This is to certify that the thesis entitled A SOCIOECONOMIC ANALYSIS OF FARMERS' FOOD GRAIN MARKETING LINKAGES AND BEHAVIOR IN EASTERN UPPER VOLTA presented by Ismael S. Ouedraogo has been accepted towards fulfillment of the requirements for Ph.D. Agricultural Economics deg/WV 0 Major professor degree in £202 7/700; 0-7639 MS U i: an Affirmative Action/Equal Opportunity Institution MSU LIBRARIES .4___. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. wast ' vafi1nal 111 A112 Armoz Oibsot 01 / .. A SOCIOECONOMIC ANALYSIS OF FARMERS' FOOD GRAIN MARKETING LINKAGES AND BEHAVIOR IN EASTERN UPPER VOLTA By Ismael S. Ouedraogo A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultural Economics 1983 ABSTRACT A SOCIOECONOMIC ANALYSIS OF FARMERS' FOOD GRAIN MARKETING LINKAGES AND BEHAVIOR IN EASTERN UPPER VOLTA By Ismael S. Ouedraogo The purpose of the study was to analyze food grain (sorghum and millet) marketing processes, and producers' food grain marketing behavior and problems in Eastern Upper Volta. The study was part of a large microeconomic research effort aimed at evaluating animal traction pack— ages and providing baseline data for future comparative studies and evaluation. The analysis follows a descriptive-diagnosis-prescriptive framework. The analytical measures include descriptive analysis, food grain and cash flow statements, timing of transactions, and inferential statis- tical tests. Primary data were collected in a region-wide farm survey of 480 farmers, from May 1978 to April l979, and a region-wide market- ing survey, from July 1979 to June l980. The analysis finds that the E-ORD grain marketing system is com- prised of house trading, marketplace-oriented trading, and farm gate- oriented assembly processes particularly adapted to the local conditions. Both public and private networks mainly operate to export grain outside the region, and neglect the back-flow of grain. Household grain flows involve small exchange volumes. Ismael S. Ouedraogo In 1978-79, there were more grain purchases than sales, most stocks were decreased, and there was hardly any marketed or marketable surplus in the villages studied. Farm gate prices received and paid were low at harvest and high at the hungry season. But average sales were larger before and after harvest than during harvesttime. Food and other cash expenses were financed more by other enterprises than crop sales. Another major finding is that advanced sales are not that important in terms of volumes and number of producers involved. The information relating to farmers' perception of market fairness and measurement on 'the farm is mixed. The data reveals farmers' strong needs for better measurement devices. Another finding is that Village Cereal Banks face serious problems necessitating a rethinking of the scheme's objectives and management. To improve the E-ORD's grain marketing performance, the study recom- mends policy actions aimed at providing better marketing services to producers, and fostering better cooperation between public and private institutions. Future research needs in the E-ORD are also suggested in areas of participants' storage behavior, marketing costs, and key mar- keting statistics. To My late mother, Bintou Kaboré and to my wife, Betty, and my daughters, Nina and Laissa in appreciation for their love ACKNOWLEDGMENTS My deepest gratitude goes to Carl Eicher who provided advice and assistance throughout my graduate program. Special appreciation is ex- tended to Michael Weber, my major professor and thesis supervisor, whose guidance and friendship were instrumental in the completion of this dis- sertation. Thanks are also given to members of my guidance committee, James Shaffer, Eric Crawford, and Peter Schmidt, and to Harold Riley and Assefa Mehretu for their insightful comments. All errors and omissions are my own. Many other individuals also contributed to this study. I thank my long-time friend, Edward Tapsoba,for his encouragements, and David Wilcock for his leadership in the field research. I am indebted to them and the other members of the MSU-Fada team, Greg Lassiter, Vince Barrett, and