those in the other Sahelian countries. Among the Sahelian countries, Senegal has the most developed modern economic and commer- cial base, the largest urban population, by far the best communica- tion and transportation system and the most advanced center-periphery extractive relationships which, at least in their classical form, seem to be becoming increasingly untenable. There are recent signs of some diversification away from almost exclusive dependence on groundnut exports and increased attempts to employ decentralized regional devel- opment mechanisms. There is also a realization that policies promoting extractive capital accumulation and subsidized food imports for urban areas were not tackling the key problems of rural stagnation and lag- ging domestic food production. As a result the government is attempt- ing to provide producers a greater degree of control over their own agricultural surplus (e.g., formation of rural administrative dis- tricts which would control up to 75 percent of collected local taxes) and assistance in more localized grain marketing and storage. How- ever, old habits of increasing dependence on government enterprise and employment, and urban expectations of lifestyle and consumption pat- terns remain strongly entrenched along with most of the structural foundations of the previous system. These evolutionary tendencies are reflected in some recent changes in ONCAD (l'Office National de Cooperation et d'Assistance au 155 Développement)5 which, until 1974, had played a very limited role in coarse grain storage and marketing except for the distribution of relief supplies in 1973 and 1974. ONCAD, created in 1966 and under the Ministry of Rural Development, originally shared its cash and food crop marketing responsibilities with OCAS (l'Office de commer- cialisation Agricole du Senegal) which it absorbed in the early 19703. In rural areas ONCAD's functions are linked to two principle elements in Senegalese rural development strategy (which as we will see are similar to basic strategies in Mali, Upper Volta, and Niger): semi- autonomous regional development authorities, usually fairly crop- specific in their programs due to historical patterns of regional crop specialization, and a very extensive network of village level farmer marketing cooperatives. ONCAD has the following wide-ranging functions and objectives: 1. Development and support of rural cooperatives 2. The organization of agricultural input supply systems and the collection and marketing of farm output (traditionally the primary focus was on marketing groundnuts), by way of the cooperatives and in conjunction with the regional development organizations 3. Provision of financing and credit for urban rice imports and, through the cooperatives, for farm-level technological inputs and marketing with funds from the BNDS (the Banque Nationale de Développement du Senegal) 4. Provision and management of the transportation and storage needed to achieve this input (including improved seed distribution) and product marketing. As further background to an understanding of the institutional position of ONCAD, we must note that government price policies, 5This section draws on material from Ediafric [1975], UN/FAO [1973e, l974e], Food and Feed Grain Institute [1970], IBRD [1975c] and other Bank reports and USAID [1975b]. 156 although more extensive on paper, have had significant impact only in three areas. These areas and their relationship to ONCAD and the National Distribution and Price Stabilization Fund are represented in Figure VII-l. First, in rural areas, ONCAD channels credit and agricultural inputs-~primarily fertilizer for cash crops--to farmers through their cooperatives. Fertilizer, sold at fixed prices, has included effective subsidies of up to 50 percent. Second, slowly rising guaranteed producer prices for groundnuts and cotton (and only recently coarse grains) provide product flows through the cooperatives which ONCAD exports or diverts to urban areas (specifically groundnut oil, rice, millet in the latter case). Third, in urban areas, ONCAD stabilizes and, if need be, subsidizes the prices of rice, wheat, groundnut oil and sugar-—all urban dietary staples. The net distribution and production incentive effects of such a system do depend on the nature of relative prices. As mentioned in Chapter III, these domestic terms of trade have often been turned against rural areas with one World Bank estimate placing a net CPSP fiscal "tax rate" of 20 percent (equal to the difference between pro- ducer and export prices less import subsidies and production credit which has fairly good repayment rates) on rural production between 1971 and 1975, including the worst drought years. In 1975 a number of relative price and other policy changes improved the situation for the rural sector; producer prices for groundnuts and grain were increased. The consumer price for rice was increased by almost 100 percent. This stimulated rural production, reduced rice imports by almost one-half and dramatically improved the rural/urban terms of trade. Part of the increase in millet consumption is attributable to 157 mmoxu am<0 nz< noon 2H monhU< Hzmzzzm>ou mmmq Hannah Acuouov I duo assessesc I :Ouuou I howsm I nuance=Ouu I Amouwauuuomv usofim moon: I . scuuouwaaooum ouwum codewonsm usury scam I w Vv uoseoum uqoouo avaonam a ecuueufifianmum ouuoasa . . oouum sossneoo II. uo>quouomooo .moou< Hausa mumm¢ can»: AAII owouOUm AAm room _ > > > mafia > > > Aocqmuo\oeouo saucy \Llr h w uswosozouez m a a .33: < _ _ .mussam detox seasoeac—o>wo esofiuouoauoo urosao~o>oa wouuooooum on sumafissusz uamuooamlaouo so Hmcowwoz me< I.mvv monom\dmwwcwm .IIII w vv Iumvv mascara so usosoaaofio>wn ueoaoaaoHo>wn so ooseuawuo<.v echo or oamsowunz ooocom um :owuouonooo up massages: ouuuuo some moze\moze mammm AA 0 i _ x - w w w i nuag sq rogues: msamusm AA a ”sneak uosvoum ”UHQHQE—JUU< AA % Ava—ah =°fluaflfiflfinmam QUfiHW we” COHU—dpfihumfioy A} ” uU’OHh HQUWH“ unsecuo>ou xwum not coauoauaauum so sequeskuwm on mooqou I ammo Nox 158 two technological innovations, the growing success of millet/wheat bread and the development of a processed millet, instant cous-cous. Whether these changes represent a new trend in persistent improved terms Of trade for rural areas or, as in past eras, a temporary improve- ment associated with the installation Of a new set of extractive struc- tures, is not clear. What is clear is the complicated role ONCAD plays as the government middleman in cash and food crop trade. Given past experiences this may not bode well for the establishment of more independent, locally-controlled grain stock mechanisms to benefit all sectors Of the Senegalese political economy. One suspects that incen- tive structures and operating procedures formed during the years when the groundnut trade was almost the exclusive concern Of the organiza- tion may be less than optimal for a grain marketing and storage pro- gram even if handled by a newly formed special section of ONCAD. Al- though ONCAD has received outside technical assistance financed by IBRD since 1969, it has been persistently criticized for lack Of finan- cial autonomy, for related financial mismanagement and poor timing in forwarding Operating funds to buying agents in the field, and persis- tent overstaffing and other civil service related problems. ONCAD's direct experience with coarse grain storage and market- ing is quite limited. Between the 1965/66 and 1973/74 marketing sea- sons it purchased total millet and sorghum ranging from only 21 tons to about 30,000 tons. Before 1974 these were simply seasonal purchase] resale stabilization Operations. With Official buying and resale prices set at 17 and 22 CFA per kg, ONCAD was estimated to have in- curred losses Of about 6 CFA per kg which were met through funds from 159 groundnut reserves. These inappropriate price structures provide partial explanation for its limited activities in grain markets. The physical means at ONCAD's disposal to engage in more exten- sive grain storage have also been limited. While it has a large, well-maintained fleet of about 250 to 300 owned and rented trucks used primarily for input supply and groundnut marketing to Oil plants and the Dakar port, its warehouse storage facilities, particularly in rural areas, have been very limited. According to available infor- mation, ONCAD largely has used rented storage both in the Port of Dakar and in rural areas. In Dakar where about 100,000 tons Of com- mercial storage is available, ONCAD has maintained rolling stocks Of about 35,000 tons Of rice for urban consumption but cannot begin large- scale coarse grain storage in these facilities without jeopardizing the cargo handling capacity Of the port. Both in Dakar and in rural areas where small commercial storage facilities are used, storage losses through excess moisture and insect infestation are reported to be unusually high. In order tO meet the storage needs associated with its commer- cial grain marketing, ONCAD in 1975 prOposed a $10 million plan to construct a three level 50,000 MT storage system for coarse grains with facilities at the producer, cooperative and center levels (25,000 MT capacity for Dakar and seven regional facilities Of 1,000 to 3,000 MT capacity each). TO a limited extent ONCAD has be- gun to implement this storage construction plan but the bulk of it awaits foreign donor financing. In addition to these buffer and pipeline stocks, preliminary plans for a 50,000 MT emergency reserve stock have been discussed. This, of course, would require even more 160 storage since the previously proposed capacity would be used to handle annual and inter-annual urban supply and buffer flows. Chad Responsibility for food grain marketing and storage in Chad is vested in the FDAR (ands de Développement et d'Action Rurale) which was created in 1967, and specifically in its Département Céréalier (D.C.) which was organized in 1969. The FDAR/DC is part of the Min- istry of Agriculture and Land Management6 and has a quasi-commercial charter and some financial autonomy which make it similar in struc- ture to the other Sahelian marketing boards. In addition to grain acquisition, storage, price stabilization and emergency distribution, FDAR has other tasks such as rural well drilling and some rice mill- ing and marketing. The drought period, particularly after 1972, and the relief efforts it brought provided a major impetus in the expan- sion of the organization and the development of its Operational capa- bility. FDAR/DC has the most limited cereals marketing and storage exper- ience in the Sahel except for Mauritania and its organizational and infrastructural capacity is equally limited. Nevertheless, it has received an increasing amount of technical assistance and fairly in— tensive study in recent years.7 For the first five seasons FDAR 6Formed by the merger of two Ministries in a government reorgan- ization in June 1976. 7These summary observations on FDAR/DC are based on a surpris- ingly rich documentation becoming available including: Auffret [1974], Borsdorf [1976], Republic of Chad [1973], Creupelandt and Corbett [1974], Levac [1974], Maxon [1976], Naim [1976] prepared for the Club des Amis da Sahel, UN/FAO [1973f and 1974f] and USAID [1975b]. 161 engaged in limited intraseasonal stabilization in N'Djamena and sev- eral regional towns. In the capital it bought cereals directly from commercial traders, stored them and resold them during the "hungry season" in an attempt to moderate the seasonal price increase. In smaller towns the administrative "sous-préfet" was the designated FDAR buyer. These latter actions were not felt to be too effective even on a local basis, because government officers are not always the most qualified grain buyers and stock managers, are not adminis- tratively responsible to FDAR, and because "hungry season" sales have Often been made only to government bureaucrats and their fami- lies, a practice which has its roots in traditions of colonial priv- ilege. In 1969 an Operating fund for cereals purchasing was constituted through the sale of food aid provided by the WOrld Food Program, later supplemented with direct capital grants and FAO technical assistance. Cereal purchases on local markets have never exceeded 2,000 MT. From 1972 through 1974 FDAR/DC largely functioned as a relief dis- tribution agency and abandoned all efforts to buy on local markets. With greater financial and material backing FDAR/DC used three mobile buying teams to buy cereals directly from farmers for the first time in 1975. In addition about a fifth of total Chadian rice production of some 35,000 MT (milled) is processed by FDAR's rice mill and in 1976, for the first time, FDAR planned to market the entire output of the mill itself. Official producer prices for rice and wheat have tended to keep most marketed surplus out of government control since better prices could be had in the private market. 162 Experiences in the delivery Of relief supplies point up some of the difficulties facing FDAR/DC in its storage and marketing programs. First, for grain imports, Chad is the Sahelian country furthest from the sea, reached by very expensive and time—consuming combinations of rail and road transportation through Nigeria and Cameroons. Creupe— landt estimates that six months is a reasonable time interval between initial donor notification concerning food shortages and the actual delivery Of relief supplies in Chad.8 This includes the concentrated use of the national transport monopoly, the CTT (COopérative des Transporteurs Tchadiens), which handles most external trucking to railheads in neighboring countries but only about 10 percent of inter- nal traffic. Second, storage facilities, with their placement and arrangement, represent another critical bottleneck. About 10,000 MT of a total capacity of 18,000 MT is located in or near D'Djamena and used for operational or pipeline stocks. Various plans have been drawn up to increase this capacity to 25,0009 and even tO 50,000 MT.10 However, this raises realistic concern that too much Of the storage may be lo- cated in N'Djamena, further emphasizing the urban bias of FDAR emer- gency stock and stabilization efforts. For example, FAO proposed, and the government accepted, an emergency reserve stock plan which would place 9,000 MT (of the total 10,000 MT) in N'Djamena and 1,000 MT in Abéché on the opposite eastern side of the country under the osten- sibly plausible reasoning that this would promote more efficient 8Creupelandt [1974, pp. 11-13]. 9Naim [1976, p. 33]. loMaxon [1976, p. 36]. 163 storage and permit easier supervision. However, in addition to ignor- ing the populous southern and chronically deficit north-central areas of the country, this seems to ignore the logistical advantages of dispersed placement of relief supplies given Chad's huge distances and very poor roads which are largely impassable after the first rains when the need for extra grain supplies can be the greatest. Third, attempting to use a small, highly centralized organiza- tion to Operate a grain stock program in a decentralized economy will, as we have seen in other economic sectors, tend to channel funds, in- frastructure, jobs and incentives to a few urban centers. This is shown in recent FDAR funding proposals made to USAID which focus on construction of new administrative structures and storage facilities and opening new sales points in N'Djamena while, in rural areas, only increasing the number of buying points to feed this urban appetite. Given the multiple objectives possible for a grain system this par- ticular set seems to be rather heavily skewed in favor of potentially profitable commercial sales and urban consumer price stabilization while ignoring producer price supports and incentives, and making stabilization and relief supplies more readily available in rural areas. USAID, to its credit, has resisted these proposals and sug- gests instead funding a pilot program of FDAR field administration of grain storage at the sous-prefecture level.ll Finally, overcentralization of storage and market actions in a few urban areas tends to increase per ton operating costs and this, coupledthfiipressures on the Chadian government to enhance its very llMaxon [1976, pp. 25-39 and Appendix H]. 164 weak political control by forcing FDAR/DC to distribute food supplies at substantial economic losses to certain urban and regional consti- tuencies, leads to increased dependence on outside financing to meet the rapidly inflating budget needs of a national food program for "self-sufficiency". Mali Mali's experiences in cereal storage and marketing are, for a number of reasons, perhaps the most important in the Sahel in terms of their pedagogic value. First, during colonial times and even into independence Mali was considered a surplus production area and made substantial exports to neighboring areas. While it still has some export trade, particularly of a clandestine nature, and has begun to export Officially by way of its grain board, it is clear that aggregate cereal production has dropped for reasons other than drought and population migration alone. Second, Mali has made the strongest attempts since independence to break colonial and neo-colonial rela- tionships, has intervened most heavily in all sectors of the economy-- with decidedly mixed results--and has tried the widest range of grain storage and marketing institutions and policies. Third, the govern- ment marketing agency, OPAM (Office des Produits Agricoles da Mali), part of the Ministry of Commerce and Finance, is the largest in the Sahel in terms of its personnel, facilities and experience in cereals marketing. OPAM, as a result of a growing realization that a combi- nation Of government policies was leading to increased stagnation in commercial food production and marketing (compounded by the adverse effects of the drought) and as a result of its inability to achieve even a part of its own stated objectives with the means available is, 165 like most of the other Sahelian cereals boards, undergoing a period of extensive study and reorganization.12 A National Commission to restructure OPAM was formed in 1976 and its policy recommendations are expected in the near future. The basic geography of local cereal surpluses and deficits helps explain a good part of the major Operational activities of each Sahel- ian cereals board. In Mali the zones of chronic cereals deficit are those involving the larger cities (e.g., Bamako, Kayes, Mopti) and the Sixth Region, a vast area abutting the Niger border around the desert town of Gao (see Figure III-1 in Chapter III). Regions of sur- plus production are generally located in the south-central part of the country. OPAM's infrastructure and personnel are distributed primar- ily to facilitate the seasonal transfer of this domestic surplus and, as we will see, are less optimally located to serve the local needs of producers in areas of balanced production and consumption. The staff of OPAM consists of administrative, warehouse, transportation, purchase and sales personnel who, numbering about 750, have been given a theoretical monopoly on all domestic marketing of coarse grain in addition to the fairly minor export of fruit and vegetables. The annual "marketing campaign" is a complex process and will be described in some detail since it is typical in many respects of pro- cedures used in the other countries. It begins with regional OPAM 12In this chapter it is impossible to do justice to the wealth of documentation on cereals marketing and storage in Mali, some of which is a product of this very recent study and reorganization plan- ning. For somewhat older background information see Food and Feed Grain Institute [1970], Panhuys [1973], Robbins and Garvey [1972] and UN/FAO [1973b, l974b]. For more recent information, see Haik, et al. [1976], IBRD [1975c], Richard and Van Den Berg [1975], Steedman, et al. [1976] and USAID [1975b]. 166 advisory councils summarizing local stock levels, likely regional cer- eal surpluses and deficits and estimated needs for Operating funds from the central government. These calculations are forwarded to OPAM central Offices in Bamako where they are to be used in the re- gional allocation of available credit and personnel and to plan needed inter-regional cereal transfers. This information, along with fore- casts Of cereal availability, market conditions, etc., are theoreti- cally used to set Official grain price spreads in the marketing channels between the producer and the urban consumer. There is very little evidence, however, that Mali, or any of the other countries, uses anything beyond the crudest notions of supply and demand, a rough national averaging of operating costs and past "rules of thum " in setting these price levels. The setting of the consumer (i.e., urban) price has usually received the greatest attention and is subject to great political sensitivity. It is Often alleged that, within cer- tain limits, the price setting process proceeds as follows: a "rea- sonable" consumer price is selected, rough cereals Office per ton Operating costs (and anticipated government subsidies) are subtracted, both of which produce, in turn, a "practicable" producer price.13 A schedule of these marketing margins, called a bareme in French, is established and published by decree at the beginning of the harvest season. The 1974-75 Malian seasonal bareme for millet is illustrated in Table VII-1. 13In Niger, OPVN officials reported a similar process and when it produced a producer price that was "too low", various categories Of intermediate Operating costs were simply reduced by 50 percent to produce a more "acceptable" price structure. 167 TABLE VII-1 MALIAN "BAREME" (MARKETING MARGINS) FOR MILLET, 1974-75 Malian Francs Per Ton Producer price Transport costs (producer's) Transport costs (cooperative's) Purchase cost to OPAM Losses Finance costs OPAM tax (management fee) OPAMis transport costs Wholesale price Retail margin Retail price 32,000 2,150 2,900 37,050 800 2,130 5,000 5,000 49,980 (rounded to 50,000) 1,500 51,000 SOURCE: Haik [1976], quoted in Steedman, et al.[1976, p. 110]. 168 The seasonal financing implied in the bareme, particularly fund- ing for payment of the producer price and local transport costs, is obtained on credit from the Development Bank (the BDM) and for- warded in segments through the administrative hierarchy (circles, arrondissements) to the local level. The timing of these funding dis- bursements is critical, particularly if the agency is trying to mount a credible price support effort at harvest time. Unfortunately, the monies Often do not arrive on time and delays in second or third dis- bursements can interrupt the rhythm Of a buying campaign. Two further elements enter into the buying campaign. First, OPAM purchases are made by using the intermediary services of two other structures which are important in Malian development policy, a series of regional development agencies--known as Operations or Actions, and the extensive system Of marketing cooperatives in rural areas. Second- ly, a system of purchase quotas are assigned to the village coopera- tives. Quotas will not be discussed here except to offer the obser- vation that some producer cooperatives have had to go into the Open market and purchase grain in order to make their required quota de- livery tO OPAM buying points. There are at least thirteen separate Operations and Actions in Mali in addition to the existing central, regional and local govern- ment administrative structures. Operations cover economic or agri- cultural regions rather than administrative areas and have a certain measure of financial and administrative autonomy. Some are more gen- eral purpose in nature and promote both food and cash crops while others focus more specifically on single crops (e.g., rice: Office de Niger, Operation Biz Mbpti, Operation Biz segou; millet: Operation 169 Mil Mbpti, Operation Mil Kaarta, etc.).14 In areas where these organ- izations are active, they act as purchasing agents for OPAM, which allows them, in somewhat traditional fashion, to deduct principle and interest payments for previous credit in kind from the farmer's gross crop receipts. There is often some reluctance on the part of the Operations to increase their cereals marketing activities since the bareme often does not permit a full recovery of costs, particularly those for transportation. This is illustrative of the kinds of pro- blems induced by the fixed cost and price margins of the bareme when account must be taken of real differences in local transport and oper- ating costs. Fortunately, there is some adjustment made in urban re- tail prices to account for higher transport costs to more remote areas. Outside the effective areas of the Operations which have strong grain production activities, cereal purchases are conducted using Mali's extensive system of rural cooperatives which have been correct— ly called "para-administrative organizations",15 rather than true cooperatives since they perform little more than marketing functions to facilitate the work of state agencies. As we have seen money ad- vances are forwarded through the administrative hierarchy to the coop- erative and the barEme includes fixed margins to cover producer trans- port costs and the cooperative's overhead and transport costs to the cercle level where grain goes to OPAM warehouses or buying points. Often the transportation allowance is apparently not paid to the farmer and the cooperatives may face difficulties in covering costs ll.See Steedman, et al. [1976], particularly pages 32-39, for a good description of the Operations. 