Thomas Stickley,for the collection and preparation of much of the data used in this study. Special acknowledgment is given to Jean Dahny who contributed invaluable services as a knowledgeable informant. Chris Wolf and Paul Wolberg provided assistance in the computer work, and James Bingen's help is appreciated for arranging with USAID part of the financing for my Ph.D. program. Special acknowledgment is also given to Cindy Spiegel, who typed earlier drafts of the manuscript, and Nancy Creed with assistance of Debbie Greer, who typed the final copy, for their tenacious work under considerable time pressure. iii I also want to thank my loving wife, Betty. Her confidence and support cannot be overstated. TABLE OF CONTENTS Page DEDICATION ............................ ii ACKNOWLEDGMENTS .......................... iii LIST OF TABLES .......................... x LIST OF FIGURES .......................... ixv CHAPTER 1 . INTRODUCTION ............ . ........... 1 1.1 Background ....................... 1 1.2 Problem Statement ................... 3 1.3 Objectives of the Study ................ 5 1.4 Organization of the Study . . . .. ........... 6 ' 2. APPROACHES IN THE STUDY OF AGRICULTURAL MARKETING IN DEVELOPING COUNTRIES ................... 9 2.1 Introduction ...................... 9 2.2 Agricultural Marketing in the Development Process . . . 9 . 2.3 Typology of Approaches in the Study of Marketing 1 Issues ......................... 11 2.3.1 Feasibility Studies ............... 11 2.3.2 Descriptive Studies ............... 12 2.3.3 Economic Diagnostic Assessments ......... 13 2.4 A Framework for Studying Rural Food Marketing ..... 16 2.5 Summary ........................ 18 3. MAJOR TRENDS IN FOOD GRAIN MARKETING IN UPPER VOLTA ..... 20 3.1 Introduction ...................... 20 3. 2 General Trends in Economic Growth and Food Availability ...................... 20 ; 3. 3 Evolution of the Debate on Food Grain Marketing . . . . 22 ’ 3.3.1 The Private Trade ................ 22 3.3.2 Historic Perspective of Government nterventions in Food Marketing ......... 23P” 3.4 The Present Issues ................... 33 4. RESEARCH DESIGN AND INSTRUMENTS ............... 35 4.1 Introduction ...................... 35 4. 2 Study Site ....................... 35 4.2.1 Main Physical Characteristics .......... 35 4.2.2 Relevant Institutional Characteristics ..... 41 4.3 Research Design and Instruments ............ 44 V CHAPTER .1 The 1978-79 Farm Survey ............ . .2 The 1979-1980 Food Grain Marketing Investigation .................. .3 The 1980 E— 0RD Village Inventory Survey ..... .4 Data Limitations ................ 4. 4 Summary ........................ 43-5 15% (.000 (04.0 SORGHUM- MILLET MARKETING CHANNELS AND PROCESSES ....... 51 Introduction ...................... 5.2 Marketplaces in Eastern Upper Volta .......... 5.2.1 Number and Periodicity ............. 5.2.2 Location and Size of Marketplaces ........ 5.3 Typology of Major Market Participants ......... l The National Wholesaler-Retailer (NWR) ..... 2 The Regional Wholesaler-Retailer (RNR) ..... 3 The Local Independent Trader (LIT) ....... 4 The Trader's Apprentice (TA) .......... 5 The Commission Agent (CA) ............ 6 The Village Resident Buyer (VRB) ........ n Marketing Channels in E-ORD ........... 1 Traditional Grain Marketing System ....... 5.4.1.1 Marketplace-Oriented Trade ....... 5.4.1.2 Farm Gate-Oriented Trade ........ 5 4 2 State Marketing Channel ............. 5 4 3 "Back Flow" of Grain .............. 5.4 4 Major Grain Physical Movements in E-ORD ..... 5.5 Standard Operating Procedures .............. 5 5 5 5 2 5 5 3 5.4 r 3. .3 3. .3 3. .3 ai 4 momma-101mm Private Agent Network .............. OFNACER Agents'Operating Procedures ....... A Total System View of the Grain Marketing Processes .................... 5.6 Private Merchants'Physical Distribution Out-of-Pocket Costs ......................... 5.6.1 Cost Components ................. 5.6.2 Illustrative Marketing Costs .......... 5.7 Summary ........................ VILLAGE AND FARM-HOUSEHOLD LEVEL ANALYSIS OF GRAIN MARKETING BEHAVIOR ..................... 