15Panhuys [1973, p. 6]. 170 if the distances involved are substantially greater than the average. Cooperatives are not allowed to store any OPAM grain at the local level for subsequent resale. Instead, it is transported to the cercle level and then back to the local level with a very small re-transport allowb ance (1,500 N? per ton) generally not adequate to cover costs. At the cercle level, stocks in excess of the local resale quota are shipped to regional OPAM storage or transshipment facilities. OPAM was able to purchase an estimated 15 to 50 percent of the marketed surplus of cereals in pre-drought years which left the rest of the market to the commercial sector which has often been forced to operate clandestinely. Experiments to license various types of trad- ers to permit them to earn a fixed margin in purchases in areas not covered by OPAM have largely failed. Among the problems which have arisen, failure to respect the official producer price, speculative misuse of advanced funds, and participation only when OPAM has pro- vided higher prices than the open market, have been most frequently cited. ‘ OPAM's own transportation capacity is quite limited and it large- ly relies on services obtained from.Mali government river, rail and road transport companies (i.e., COMANAV, Cbmpanie Malienne de Naviga- tion; RCFM, le Régie des chemins de Fer da.Mali; CMTR, Cbmpanie Malienne des Transports Routiers), as well as licensed private truck- ers. Again, it is often difficult for both the private and government hauler to adjust his costs for variable road and load conditions given flat kilometer per ton rates for paved and dirt roads. OPAM storage, the primary infrastructural component of its mar- keting and stock system, is the most extensive available to any 171 Sahelian cereals board. In addition to the relatively minor use of rented warehouse space, OPAM, in 1975, had 91,500 MT of storage ca- pacity available, most of it in the major urban deficit areas and in about half the cercle seats. (The other half of the cercles must relie on rented facilities often not entirely suitable for grain storage or incur costs by exporting and then reimporting cereals into the local region.) There are few if any official inducements to hold farm-level or village (cooperative) stocks which, for certain stock objectives, would provide a much cheaper alternative to increasing the number of centralized government bag warehouses. However, it is also clear that if OPAM is going to attempt to hold emergency, buffer and pipeline stocks, its storage capacity and utilization procedures will have to be improved in order to meet at least some part of the various combina- tions of storage requirements that differing aggregate supply and de- mand conditions would produce. It has plans to approximately double its capacity within the next few years, again with disproportionately large amounts of the new capacity going to urban areas. Niger Until 1970 the Niger government attempted to augment sorghum and millet supplies and stabilize prices in Niamey, the capital, through a limited buying program which was hampered by insufficient operating funds and overly restrictive margins between producer and consumer prices which resulted in chronic financial deficits. Since 1970 cereal storage and marketing in Niger has been the responsibility of l'Off‘ice des Produits Vivriers du Niger (opvu) .16 16Information in this section comes from a review of numerous Government of Niger and USAID documents and the following: Becker 172 The primary objectives of OPVN have been price stabilization, inter-regional grain supply management and collaboration with other cereals boards in neighboring countries. The creation of OPVN was made possible by a multinational technical and capital assistance program.ohanneled through the conseil de l'Entente, one of the re- gion's relatively weak political unions.l7 In recent years, due to a relative improvement in producer prices for cereals versus cash crops (groundnuts primarily), a larger percentage of OPVN's cereals purchases has been made through the assistance of the marketing coop- eratives of the UNCC (Onion Nigerienne db Credit et de Cooperation) which until a recent decline in world prices had concentrated its efforts almost exclusively on groundnuts. Niger's geography strongly affects the location and logistical requirements of any cereals storage and marketing program. Most cer- eal production and marketing is spread over a relatively narrow east- west belt which follows the country's southern border between Chad and Upper Volta. The recently completed paving of the "highway of nation- al friendship and unity" has effectively linked the major towns across this belt but transportation is still very difficult in the isolated deficit areas in the northern part of the country. Niger would clear- ly be the country most favorably located to take advantage, on a decen- tralized basis, of cereals trade with northern Nigeria. [1974], Burke, et al. [1971], IBRD [l975c], Jomni [1975], Republic of Niger, OPVN [1972, 1975], Pattinson [1975], Reed [1974], UN/FAO [l973d, 1974d, 1975b, l975d], USAID [l975b, 1976b] and Wilcock [1976a]. In addition, many useful insights were gained in unpub- lished reports and in personal interviews with OPVN officials con- ducted in Niamey in November 1975. 17See the following section on regional stock efforts for more detail on this program. 173 Niger, among Sahel countries, was one of the most severely affected by the recent drought and, along with Mali, it received the largest total shipments of relief grain. After one year of operations, when it worked quite successfully with a very limited infrastructure of several trucks and almost no storage capacity, OPVN mushroomed in size as it became primarily a relief distribution agency to the rela- tive abandonment of its cereal price stabilization efforts. Relief distributions were handled in cooperation with the gendarmerie nation- ale and the magnitude and urgency of these operations led to a general erosion of accounting accuracy, vehicle maintenance and the ability to stabilize local markets through selective buying and selling. OPVN's truck fleet grew rapidly to about 80 vehicles by 1975 but it has been hampered by repair problems. Similarly, storage capacity grew from about 4,000 MT in 1971 to 35,000 MI of warehouse storage in 1975. (Currently modern warehouse capacity is being increased and, in addi- tion, there are about 3,000 MT of mud-brick local storage facilities which are used but inadequate.) In general there is a strong ten- dency to increase physical infrastructure even before a consolidated buffer and emergency stock program has been devised. OPVN's first buying campaign, 1970-71, was fairly successful de- spite deficiencies in equipment, infrastructure and personnel. This was due to the fact that the agency began with a sizable and readily available operating fund created through the sale of PL480 grain and the fact that it employed a flexible system of regional pricing which permitted variable transport and operating costs to be reasonably covered. There is good evidence that the limited purchase/resale operations it conducted in 1970/71 had observable impacts in 174 increasing producer prices at harvest time and decreasing "soudure" prices, particularly in urban areas.18 Over the next four years, however, a combination of drought relief activities and erratic, politically-based official price policies produced institutional repercussions from.which OPVN is perhaps only beginning to recover and which frequently seemed to have had a destabilizing and disruptive effect on national grain markets and production patterns. First, other OPVN market activities were overshadowed by the magnitude of the relief distribution task which handled 10 to 20 times as much grain over the 1972-73, 1973-74 and 1974-75 seasons. Second, official prices, set by an interministerial Gomité de Céréales, were fixed at uniform national levels which elimdnated previous regional differentials. This, com- bined with chronically insufficient margins, caused OPVN to accumulate operating deficits on its nonrelief market activities. A third by—product of inappropriate and ad hoc policies and ‘widely fluctuating relative prices were the effects of these factors on the allocation of farmer production resources and the distribution of benefits. For example, during the 1973-74 season when grain supplies were shortest, producer prices were increased to 25 CFA/kg in order to try and attract greater supplies, largely to no effect. Then during the 1974-75 season aggregate levels of production were greatly increased but prices remained at 25 CPA/kg. Traders who were acquiring most cereal supplies at the village level at 12 to 15 CPA/kg resold them to OPVN at the official price. As a result it has been estimated that about one-half of the three billion CFA operating fund was paid to private traders who realized an estimated net profit of 18new: [1976b, pp. 27-28]. 175 about 700 million CFA.19 Reduced harvests in the next season, 1975- 76, reversed this pattern with traders making very few deliveries to OPVN since unofficial market prices were running substantially higher than 25 CFA. This was compounded by the fact that much of the cereal trade had moved out of public market places because of police sur- veillance of prices making OPVN's consumer protection task virtually impossible to accomplish. Even here there was an exception as offi- cial prices for Niébé (cowpeas, profitably exported for several years to northern Nigeria) were far higher than prevailing domestic market prices with some evidence of substantial diversion from cereal produc- tion. These costly and destabilizing actions and the growth of a larger and less productive OPVN bureaucracy illustrate the risks of overly rapid institutional growth which, under highly unstable market con- ditions, tend to benefit traders and selected groups of powerful urban consumers much more than producers. This is confirmed in the stern conditions being imposed on OPVN operations associated with additional FAO coordinated assistance for the institution of an emer- gency stock system and the reestablishment of a reduced but more finan- cially viable and efficient buffer stock program. 19Jomni [1975, p. 3]. These negative distributional impacts are almost completely opposite to the conclusions reached by USAID [1976b], which felt that most of the benefits had gone to producers. 176 Considerations for Sahel-Wide Grain Storage Efforts There are good arguments to be made for Sahel or west African regional cooperation in grain storage and marketing programe. These are due to similarities and complementarities in production patterns, high transport costs for imported grain, large volumes of grain smuggled across borders to take advantage of significant differences among market and official prices, similarities in institutions and types of government intervention in grain markets, and a strongly expressed desire to enhance food self-sufficiency. Patterns of corre- lation in production variation reported in Chapter IV indicate that, in normal years, there are intra- and inter-national variations in grain supplies which are potentially off-setting. In addition, rela- tive comparative advantage and traditional patterns of molti-national grain trade indicate strong possibilities for cereals trade for mu- tual benefit. Exchanges among cereals offices have occurred on a small-scale, ad hoc basis in the past few years but there are very few operational mechanisms to facilitate the necessary flow of information and grain supplies. There are, however, a number of new institutions which might serve as vehicles for this type of regional economic and institutional cooperation. For example, a francophone West African economic union, CEAO (Commauté Economique de l 'Afrique de l 'Ouest) with headquarters in Ouagadougou, has recently been formed, joined by an even newer and larger organization, ECOWAS (the Economic Community of west African States) which also includes the anglophone states, most importantly Nigeria. In addition, the Sahelian organizations of CILSS and the Club des Amis du Sahel (described in Chapter I) 177 might be in the position to promote regional cooperation through coor- dinated technical and capital assistance programs. we should not, however, be overly optimistic since there are strong traditions of lack of cooperation in West Africa even where mutual interest is fairly obvious. Some of this may be casually attri- buted to differences between anglophone and francophone traditions and to the seemingly necessary development of competitive nationalism. In contraposition, we have also seen strong inherited patterns of deliberate or functional/structural fragmentation of the Sahelian per- iphery in order to enhance colonial rule and trade. These traditions reinforce the fact that boundaries facilitate differential flows of productive resources which are vital to the process of unequal capital accumulation in center-periphery structures. This certainly has been true in west Africa with the materially advanced states such as Sene- gal and the Ivory Coast being more hesitant to participate in any structure which would impinge on their position as regional economic centers. This bleak assessment, however, need not be that discourag— ing. The most important lesson which can be drawn from.the relative success of economic unions in different parts of the world is that functional integration for mutual gain is much easier to achieve and lays the needed groundwork for considering political union at a later date. Grain storage and marketing may be one of these functional areas where a series of flexible bilateral and multilateral arrangements ‘might be built between cereals offices, under the auspices of a larger, less-focused regional grouping. We may benefit from an examination of the experience gained in one "regional" grain stabilization project and Nigeria's plans for 178 a comprehensive grain marketing and storage system. We must, of course, also evaluate factors such as the effects of road improve- ments on international transport costs and the desirability and fea- sibility of coordinating stock movements and producer prices. These will be covered in later chapters. Oonseil de l'Entente Stabilization Project As earlier mentioned, Niger and Upper Volta have received tech- nical and commodity assistance from.a multinational aid project (con- tributions were primarily from France, Canada and the United States) channeled through the Cbnseil de l'Entente, a regional organization ‘with its headquarters in Abidjan, Ivory Coast. The following summary observations on this regional grain marketing and storage project may be instructive for future efforts. First, the choice of regional organization is critical and can greatly influence the nature of the interaction possible to undertake. The Cbnseil de l'Entente includes only two Sahelian countries in its five members (Benin, Ivory Coast, Niger, Togo and Upper Volta) and has no anglophone participation. There seems to be very little rea- son for its existence other than historical circumstance, particularly given the limited range of activities it promotes (primarily sponsor- ing infrequent Presidential summit meetings and operating a loan guarantee fund) and particularly with more broad-based, functionally coherent alternate regional organizations now taking form. Second, the Entente Stabilization Project was originally designed to include all five Entente partners and Ghana which they encircle.20 20See, for example, Checchi and Company [1970], Food and Feed Grain Institute [1970] and USAID [n.d. and 1976b]. 179 Although there was substantial attention devoted to functional inter- state exchanges this admirable regional focus was quickly abandoned. Ghana never seriously participated in the scheme and the three coast- al countries apparently did not have as great a need or interest in the establishment of food grain stocks and new patterns of food mar- ket intervention. This left Niger and Upper Volta in the picture. Several Entente personnel admitted that virtually the only remaining rationale for channeling funding through that organization was to circumvent limitations on the number of bilateral aid agreements in- dividual countries, particularly the United States, could enter into at that time. Third, patterns of "inter-office" interaction established under the project have not always been the most productive. Virtually no functional contact or systematic information exchange between the two offices was promoted or, seemingly, even contemplated. Contact and coordination, when it occurred, were often through the intermedi- ary of the Entente office in Abidjan. These procedures did not pro- mote functional interaction and seemingly only added another layer of costly administration between the cereals offices and sources of needed assistance. Fourth, the Entente Regional Cereals Office, which received much of the available project technical assistance, was placed in Niamey and, as a consequence, OPVN received a highly disproportionate share of the training, planning and operational assistance which certainly did little to strengthen OFNACER, already hampered by inappropriate institutional and financial structures and arrangements. While the location of regional institutions lends itself to the host country 180 receiving a larger share of total benefits, other mechanisms might have helped diffuse available assistance more evenly. Finally, donor dissatisfaction with certain national policies-- particularly with Upper Volta's granting of monopoly buying powers to the ORD's-hampered the operation of the entire program.hy delay- ing the project's second phase funding in an attempt to pressure Upper Volta to rescind this objectionable policy. The resulting delays in obtaining needed material and technical assistance further weakened the capacity of both offices to begin to adjust to more normal oper- ations in the years immediately following the drought. These con- flicts have been resolved, however, and the project has been continued under a reduced level of funding. In summary, the Entente Grain Stab- ilization Project was one born of admirable intentions whose effective- ness has been blunted by attachment to a weak, highly political re- gional organization, by the participation of only two Sahelian coun- tries in the scheme, by the disproportionate concentration of project resources with one board, and by the failure to build more functional or operational interboard contacts. Proposals for Nigerian Food Security In the past five years, Nigeria, particularly in its northern areas, has faced food production and marketing problems similar to those in the Sahel. Food production, particularly sorghum and millet was hard hit in 1973, the worse drought year, but this only represents an exaggeration of prevailing trends in declining per capita output. Similar phenomena of rapidly growing urbanization, aggregate market food demand and food imports have also been quite apparent. 181 The federal government of Nigeria has decided to try and meet this problem in several ways. First, it is placing increasing empha- sis on augmenting private and public sector food production through efforts in its university agricultural research centers and through programs such as the National Accelerated Food Production Program (NAFPP). Second, it is establishing a National Grains Production Company (NGPC) with wide terms of reference to stimulate private and begin public production of grain as well as engage in grain procure- ment, storage, distribution, processing and marketing. Recently, a Nigerian Consultative Group and an FAO Food Security Mission completed a major joint study of the situation with recommendations concerning the operation of the NGPC, other facilitating agricultural sector policies, and specific measures which need to be taken to maximize the viability of the national grain storage program. In the latter program the federal government is committed to the installation of a minimum of 250,000 Ml? of grain storage facilities to be coordinated with those planned by state governments amounting to an additional 350,000 MT.21 In terms of an overall food security program the report first recommends that the state and federal emergency and buffer stock actions be carefully tied to each other, and to the NAFPP and NGPC programs to provide farmers with increased assistance in production technology and in farm level storage. Second, it recommends that pre-planting minimum producer prices be instituted in certain regions on a pilot basis. These prices would be set on a flexible regional 21Nigeria, Federal Ministry of Agriculture and Rural Develop- ment [1976]. 182 basis following the recommendations of a proposed Federal Agricultural Price Commission and would be payable to farmers at numerous state and federal grain depots and storage points. Intra-seasonal price stabilization would be largely the respon- sibility of state storage and marketing authorities who could call on federal backeup support if need be. Inter-annual price stabili- zation is a much more demanding task and would require much larger amounts of storage capacity, some of which is now slated for emergency stocks; this activity is therefore largely confined to the future to be built on experience in handling yearly stabilization and emergency stock turnover and storage losses. In addition to very detailed recommendations on the structure and functioning of the MGPC, the report makes several specific recommenda— tions relevant to this study. First, it recommends that the agency's storage functions be considered by the federal government as a service whose full operating cost should be covered by appropriate prices and margins. Second, it recommends that new stock facilities be of proven, uniform design for bag storage (except for two bulk storage facilities to be maintained for training and experimental purposes until the costs of this system can be effectively reduced), and that they be built by one contractor carefully supervised by inspection and auditing teams. It also recommends the installation of centralized grain drying facili- ties at several key locations so that grain can be dried, stored and profitably salvaged. The training of decentralized storage facility management and warehouse personnel receives heavy emphasis. A third major set of recommendations deals with improving farm level storage through improved extension services and an expanded 183 program of providing technical storage inputs to producers. A fourth and final point is a strongly expressed desire to promote a regional emergency and buffer stock coordination program with "neighboring states, especially those in the drought-prone Sahelian zone". It seems clear that Nigeria will offer a major opportunity for the Sahelian cereals office to profit both from a strong regional supply source and from the lessons of a food security system conceived and executed along somewhat different lines than their own. Sahelian Food Grain Marketing and Storage Institutions and Policies: A Summary In this review of Sahelian cereals marketing and storage institu- tions and policies, we have seen a number of common recurrent features which are summarized in this section. We have also seen that the phy- sical environments of the different Sahel countries, while sharing certain common characteristics, also differ substantially in detail which helps to define specific physical, logistical and institutional food security problems and possible structural and functional remedies. Among these cereals offices the oldest has been in existence for only about ten years and their collective recent history has been marked by the destabilizing effects of massive drought relief activities. The overall performance of grain boards in the Sahel has been quite poor to date. The offices have not exercised much real influ- ence in cereals markets and their actions have often contributed to the misallocation of production resources leading to negative dis- tribution consequences as urban consumers and private traders have often been the chief recipients of program benefits. In contrast to this bleak assessment, however, the potential of cereals marketing 184 and storage institutions to promote greater food production, reduce producer risk as well as stabilizing urban grain supplies has been selectively demonstrated. The objective now should be to help identify the structural and functional elements of the present sys- tem of grain marketing and storage programs which are amenable to change. The following summary is designed to provide insights and lessons which can be used to improve food security institutions in the Sahel. Institutional Structure The grain boards are often hampered in the accomplishment of their various objectives by inappropriate institutional structures. In some cases marketing and storage responsibilities are divided among various agencies with coordination problems and duplication of effort resulting, particularly if these agencies are in diff- erent ministries. This may also occur if there are a number of agencies along the food channel between producer and consumer leading to unnecessary and costly grain exchanges. Further, the negative effects of suboptional structural arrangements are exac- erbated by a lack of coherent food policies. All the cereals offices, like other Sahelian institutions, are structured administratively or "from the top down" leading gen- erally to disfunctional overcentralization in what are still decentralized economies. Attempting to force such structures onto the localized cereal economy leads to very high operating costs and negative distributional effects in favor of the urban population. 185 Several of the offices are weakened by having to fill multiple and sometimes conflicting missions. This is particularly true for those organizations which are responsible for cash crop as well as food marketing and storage and the provision of credit and inputs to producers. While there are often savings in logistical costs when one institution handles multiple roles, problems stem largely from differing institutional behavioral norms associated with accomplishing these tasks. In some cases, the structures and patterns of incentives which characterize intermediate regional development organizations and local cooperatives are largely extensions of patterns of top-down, administrative bureaucracy which serve to channel up- wards expropriated agricultural surplus to support urban sectors, functionally linked to the developed Center of the internatonal economy. Again, what one could call "cash crop thinking" is applied to food crops, particularly as supplies increase with the "deepening" of urban food markets. Ineffective official monopoly powers only make governments look foolish, anger certain donor agencies and often put even greater financial benefits into the hands of the private commercial sec- tor. I Similarities in structure, function, objectives and statutory powers of cereals offices should serve to facilitate direct re- gional inter-office contacts and cooperation although very little of this has yet occurred. This can largely be attributed to the functioning of "feudal interaction structures" (see Chapter II) focused on each country's frail center and thus linked to the 186 outside world. There is a tendency for any formal international contract to pass through and be limited by the common denomina- tor of the degree of inter-state political cooperation which is generally very law. There is room for the promotion of direct functional contacts among cereals offices particularly as infra- structural improvements take place. The coastal countries are vital to the structure and functioning of evolving West African regional economy and thus must be included in any viable regional food-security strategy, particularly Nigeria which produces about one-half of west Africa's millet and sorghum. Objectives All the boards have similar formal objectives for their cereals marketing and storage activities. As we have seen, these include establishment of readily available emergency reserve stocks, facilitating internal grain transfers (including development of market infrastructure as food markets "deepen" with urbaniza- tion), and the intra-annual and inter-annual stabilization of both producer and consumer prices and incomes. We have also seen that particularly within a center-periphery framework there are inherent contradictions among these objectives. Given these conflicts in stated objectives, some "modernizing" institutions, controlled by the urban sector have largely looked to the top of vertical marketing channels. In some respects the offices have functionally been a part of an integrated system which results in the transfer of agricultural surplus to urban areas in a manner which does not promote balanced economic growth. The activities of the offices have in some instances specifically 187 benefitted only government officials and their families in urban centers and smaller towns in ways reminiscent of colonial pat- terns of forced food deliveries. The most fundamental conflict of objectives is between support- ing producer (rural) and consumer (urban) interests. This con- flict is seen in two forms in the functions of Sahelian cereals offices and related food policies. One is the choice between stabilizing (and often subsidizing) urban food prices versus stabilizing and " subsidizing" producer incomes through localized food stocks and higher prices for food transferred to urban areas. The second is the use of a food marketing and storage system for revenue generation versus its use as a risk-minimizing pro- ductive service provided to different sectors of the economy "at cost". This conflict is seen, for example, in Senegal's ONCAD making money in its food transactions and, as we will see in the next chapter, the 0RD's in Upper Volta hoping to generate operat- ing revenues through their agricultural marketing activities. Finally, there has been some tendency for governments to use food marketing and storage for direct political ends through urban food subsidies and special belowbcost food deliveries to speci- fic urban and regional constituencies where political support is weak. This undermines the operational viability and integrity of the cereals offices. Means to Implement Objectives There are four main ingredients to the operation of a cereals marketing and storage program: storage facilities, transporta- tion, trained personnel and operating funds. In all respects, 188 the Sahelian systems are urban-oriented, overcentralized, and top heavy in terms of being able to achieve some balance in the standard range of objectives assigned to them. Sahelian cereals offices have often increased in size and infrastructural capacity at a rapid, discontinuous pace so that very little "learning by doing" (which characterizes the private grain trade) takes place. This leads to excessive waste, equipment maintenance problems and chronic budgetary reliance on outside funding. Storage facilities are generally centralized in the capital city and a few larger urban centers. This facilitates the commercial sale of grain to the urban consumer and will be useful in the handling of emergency stocks donated or purchased from the out- side. It is less optimal in helping to dampen intra-annual and inter-annual supply and price variation for the majority of the population, rural producers. Large-scale bag storage facilities have been favored over improvements in farmrlevel or village storage technology. Modern bulk storage is generally not used although it has been advocated for the holding of permanent emer- gency stocks in centralized locations. Transportation is an obvious problem with all of the offices. The centralization of stocks can lead to unneeded rehauling to the village level if resale to rural people is part of the pro- gram. Improvements could be made in the use of available trans- portation facilities. Farmer transportation to grain office buy- ing points is a major constraint on a viable price support, stor- age and inter-regional transfer system. 189 There has not been adequate training of personnel in techniques of grain storage and in market intervention strategies. More seriously, the structure of incentives tends not to support operating efficiency. Most office employees are "civil ser- vants" whose loyalties and career incentives are not functionally dependent on operational success. Further, at the local level, there are virtually no means of producer or other local control over the hiring and promotion of office employees at that level; all of these decisions "descend" administratively. Operational funding is controlled at the center and "allocated" to local areas for a given purpose. While theoretically having financial autonomy, the offices are often dependent on a flow of funds from agencies (development banks, etc.) frequently con- trolled by other ministries. Funds are chronically late in reaching the field, allowances for producer or local transport are often not paid, and the level of funding is usually unsuffi- cient to purchase available supplies. Very often this situation plays into the hands of the commercial sector which is more cap— able of efficiently managing its operating capital. Price Policy Price policy is undoubtedly one of the biggest constraints on the successful operation of grain marketing and storage programs. Deliberate buying experiments and the occasional high producer price effectively administered at the village level clearly demonstrate that farmers will respond to price incentives. The opposite situation is more often the case, however, and when 190 conditions are bad enough farmers will simply withdraw from the market to the extent possible. The timing and composition of the annual cereal baremes are of utmost importance. Cereals offices are forced to accept margins between producer and consumer prices which often cannot cover costs, particularly when costs are higher than need be through bureaucratic bloat and operational inefficiency. The rigidity of uniform baréme price levels (themselves function- ally more necessary with a high degree of institutional and fi- nancial centralization) is compounded by fixed transportation and other cost allowances which do not allow for reasonable re- gional variation. This can lead to a misallocation of productive agricultural resources and it hamstrings the offices by denying them the flexibility to take quick market actions and to rene- gotiate prices (to institutional buyers for example) under chang- ing supply conditions. It also encourages smuggling to take ad- vantage of wide official price differentials (e.g., in the 1976-77 buying season substantial quantities of grain were reported mov- ing from Niger to northern Nigeria to take advantage of official prices which were about twice as high, most likely due to Niger- ian purchases for emergency stock constitution). Various "inter- mediate" pricing systems such as offering uniform producer prices at government depots but covering actual kilometer transportation costs for village cooperatives might be administratively feasi- ble. The timing and support of offical prices is also critical. Prices are generally set well into the harvest season, thus providing 191 little incentive to the producer and injecting needless delays into operations. When prices happen to be attractive to farm- ers, the marginal credibility of government organizations may be further eroded by the lack of funding to purchase all offered supplies or by not getting funds to buying points until too late in the season. Relief Efforts The international drought relief effort which centered on the years 1973 through 1975 provided the Sahelian countries with almost 1.24 million metric tons of pledged and/or delivered grain worth an estimated $410 million.22 This effort undoubtedly saved many persons from.atarvation and helped reduce the risk of severe malnutrition for an even larger population. It also pro- vided the Sahelian governments and donor agencies with some val- uable lessons in relief coordination and management, and helped provoke international responses such as the FAQ World Food Con- ference in 1974, FAO food security emergency stock programs and improvements in global early warning systems for food shortages. The relief effort, however, was not entirely without controversy both in the "developed world"23 and in the Sahel, some of which has been reflected in the cereals offices. The lack of usable transportation and storage infrastructure provided very serious bottlenecks in the efficient distribution of Sahelian relief supplies. Major problems occurred in: 22For more detail see USAID [1976a]. 23For example, see Sheets and Morris [1974] and Glantz [1976]. 192 a. Keeping grain traffic moving through coastal ports (primar- ily Dakar, Abidjan, Cotonou, Lagos, Port Harcourt) particu- larly those in Benin and Nigeria; b. Time consuming sacking and handling in the ports of bulk- shipped grain; c. Transportation to the four inland countries (and interior distribution in Mauritania) which absorbed a very large part of available government and private hauling capacity; d. Transportation from internal distribution points to village consumption areas. Measures such as increasing port capacity, finally paving all of the key west African road arteries, and the current installation of greater storage capacity should help alleviate these problems in the future. Distribution costs were extremely high for some of the relief supplies due to the use of ad hoc methods such as air lifts and truck convoys across the Sahara from Algerian ports. This provides a strong argument for improving infrastructural capacity and for financing the recurrent costs of properly managed emer- gency stocks. This will be discussed in greater detail in the next chapters of this study. Various institutional combinations were used to carry out relief supply distribution. In some cases this was done by the Army and the Gendarmerie Nationale using their own logistical capacity supplemented with donated trucks and airlifts. In other countries, some or most of the relief distribution responsibility went to the cereals office which was then converted from a market 193 participant into a distribution system which required and pro- duced a new set of cost structures, standard operating proce- dures and performance norms. This seemed to have a long-run adverse effect on several of the offices suggesting that special use of predesignated organizations such as the army or the police (or the Federal Government in the case of Nigeria) may be more appropriate but only if carefully coordinated at the national level with the cereals office so that confusing and destabilizing situations do not occur. It is important that emergency distri- bution mechanisms be seen as temporary and that they be dis- mantled as soon as possible so that cereals offices can resume their normal functions. In addition to "destabilizing" the cereals offices relief sup- plies also served at times to disrupt what remained of local mar- kets, particularly when unusally large grain supplies were dumped at low prices or provided free of charge (which was required of some donated grain). There have also been numerous instances cited in which farm level planting patterns were substantially modified due to the presence or expectation of continued low cost grain availability. Many of these instances can, of course, be eliminated through improved market intelligence and crop report- ing services and better stock flow management. CHAPTER VIII UPPER VOLTA: A CASE STUDY IN THE IMPLEMENTATION OF GRAIN MARKETING AND STORAGE PROGRAMS IN THE SAHEL Introduction This study has moved forward developing and applying a center- periphery structural dependency model to general trends in economic change and to more specific food grain marketing and storage problems in Sahelian West Africa. This process continues in the present chap- ter as Upper Volta is examined as a case study of the complex problems faced in the development and implementation of national food grain storage and distribution programs. The analysis here will focus on the practical constraints faced by administrators attempting to build new structures and operating procedures on the existing infrastructur- al and institutional base and within limits permitted by the political- economic environment. To accomplish this end the chapter will incorporate the following topics: (1) a brief review of the major elements in Upper Volta's rural development strategy, (2) an analysis of the institutions in- volved in government grain marketing and a preliminary assessment of their performance over the past several years, (3) a description of the first two years of application of FAO food security assistance to the holding of Voltaic emergency stocks, (4) a review of current 194 195 proposals for the institution of a village buffer storage program and the linkages between that program and urban supply stabilization, (5) an analysis of the cost of alternative storage and transportation components in a national system, and (6) an outline of an integrated national foodgrain marketing and storage program for Upper Volta which will synthesize the other topics and offer several suggestions for institutional reform. Elements of Voltaic Rural Development Strategy As stressed in previous chapters, Upper Volta has neither the mineral wealth, the large-scale irrigation potential nor the strategic coastal position of some of its West African neighbors. It does, howh ever, have a considerable potential in its large population of small- scale farmers many of whom have migrated to coastal areas on a sea- sonal or permanent basis in recent years. Several key elements in the country's emerging rural development strategy have focused on improv- ing the productive resources and product markets available to this work force within Upper Volta. Since independence the overall direction of Voltaic development politics has been a moderate one, falling between the attempts at more rapid radical transformation practiced in Guinea, Mali and Ghana and the more neo-colonial policies of neighboring Ivory Coast. Re- flecting the ambivalence of those persons and institutions "caught in the middle" of center-periphery relationships, Voltaic rural develop- ment strategy is a combination of imaginative attempts to stimulate local participation in self-directed change and policies which have survived virtually unchanged from the colonial post. 196 The most important component in the overall effort to increase agricultural output and promote rural change is the national system of Regional Development Organizations (0RD's) created in 1966; the 0RD's divide the country into eleven geographic regions and have very wide powers and responsibilities including all farm extension activi- ties, livestock veterinary care, agricultural credit and rural coop- erative formation, rural education and adult literacy, and agricul- tural marketing as well as being empowered to operate a wide range of commercial enterprises such as agricultural processing plants and gas stations. One of the rationale for the system is to bring under one controlling body the various rural development services which tended to be competitive rather than cooperative at the local level; another is to decentralize rural development program planning and implementa- tion so these would would be better adopted to particular local con- ditions. This is consistent with patterns of regional decentralization observed in the other five countries but represents the strongest attempt to combine both rural development administrative coordination and economic development actions into one regional organization.1 A potential criticism of the 0RD or any of the other regional development schemes is that the decentralization of decision-making may not extend much beyond the regional capitals, simply creating a larger number of jobs in a bigger, center-dominated administrative system. Partly in recognition of the inherent limitations to a "top- down" form of rural transformation, the Government of Upper Volta has adopted a specific policy of promoting "community development" at the lSee Gregory [1974] and Eicher, et al. [1976] for more informa- tion on the 0RD system. 197 village level. The village as a whole or voluntarily constituted sub- groupings would, at the provocation of the 0RD extension agent, decide what its goals, needs and capacities are and then be able to call on the agencies of the central government and its regional branches to provide the complementary technical and material support needed to complete its rural development actions. In essence this would be an attempt to build upon and "modernize" traditional structures of labor mobilization and discipline. This is a new policy for Upper Volta with its initial implementa- tion coming in 1974. Each 0RD was asked to select an initial five "community development villages" to receive more intensive study and extension services. This is in addition to the normal charge of pro- moting cooperative on "precooperative" groups of farmers for joint production and marketing ventures. While the implied decentraliza- tion of decision making and the "federal" cost sharing structure is laudable there have been as yet few other basic changes in government structure and incentives which would functionally reinforce the commun- ity development philosophy. Experience in various parts of West Africa has shown that one of the tasks a system of village groups can first engage in is input and product marketing. These marketing activities, if successful, can then be incrementally followed by other collective activities which have greater organizational, logistical and distributional complexity. This has been reflected in Voltaic development thinking and will be explored in later sections of the present chapter. A number of specific policies can be grouped under the general strategy of providing greater resources to the Voltaic farmer. First, 198 and of great importance in a semi-arid country, are efforts to increase the supply of water available over the entire year. This is being accomplished through effective local labor, food-for-work programs of well-digging and the building of small dams for surface water storage undertaken by the Voltaic government and private aid agencies. Second, a major effort is being made to provide more land to farmers through population resettlement from the densely populated central portion of the country to "new land" in major southern river basins which had previously been sparsely or totally uninhabited due to onchocerciasis (river blindness) infestation. Finally, there is a general acceptance in numerous government programs of the somewhat controversial idea of concentrating resources in particular areas or activities in order to constitute a minimum "critical mass" needed to achieve self-sustaining change. This can take place on a geographic basis through the crea- tion of special rural development zones or it can take the form of pro- viding a range of complementary inputs and services required in the production and promotion of particular high-value crops, usually for export. The latter example, of course, risks falling into patterns of export crop production which do little to encourage rural trans- formation but rather rely on the exploitation of a very low effective rural wage rate maintained through the general control of prices, input and consume good supplies, and output marketing services. A final element of Voltaic rural development strategy which is shared at least in part with all of its Sahelian and West African neighbors is a commitment to participate actively (if not dominate) agricultural markets, ostensibly to benefit the rural population. The net effects of this market participation are open to specific 199 empirical analysis but there seems to be little doubt that government agencies will continue their marketing activities. This is explored in greater detail in the following section but first, however, it should be stressed that the reader should not overestimate the impact of these government programs on the lives of the majority of the Vol- taic rural population. While there are few people who are not reached by private markets and government tax collection mechanisms, it is safe to say that large numbers of rural citizens have had no contact with 0RD extension agents nor benefitted from government price support pro- grams. This is another indication that a truely widespread effort at rural transformation will have to further abandon costly patterns of "administrative rural development" and concentrate on employing the existing institutional structures and resources readily available to the rural population. Government Grain Marketingiin Upper Volta Much of what the Government of Upper Volta has attempted in agri- cultural marketing, most specifically in the field of grain marketing and storage, is very similar to actions undertaken in neighboring countries. However, while many of the institutional structures and price policies seen in Chapter VII in the other Sahelian countries apply to Upper Volta, there are specific differences in application and performance which merit our attention. As in Niger, Upper Volta engaged in limited grain marketing and stabilization projects in the 19603, most of which met with little 200 success.2 As part of the multilateral Entente Grain Stabilization and Marketing Project, described above in Chapter VII, the government of Upper VOlta created in 1971, within the Ministry of Commerce, l'Office National des Caréales (OFNACER), and gave it semi-autonomous financial and administrative powers and functional objectives similar to those of other Sahelian grain boards.3 Through drought relief ac- tivities undertaken in Upper Volta, OFNACER has also benefitted from aid in addition to that provided under the Entente project. Its stor- age capacity has increased in the past few years to about 50,000 MT (and new capacity is being financed by the German Federal Republic) and its vehicle pool has also been strengthened. In addition to OFNACER, there are a number of other government institutions involved in food grain marketing in Upper Volta. The roles of these various institutions have evolved rather haphazardly in response to changing circumstances in government policy and in na- tional cereals markets. First, and of most importance, the above-mentioned national system of 0RD's has participated actively if erratically in food grain mar- keting and has been given the general task of making initial cereal 2Local level grain stabilization and forced cereals collections were practiced in the colonial era with the empty concrete silos which dot rural Upper Volta standing as mute reminders of these abandoned practices. 