6.1 Introduction ...................... 6.2 Concepts and Operational Definitions of the Grain Flow Statement, Marketed and Marketable Surplus . . . . 6.2.1 Grain Flow Statement .............. 6.2.2 Food Available to the Farm-Family Unit a . . . . 6.2.3 Grain Disposals .................. 6.2.4 Grain Marketed and Marketable Surplus ...... 6.3 Sources and Limitations of the Data and Estimation Procedures ...................... 6.3.1 Sample Selection ................ vi 118 Page CHAPTER 6. 3. 2 Yield, Acreage, Production, and Seed Use Data ...................... 122 6. 3. 3 Consumption, Stock, and Storage Losses ..... 123 6. 3. 4 Processed Food Grain .............. 127 6.4 Gr ra1 n Flow Statement .................. 129 6.4.1 Annual Grain Flow Statement ........... 129 6.4.2 Seasonal Grain Flow Statement .......... 136 6. 4. 3 Stock Change and Minimum Consumption Requirement ................... 139 6.5 Marketed and Marketable Surplus ............ 143 6.6 Timing of Farm Sales and Household Purchases ...... 147 6.6.1 Producers Grain Sales and Repurchases ...... 147 6.6.2 Timing of Farm Sales and Household Purchases .................... 155 6.7 Farm Gate Prices .................... 162 6.8 Cash Flow ....................... 169 6.9 Summary ........................ 177 7. GRAIN PRODUCERSI MARKETING LINKAGES AND PROBLEMS ...... 180 7.1 Introduction ...................... 180 7.2 Marketing Problems Felt by Producers .......... 181 7.2.1 Producers' Awareness of Marketing Problems . . . 181 7.2.2 Producers' Attitudes Toward Public Intervention in Grain Marketing ............... 183 7. 2. 3 Producers' Suggested Solutions ......... 188 7. 2. 4 Problem Diagnosis ................ 190 7.3 Measurement of Farm Crops ............... 192 7.3.1 Issues and Hypotheses .............. 192 7. 3. 2 Measurement on the Farm ............. 195 7. 3. 3 Measurement by Producers on Farm and Perceived Market Fairness ................. 196 7. 3. 4 Farmers' Preferences for Measuring Devices . . . 201 7. 3. 5 Price and Volume Differential Related to Measuring Devices ................ 204 7. 3. 6 Synthesis .................... 206 7.4 Market ting Credit and Advanced Grain Sales ....... 207 7.4.1 The Issues 7. 4. 2 E-ORD Areas with Advanced Grain Sales ...... 209 7. 4. 3 The Magnitude of Advanced Grain Sales ...... 211 7. 4. 4 Some Patterns Related to Advanced Sales ..... 212 7. 4. 5 Grain Prices in Advanced Sales ......... 214 7. 4. 6 Inputed Nominal Interest Rate of Marketing Credit ..................... 217 7. 4. 7 Synthesis .................... 223 7.5 An Evaluation of the Village Cereal Bank Program. . . . 224 7.5.1 Objectives of the Village Cereal Bank ...... 224 7. 5. 2 Performance Dimensions ............. 226 7.5.3 Evolution of the Village Cereal Bank Membership ................... 227 vii CHAPTER 7.5.4 Procurement and Sale of Grain .......... 7.5.5 Viability of the Village Cereal Bank System. . . 7.5.6 Problem Diagnosis ................ 7.5.7 Synthesis .................... 7.7 Summa ........................ 8. SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS .......... 8.1 Study Objectives .................... 8.2 Major Research Findings ................ Historical Perspective of Food Grain Marketing Issues in Upper Volta .............. 8.2.2 Characteristics of the E-ORD's Food Grain Marketing Channels ............... 8.2.3 Farm-households' Food Grain Flows and Transaction Patterns .............. 8.2.4 Timing and Farm Gate Prices of Farm Grain Sales and Purchases ............... 8.2.5 Farm-households' Cash Flows ........... 8.2.6 Diagnosis of Food Grain Marketing Problems and Issues in the E-ORD ............... 8.3 Recommendations .................... 8.3.1 Improved Sorghum-Millet Varieties and Farming Techniques ................... 8.3.2 Physical Infrastructure ............. 8.3.3 The Marketing Role of the E-ORD Management . . . 8.3.4 The Role of OFNACER ............... 8.3.5 The Village Cereal Bank Program ......... 8.3.6 Standardization of Measurement ......... 8.4 Further Research Needs ................. APPENDIX A PHYSICAL DISTRIBUTION COSTS OF GRAIN MERCHANTS OPERATING IN EASTERN UPPER VOLTA: AN ILLUSTRATION ............ A.1 Local Independent Trader (LIT) ............. A.l.1 Case 1 ..................... A.1.2 Case 2 ..................... A.1.3 Case 3 ..................... A.2 Regional Wholesaler-Retailer (RNR) ........... A.3 National Wholesaler-Retailer (NNR) ........... APPENDIX B KG CONVERSION FACTORS FOR SELECTED MILLET-SORGHUM MEASUREMENT DEVICES. . . .................... APPENDIX C EXAMPLES OF INTERNALLY INCONSISTENT PERIODIC STOCK ESTIMATES OF SORGHUM AND MILLET IN SELECTED VILLAGES ........... viii 270 271 APPENDIX D PRODUCERS' ESTIMATES OF DAILY GRAIN CONSUMPTION (ONE-THIRD 0F 1978-79 FARM SURVEY SAMPLE) ................. APPENDIX E SOURCES AND USES OF SORGHUM AND MILLET IN SELECTED ANTRAC ZONES ......................... APPENDIX F FARMERS' RECALL OF THE 1978 GRAIN CROP USES (IN PERCENT OF TOTAL PRODUCTION) ..................... APPENDIX G SEASONAL GRAIN FLOW STATEMENT AND PERFORMANCE MEASURES FOR ANTRAC AND TRADITIONAL HOUSEHOLDS (SELECTED ANTRAC ZONES). . . . APPENDIX H CHANGES IN STOCK AND MARKETED SURPLUS UNDER ALTERNATIVE GRAIN CONSUMPTION REQUIREMENTS .................... APPENDIX I SEASONAL GRAIN MARKETED AND MARKETABLE SURPLUS BY TECHNOLOGY IN SELECTED ZONES (AVERAGE KG PER HOUSEHOLD) .......... APPENDIX J ' ' RELATIONSHIP BETWEEN ACTUAL GRAIN TRANSACTIONS AND LEVEL OF TECHNOLOGY, ECOLOGICAL ZONES, MARKET, AND ROAD ......... APPENDIX K MONTHLY AVERAGE FARM GATE PRICES ESTIMATED FROM OBSERVATIONS OF REDUCED SAMPLE (196 HOUSEHOLDS) IN ANTRAC ZONES (WITHOUT TRIMMING) ........................... APPENDIX L RETAIL PRICES FOR SORGHUM AND MILLET BY MONTH IN ZINIARE AND NABITENGA, 1980-81 (CALCULATED AS MOVING SIX WEEK AVERAGE) . . . APPENDIX M MONTHLY CASH FLOWS ...................... APPENDIX N MARKET SHARES .......................... APPENDIX 0 DISTANCES TRAVELLED TO SELL FARM GRAIN AND BUY GRAIN FOR HOUSEHOLD NEEDS ......................... BIBLIOGRAPHY ........................... Page 273 279 280 281 282 283 285 287 293 295 296 -h -P 00 00 w 0 I I i 0 2a 2b LIST OF TABLES OFNACER Cereals Operations, l971/72-—1978/79 ....... OFNACER Grain Purchases, 1978-79 ............. OFNACER Grain Purchases in Eastern 0RD, 1979-80 ..... Agroclimatic Characteristics of Surveyed Zones ...... Distribution of the 480 Samples Agricultural Households by Agroclimatic Zone, Village, and Technology ...... Location of Sampled Villages of Surveyed Farmers, 1979-80 Farm Survey ................... Sample of Markets and Producers Surveyed in the 1980 Marketing Survey ..................... Market Periodicity: Percent of Village Markets With Varying Periodicity in Each Sector ............ Rural Markets in the Eastern Region That Serve More Than Five Villages .................... Distribution of Village Markets in the Eastern Region by Mean Radius of Market Area .............. Major Agricultural Produce Markets in Eastern Upper Volta .......................... Criteria of Categorization and Types of Major Middlemen ........................ OFNACER Agents' Network in 1979-1980 ........... Illustrative Marketing Costs Incurred by Private Traders ......................... Grain Transport Costs Along Selected Transport Routes . . . Sorghum-Millet Acreage, Production, and Seed Use, and Total Household Income by Zone and Technology in Selected Zones ...................... Page 31 32 32 38 46 50 52 60 61 62 67 72 89 120 6.14 6.15 6.16 Page Millet-Sorghum Annual Storage Loss Rate (Selected Households' Estimates) ................... 128 Annual Grain Flow and Performance Measures by Zones and Technology (Selected ANTRAC Zones) ........... 130 Seasonal Grain Flow and Performance Measures by Zones (Selected ANTRAC Zones) .................. 137 Projected Ending Stock (May 79) Under the Assumption of Zero Stock Prior to Current Harvest (kg.) ........ 142 Annual Grain Marketed and Marketable Surplus by Zone and Technology in Selected Zones (Average kg per Household) ......................... 144 Seasonal Grain Marketed and Marketable Surplus in Selected Zones (Average kg per Household) ......... 146 Distribution of Grain Production, Grain Yield and Farm-Household Income According to Grain Transaction Patterns .......................... 