3Information on Upper Volta cereals storage and marketing is based on interviews conducted in Upper Volta in November and December 1975, and October 1976, consultation of numerous unpublished govern- ment memoranda and reports, and on the following published materials: Aithpiard [1974], Eicher et al. [1976], Gregory [1974], IBRD [l975b, l975c], UN/FAO [l973g, l974g, l975e], Republic of Upper Volta, Minis- try of Rural Development [l975a, l975b, 1975e, l976a, l976b, l976c, l976d], SEDES [1973], USAID [l975b, l976b] and Wilcock [1976]. 201 purchases directly from.farmers or from first handlers. The 0RD's on two occasions have been granted total monopoly cereal buying powers and these are still legally (but not practically) in force although several government statements have indicated that this was intended to only be a "monopoly in principle", designed to protect producers from the alleged predatory practices of commercial traders. The total failure of the first granting of monopoly powers in the late 19603 led to various attempts to license private traders to buy grain at offi- cial prices for resale to the 0RD's and to OFNACER. In 1974, with con- siderably more planning, the primary purchasing monopoly was given to the 0RD's for a second time and, due to a new set of unforeseen cir- cumstances, this again produced unsatisfactory results for reasons which will be analyzed below. In addition to lack of success this attempt also had the further negative effect of angering foreign donors pro- viding technical and material support to OFNACER which resulted in aid cutbacks and delays. In fact, by 1976 the general results of ORD par- ticipation in national food grain marketing campaigns had been so bleak that many 0RD's were no longer interested in large-scale cereals marketing. Second, in addition to the 0RD's, there are at least three other institutions within the Ministry of Rural Development which have roles in food grain marketing and storage. There is, first of all, Le Cbmité National pour la Constitution et la Gestion des Stocks de Céréales de Réservesf' created in 1976 as a precondition for the receipt of FAO- coordinated aid in creating a national system of emergency grain stocks. 4The National Committee for the Constitution and Management of Reserve Cereal Stocks. 202 The functions of the Cbmité National will be discussed in the next section on emergency reserves. Also within the same Ministry are the Direction of Agricultural Services, responsible for the recondition- ing and treatment of the grain in emergency stocks, and a previously created Sbus-Cbmité de Lutte contre les Effats de la Sécheresse5 which is charged with distributing grain stocks under certain severe emer- gency conditions. The "Sbus-Cbmité", primarily composed of military personnel and logistical support, was created during the period of massive distribution of drought relief supplies to supplement the meager logistical capacity of OFNACER. Each year major financing for cereals purchasing, aside from an initial OFNACER operating fund provided by the sale of US PL480 grain, comes from a third major institution, the Banque Nationale de Develop- pement (BND) in the Ministry of Finance. Funds have gone directly either to the 0RD's or to OFNACER for distribution to the 0RD's to buy grain on its behalf. Delays and problems in the complex finan- cial and accounting arrangements between the various cereals institu- tions and the BND have caused the latter to withhold needed financing at key points in several buying campaigns. ' Fourth, Upper Volta also has a stabilization fund (caisse de Stab- ilisation des Prim des Produits Agricoles) which provides the 0RD's 'with cash crap purchasing funds which are also channeled through the END. The fund fixes a reference export price and requires that li- censed exporters repay the difference between the actual world market price and the reference price to the fund. Even though Upper Volta's 5"Subcommittee to Combat the Effects of the Drought". 203 cash crop exports are relatively small, since 1968-69 the differences in the two prices for groundnuts, sesame, sheanuts, and cotton have produced relatively large surpluses for the fund even during the worst drought year of 1973-74. Contrary to the ONCAD pattern in Senegal, the fund does not deal with staple foods, but there is a srong rela- tionship between the two because they are both handled by the BND (which has blocked cash crap advances to 0RD's due to failure to repay cereals advances from the previous season) and both are administered at the village level by the 0RD's in competition with private traders. In conjunction with several of the above described institutions, OFNACER, in at least the first four buying seasons, was notably unsuc- cessful in filling its assigned program, primarily that of urban grain supply and price stabilization. The reasons for this failure are many and complex, often not the fault of the institution itself. A number will be specifically highlighted. First, the agency received only a small part of the technical assistance it needed to correctly begin operations, partly due to the malfunctioning of the Entente Cereals project as described in Chapter VII. Second, and most important, OFNACER.was caught in a complex web of institutional duplication and conflict with no distinct national food grain market intervention strategy or authority available to help resolve these conflicts. For example, three ministries had to make a coordinated input into the annual buying campaign: Commerce (OFNACER), Rural Development (the 0RD's, the Cbmité National, the Sous-Oamité, etc.) and Finance (the BND). Differences of opinion and political loyalty and a lack of policy guidance helped sabotage buying campaigns by delaying the arrival of funds at buying points, 204 by the 0RD's selling to special relief organizations (Sous—Cbmité, World Food Program, FAO/ORSO, etc.) at prices higher than what OFNACER could legally pay under poorly designed official price levels, and finally by OFNACER not being able to sell grain in chronic deficit areas at the official retail price because the above-mentioned drought relief organizations were releasing subsidized grain at lower prices. Some recent measures, such as the 0RD's buying grain on OFNACER account and the allocation of grain purchase quotas, have alleviated these con- flicts somewhat but there still remains the overarching problem that too many distinct institutions are involved in grain transactions leading to superfluous exchanges between producer and ultimate con- sumer, each of which adds to total cost in grain marketing channels. In addition to increasing costs, institutional conflict and lack of organization, when they prevent any purchases from taking place, contradict the objective of assisting producers with better prices. In effect, the "monopoly in principle" is handed to private traders who are in an even more advantageous position due to the disorganiza- tion of government agencies. A third problem area, related to the second, has revolved around the composition and timing of the price bareme (or schedule) for the buying season. There has been no regional differentiation in buying prices and these prices seem to have been set with little regard to either market conditions or the financial constraints of the various participants in grain marketing channels. This is particularly crit- ical for the 0RD's which are under pressure to try and generate oper- ating funds from their cash and cereal crop marketing. The timing of price decisions is another area of difficulty. First, producer 205 and consumer prices have been announced at the same time (often well after the harvest has begun) and the buying and selling activities of grain agencies overlap. Thus, grain is being sought and sold concur- rently at two different price levels. Second, if the "soudure" con- sumer price were fixed later in the year OFNACER and other institutions would have the chance to calculate their real operating costs per kilo and have that information enter into the price setting process. This would help improve their financial viability as well as help get funds to the field for purchasing on a more timely basis. Finally, a general problem is that of overly rapid organizational expansion which causes personnel to be poorly trained, equipment to be misused and not adequately repaired, financial controls to be sac- rificed to the urgency of the moment, and elimination of the idea that cereal office expansion would come about incrementally as a result of its proven capacity to intervene successfully in the cereal market place. This was perhaps inevitable under the circumstances of the drought relief program and the growing flood of foreign aid and tech- nical assistance which followed. As we will see, some moves have been taken to tighten operating controls and institute retraining programs for existing personnel. Emeggency Grain Stocks in Upper Volta In Upper Volta, as in most of the other Sahelian countries, the decision to hold separate grain stocks solely for emergency use stems from.the 1972-75 program of drought relief activities, from the 1974 World Food Conference and from the resulting "International Undertak- ing on World Food Security" and the FAO Food Security Assistance Scheme, 206 described in Chapter VI. The government of Upper Volta has adopted emergency stock program objectives similar to those outlined in Chapter VI and, by late 1976, it had reached provisional agreement with FAD/FSAS concerning operational rules and the provision of technical assistance to the program. Emergency Stock Implementation: The First Two Years Upper Volta's current emergency program thus traces its origins to the second PAD Multi-donor Mission to the Sahel in 1974 when the government presented a proposal for a 20,000 MT emergency stock. While the rationale for this stock level may have been derived by ad hoc calculations of the type illustrated in Chapter VI, this was not detailed in any FAD/FSAS reports on Upper Volta or in any of the emergency stock documents available to the author.7 The WOrld Bank investigation of Sahelian emergency stocks accepts this figure probably because it seemed to represent a "reasonable" percentage of estimated normal sorghum/millet production of about 950,000 MT and the estimated quantity of marketed surplus of between 105,000 MT to 122,000 MT.8 6UN/FAO [l976a]. 7See UN/FAO [l973g, l974g, l975e, l976a] and Republic of Upper Volta [l975a, l975b, l976a, l976b, l976d]. The last report does dis- cuss a partial explanation for the stock level in which the popula- tion at risk is defined as that approximately 1.8 million population living in the northern one-third of the country above the "normal" 800 mm rainfall isohyete. Using an average consumption of 180 kg per capita and an estimated 5 month lead time required to receive out- side emergency supplies, a gross shortfall of 135,000 MT was cal- culated. It was then stated, however, that actual practice has shown that a total crop failure and no stored "carry-in" are extremely un- likely and that, therefore, a substantially lower figure would be "reasonable". 8IBRD [l975c, pp. 26-32]. 207 Thus, once launched, the 20,000 MT figure took on a reality of its own with later reports quoting earlier ones to substantiate its authenticity. In any case, a formal request for the funding required to pur- chase this amount of grain (estimated at about $2.7 million) was made to FAO/ORSO in early 1975. Acting very quickly, the German Federal Republic provided about one-third of this amount and some technical assistance and buying began in March of 1975. Even though the buying program got off to a late start, 5,802 MT were accumulated from.March through July of 1975. The average purchase price--reflecting seasonal price rises-dwas about 30 CFA per kg compared to the 1974-75 offi- cial buying price of 22 CFA per kg. Average transportation, handling and storage costs over the first few months added an additional seven CFA per kg and this exhausted available funding. Coming directly after the drought relief distribution experience of 1973-75, these purchased tons were placed in those areas felt to be of potentially greatest need, an arc of small villages in Northern Upper VOlta between the town of Ouahigouya and the isolated village of Sebba (see map in Figure VIII-l which follows). In this first phase, several of the villages were chosen because they are cut off from road transportation during the rainy season. Older bag warehouses of small small capacity (mostly of 300 m2 surface area which can be filled with l to 1.6 MT per m2) were available and put into use but losses have been quite high due to leaky roofs, poorly fitting doors and other openings which permitted rain water, termites and other in- SECtS easy entrace . 208 HuHHH> shaman . .. a ,. 550 :52 / . /.. .1/ u 22; v (I 0.23\\ (.\/ L. r a. . e. .. I... .. x :2... .. . 5:1 / 630,033.3805' .. 8858.12]. 9.85 62355 as»... 8 Ha . 0005. M Ozm0mw2m 45.52. no as... 3.s\ 20:.(004 OZ< mmmhm<300mo 4(20.0mm H<._...O> awn—n5 i0 0.4m3imm 209 The second phase of the FAO/ORSO project planned to buy the remaining 14,198 MT and consideration was given to storage of most of the grain in Ouagadougou to facilitate stock rotation. Early plan- ning in 1975 called for a total expenditure of about $2.5 million to acquire, transport, handle and store these remaining tons at a cost of about 40,000 CFA per MT. An appeal for funding was launched to donors but no operating funds reached the OED purchasing agents until late March 1976, by which time an estimated 90 percent of surplus grain was already in the hands of the commercial sector. Due to the untime- ly arrival of funds, a smaller harvest than the previous year, etc., only about 2,900 MT (of a reduced target of 7,000 MT) were delivered to the Comité National and OFNACER by mid-July 1976. Flexible trans- portation reimbursement procedures helped facilitate cereal movements since the 0RD's could contract with private haulers who were reimbursed at the official government rate of 17 CFA per ton per km. Thus, by July 1976 the emergency stock had grown to 8,700 MT. However, of this amount 1,400 MT had physically deteriorated to the point where it had to be disposed of in order to prevent total loss. This quantity was to be transferred to OFNACER to sell on the open market in one sous-prefecture to stabilize soudure prices. As required in FSAS programs, the receipts from these sales are to be deposited in an account to be used to defray the costs of stock reconstitution.9 To end phase II of the project, an additional 5,000 MT was proposed for purchase from the harvest in the fall of 1976. This would 9It should be pointed out the Upper Volta has adhered fairly closely to the FSAS emergency stock operating principles outlined in Chapter VI. This will presumably be further reinforced when a formal assistance agreement is signed with FAQ/FSAS. 210 leave at least 7,700 MT to be purchased from the 1977 harvest in Phase III for the emergency stock project and would bring the total stock to the desired 20,000 MT with about 5,000 MT held in the north- ern villages and 15,000 MT in Ouagadougou. In actual fact the 1976 harvest was below "normal" with parts of southern and eastern Upper VOlta and northern Ghana experiencing drought-like rainfall conditions similar to, and in some cases worse than, their 1973 levels. Due to actual shortages, intense specula- tion and some trafficing of grain to markets in neighboring countries, grain prices in Voltaic urban and regional markets soared to 60 to 90 CRA per kg even when purchased by 100 kilo sack. Because of these conditions the Comité National requested OFNACER to first sell the emergency reserves held in the Sahel and in July 1977 to then sell the 2,900 MT Ouagadougou reserves. As a result of these actions, OFNACER provided the Cbmité 35 CFA per kg for the Ouagadougou stocks, 31 CFA per kg for the Sahel stocks sold where stored, and 27.50 CFA per kg for those which were transferred before sale. These funds will be added to monies unspent during the 1976-77 harvest for an attempt to reconstitute the entire emergency reserve during the 1977- 78 harvest. In addition, OFNACER disposed of its own remaining sor- ghum and millet stocks and three year old maize stocks originally purchased in Ghana. Lessons and Possible Changes A number of lessons can be drawn from the first several years of emergency stock holding in Upper Volta. One, this has been an ad hoc, iterative process which has been slower and more costly than foreseen partially due to the late arrival of funding, unexpected 211 variation in harvest size and unavailability of supplies on local markets for 0RD purchase. Second, however, even within this period of less than three years, there is evidence of significant institu- tional and personal "learning" with respect to the adoption of pro- cedures for minimizing storage losses with existing infrastructure and the development of viable normal operating procedures for buying, transporting, exchanging title, storing and recycling millet and sor- ghum. Third, total costs per MT have been fairly high due to this learning process, the use of inappropriate facilities and the costs of constructing more adequate storage. A fourth conclusion is that although progress has been made in diminishing structural conflict through better coordination of OFNACER, ORD, and Cbmité National activities, there is still great room to eliminate costly institutional duplication and overlapping areas of responsibility. This is discussed further in the final section of this chapter. Fifth, the initial strategy of stock location may have been called into some question during 1976-77 harvest and "soudure" per- iod as shortages in some southern and eastern parts of the country have been more severe than those in the north. Further, although limited in terms of total tonnage, some grain supplies in the isolated villages of Sabba and Aribinda have not been sold due to adequate local supplies but cannot be moved further south since the roads are not passible during the rainy season. In any case, it is clear that these conditions call for a further, more comprehensive study of (l) the optimal size and location of emergency stocks under a wider range of possible sub-national shortfalls, (2) a more precise 212 identification of the population at risk, and (3) defining more detailed, flexible trigger mechanisme to be used in conjunction with a practi- cal "early warning" system which can provide relevant data to deci- sion makers in Ouagadougou. A sixth and final point relates to the type of storage recently constructed and proposed for future construction. In both cases, this is a combination of prefabricated bag warehouses and silos. In almost every project document there are statements concerning the longer term desirability of using bulk storage silos. For example, the latest progress report notes that, Upper Volta Government and project personnel share the opin- ion . . . that silos to minimize attacks by insects, mold, rats, etc., as well as the transmission of heat to the grain, should be solid, hermetic structures, of either a subterranian or semi- subterranian type . . . Sack storage silos are only considered an intermediate solution . . . A program of long-term storage will necessitate, after detailed study, the construction of 18,000 MT of silo capacity in Ouagadougou (for example, 12 semi-subterranian silos of a type used on Cyprus and in Kenya which each have a 1,500 MT capacity). In the areas not accessible during the rainy season (e.g., Sebba, Oudalan, Aribinda) it will be sufficient to create a 2,000 MT capacity in each location by putting 200 MT metal silos inside exist- ing bag warehouses. In this way, the advantages of both kinds of silos will be combined.10 While this statement contains much of interest, it completely ignores the financial and economic costs of alternative storage sys- tems and the integration of the emergency stock program into an over- all system of food grain marketing and storage. See the final sec- tions of this chapter and particularly Appendix B for more details on relative costs of alternative storage systems. 10Upper Volta [l976d, p. 19]. 213 Buffer Grain Stocks Operations to Date As we saw earlier in this chapter, Upper Volta's efforts to play a stabilizing role in urban and rural cereal markets on either an intra- or inter-annual basis have had very little systematic success to date. Where producer prices have been increased or consumer prices held down, it has been on a very localized basis and largely attri- butable to the chance coincidence of available funds or supplies and an appropriate official price structure. In Chapter VI the objectives of buffer stock holding were discussed and we noted that in the Sahel there are three principle target groups for stabilization actions: sedentary farmer producer/consumers, rural non-producer/consumers (largely nomadic herders), and "urban" non- producer/consumers who live in the country's cities and small towns. Special actions are required to provide quantity or price stabilization "services" to each of these groups. Upper Volta is moving toward an increased ability to provide such services through proposed village storage programs and improving overall OFNACER/0RD coordination to facilitate intra- and inter-regional grain transfers. Proposed Village Grain Stock Program In this section, I discuss the major elements in a proposal for the creation of a cooperative village grain stock system which was developed in the Office of the Permanent Secretary of the Rural Development Coordinating Committee of the Ministry of Rural Develop- ment and presented to FAQ for consideration in late 1976.11 The 11See Upper Volta [l976c]. 214 Rural Development Coordinating Committee and its Permanent Secretary are responsible for the policy, infrastructural and material coordi- nation of the 0RD's discussed earlier in this chapter. The 0RD's, in turn, are charged with coordinating and assisting the second basic unit in VOltaic rural development strategy, the groupement villageois, or local village "pre—cooperative". The proposal expresses concern with three main factors affecting rural life: 1. The distributional effects among farmers facing relatively free local labor and product markets who have differential ability and motivation to deal in those markets 2. Farmers' obvious need for a source of credit in kind or in cash, particularly if they are to adopt new technologies to increase productivity 3. The connection between local grain stocks and national programs to constitute emergency and buffer stocks. What is proposed is a system of village stocks, employing tradi- tional stock technology, owned and managed by the village group or later by a more institutionally advanced service cooperative. The proposal states that the stock program will have three objectives: 1. To buy millet and sorghum from group members who have surplus to sell at prices which will be between commercial market prices and the government's official buying price (the latter apparently is assumed to be higher) 2. To offer at least part of these supplies to national grain mar- keting institutions (or a higher bidder) which will be made possible by bulking the supplies of individual farmers 3. To use the stored value of grain to offer credit in money or in kind to members during the soudure, repayable in kind at the next harvest. The program is based on the use of the groupments villageois, and their number is to be increased to 3,000 over the five year life of the initial program. This means that the 0RD's will have to provide 215 the groupements with greater technical and financial assistance to facilitate accomplishing this and other "self-directed" tasks. The program proposed using and perhaps improving the various types of traditional farm storage structures which are estimated to be capable of holding about 2 MT on the average. This will promote local employment of available resources and reduce construction costs. The program has an initial goal of regularly marketing 12,000 MT of cereals from the village stocks through the 0RD's to OFNACER. This will require some expansion of 0RD bulking warehouse facilities and perhaps the regional implantation of grain drying and treatment facili- ties. Financing for silo construction will initially be provided to each village group as well as an initial operating fund for use in making stock purchases. External funding will be sought for these two functions under grant or IDA loan terms. It will be possible for the village stock to loan grain at a nominal interest rate which would be repaid in kind at the next harvest. Surplus stock sold to OFNACER by way of the 0RD's will earn the village group a portion of the marketing margin between the producer price and the OFNACER acqui- sition price. The proposed program is based on a system of hierarchical coordi- nation where the village groups would hold the stocks and receive assistance from 0RD 'Community Development Agents" (0RD sub-sector and sector chiefs, with each sub-sector chief supervising a number of village extension agents) whose activities would be under the supervision of eleven regional 0RD Cooperative Supervisors who, in turn, would be coordinated by the national head of the Rural 216 Institutions and Credit "Cell" in the Permanent Secretariat of the Rural Development Coordinating Committee. The 0RD's would thus play a critical role in the success of the entire program with their ex- tension activities aimed at "sensitizing" farmers to potential pro- gram benefits, facilitating group formation, providing management assistance, bulking surplus supplies for sale on the account of the village groups, and holding credit funds at the regional level for program operation. Other national cereals organizations would also have key support- ing roles in the program. OFNACER would be obliged to buy grain supplies first from the village stock system before making commer- cial purchases from traders. To the extent that the Obmité National controls national emergency stock operations it would support the village stock program by replenishing those stocks by way of purchases made through the 0RD's. The proposal stresses the need for systematic rural mobilization in order for this program to succeed in altering patterns of grain marketing and storage to the benefit of all segments of the rural population. This emphasis is supported by the analytical conclusions reached in application of this study's conceptual framework. If multi- sector structural and behavioral change is not promoted simultaneously, well-meaning programs in one area may deviate from their stated ob- jectives. This emphasis calls for major efforts to educate govern- ment cadre as well as farmers and some of these training requirements are detailed in the village stock proposal. Increased extension ac- tivities would be supported by a system of audio-visual aids including 217 radio broadcasts, simple films, printed tracts and posters, all in local languages. The cost structure of the proposed program, while probably meant to be illustrative, is oversimplified and unrealistic given other West African stock experience. It includes over a five-year period: 1. 12,000 grain storage facilities (in 3,000 village groups) at a cost of 4,000 CFA per facility = 48,000,000 2. Operating funds: 3,000 groups x 144,000 CFA (8 MT grain at 18,000 CFA per MT) = 432,000,000 3. Training and propaganda costs = 67,550,000 Rounded Total $50,000,000 CFA12 A more complete assessment of alternative storage cost structures is presented below and, in the final section of this chapter, the inte- gration of the village stock program into a national grain marketing system is discussed. Grain Price/Supply Stabilization for Non-Producers Grain supply stabilization for nonrproducers in Upper Volta will be simplified if any or all of three facilitating conditions are met. First, this task will be easier to accomplish if the village stock program is successful in promoting increased grain supply stabiliza- tion at the village level and if it leads to a net increase in the aggregate amount of grain held under the control of producers who make up about 80 percent of the Voltaic population. In addition to more grain being disposable by producers in local market and 12$2.2 million U.S. @ 250 CFA per $1 U.S. 218 nonmarket exchanges, there would also be relatively greater supplies available to the 0RD's for intra—regional stabilization transfer and to other 0RD's and to OFANCER for intervention in urban grain markets and for emergency storage. The second facilitating condition would be met through control- ling the growth of the non-producer population largely through slow~ ing the rate of urbanization. Meeting this condition, of course, in- volves a range of policies much wider than just those affecting grain marketing. In general such a set of policies would serve to refocus part of the social and economic reward and incentive structure on the rural rather than the urban population. The third facilitating condition would involve diminishing the destabilizing effects on Voltaic grain supplies of divergent price and stock policies and other governmental market interventions in neighboring countries through efforts at improving grain policy coordi- nation, information flows and actual grain supply transfers among the Sahelian countries and their coastal neighbors. In general, Upper Volta's major surplus grain supply areas are in West and Southwest (around Dédougou, Bobo Dioulasso, Banfora, Diébougou) which are in— fluenced by supply and price conditions in.Mali, the Ivory Coast and Ghana or in the East and Southeast (around Zambré, Tendodogo, Koupela, and more remote parts of the Fada N'Gourma 0RD)l3 which are similarly linked to northern Ghana, Togo and Benin and to major grain markets in Western Niger (see Figure VIII-l). This is not to say that Voltaic grain surpluses should not leave the country, rather that these l3See Eicher, et a1 [1976] for a more detailed description of these conditions in the Eastern (Fada N'Gourma) 0RD. 219 movements could be anticipated and planned for with greater coopera- tion among the various national grain boards. This type of flexibility has been recently attempted in Upper Volta with the Eastern 0RD offi- cially permitted to directly export grain surpluses to OPVN, the Niger grain board. Direct program actions to enhance supply stabilization for non- producers can be divided into two groups, those which are directed at rural, pastoral non-producers and those aimed at the urban and small- town populations. Initially, at least, these actions must be seen as enhancing or supplementing the performance of the private grain trade simply because management of a total grain distribution system is far beyond the institutional capacity of the Upper Volta Government. With respect to the nomadic, pastoral population the major feasi- ble actions would largely be within the jurisdiction of the 0RD's. Each 0RD, through its coordination of intra-regional grain transfers, storage and stabilization, could attempt concerted actions in those geographical areas and markets most heavily used by the pastoral popu- lation. This could involve three types of actions: supplementary stabilization efforts during the "soudure" in those regional grain markets most heavily used by pastoralists; programs to facilitate imr proved patterns of economic exchange between pastoralists and seden- tary farmers (e.g., increased livestock or livestock herding services in exchange for grain in conjunction with increased groupement activi- ties among both groups); and the longer term facilitating organization of pastoralist groupements villageois. With respect to the urban non-producer population, grain supply stabilization actions would be the responsibility of OFNACER and the 220 0RD's working in close collaboration. Again stabilization efforts should supplement the working of urban grain markets. OFNACER should be capable of intervening, under certain supply price conditions, in the largest urban markets. For example, OFNACER could potentially be responsible for supply stabilization in the three largest cities in Upper Volta, Ouagadougou, Bobo-Dioulasso and Koudougou, which are all served by the Abidjan-Niger railroad which might permit the use of a relatively cheap, efficient bulk-handling sub-system in conjunction with the recycling of the major portion of the centrally held emergency stock and larger inter-annual buffer stocks. Such a limited stabili- zation program is within the foreseeable institutional and financial capabilities of OFNACER and would be capable of affecting market prices faced by at least 10 percent of the total papulation or about half the nonrproducer population. Intervention in other urban areas and smaller towns could come under the jurisdiction of the various 0RD's whose grain market actions and transfers would be coordinated with those of OFNACER. In surplus regions the major 0RD actions would focus at harvest-time market inter- vention to influence producer prices and facilitate the transfer of grain to deficit regions. In deficit regions, particularly the north- ern 0RD's of Kaya, Ouahigouya and the Sahel, the 0RD and/or OFNACER would have greater responsibility for stabilization intervention to dampen the "soudure" price increase and maintain adequate supplies in regional markets. 0RD ability to function in this capacity will be strengthened if they receive greater financial and operational autonomy in grain market intervention and pricing. The entire system of rural and urban grain supply stabilization will be facilitated through: 221 (1) improved roads and market infrastructure; (2) improved communica- tions, particularly a simple market price information system which is critical to both buffer and emergency stock actions; (3) increased price policy flexibility (including a quota and supra-quota price system) and regional autonomy; (4) the introduction of limited, low cost bulk handling components where appropriate; (5) more adequate provision for surplus grain disposal through international trade, concessionary sales, livestock finishing, etc,; and (6) improving the planning of minimum grain transfers and quotas. These will be examined in the final section of this chapter after a brief description of some of the economic and financial costs to alternative storage and transportation components in a complete grain marketing and storage system. Costs of Alternative Storage and Transportation Components In formulating practical plans for an integrated grain marketing and storage system two of the key variables to consider are the costs of alternative storage infrastructure and transportation services. While political and economic objectives, indirect costs and benefits, employment and distribution effects and so forth may ultimately be more important in deciding what to store and when to do it, once these decisions are made there are still significant economic and financial cost considerations regarding alternative technologies. In what follows I provide a summary of available information on major alternatives. With some refinement this type of information could provide the cost coefficients to be used in conjunction with price and non—price ob- jectives and constraints in a linear programming, cost minimization modeling exercise. 222 In Sahelian grain storage there are three major technologies which are or can be used within a total grain marketing system. These are (l) farm level storage which generally involves small, locally built silos which hold grain either on the head or in threshed farm; (2) the traditional commercial storage system of masonry bag warehouses; and (3) bulk storage systems which can involve numerous alternative silo structures. The cost of the latter two alternatives are compared in detail in Appendix B. Here observations on the relative costs and advantages of all three systems will be summarized. In terms of costs we must consi- der the initial investment costs associated with a particular type of structure, its annual operating cost (per metric ton for purposes of rough standardization) and the associated costs incurred in integrat- ing the particular type of structure into the total grain marketing system. Benefits are assessed both in terms of the direct protection of stored grain from deterioration and its availability to meet demand on a timely basis and the indirect effects stemming from employing a particular type of structure (employment, use of local resources, divisibility and other locational advantages, etc.). Evidence from recent but geographically scattered economic stud- ies on two key items in the financial cost of storage, initial con- struction costs and annual per ton operating costs, are summarized for the three systems in Table VIII-l. This secondary information, standardized to extent possible, is quite fragmentary but still indica- tive of the magnitude of relative costs. Most evidence points to the initial costs of traditional storage as being at least two to three times cheaper per ton than either the 223 TABLE VIII-1 COMPARISON OF PER TON STORAGE CONSTRUCTION AND ANNUAL OPERATING COSTS Construction Costs Per MT Annual Operating Costs Per MT 1. Traditional farm level storage a. Northern Nigeria b. Northern Ghana c. Upper Volta d. Chad - CARE program 2. Bag warehouse a. Chad b. Mali c. Niger d. Sahel 3. Silo storage a. Mali b. Mauritania c. Niger $22.80/MT $15.50/MT $12.00/MT $73.00/MT $81-$192/MT $60/MT $48-$68/MT $110 $240/MT $82/ME $40, $107, $227/MT $10-$23 $17 SOURCES: Traditional storage: storage: Appendix B. Hays [1975], Nyanteng [1972], Repub- lic of Upper Volta [l976c] and Maxon [1976]; bag warehouse and silo 224 bag warehouse or silo systems. Similarly bag warehouse storage is gen- erally somewhat cheaper to build per ton than silo storage when amor- tized over equivalent time periods. However, there are smaller, less mechanically complex silo systems with lower initial costs than some more elaborate warehouse systems. Traditional graneries also have substantially cheaper annual operating costs than the other two sys- tems if we count the same number of grain movements into and out of each system. This is partially due to the need to include at least part of the cost of a permanent salaried staff, chemical treatment, electri- city and the cost of sacking and handling in the latter two systems.14 The costs of integrating the different storage technologies into an overall grain system depend upon the ease with which each technology can be employed in the performance of the three major tasks required of the total system, storage of local production for local consump- tion, annual storage and interregional transfer to deficit areas, and interannual stabilization and long-term emergency stock storage in a limited number of centralized locations. It is clear that each of these technologies, under the right conditions, can provide the least cost solution for the accomplishment of at least one of these system activities. Traditional storage is, and will most likely remain, the 14Grain losses are included in these annual cost estimates with the general observation that each technology, when correctly employed, will generate approximately the same small percentage of annual grain loss. Under ideal conditions, hermetically sealed underground bulk silos should be able to produce the smallest losses but actual losses under these three technologies, particularly when they are employed suboptimally or under stress conditions, need to be further studied (e.g., using a traditional storage technique under different eco- climatological conditions, use of bag warehouses without correct palettes and fumigation, or use of bulk silos with abnormally high temperatures, overly moist grain, etc.). 225 most efficient, least cost method for grain storage at the farm and village level when significant bulking and transfer is not required. When these latter conditions hold bag warehousing will remain most efficient given the large number of intermediaries, the variety of transportation methods used, the acceptance of the sack as a unit of measure and its easy use in transport and bulk breaking operations. When grain has to be stored in sizeable quantities over a number of years with minimum storage losses, bulk silos become more cost effi- cient particularly if they can be linked to bulk transportation and handling components to and from wholesale buying and selling points. The benefits of alternative storage technologies, in addition to time, place and quantity utility, can be assessed in terms of direct price stabilization and income effects and indirect effects associated with employment generation, use of local resources (or the saving of scarce foreign exchange) and the type of "learning" associated with alternative ownership and management structures. The cost and availability of transportation infrastructure and services will directly influence the relative cost of incorporating alternative storage technologies into a national grain system. In Upper Volta, as in the rest of the Sahel, most farmers do not have motorized transportation available at the "farm gate", necessitating the use of headloading or transport by animal, cart, bicycle or motor- bike. From those points accessible by motor vehicle the major determi- nants of cost are road conditions and vehicle carrying capacity. Even under the best of conditions truck transport costs are high due to the use of entirely imported technologies and sources of energy. 226 Upper Volta currently provides reimbursement to private trans- porters for grain hauling at the official rate of 22 CFA francs per kilometer-ton. This is approximately equal to most other recent estimates for large volume truck (10 MT and over) transport costs on primary Sahelian dirt roads. In neighboring Mali the government re- cognizes the heavy cost of using vehicles on dirt roads by permitting freight charges approximately equal to 13 CFA francs per kilometer- ton on paved roads and 26 CFA francs on dirt roads. Further, estimated economies of size are striking in truck transportation. For example, on good roads in Mali the breakeven point per kilometer-ton was esti- mated for different truck sizes as follows: 5 tons or less, 24 CFA per kilometer-ton; 10 tons or less, 17 CFA and for 25 tons and over, 12 CFA.15 Several general propositions stem from the high cost of transport. First it increases the economic viability of a diffused grain storage system and would cause more grain to be optimally stored than in a sys- tem with better transportation possibilities. Similarly, improvements in transportation infrastructure and services will increase the economic return to the more centralized storage of larger quantities of grain. Third, the optimum storage/transportation "mixture" facing farmers, commercial traders and government institutions along major road and rail links will be substantially different than that faced along secondary axes or in areas with no modern transportation at all. These facts should be actively incorporated into the planning or a grain marketing and storage system. Thus, in Upper Volta this means taking lsSteadman [1976, pp. 51—53]. 227 better advantage of the existing road and rail infrastructure16 in planning the location of central stabilization and emergency stocks, in programming recurrent shipments among regions, and in the use of a more flexible transportation schedule which would more accurately reflect real costs. These generalizations concerning transportation and storage costs are reflected in the following section. Toward a More Integrated National Food Grain Marketing and Storage System Overview of a Possible System An integrated food grain marketing and storage system can play an important role in Upper Volta's attempts to increase food self- sufficiency and promote a more self-directed, independent path toward economic modernization. To counter inevitable weather-induced varia- tion in supply, direct action must be taken to place more grain effi- ciently in storage to meet a constant or increasing per capita demand over time. Indirectly supply can be increased as farmers will have greater incentive to produce when facing a more stable market provid- ing them with higher effective floor prices. This type of system is attainable through the progressive and incremental modification of current Voltaic private and state institutions. Further, this can be done in such a way as to increase the use of locally available re- sources (both physical and human), increase the participation of the rural population in the development of community controlled institu- tions contributing to modernization, increase or at least maintain l6Estimated kilometer-ton rail freight charges on the Abidjan- Niger line are approximately one-half truck freight rates but to this point in time recurrent OFNACER interregional grain shipments (e.g., Bobo-Dioulasso-Ouagadougou, etc.) have been almost exclusively made by truck instead of rail. 228 distributional equity through allowing all segments of the population to enjoy the benefits of increased food grain availability, and counter the trend of Upper Volta's increased political and economic dependence as it is gradually incorporated into the world economy. Clearly, imr proving the grain marketing and storage system will not in itself bring all of this about but making more effective use of existing and poten- tial food resources is a good place from which to begin. To do so we must look for those feasible changes in economic infra- structure and method which will minimize the cost of additional grain storage and transfers in the accomplishment of: (1) improved storage of local production for local consumption, (2) facilitating food trans- fers to permanently deficit producer and non-producer populations, and (3) most efficiently storing emergency stocks to offset severe pro- duction shortfalls. To meet these objectives in the Voltaic economy a multilayered system involving local, regional and national components is required. However, a system which is ultimately dependent on the productive output of its peasant producers cannot hope to meet these objectives if official food marketing and storage institutions continue to focus on the interests of a small proportion of the urban, non- producer population. Thus, the base for building an improved Voltaic grain system.must rest on the further development of on-farm and village storage and increasing the access of all farmers to the benefits of increased stor- age capacity. Marginally improved traditionally storage technology can be owned and managed by village groups in a "grain insurance system" by which the costs and risks of holding larger amounts of grain are collectively borne. 