150 Farm Sales and Household Purchases of Grain in Relation With Farming Technology and Grain Transaction Patterns. . . 151 Farm Sales and Household Purchases of Grain in Relation With Household Income Distribution and Grain Transaction Patterns .......................... 153 Relationship Between Farming Technology and Sale/Purchase Volume Differential for Households with Grain Sales and Repurchases ........................ 154 Relationship Between Household Income Distribution and Sale/Purchase Volume Differential for Households with Grain Sales and Repurchases ................ 156 Timing of Grain Farm Sales and Household Purchases by Actual Transaction Category (Monthly Percentage of Annual Volume), and Average Farm Gate Prices Received and Paid (in FCFA per kg) ...................... 158 Monthly Percent of Farm Grain Sales and Household Purchases by Household Income Classes ........... 164 Annual Cash Flow and Performance Measures by Ecological Zones, Technology, and Grain Transaction Pattern ...... 171 Relationship Between Cash Generation Through Crops and Animal Sales and Patterns of Grain and Animal Transactions ........................ 173 xi 7.5a \l \l \l \l a o . o (D 0' 7.12 7.13 Relationship Between Cash Expenditures on Foodstuffs and Patterns of Grain Transactions ............. Relationships Between Crops and Animal Sales Contribution to Cash Inflow ................ Relative Importance of Marketing Problems Producers Felt the E-ORD Should Pay Attention To ........... Perceptions of E-ORD Grain Marketing ............ Major Reasons Why Producers Did Not Sell to OFNACER n198O Producers' Suggestions for the Improvement of the Performance of the Marketing System ............ Frequency and Reasons for Grain Measurement ........ Do Producers Measure Grain at the Farm Before Market Sale? ........................ Reasons Why Households Usually Do Not Measure at the Farm ............................ Measure of Produce on the Farm and Perceived Market Fairness (Chi-Square Tests) ................ Measure on the Farm and Perceived Market Fairness (T-Tests of Grain Values Expected, Actually Received, and Differential) ....................... Reasons of Preference for Measurement Devices ....... Heads of Household (1980 Farm Survey) ........... Farmers Interviewed at Selected Marketplaces. .' ...... Village Locations Where Merchants Usually Buy Millet-Sorghum in Green .................. Producers' Transaction Patterns Related to Green Sales of Grain .......................... Price Quotes for Advanced Sales of Millet-Sorghum Aggregated Across Zones and Periods ............ Exchange Rate Quotes for Repayment of Food Grain Bought on Credit in the Hungry Season ............... Advanced Sale Prices and Nominal Annual Interest Rates. . xii Page 174 176 182 185 189 197 197 197 199 200 202 202 203 210 213 216 218 219 \l \l \J a \INNNN .13a .13b .15 .16 .17 .18 .19 Actual Borrowing ...................... Actual Lending ....................... Inputed Interest Rate Charged to Producers Who Paid Back and Received By One Selected Grain Merchant ...... Number of Active VCBs (1977-78 to 1979-80) ......... Evolution of the Membership of the VCB From 1977 to 1980. Millet-Sorghum Bought and Sold by VCBs in 1978-79 ..... Number and Scheduling of Loans to Village Cereal Banks. . . Loan Repayment Rate of VCBs ................ xiii Page 219 220 222 228 229 231 232 234 "T 6.4 6.4a 6.4b LIST OF FIGURES Upper Volta: Proximity Within West Africa and Departmental Boundaries .................. Eastern Region of Upper Volta: Climatic Characteristics ...................... Eastern Region of Upper Volta: Population Density by Canton ........................... Eastern Region of Upper Volta: Surface Transportation. . . Map of Sampled Villages: Eastern ORD--Farm Survey, 1978-79 .......................... Location and Size of Markets in the Eastern Region ..... Traditional Grain Distribution Channels in Marketplace- Oriented Trade ...... , ................ Traditional Marketing System Showing House Trading, Assembly at the Farm Gate, and Merchants' Transactions in Marketplaces ........................ OFNACER Marketing Channel ................. Conservation of Flow Principle ............... Coverage of the 1978-79 Farm Survey and Related Marketing Seasons ..................... Sources of Sorghum and Millet for ANTRAC and Hoe Farmers. . Sources of Sorghum and Millet for ANTRAC Farmers (Carryover Stocks Not Included .............. Sources of Sorghum and Millet for Hoe Farmers (Carryover Stocks Not Included) .............. Uses of Sorghum and Millet by ANTRAC and Hoe Farmers. . . . Uses of Sorghum and Millet by ANTRAC Farmers (Ending Stocks Not Included) .................... Uses of Sorghum and Millet by Hoe Farmers (Ending Stocks Not Included) .................... ixv Page 36 37 4O 43 45 63 75 77 80 106 121 132 132 132 133 133 133 6. 6. 01010101010» 5 6 .6a .6b .6a .7a 7b 8a Assumed Grain Stock Behavior of Non- Self-Sufficient Producers ................. Timing of Farm Sales and Household Purchases by Technology and Grain Transaction Patterns ......... Timing of Farm Sales and Household Purchases (Households with Grain Sales and Repurchases) ....... Timing of Household Purchases (Farmers with Household Grain Purchases Only) .............. Timing of Farm Sales (Households with Farm Grain Sales Only) ........................ Grain Sales and Purchases by Grain Transaction Categories ......................... Timing of Farm Grain Sales ................. Timing of Household Grain Purchases ............ Timing of Grain Sales and Purchases by Income Classes . . . Timing of Farm Grain Sales by Income Classes ........ Household Grain Purchases by Income Classes ........ Sorghum and Millet Farm Gate Prices (Trimmed) ....... VX Page 141 159 159 161 163 163 163 165 165 165 168 CHAPTER 1 INTRODUCTION 1.1 Background Going into the 19805 and the third decade of their independence, most sub-Saharan countries still face the biggest challenge of economic development: How to provide to all segments of their growing populations with abundant, reliable, and nutritious food supplies. On the whole, it appears that the challenge is not being met. A recent USDA report (1981, p. 1) shows that "sub-Saharan Africa is the only region of the world where per capita food production declined over the past two decades." Clearly, numerous factors account for the precarious food situation in sub-Saharan Africa. Strong population growth certainly plays an im- portant part. Natural disasters, undoubtedly, have also taken their heavy toll. (Caldwell, 1975, estimated that the drought that plagued Sahelian countries in 1968-1974 reduced total food production by a third or more and was directly responsible for the deaths of up to 100,000 people.) Even barring prolonged droughts, the unreliability of annual rainfall plays havoc with dry land farming systems. There is also a lack of readily available improved production technology, effective ex— tension services, and functional rural infrastructure. However, the food problem is not solely caused by low per capita food production; it is also caused by inadequate food distribution sys- tems. In fact, it is the whole food production-distribution-consumption 1 system that is not performing well. That is, even when production is abundant, food may not flow adequately within the rural area and from the producing areas to urban consumers. Many food production projects have been thwarted by the neglect of the input and product distribution system. The neglect of the agricultural marketing system by policy makers is rooted in misconceptions and misunderstandings. First, economic planners sometimes believe that the marketing system is self-evolving and that only farmers need assistance. Second, decision makers often think that private traders are exploitative or parasitic, and are thus led to the belief that government agencies could replace private mer- chants and provide all necessary marketing services at lower costs. Third, the knowledge of the behavior of so-called subsistence farmers with respect to marketing is very shallow. For example, development projects sometimes assume simplistically either that producers