229 With 80 percent of the population facing a larger, more stable supply of grain, recurrent transfers to deficit regions and the urban population and surplus storage against unusual shortfalls become more manageable tasks. At the regional level the 0RD's can potentially provide village groups with training and technical assistance in imp proved storage methods and in planning as well as incrementally increas- ing their own ability to transfer grain within and beyond each region. The particular 0RD actions required will depend on the relative sur- plus or deficit nature of the region, its existing road and storage infrastucture and the size of the non-producer population. Interregional transfers, the supply of urban areas, centralized holding of stabilization and emergency stocks, and coordination with similar institutions in neighboring countries should be the responsi- bility of one national institution, such as OFNACER, whose activities must be closely coordinated with those of the 0RD's. Overall policy must be coordinated to support each institution in the accomplishment of its "federated role". A flexible price policy supporting these actions is perhaps the most important required modification of national policy. Other incentives must be logically structured to promote a stable, self-sufficient production and consumption base which can afford a certain level of food grain transfers to non-producers and inter- annual storage adequate to meet a fixed percentage of maximum likely shortfall. This means that there will have to be some modification in existing urban-oriented incentive structures which are basically in conflict with the promotion of increased farm production and the stable evolution of the Voltaic rural population. 230 Practical Measures to Improve the Existing System Assuming that the system described above is desirable and can be approached through incremental change in existing practices and insti— tutions, the following are specific recommendations for improving Voltaic grain storage and marketing. Overall System Development Strategy With commitment to the basic idea of holding increased grain stocks for the common good it will be possible to realistically develop an improved grain marketing and storage system over time. Technically efficient, cost effective methods can be discovered and perfected if the system.is developed in a phased, incremental way, adopting and refining proven techniques. This means that the collective methods used will have to begin (and may well continue) in competition with private commercial trade. Emphasis must be placed on the elimination of major system contradictions and on the structuring of incentives to functionally support practical food security. Promotion of Local Grain Storagg_and Food Self-Sufficiency The base of an improved system must rest on the decentralized increase in local collective storage, primarily in rural areas but also in urban neighborhoods. The collective or cooperative nature of the storage system is required to increase the participation of all segments of the population and to maintain or improve distributional equity which is not the case under either private trade or selective government participation in grain marketing to the almost exclusive benefit of a small prOportion of the urban population. In Upper Volta this could be done through: 231 Expansion of the program of village grain storage described earlier in this chapter. This should be done incrementally, adopting marginal improvements in traditional storage techniques and vil— lage group structures. Emphasis should be on village ownership and control of grain supplies (to avoid the almost inevitable inefficiency and corruption which accompany the decentralized management of centrally owned grain stocks) and guarantees of equal access of the entire rural population to this system. Implementation of the village stock program could be coordinated through a special working group or office in the Ministry of Rural Development which would work closely with the 0RD's and OFNACER.(which should be transferred to the Ministry of Rural Development as described below). This group would channel capi- tal resources, coordinate technical assistance and conduct local- ized research to promote: a. Locally adapted, collective least-cost structures for increased grain storage. b. Incentive structures (material and non—material rewards, etc.) to promote such a system. c. Studies of traditional storage technology to isolate the causes of storage loss and make marginal corrective changes. d. Provide over time a flexible, locally controlled system of personal credit against the collateral of stored grain and administered by the village group. e. Strengthen village groups focused on self-directed moderni- zation, including a guarantee of equal access to these struc- tures . 232 f. Investigate the feasibility of neighborhood groups in urban areas holding similar small-scale stocks of grain if assisted in storage construction and group formation. Key support for the village group stock program must be provided by the 0RD's and OFNACER as described below. The latter should give priority to making minimum quota purchases from village groups in surplus production regions in order to provide greater finan- cial incentives to participate in the system and increase surplus grain production. Strengthen 0RD Capacity to Act at the Regional Level 1. The 0RD's, with coordinating and planning assitance from OFNACER, should assume an increased role in intra-regional transfers. Their precise role will be determined by the aggregate net produc- tion capacity of the region, its transportation infrastucture, etc. More intensive marketing activities will be required in both surplus and deficit regions than in regions where produc- tion and consumption areUin closer balance. The 0RD's, through their marketing, credit and extension services, can assist village groups with technical problems, transporta- tion services, investment credit and the planning of local cereal needs and surplus production potential given fixed quota and supra-quota price structures and likely alternative supply situa- tions. The 0RD's would also produce aggregate regional grain supply projections and transfer plans. The 0RD's must be provided with financial and reward structures which are keyed to their support of the village grain stock pro- gram and to providing agreed upon quotas into the national system. 233 This will require additional training of 0RD marketing staff and it may necessitate increases in 0RD warehouse storage and trans- portation capacity as well as the provision of centralized re- gional grain drying and treatment facilities to reduce aggregate losses within each region. Strengthen OFNACER and Free It From Institutional Conflict 1. To complete the layered national system, one national institution is needed to coordinate interregional transfers and supply quotas, manage stabilization, pipeline and emergency stock storage facili- ties, and pursue facilitating contacts with similar institutions in neighboring countries. Structural conflict with the 0RD's and other national institutions could be minimized by choosing OFNACER as the sole national cereals institution and placing it within the Ministry of Rural Development. This would permit abolition of the superfluous Cbmité National and other food-security organizations such as the Sbus-Cbmité. All foreign food-aid assistance should be provided through OFNACER. A standard methodology for assessing regional production, surplus and storage potential should be developed by OFNACER working in conjunction with the 0RD's. The adequacy of regional storage infrastucture can thus be assessed and priorities for investment determined. The annual program of interregional transfers, storage, stock turn- over and reconstitution and minimum quotas would be developed with the 0RD's. The full costs of the system should be covered by an adequate consumer price with minimum.transfer or stock quotas assured at a reasonable floor price. Under shortage conditions 234 additional supplies could be encouraged with a supra-quota incentive price while, in surplus years, stock disposal arrange- ments (transfer to neighboring countries, livestock finishing, etc.) must be provided. It is essential that an agreed upon, quota planning feedback mechanism be instituted to prevent chronic under or over-production. Specific Actions to Improve Emergency Stock Operations 1. As we have seen the national program to hold emergency stocks has operated reasonably well given inadequate storage facilities, adverse supply conditions and the late arrival of funds in the first two buying campaigns. The operating rules are basically conservative and sound and the entire operation would be more easily assured if placed under OFNACER in the Ministry of Rural Development so as to be coordinated with the ORDs which would supply a large portion of the grain to be stocked. The first several years of experience have demonstrated that the emergency stock system could benefit from a more thorough and sys- fl tematic re-examination of the population at risk and the optimal size and location of stocks given improved information on surplus and deficit regions, transport costs, available markets for grain turnover and changes in the overall system.which might eventually result from the village stock program. Further, there is need for greater precision in the definition of stock release trigger mechanisms. At the aggregate wholesale level it may make economic sense to employ bulk handling and storage techniques in the receipt of 235 emergency grain imports, the supply of the largest urban whole- sale markets on the rail line and principle paved roads, and in storage over periods of several years or more. This should re- ceive careful technical and economic analysis. Grain Supply Stabilization l. Improvements in local supply stabilization, in the coordination of interregional "pipeline" supply transfers and in the manage- ment of the emergency stock will go a long way toward introduc- ing greater stabilization in overall grain supply. Further, experience in the component activities involved in intra-annual stabilization will improve the capacity to engage in inter- annual stabilization. The capacity of Voltaic institutions to rationally increase their participation in national grain markets will be facilitated through the collection of better data on regional production, supply and market prices. This would be a major assignment for a sorely needed national office of agricultural statistics. It is important to note that grain supply stabilization must be developed incrementally through trial and error methods in order to be self-sustaining and avoid the massive failures of the past. This means that specific actions must be mounted in competition with private trade with the latter providing valuable clues as to the likely limits of possible intervention. Only over time can collective local, regional and national institutions begin to replace the private grain trade if that is the course of action chosen by the Voltaic political process. However, in that 236 case specific attention should be given to the ease with which partial solutions can be undermined for private gain. Price Policy: A Key Variable l. Intervention in grain markets will be facilitated under a more flexible price policy not manipulated to maintain the support of the politically vocal urban consumer population. Producers must be provided adequate price incentives to increase food grain pro- duction and the economic costs of storage and interregional trans- fers must be fully covered in the long run. A commitment to these principles will probably of necessity imply an increase in consumer prices which will provide a true test of political resolve. It is clear that counterproductive situations such as allowing different arms of the government to be currently selling grain at different prices should not be allowed to continue. The key political-economic decisions involve the setting, timing and rigidity of the total price and expense reimbursement schedule or baréme. It is strongly recommended that an incentive price applicable to predetermined minimum surplus quotas be fixed before the planting season with flexible authority to set higher supra- quota prices if needed at harvest time. It also makes sense that the ultimate setting of the basic consumer price for non-producers be set as a function of expenses incurred during the buying season within a politically predetermined range and, if necessary, in conjunction.with triggered releases from emergency stocks. Con- sumer prices in insolated areas should reflect a reasonable trans- portation premium. 237 Additional Facilitating7Conditions Although it is well beyond the scope of this section to discuss these in detail, it should be clear that an improved grain system such as proposed here for Upper Volta will be easier to realize if accompanied by factors such as a believable, locally-adapted ideology of self-directed rural modernization, supporting practical actions which demonstrate a commitment on the part of the government to restruc- ture incentives in favor of rural areas, and concrete measures taken to slow the attraction and rate of growth in urban areas. In short, the system proposed here will be more likely to succeed under a political system committed to eliminating or hampering the further development of the structures of political and economic dependence described through- out this study. Finally, it goes without saying that a stabilization and storage program will be facilitated by agronomic advances in grain varieties with greater drought resistance and higher productivity. Useful Role for Foreign Aid While the success or failure of this program would largely depend on Voltaic commitment and political determination, there are useful tasks which could be played by foreign donor agencies. For example, of greatest use could be: 1. Monetary grants or food-aid to be sold to generate operating funds to finance village storage construction and initial local, regional and national stock constitution. Recent aid from the German Federal Republic to the Emergency Stock Program under - strict FAQ/FSAS rules seems to have worked well. Another possi- bility would be to provide food-for-work assistance to village storage construction during the dry season. to 238 The entire system can benefit from infrastructure construction. Of greatest impact would be paving of several of the major Voltaic primary road axes (Ouagadougou and Bobo-Diolasso, Ouagadougou- Ouahigouya, Ouagadougou-Dori, Ouagadougou-Niamey), providing exper- imental bulk storage and handling facilities and increased bag warehouse transit and storage facilites at the regional level. There is a continuing need for operational technical assistance and training. OFNACER and 0RD grain marketing and storage per- sonnel need to be identified for various levels of career train— ing. Of particular importance would be managerial training in logistical planning and coordination, technical training in ware- house and bulk silo operation and management, and launching a full system of recurrent agricultural statistics, particularly providing timely, disaggregated production and yield estimates, and information on grain flows and market prices. The statistical system would be enhanced by a program of periodic economic stud- ies of farm level grain storage and marketing patterns and the behavior over time of traditional private grain marketing channels. The program of foreign aid in technical assistance and training should be for a restricted time period after which time it should be nationally or at least regionally operated. The Entente Grain Stabilization training program provides a useful beginning for promoting badly needed regional functional cooperation on grain storage, marketing and price policy. This model can be substantially strengthened through the choice of a more important regional institutional vehicle (CILSS, CEAO, or ECOWAS for example) and through promoting greater real functional 239 exchange of information and promoting mutually beneficial inter- governmental grain trade. CHAPTER.IX SUMMARY CONCLUSIONS AND POLICY RECOMMENDATIONS This study has focused on meeting two major objectives: the application of a structural theory of political-economic dependence to Sahelian West Africa, and, within this framework, the description and analysis of practical policy issues in the area of food grain mar- keting and storage. The structural center-periphery approach has been used because it was felt to provide a useful conceptual "window" through which we could obtain a clearer picture of Sahelian underdevelopment. A concern for development policy demands that we have some understanding of the broad range of social, political and economic variables centrally involved in the past, current and possible future structural transformation of west African economies. I contend that this framework, with its emphasis on socio-economic evolution and the analysis of capital accumulation and socio-economic conflict through the dynamic interaction of struc- tural rigidities and market exchange mechanisms, helps to illuminate broad differences in possible sectoral policy alternatives. I have attempted to draw linkages from this macro-level framework to the specific policy area of food security which, in the Sahel, is fairly congruent with the resolution of problems of grain marketing and storage under conditions of high supply variability. Coming to grips 240 241 with the food supply issue will be one of the cornerstones of any attempt in the Sahel to promote self-sustained growth through an egal- itarian basic needs strategy. . There is a certain urgency to undertaking a wide~ranging review of these policy issues as well as badly needed, complementary micro- level studies at this point in time. First, West Africa has been under- going over the past several years a second wave of rather pragmatic "decolonization" as some countries attempt to begin to exert greater economic as well as nominal political independence. Second, for the Sahelian countries there is currently a great increase in the availa- bility of external rural development financing, partly due to humani- tarian concern for the effects of the recent drought and partly due to a rejuggling of international political-economic alliances and spheres of influence. The challenge is to examine alternative strate— gies for the judicious use of scarce capital to promote the realistic attainment of the stated Sahelian goals of "food self-sufficiency and autonomous economic development" and avoiding the very real pitfalls of systematically increasing structural economic dependence and indi- rectly promoting rural stagnation. To practically approach the major objectives of this study a set of more specific, operational objectives were enumerated on page 10 of Chapter I. These correspond, in a general way, to the division of the study into its subsequent component chapters. It has been stressed that the analysis in this study moves from the broad macro-level conceptions and descriptions of Chapters II and III through an exami- nation of Sahelian grain production, consumption and traditional trade in Chapters IV and V, to the ultimate focus on the more restricted 242 topic of government participation in grain marketing and storage in Chapters VI through VIII. At the broadest level in Chapter II a conceptual framework for addressing the overarching problem of underdevelopment in the Sahel was developed through the application of a general theory of structural sys- tems dependency to the West African region. Within the general theory, a dynamic international political economic system, characterized by pervasive inequality in the distribution, control and ultimate consump- tion of human and physical resources, is composed of shifting alliances or coalitions of system.components (their interaction regulated through structures such as national boundaries, national and international in? stitutions, laws, formal and informal operating procedures and rules, etc.) which, over time, enhance and extend their differential access to power and the derived material benefits which flow from domination over other components. Thus, nations, classes, groups and individuals are linked together in complex relationships of domination and subordi- nation which are structurally legitimized and institutionalized. In geopolitical and power terms the dominant areas in the system are the centers (of the system as a whole or of various sub-systems) with the subordinate, dependent areas comprising peripheries. Applied to West Africa a central theme is that the current poli- tical economic relationships between that area and the more developed world, while structurally more complex and sophisticated than those of colonial administration, functionally continue to promote the relative development of Center country economies while continuing the greatest relative exploitation of the West African periphery and its peasant farmers. 243 In its simplest presentation, the center of the Sahelian Peri- phery--composed of small, westernized "civil servant" and commercial business elites and their institutions (education, laws, methods of doing business, etc.)-serves as the key link between the large mass of peasant producers (the periphery of the Periphery) and principle busi- ness and governmental interests in individual developed countries and in various international organizations (collectively, the center of the Center). This is in no way a simple theory of one group of countries systematically exploiting another group. Rather it depicts a multi- layered coalition of structures and groups at various levels of the world system.which systematically serve the interests of monopoly concentra- tions of capital and indirectly the interests of those groups and in- dividuals variously attached to public and private institutions which support and service those capital interests. The small Sahelian agricultural surplus is aggregated and exchanged for mass consumption items (plastic shoes, cloth, radios, beer, etc.), a very expensive imported life style for small local and expatriate elites, and for imported, facilitating means of communications and transportation. Very little surplus goes into locally-controlled pro- ductive investment. The symptoms of economic dependency-rapid urbani- zation with high unemployment, increasing breakdown of traditional social cohesion, rural stagnation, and recurrent deficits in food production as export cash crop production is maintained-—are present throughout the Sahel. Much of the greatly increased development assistance to the area, designed to combat one or another of these symptions, too often serves, instead, to perpetuate and strengthen these dependency relationships and not to promote self-sustained economic change and 244 development. Thus "underdevelopment" is not the first stage in a possi- ble assent to "development" but it is rather a semi-permanent charac- teristic of whole segments of the world political economy caught in complex patterns of center-periphery domination and dependence. Also in Chapter II, a number of exchange mechanisms, which are related to system structural characteristics and through which accumu- lation of property wealth and capital proceed, were introduced and developed. When narrowly viewed (such as in most conventional presen- tations of international trade "comparative advantage"), these exchange mechanisms seem to represent normal and equitable trade relationships. However, a broader analysis of their operation in economic trade with rural, agricultural areas in the third world (the periphery of the Periphery) is more instructive. In the hypothesis of unequal trade the Center and Periphery engage in vertical trade across a "processing gap" or gap in levels of technology with the center providing "finished goods" and the Periphery primary products and raw materials. Even if the simple terms of trade are relatively equal, over time, capital and wealth accumulate in the Center due to higher rates of increase in labor productivity in the larger, more integrated economies and due to the fact that the Center generally has multiple, vertical relationships of this type. The Periphery, in contrast, sees little change in labor productivity in its rural production areas and its trade relationships are generally restricted to a few major partners. The second "ratchet" in the system is represented by the internal trade relationships within the Periphery in which domestic terms of trade run against rural areas except.during those periods when major structural changes are occuring, often associated with the introduction 245 of new production technologies or new cash crops. Periphery rural areas tend to remain materially unchanged and stagnant (de Janvry's "marginalization" of third world rural labor); urban areas in con- trast tend to grow rapidly, feuled by the creation of an administrative government aided directly by the Center and through the introduction of capital intensive, "imported" import-substitution consumer goods industries or, in more advanced areas, manufacturing for re-export using domestic cheap labor. It was clear that the more formal two country, two socio-economic group model of center-periphery economic interaction and change was not adequate to represent the full complexity of West African underdevel- opment. In increasing the complexity of the basic model the most important additions suggested are: 1. Including an intermediate or "semi-peripheral" layer of countries between the developed world and underdeveloped areas. This sim- ply permits greater system complexity and flexibility in applying mobile capital to less-mobile labor. (Thus, the Ivory Coast and Nigeria are rapidly developing into intermediate countries in West Africa which can facilitate applying imported capital to imported labor and gaining for themselves some of the material benefits). 2. In similar fashion, including an intermediate socio-economic group to facilitate the extraction of agricultural surplus from rural areas. (This function has been played in West Africa by Lebanese traders, various religious groups, particularly the Moslem Mourides, and indigenous ethnic groups specialized in long-distance trade: the Bambara/Dioula, Haoussa, etc.) 246 3. Factors such as demographic trends, natural resource position, and geographical location modify the forms some of these mechan- isms assume which means the model must be specified for each country or region in question. 4. The expected presence of contradictory patterns within the main path of system change since systems in evolution contain vesti- gial pockets of former systems and the conflicting precursors to alternative future systems. It should also be noted that just as these contradictions exist at the macro level they also exist at the micro and even personal level. The application of the center-periphery model to Sahelian West Africa was completed in Chapter II through a brief description of the historical phases in the evolution of West African structural depen- dency, a comparison of this model with other trade and economic devel- opment modelsl (particularly with "dual-economy" models), and finally an exploration of two general alternative paths for possibly future evolutionary change in the Sahelian political economy. These latter two possible models were rather extremely stated in order to point up major possible structural differences. One model envisions an evolu— tionary path of increasing dependency and the other a path toward greater independence and self-reliance. Suggestive sectoral charac- teristics for the two models were spelled out in a final table to begin to shift our focus to application of this paradigm to one important sub-sector, food grain marketing and storage. 1These are more conventional models which deal with the same inescapable observable reality but which are more limited in the system variables chosen for analysis. 247 It is perhaps easier to see the application of the center-periphery model in West Africa than in other areas because of the overwhelmingly rural nature of those societies, particularly in the Sahel; the recent colonial past; the small political-administrative, socio-economic center groups using imported cultural and linguistic norms, administrative pro- cedures, energy and technologies; and the very low level of material economic development which leaves little short-term choice for this interdependent system but to generate foreign exchange through the ex- port of mineral wealth, cash crops and labor on a regional basis. This is explored particularly in Chapters III and V. In the economic overview of the Sahel presented in Chapter III a number of topics were briefly discussed: 1. The sectoral composition and low per capita level of Sahelian GNP. 2. The physical domination of the region by broad ecological, clima- tic belts which place severe constraints on the possible nature of agricultural transformation. 3. The predominant role of trade in Sahelian GNP and its dependent nature as illustrated by its composition (a few export cash crops and/or mineral resources) and its limited number of trading partners.2 4. Specific symptoms of center-periphery dependency in the Sahel including: a. Extremely wide differentials between non-agricultural and agricultural wages (differentials for the Sahel are three times greater than LDC averages and five to six times great- er than developed country averages). b. A brief discussion of the terms of trade "double ratchet" working consistently against Sahelian rural areas. c. Patterns in development assistance, much of which goes to maintain central bureaucracies. 2It should be stressed that diversification of trading patterns is not a sufficient condition for a more independent trade. 248 In narrowing the focus of the study still further in Chapter IV, trends in Sahelian grain consumption, production and imports were analyzed with the following conclusions: 1. Grain--particularly millet and sorghumr-dominates Sahelian diets although there are some significant differences from one country to another and within each country, particularly by ethnic group, place of residence, and access to animal protein. There are significant differences between "center and periphery diets" with the former reflecting a life style which includes increased amounts of rice and imported wheat convenience foods. There is strong evidence, however, that substitution to domestic from imported food can occur much more rapidly than previously felt under the influence of fairly modest shifts in relative prices. Urbanization does not mean improved diets for all parts of the population as poor urban migrants may have diets which are nutri- tionally poorer than the semi-subsistence ones they left behind in rural areas. The rate of growth of Sahelian urban areas is among the highest in the third world and most experts predict that this trend will continue. This is a classic pattern which is due partially to the lure of accumulating wealth and a number of very high paying jobs in urban areas, stagnation in rural areas, compounded by the dis- ruption of traditional rural life and agricultural patterns partly caused by the drought and, in some instances, by efforts to provide development assistance. Aggregate levels of cereal production have not kept par with popu- lation growth, particularly since 1964. Regression analysis of 249 aggregate production data showed a general pattern which covered most grain production in Mali, Niger, Senegal and Upper Volta. In it we saw that production levels increased fairly consistently from 1951 to 1964, oscillated around this level for six or seven years, dropped very sharply in 1972-73 at the height of the drought, and climbed back up to about their 1964-70 average in 1974 and 1975. There were, of course, significant deviations from these trends which are explored in detail in Chapter IV. While the drought did cut sharply into grain production and had an even more devastating impact on nomadic livestock production, we can also point to government policies, particularly price policies, which discouraged the production of food crops and encouraged greater output of cash crops. There were no major upward trends in the evolution of estimated yields for grain crops except for rice which is increasingly influenced by yield increases under government programs for its production with controlled irrigation and high fertilizer use. Analysis of variation in production levels from country to country shows somewhat conflicting patterns. However, in sum, it is pro- bably safe to say that in most years severe shortages will pro- bably be fairly localized and it should be possible to shift por- tions of relative grain surpluses to adjoining regions and coun- tries (assuming, of course, that transport costs do not far exceed the spread in relevant prices between the two areas). It should also be clear that in unusually bad years it may not be possible to count on much assistance in this respect from neighboring 250 production zones, particularly if two bad years in a row, such as occurred in 1972 and 1973, hit the region as a whole. Food imports have increased dramatically in the Sahel over the past ten years with the self-sufficiency percentage falling to 70 percent in Senegal and 60 percent in Mauritania. Senegal has taken fairly virgorous steps to attempt to reverse this pro- cess but the situation continues to erode in Mauritania. The trends in the other four countries indicate a drop below self- sufficiency with increased dependency on commercial food imports and donated relief supplies. In several instances, this has been complicated by inappropriate pricing policies or domestic politi- cal instability. It is fairly clear that all four countries could return to self-sufficiency and even produce surplus cereal for export under apprOpriate policies for food production and mar- keting. In the review in Chapter V of existing literature and empirical studies on Sahelian private grain trade we saw that: l. The vast majority of grain produced stays on the farm for local consumption or later sale with smaller percentages entering trad- ditional non-market exchange patterns or taken to local markets for sale. Under widely varying circumstances complex patterns of periodic markets permit "thin" supplies to be bulked and ex- changed for consumer goods. A variety of participants engage in grain marketing activities and, on the local level at least, all studies point to competi- tive market performance under the many constraints in operation. Longer distance marketing, facing severe transport, credit and 251 communications barriers, tends to be much more subject to oli- gopolistic cartelization, a fact which is also confirmed by empirical economic and anthropological case studies. There is a strong tendency to concentrate attention sequentially on the horizontal patterns of market competition at various levels in the market channel and to look at vertical organization and inte- gration in fairly simple terms. This can miss important patterns of concentration and trade restrictions along the dendritic inter- action structures which link freer market "islands" of difference sizes. These spatial and temporal structures are dotted with ad hoc and permanent, traditional and modern institutions, rules, borders, boundaries and property rights which only the special- ized few can master and which are not completely captured in simple comparisons of inter-market price and cost differentials. The relationship between institutional organization and the arrangement of economic activity over space is of particular imr portance. A summary conclusion for much of West Africa is that the organization of space for agricultural production and market- ing is not optimal for modernizing yet independent, more self- sufficient and locally-controlled economic system. The distributional impacts of staple food grain trading in con- junction with more important cash crop trading are significant in the growth of socio-economic differentiation (i.e., class for- mation) as rural incorporation into the commercial economy pro- gresses. We saw that, in some areas of communal land holding and small-scale agriculture, relatively better-off farm units (those with more advantageous traditional kin obligations and 252 relationships, a larger number of active adults, or access to better land, etc.) are more able to store and buy grain at har- vest time, engage in supplementary activities (e.g., prepare meals for sale) and sell more grain later in the year. Poorer farmers in order to meet cash needs--head taxes, small consumer items, etc.—-have to sell more of their grain at harvest time and earlier in the crop consumption year. This can lead to a deple— tion or exhuastion of farm unit grain stocks and the need to sell labor services (in the local area or in other regions) for wages or for payment in kind in order to meet food needs at higher ”hungry season" prices. These effects can be cumulative with more prosperous farmers (or, increasingly, outside traders or salaried urban residents) better able to take advantage of modern credit and technical assistance packages to produce cash and food crops with improved technology and a greater use of wage labor. It would be safe to say that at least 75 percent of Sahelian grain storage capacity is on farm, that its efficiency in terms of loss prevention is generally good, particularly in less humid areas, and that much higher costs and losses are incurred when grain moves into commercial and government bag storage facili- ties unless very rigorous management and fumigation procedures are used. Bulk storage, while offering some technical advantages for handling large volumes of grain, is quite costly and ineffi- cient since bulk handling components are isolated in the larger grain marketing system. 253 In Chapter VI the emphasis of this study moved to grain storage and marketing policy as issues in Sahelian government participation in grain markets were analyzed. After a brief discussion of a concep- tual schema for looking at an entire program of possible grain market- ing and storage activities, the chapter examined both macro and micro level grain stock policy issues felt to be relevant to Sahelian policy makers. In order to focus on those issues, it was necessary to "hold constant" many other factors through the following simplifying assump- tions: 1. National governments in the Sahel are going to engage in grain marketing activities, particularly: a. Supplying a part of urban grain needs. b. Making physical infrastructural improvements, particularly in storage and transportation. c. Holding emergency stocks. d. Employing buffer stocks along with other measures to act on supply and/or price variation. e. Generally intervene quite extensively in food markets. Many of the conditions required for competitive markets are met, particularly in localized areas. Less competitive conditions, however, often exist in longer distance trade and at higher levels of product aggregation in vertical channels. Short-term policy choices are quite limited but in the longer run, the Sahelian countries can engage in development policies which will promote greater structural and functional dependence or independence. Higher levels of outside funding (with some implicit constraints) will be available for infrastructural improvements, improved data collection and even to cover much of total operating ex- penses. Objective functions will stress "autonomous economic development and food self-sufficiency". While some trends can be identified which are moving toward the attainment of these goals, the major thrust of Sahelian evolution has been toward increasing economic and food dependency. 254 Given these simplifying assumptions the following set of fairly pragmatic grain policy issues were discussed: 1. "Macro level" policy issues a. g. The degree of food self-sufficiency to pursue The costs and benefits of supply stabilization through trade versus storage Definition of the "population at risk" Specification of a general food grain intervention 1) The optimal degree of decentralization 2) Specification of an integrated food policy 3) The size and timing of grain stock operations The nature of personnel training appropriate for such a system Technical issues concerning physical and institutional infrastructure Realistic expectations of assistance, cooperation from in- ternational and West African regional grain stocks Emergency stock issues Level of stocks Location of stocks Methods of stock constitution and replenishment Price and non-price trigger mechanisms Degree of independence of emergency stocks Strategies for surplus stock disposal Operation of emergency stock distribution Regional stock coordination and cost sharing Buffer stock issues a. b. Ownership and operational control of buffer stocks Reform in price policy 255 c. Structural reform to minimize bureaucratic conflict and waste d. Definition of an intervention strategy including: 1) Determination of population at risk 2) Adoption of appropriate trigger mechanisms 3) Balancing intra and inter-annual stabilization efforts. These issues were further discussed in the concluding portion of Chapter VI which reviewed alternative methods of analysis used in addressing the practical mechanics of grain stock holding. This dis- cussion distinguished between three general levels of analysis: the macro-systemic, the intermediate/p0licy/oriented and the operational levels. Most modeling efforts reported in the literature have been at the macro-systemic level and most have dealt with major producers of grain for the world trade or alternative schemes for international reserve stock systems. A few modeling efforts have concentrated on specific LDC stock holding problems, particularly in the India sub- continent, and based on preliminary work done in this study. It is quite likely that such modeling research projects could be useful to Sahelian policy makers if carefully designed to meet the environmental and policy constraints of specific countries. It should also be noted that there are significant costs associated with such efforts both in terms of data collection, modeling time and the effort required to "recalibrate" the model to adjust for changing policy and environ- mental conditions. In contrast to more rigorous approaches, the analytical methods used in determining what levels of emergency and buffer stocks to hold are crude, pragmatic and reflect the pressures under which policy makers must work and take action given very incomplete information. 256 For example, at their best, in determining emergency stock levels, these methods involve probability calculations of the magnitude and distribution of shortfall, estimation of a population at risk, deter- mination of a net shortfall (by allocating gross shortfall to the population at risk by way of food balance sheets, average consumption levels, estimated arrival time for outside relief supplies, and domes— tic distributional capacity), and arbitrarily comparing these figures to political and financial constraints to fix a percentage of net shortfall as the national emergency stock level. The operational world of successful cereals boards is similar to that of the private grain trader; it involves "learning by doing" and the establishment of ad hoc and "rule of thumb" procedures. The inadequacy of the policy level analyses described above can be corrected in two ways, either by expending the resources needed to permit more sophisticated analyses to be conducted or by beginning to partially "solve" the problem through a redefinition of the unit of analysis and accounting which would, of course, have to be preceded by reorganization of collective grain storage and marketing activities to that level. We have seen the functional difficulties inherent in trying to operate a centralized grain market operation in a decentral- ized economy. At least a partial solution could involve a progressive structural and functional decentralization and redefinition of the basic operating and accounting units so that they are more congruent with the underlying base structure of rural production and consump- tion, the available physical and human resources of the country and stated rural development and distributional objectives. The two solutions, while somewhat contradictory at the theoretical level, are 257 not incompatible in actual practice as an integrated grain marketing and storage system must operate at different levels to meet multiple objectives. Thus, while the basic food security operational and ana- lytical evaluation unit can be decentralized to the village or village cluster level, other transhipment, commercial sales and storage acti- vities can be more narrowly, but more appropriately, analyzed and instituted on a more aggregate, federated basis. In Chapters VII and VIII Sahelian governmental grain marketing and storage institutions and their objectives, structures, means of implementation and initial performance over the past three to four years are described and analyzed. Upper Volta is covered in consider- able depth in a separate case study in Chapter VIII in which special attention is given to a review of the first several years of the FAO/ FSAS emergency stock program and an Upper Volta Government proposal to institute a program of collectively held village level grain stocks to increase local food security. The experiences of the other five countries are covered in a more summary fashion in Chapter VII. The following conclusions and policy recommendations represent a synthe- sis of the more detailed results presented in those two chapters. Summary Characteristics of Grain Marketing and Storage Institutions Objectives The formal objectives set for the various grain boards are quite uniform from country to country and include establishment of readily available emergency reserve stocks, facilitating internal grain trans- fers (including development of market infrastructure as food markets "deepen" with urbanization), and the intra-annual and inter-annual 258 stabilization of both producer and consumer prices and incomes. In contrast to these stated objectives more informal goals for cereals boards have revolved around the selective use of government resources to benefit already privileged urban groups contributing to patterns of disfunctional urbanization, rural stagnation and declining food production. These contradictions among objectives are predicted and high- lighted in the center-periphery framework. The most fundamental con- flict is between supporting rural producer (periphery) and urban con- sumer (center) interests. This conflict is seen in a number of forms in the functions of Sahelian cereals offices and related food poli- cies. One is the choice between stabilizing and "subsidizing" urban food prices versus producer incomes, the latter possible through the promotion of localized food stocks and higher prices for food trans- ferred to urban areas. The second is the use of the food marketing and storage system.for revenue grneration versus its alternative use as a riskrminimizing productive service provided to different sectors of the economy "at cost". Finally, there has been some tendency for Sahelian governments to use food marketing and storage for direct "micro-political" intervention through urban food subsidies and spe- cial belowhcost food deliveries to specific urban and regional con- stituencies where political support is weak. This of course under- mines the operational viability and integrity of cereals offices. Institutional Structures This study has placed major emphasis on ways in which inappro— priate institutional structures hamper the accomplishment of reason- able Sahelian grain marketing and storage objectives. This problem 259 can be seen at various levels. First, at a general level, the cereals offices, like other Sahelian institutions, are structured administra- tively or "from the top down", leading generally to disfunctional over- centralization, ineffective incentive structures, very high operating costs and negative distributional effects as such structures are forced onto the local, decentralized cereal economy. This pattern is also often observed in the growing number of regional development organiza- tions within each country and while this may facilitate the upward channeling of expropriated agricultural surplus to support urban sec- tors functionally linked to the developed Center of the international economy, it does not promote the development of viable local institu- tions and farmer controlled agricultural progress. On a more operational level, two major patterns were observed in the misallocation of national marketing and storage responsibili- ties. In the most common pattern, there are too many institutions, often in different ministries, participating in grain marketing channels leading to lack of coordination, duplication of effort, and unnecessary and costly supplementary grain transactions. This pattern was examined in detail in the Upper Volta chapter. We have also seen the opposite pattern where one institution is weakened by having to fill multiple and often conflicting objectives. This, for example, is the case for ONCAD in Senegal which is responsible for cash crop as well as food crop marketing and storage and for the provision of credit and inputs to producers. Here the problem focuses on the greater stress placed on cash crop marketing and the inappropriate application of these extractive techniques to food crops. 