are cur- rently satisfying all their consumption needs so that all increased sur- plus production generated would be available to urban consumers, or that the food disposed of is a fixed proportion of total output. The farming system in developing countries is neither a subsistence (autarchic) nor a commercial agriculture, but rather one in transition (R. Krishna, 1969; Wharton, 1969) in which food is not only produced for consumption, but also for sale. Food is also disposed of and acquired through many exchange systems. Thus, there is a need for a better under- standing of the food availability at the farm level and the functioning of the rural food distribution system. Such understanding is the key to improving the food production and marketing situation in sub-Saharan Africa where the bulk of the population still lives in rural areas. 7'? 3 Unfortunately, there is often little available research on market- ing. Hence, when reform is contemplated, policy makers introduce it in ignorance of the existence situation. Jones (1972, pp. 2-3) emphasizes that they should be "provided with the soundest possible information about the present operation of the system so that they may identify where government intervention might improve the system's operation and where it might be harmful." 1.2 Problem Statement In Upper Volta, a land-locked sub-Saharan country with a population of 6 million and the size of 274 square kilometers, the Regional Develop- ment Organizations (0RD for Organisme Regional de Developpement), whose mandate is to promote agricultural development at the regional level, have had very little relevant marketing information to rely upon. The .knowledge about food marketing in Upper Volta is sparse and is mostly, provided by anthropologists and sociologists (Meillassoux, 1969; Kohler, 1971), and by short-term consultants whose reports may suffer from lack of in-depth analysis due, in part, to an unreliable data base (Morel, 1973; Bollinger, 1974). Among all 11 ORDs in Upper Volta, the Eastern ORD (E-ORD) is per- haps the one for which this lack of information is the most dramatic. Yet, the Eastern Region is considered an important exporter of the staple food crop (sorghum and millet) and is thought to have even greater pro- duction potential. In 1974, an MSU consultancy mission found that the E-ORD lacked the most basic agricultural statistics, demographic data, and economic information (Eicher, et a1., 1976). The mission recommended an Integrated Rural Development (IRD) project aimed at strengthening the '®~ 4 infrastructure of the regional agricultural extension organization, de- veloping its institutional capability, and implementing an experimental extension program. Within this IRD project financed by USAID, MSU was selected to carry a microeconomic research at the village level. From 1977 to 1981, the main thrust of MSU's effort in the E-ORD was to design and carry a large-scale farm survey1 designed to provide socio- economic data to help evaluate the farm level impact of the current tech- nical package (animal traction), provide basic information for regional planning and project design, and serve as a baseline for future compara— tive studies and evaluation.2 As part of this farm survey, micro-level market research was to provide a basic understanding of the components of the rural food grain marketing system with particular emphasis on two questions, "...a systematic understanding of farmer production and mar- keting decisions; ...and (an assessment of) the degree of integration, competitiveness, and efficiency in various marketing systems" (Eicher, et al., 1976, p. 38). The author helped carry out part of this market research from 1979 to 1980. At the start of the MSU farm survey in 1978, the E-ORD was confused about its actual and future role in grain marketing in the region. Be- cause of financial difficulties, it had stopped buying producers' grain 3) to sell to the cereal marketing board (OFNACER as the arrangement 1 2 See Chapter 4 for details on the farm survey design. See Lassiter, et al., 1982. 3OFNACER is the French acronym for Office National des Cereales. 