260 Finally, similarities in the structure, functions, objectives and statutory powers of cereals offices should serve to facilitate direct regional inter-office contacts and cooperation in many cases but this has been structurally blocked by the operation of "feudal interaction structures" which provide incentives to maintain uneconomic links be-' tween each country's frail center and the outside world. Consistent with administrative "formalism”, international contact is usually restricted to the lowest common denominator of political cooperation while there is substantial room for developing direct functional con- tacts among cereals offices. It has also been repeatedly stressed that the coastal countries are vital to an understanding of the struc- ture and functioning of possible evolutionary paths for a West African regional economy. This means that they must be included, more than they are now, in any viable regional food security strategy, particu- larly Nigeria which produces about one-half of West Africa's millet and sorghum. Means of Implementing Program Objectives There are two main types of ingredients required in the operation of a cereals marketing and storage program: the physical and human components of the system (such as storage facilities, transportation infrastructure, trained personnel, and operating funds) and opera- tional policies (such as price policies) which put these elements together over time and space. These specific elements can only read- ily be addressed in detail on a country-by-country basis, as they are for Upper Volta in Chapter VIII. A few general patterns, however, can be underlined here. 261 Storage and transportation infrastructure is generally inade- quate for the tasks of holding increased stocks of grain and promot- ing more self-reliant patterns of rural development. Many infrastruc- tural investment strategies for the Sahel have placed inordinate em- phasis on the construction of capital intensive, large-scale storage and transport infrastructure which corresponds to general patterns of increased dependence on imported capital, technology and food supplies since they are not accompanied by serious attention to the improvement of infrastructure and institutions at the heart of the system, the farm and village level. The major grain shortfalls and ensuing drought relief actions, centered on the years 1973-75 and described in detail in Chapters IV and VII, demonstrated both the inadequacy and inefficiency of the macro-level distribution and stor- age system and the increasing vulnerability of an eroding, stagnant food production base. In terms of operational policy issues there are two main areas which have received substantial emphasis in this study, price poli- cies and those dealing with the degree of farmer participation in, and control over, grain marketing and storage mechanisms. Price pol- icy is undoubtedly one of the biggest practical constraints on the successful operation of grain programs. Deliberate buying experi- ments and the occasional high guaranteed producer price have effec- tively demonstrated that farmers will respond to price incentives. The opposite situation is more ofen the case, however, and when con- ditions are bad enough farmers will simply withdraw from the market to the extent possible. 262 The composition and timing of annual cereal price barémes are of utmost importance. Cereals offices are often forced to accept politically influenced margins between producer and consumer prices which cannot cover costs. The rigidity of uniform baréme price levels is often compounded by fixed transporation and other cost allowances which do not allow for necessary regional variation. This can lead to a misallocation of potential agricultural resources and its hamstrings grain institutions by denying them the flexibility to make quick market actions and to renegotiate prices (to institutional buyers for example) under changing supply conditions. It also encour— ages smuggling to take advantage of wide differences in official prices among neighboring countries. Finally, prices are generally announced well into the harvest season providing little incentive to the pro- ducer and injecting needless delays into marketing operations. When prices happen to be attractive to farmers, the marginal cedibility of government organizations is often further eroded by the lack of necessary funds to purchase all supplies offered or by not getting funds to buying points until too late in the season. The policies which govern the structure of grain board personnel incentives do not support the key area of local operating efficiency. Most institutional employees are "civil servants" whose loyalties and career incentives are not functionally related to operational success. Further, at the local level, there are virtually no means of pro- ducer or collective village control over the hiring and promotion of grain board employees at that level; all of these decisions "descend" administratively. 263 Performance of Government Grain Institutions As we have seen in this study the overall performance of grain boards in the Sahel has been quite poor to date. The summary assess- ment offered in Chapter VII still holds and bears repeating: The offices have not exercised much real influence in cereals markets and their actions have often contributed to the misallo- cation of production resources leading to negative distributional consequences as urban consumers and private traders have often been the chief recipients of program benefits. In contrast to this bleak assessment, however, the potential of cereals market- ing and storage institutions to promote greater food production, reduce producer risk, as well as stabilize urban grain supplies has been selectively demonstrated. The objective now should be to help identify the structural and functional elements of the present system of grain marketing and storage programs which are amenable to change. General Poligy Recommendations The detailed analysis of cereals institutions in Upper Volta in Chapter VIII and its concluding section on suggested structural remedies constitutes an example of such an attempt to propose modi- fications in existing institutions and behavior to more fully meet national objectives. While it is clear that these prescriptions must be tailored to the specific physical, logistical and institution- al environment of each country, the following general set of policy reommendations may be of use in this process. General Strategy .Pursuit of the objectives of "food self-sufficiency and auto- nomous economic development" (or more specific, operational formulations such as those suggested on page 14 of Chapter II) given the current environment of Sahelian center-periphery dependence, necessitates following a basic strategy of reducing external cultural, economic 264 and political dependence and building self-reliance through the satis- faction of basic needs by more fully employing the human and physical resources available at the local level, and by reintroducing more bal- ance between production and consumption at both the local and national levels. Increasing food security both by increasing grain production and by more rationally marketing and storing the grain must be faced at a number of levels at once in order to meet different objectives and to fit into a process of general system transformation. The strategy must begin where most of the Sahelian population lives and produces most of the food--in rural areas. The foundation of the general strategy is to promote the storage of greater quantities of grain by strengthening village level grain storage capacity to serve the needs of all segments of the rural population for food security and to provide a form of seasonal credit against this store of agricultural value. The village stock system proposed for Upper Volta and described in the previous chapter serves as a useful, if somewhat imperfect, model. The implementation of; this model will have to vary from surplus to deficit production areas with the former providing greater average percentages of production to national grain marketing channels. Above the village level, three additional functions must be in- corporated into an integrated system: normal supply of grain to the non-producing population, storage for supply stabilization in urban and rural areas, and the maintenance of strategic emergency stocks which will be available on short notice under predetermined condi- tions. The precise mixture of all four system components must be specifically determined for each Sahelian country. 265 It is important that this overall system by structured in feder- ated form so major responsibility for the constitution, ownership, operation and accounting of local stocks can be put under local con- trol with regional and central government authorities providing tech- nical assistance, seed capital and the planning and coordination of transfers from surplus to deficit areas, holding of strategic reserves and some buffer-stocks, and promoting functional contact with similar organizations in adjoining countries. In this way, increased local self-reliance will help reduce one aspect of dependency and promote reasonably-sized village geo-political units linked together to pro- mote self-sustained development from.below. Past experience has shown that attempting to transform top-down administrative, political-economic structures such as the ones de- scribed in this study is a perilous undertaking with many seeming re- forms turned to further reinforcement of objectionable trends. Thus, change must be coordinated at all levels if success is to be realized and sustained over time. Technically efficient, cost effective methods of grain storage and distribution can be locally perfected if the system is developed in a phased, incremental way, adapting and refining proven techniques. This means that collective methods will have to begin (and may well continue) in competition with private commercial trade. The type of food security system proposed here will be easier to accomplish if accompanied by facilitating factors such as a believ- able, locally-adapted ideology of self-directed rural modernization, government actions which demonstrate a commitment to restructure in- centives in favor of farmers and rural areas, and concrete measures 266 taken to slow the attraction and rate of growth of urban areas. In a similar fashion, a locally-controlled system will demand greater flexibility in pricing policies and prices which move in favor of rural areas even if this leads to a temporary drop in outward- oriented indicators such as GNP. Investment Priorities While the success or failure of Sahelian grain programs ultimately depends on national commitment and political determination, this is not a proposal for autarchy. Foreign capital and technical assistance can still play a valuable role in supplementing national resources, particularly in the areas of: . 1. Village Stabilization. Fairly moderate investment costs will be needed for constructing village storage facilities based on possible modifications in traditional storage technology so that maximum use of local resources and labor can be realized. MOnetary grants or food-aid to be used as pay or to be sold to generate operating funds could help finance local construction costs and the constitution of initial stocks. 2. Market and Transportation Infrastructure. The evacuation of village surplus production (and supply of needed inputs and services) will be facilitated through improvements in local road networks and regional grain storage, drying and disinfesta- tion equipment. Much of this infrastructure may be built using labor-intensive, food-for-work type programs. warehouse facili- ties should be placed so that shipping costs are minimized and effective service can be offered to the village storage units. 267 Igaigigg. Training for management of a coordinated, decentral- ized grain storage system should be conducted in pilot projects where personnel will be working, be it at the village, regional or national levels. Emphasis should be on a well managed "learning by doing” expansion of grain storage actions operated as "government" (includes local co-ops, etc.) services. Surplus Disposal. Major attention should be given to alterna- tive uses of possible surplus production of grain (particularly until a system of local-determined supply management can evolve). Here there are major possibilities of functional cooperative arrangements with emergency and buffer storage agencies in neigh- boring Sahelian and coastal West African countries which might be most effectively promoted under an international technical assistance program. Enhancing complementary production systems such as livestock marketing could be encouraged through experi- mentation in dry season feeding and finishing diets. Emergency Storage. The tightly-run management of limited emer- gency stocks, operated as a public service and coordinated with village grain stocks, urban supply and stabilization programs, will require a continued infusion of capital from domestic or foreign sources. Hopefully, improvements in storage technology, management procedures for stabilization stock turnover and so forth will continue to reduce the subsidy required for such a system. As described below further applied research into defin— ing the "population at risk", determining optimal emergency stor- age levels and locations, and the potential use of bulk handling 268 and storage components, particularly at the wholesale level, should be pursued under an emergency storage program. Research Priorities Mest of the research priorities are natural by-products of the general strategy and investment priorities listed above. Research priorities should serve the interests of national development strate- gy, for as Mao Tse-tung noted, "The most important problem does not lie in understanding the laws of the objective world and thus being able to explain it, but in applying the knowledge of these laws actively to change the world." At the national or macro-economic level the following topics could be usefully explored: 1. The structure and operation of traditional grain marketing sys- tems, with particular attention devoted to patterns of concen- tration, marketing margins over extended time period, and mar- ket components which could be effectively incorporated into a national system. A grain marketing system modeling project which would require country specific data collection and model specification in order to take into account probabilistic sub-national patterns of production, distribution and consumption under alternative sets of constraining policy rules. Once a working model of a particular grain economy is established the effects of alterna- tive price policies, non-price constraints, storage rules could be tested. Within the context of emergency grain stocks, this review of the first several years of Sahelian experience has shown the need 269 for a more thorough and systematic re-examination of the "population at risk" and the optimal size and location of stocks given improved information on surplus and deficit regions, trans- port costs, available markets for grain turnover and changes in overall grain system performance which may result from a more widespread village stock program. Top priority should be given to the development of a standard grain planning methodology for the annual assessment of regional production, surplus and storage levels. This will be greatly facilitated through the achievement of badly-needed agricultur- al statistics units in each country. At the local operational level there is need for applied micro- economic or anthropologic research and various types of technical experimentation. For example, there is need for greater knowledge of farm level grain marketing decision making and how this behavior fits into the individual farm and village economies. Cheap methods for simplified reproduction of these investigations should be developed for the use of regional program administrators, extension agents and village groups themselves. Further, there is great scope for local level applied "social-engineering" research in devising effective incentives to promote increased local grain storage in urban as well as rural environments. Technical field research should be focused on the use of energy efficient techniques for the employment of local labor and materials in improved grain storage technology, road and river transportation infrastructure and methods, grain processing, drying, and the 270 high-nutrition use of traditional grains to promote improved diets and decrease reliance on food imports. APPENDICES APPENDIX A STATISTICAL TABLES TO ACCOMPANY CHAPTERS III AND IV 271. an henna uo snowshoe Amuse: Home umauo>q no woven any mo coauuonoua - no unsavenoum us new»; onusouan osu mo woven .unco quad unease .nusnvcsoum uo uuonncuuu can nouuusvoum ecu :« huu>wuu=voun uo novcw 0:» we scuuuonoun a so some: as aspen nu sous: mean one we wanna seas .uucaum a .cca - omega .nn .a .auauu< an»: as aaqauaeonoonooz .aaa< uuaam "mumaom o.nm o.o- c.n~ cc I: nod no~.~ o~H.N can acousmooa o.ma o.enu q.nn cc II c~n muo.~ o~e.~ ous comanooaa o.oh o.ec q.n~ me an con ooo.u con.N cue ammannmma n.~n o.oe a.o so as on” cn~.H ca~.~ one snmauommd ~.o~ o.e~ m.~ ea ace um ~mc.~ com «an aqafianqmn ~.c o.n~ m.e Nun mum an own Nod ecu ccmfisOQma c.5u c.o~ c.m on acd mm. mm mu cec amauncmmm n.~ n.m n.~ «nu nun. won and «HM can mNmHIONma .. an.sv Am.ev cammnv aaaanv dog as an cam aaaa-oamd ".0 o.m m.~ caN ann nnu mu Nu ecu moausooma o.c v.u Q.“ nmN wen m- nu ma no mwmdacamu I: In I: II :1 mad nu c~ um ommaacmmu museum Annoy onuuu>< Hunnc< Oman coo .amsuo>< cuoxsa ocov was: Asocnum <50 .uocaum < umsuo>< dancs< mo canoe mouuum onou we wanna «some :« mousse no cooaum .mx OOH guacc< vuuuouocsua amculowmu as an noun unauOuuam nouuum uo alums ausuoz no uufium .uuuuoaxm annnusm echo «0 onao> no osao> amazon «mom huuvoaaoo «a sauna no woman noduuswoum caaaummma .mn<¢a no magma hbznz=omu mmMAm ~< mand? 272 TABLE A2 CONSUMPTION OF ANIMAL PROTEIN BY FOOD CLASS AND BY COUNTRY, 1970 ( GRAMS PER PERSON PER DAY) Food Classes Country Meat and Eggs and Fish' Total Milk Products Poultry Chad 7.9 0.3 5.7 13.9 Mali 10.3 0.8 3.6 14.7 Mauritania 31.5 1.0 4.6 37.1 Niger 12.6 1.1 0.5 14.2 Senegal 10.3 0.9 12.8 24.0 Upper Volta 2.8 0.6 0.6 4.0 SOURCE: FAO Nutrition Division. .Hekmfl_ uaummaa “summon pmum>oo was Hma Hod see some to unmouom maq.~ Hem.~ oas.~ ma~.~ maeuoamu sauce Hm nae moH we mmapm»m»o> om NHH me on muamflsumm Ho «NH so we nauseous Hfio moH one so on saunas Mu AMNsV Amway Aeomv Amuse Assesses can “masses 9. was Nmm Ham Hes Hmwumo sauce mH\m . mH\~H mH\~H . nH\a mH\m . mH\o maxo u mH\m condom nomomm :aommom common mun .vaou umm>um= hum::=: hum .uom mama room nnompom awed mum mzamuv mesa .namo ammzumesom A qazomamm m< mgmo cannons .on conuun>oa nonusu>oa cowusn>on sonuan>on nonuun>oa soaunn>on unavasum noon unaccoum can: vuavcsum coo: ouuvsuum can: vuavcuum new: vanadium ado: vunvsnum use: unannounuaa: acne .vanu woody auuo> noun: uamocam noun: announusux was: omen Huron an: occ.c Sean «on nupexnanu zonaoaaomm nz< .mhmaluoan :24 whoalanma .ruhzaoo km 924 mcao we onhuaanm 2Hs wanna .SWD olflflhuflaou Hfln—UO HON .U.‘ nah 3Q aflH olflEUdH-ulz BI ‘06 MOM 6869 umNQflg nu un nn New nu «ea nu can un gnu nn ans nu mun «sad nn nu ma «no one new and «we no «NM sad «no and «No nmsannaeaa nn nu on nuo cad men men cum nn nu Ned con un nn msmdnanan . unannou annoy nn wand nu econ nu awed nu meON nn nn nn Hana nn nn. «man can ncwn can can “nu damn new news nn nu can can nn un nsmnnaoma cos mace and «on was «and new amen nn nn can Hem nu nn nsaannmmn «and nn can nn mNo nn new nu ace nn nn nn moo nn nn chad can «on and cow and now «an own nn nn com hum nu nu nnmnnuomn HNH nus snu eu~ and sun had one nn nu nun nun nn nn muaannnafi owns: nu an nn nn nu V on nu nn nn nu nn nn nu nn :3 on nan nn nn «on nun nn nu nn nn nn nn nn nu nnmnnnnmn on non nn nu nu «we nn nn nn nn nn nn nn nn nnmnnnnmn .uuomxuuunnz nn nu nn com nn nn nn «we nn nn nn nn nn nn «sad nu nn med mnn nn nu «my own nn nn nn un nn nu msmanaoau nn nu am can nn nn and «Na nn nn nn nn nn nn msaanummu ssnmuom nn nu nn nae nn nn nn can nn nn nn con nn nn «and nn nu no c~o nn un he «on nn nn «NH coo nn nn nsnnnnoau nn nn no mo: nn nn . no nos nn nn «on one nu nn nsmnnunoa uoHan nonusu>oa nonuaw>oa sonuou>on nouuan>oa nonuuw>on nonuon>on soauanpua unaccuum can: vuovssum can: unavauum can: vuuvcsum can: eunuaaum coo: unaccoum use: vusuasum can: 13532.6: 85 .vosu macaw ouuo> some: Humosom noun: announunqx «an: vano Assam Amm .25!) SUMMARY OF REGRESSION ANALYSIS PART A: 1961-1974 ONLY (DATA: IBRD) Country Cereal(s) Datab Regression Standard 32 Transform. Coefficients Errors In '000 MT, KG/EA, or Logs A 8 ~ A 3 Production 1. Chad Sorg./Mill. -— 1,021 -44 38 4 .91 2. Chad Total - 1,066 -43 35 4 .90 3. Mauritania Sorg./Mill - 111 -4 12 1 .43 4. Mauritania Total - 116 -4 l3 1 .42 . Yield 5. Chad Total -— 833 -22 4O 5 .64 6. Mauritania Total - 429 -13 35 4 .47 PART B: 1951-1975 AND 1961-1975 (DATA: USDA) Country Cereal(s) Time Data Regression Standard R2 Period Transform. Coefficients Errors ' In '000 MT, KG/EA, or Logs A B A 3 Production 7. Mali Millet 61 - 868.0 -17.00 53.00 6.00 .39 8. Mali Rice 61 logs 5.0 -15.00 .15 .07 .25 9. Mali Total 61 - 1,123.0 -20.00 57.00 6.00 .44 ll. Niger Millet 51 logs 5.9 .14 .11 .04 .30 12. Niger Rice 51 - -6.0 1.90 2.70 .20 .83 14. Niger Rice 61 logs 1.9 .70 .20 .10 .77 17. Senegal Maize 51 logs 2.5 .39 .17 .07 .59 18. Senegal Rice 51 - 57.0 2.00 11.00 1.00 .29 19. Senegal Total 51 - 342.0 17.00 48.00 3.00 .54 20. Senegal Total 51 logs 5.7 .26 .12 .05 .55 sxnnm. CEREALS Paomncrxou AND YIELD mum ANALYSIS: (mewsmnn or R2 > .258) PART B: 277 TABLE A6 - CONTINUED 1951-1975 and 1961-1975 (DATA: USDA) - CONTINUED SUMMARY OP REGRESSION ANALYSIS Country Cereal(s) Time Data Regression Standard R2 Period Transform. Coefficients Errors In '000 MT, KG/RA, or Logs A 8 A 8 Production 21. Upper Volta Millet 51 - 192.0 7.00 22.00 1.00 .48 22. Upper Volta Millet 61 logs 5.5 .13 .12 .06 .25 23. Upper Volta Sorghum 51 - 336.0 13.00 50.00 3.00 .41 24. Upper Volta sorghum 51 logs 5.6 .27 .13 .05 .52 25. Upper Volta Rice 51 - 15.0 1.00 3.00 .20 .65 26. Upper Volta Rice 51 logs 2.2 .48 .11 .04 .85 27. Upper Volta Total, 51 - 639.0 21.00 74.00 5.00 .43 28. Upper Volta Total 51 logs 6.3 .22 .12 ' .05 .49 29. Sahel (0° SorgJMill. 51 -- 2,303.0 35.00 138.00 9.00 .38 30. Sahel (4) Sorg./Mill. 51 logs .45 31. Sahel (4) Rice 51 - 230.0 4.00 18.00 1.00 .34 32. Sahel (4) Total 51 - 2,707.0 41.00 162.00 11.00 .38 33. Sahel (4) Total 51 logs 7.8 .12 .07 .03 .45 Yield 35. Mali Maize 61 - 1,151.0 -28.00 116.00 13.00 .27 36. Mali Total 61 - 958.0 -18.00 49.00 5.00 .45 37. Niger Sorghun 61 - 695.0 -17.00 53.00 6.00 .41 38. Niger Maize 51 - 851.0 -15.00 64.00 4.00 .34 39. Niger Maize 61 - 791.0 -26.00 56.00 6.00 .57 40. Niger Rice 51 - 515.0 71.00 187.00 13.00 .58 41. Niger Rice 61 logs 7.0 .26 .18 .09 .38 42. Niger Total 51 - 614.0 12.00 51.00 3.00 .35 43. Niger Total 51 logs 6.3 .14 .09 .04 .40 44. Senegal Sorg./Mill. 51 - 383.0 8.00 32.00 2.00 .35 45. Senegal Maize 51 - 1,096.0 -15.00 62.00 4.00 .37 46. Upper VOlta Rice 61 logs 6.5 .16 .10 .05 .44 47. Sahel (4) Rice 51 - 829.0 20.00 58.00 4.00 .53 48. Sahel (4) Total 51 logs 6.5 .06 .05 .02 .24 SOURCES: IBRD, USDA. aLog transformations included only if R2 greater than that for nontransformed. b cLess Chad and Mauritania. Production, yield and time transformed to logs. standard errors, for any particular year, can be retransformed by anti-logs. duction equation would therefore be in form: Prod - anti-log (a) * time. Regression coefficients and Pro- 278 .omuumo to no. no oaoonmnamnm =«= nuns muamnunoomoo "mooz .monmmm some Hoooo .on oo.n «no. «on. non. son. on. «on. No. no.n .muom sunoo swoon .o oo.n «on. son. «on. an. «an. an. no. .nnnz sono> Home: .o oo.n «no. «ms. no. nos. an. no. Hosea nomoomn .a oo.n «on. so. «ms. NH. no. .muon\.nnn= Hammomn .o oo.n «an. inn. snn. «nn. fiasco ummnz .n oo.n «on. nos. nos. .muom sownz .s oo.n su. on. .Hnn: “monz .n oo.n «no. Hobos ens: .N oo.n .muom\.nnnz nan: .n .on .n .o .5 .n .n .s .n .N .n onaoouoo huuoooo ouHo> some: Hmwocmm nomnz «an: neonnnnon .zonooooomo o some: .on co.~ saw. on. an. nu. mu. «N. .n au.n an. ON. mg. on. .nuom euuo> Home: .nH co.~ on. an. son. an. eco. «a. #0. na. an. e~.n an.n .Hnwz suao> Home: .NH Ocnn cam. om. ed. ON. on. ha. On. an. oc.u mo.n Hausa Homomom .dH oo.~ NH. we. 0H. NH. ma. cu. my. m~.n m~.u, .uuom\.~dqz Howocmm .OH 8; son. uso. no. «no. so. «no. 2. an. :33 .8on a oc.d emu. emu. sac. emu. one. «so. «so. .muom mownz .m oo.u «as. ewe. «on. man. 0N. «N. .nuuz sown: .n cc.u «mm. «mm. ems. ens. eds. deuce «Ho: .o 84 an. .2. .oo. to. $3332: 23. .n oc.~ «am. «we. «be. Hmuoa odcuuuuoo: .c 8. n «so . Lo. .983 .23. 38326: .n oc.u toms Hooch coco .N oo.n .962 4:: osoo .n .eu .nu .Na .du .on .a .a .n .0 .n .c .n .N .H ouHo> some: Heuocem nowmz «do: encounuooz coco Aevaooueu anacooo nNaHnncOd .ZOHHoaanm A¢fldmo AdHOH :24 zamuzom .HNAAux zs sonata: swoon inane. oooo Hz\oow mononouos won Hoconufivmuh ammou hho>H one: .oeoni ans: Hz\omm owouo>o wmnxuoz omnu mommHH upcmaoeoouo wz\aww Hz ooo.H Hz\ooo a: ooo Hz\Hmm Hz ooo “omsonouma mo ounm mono mommHH mouwmo< mwwuoum mmoosouoz mum Hz\osm Homnz How moumoo .oxoouo zoammuoao pow chasm on com: :oanm mshdho: .n Hz\m-w unwaennvo wcnpmoa Jase aaow suns oHHm Hausa venom» o>ono AHZoHo manpooa can; moans Hausa vmsouw o>on< was: momma. xnm: n ownuoum oanm Hz\umoo aoumnm mo mans humanou mouoom mMH