5 between the two institutions called for. However, it was still managing the Village Cereal Bank, a project providing food grain consumption cred- it to farmers grouped in a precooperative association called village group (Groupement Villageois). As the foregoing comments illustrate, there is a need for research on the staple food grain marketing in the E-ORD. In the state of con- fusion described, the questions to be asked are basic ones. Elliot Berg's following comments provide a good summary of the issues faced by the E-ORD: "What we need is to know more about the structure and func- tioning of grain markets. More specifically, we need a set ofin- depth, village-level studies of crop disposal. This would in- volve closely linking marketing with production and labor use studies and would basically start at the harvest. The questions to be asked are: Who buys and who sells, when, where, to whom, and at what price? The link to the first market--whether house trade, local periodic market, or other—-should be explored in depth? There is also need for more standard types of marketing studies, following the flows of grain through the distribution channels from producers or local markets to major consuming centers. The important point is that the basic structure of these markets is so poorly known that studies of this general type would seem to have first priority. 1.3 Objectives of the Study The overall objective of the study is to provide a better under- standing of the processes of the rural food grain marketing system, par- ticularly farmers' market linkages and decisions with respect to food grain marketing in E-ORD. The study is aimed at complementing the other studies5 generated by the applied microeconomic research effort conducted 4Elliot Berg, "Discussant's Connents," p. 291, to B. Harriss' article (l979a). 5A few of these studies are Ph.D. dissertations: Tapsoba, 1981; Lassiter, 1982; P. Fotzo, 1983; Kifle Negash, 1983 (forthcoming); F. Sands, 1984 (forthcoming). 6 by MSU in the context of the E-ORD Integrated Rural Development Project. The study has the following specific research objectives: (1) To review the major food grain marketing issues in Upper Volta over the two decades following the country's in- dependence (in 1960). (2) To describe and analyze the food grain market channels and processes in terms of the organization and standard opera- tional procedures of major market participants. (3) To analyze producers' grain marketing behavior at the vil- lage level. (4) To provide a diagnostic analysis of farmers' market link- ages and problems. (5) To suggest policy recommendations and an outline of research needs for analyzing over the longer-run food grain market- ing problems in the E-ORD and Upper Volta. 1.4 Organization of the Study The balance of the study is organized into seven chapters. Chapter 2 reviews the literature on alternative approaches to the study of food marketing in the context of economic development. It identifies the role of marketing in development, develops a typology of major research approaches, and outlines the conceptual research frame- work used by the study. In the descriptive-diagnosis framework used, the study considers three levels: The national level (Chapter 3), the regional market channel level (Chapter 5), and the producer level (Chapters 6 and 7). —. 7 Chapter 3 provides a historical perspective of the major food mar- keting issues in Upper Volta in terms of the trend in food availability; this includes a discussion of the evolution of the debate on food grain marketing and particularly of the pattern of government interventions in grain marketing. Chapter 4 reviews the research design and instruments. In par- ticular, the main physical and institutional characteristics of the study site are described, along with the characteristics of the various empiri- cal data sources which support the study. Chapter 5 is a descriptive diagnostic analysis of the Eastern Region food grain marketing channels and processes. It describes the character- istics of the marketplaces, the major market participants, and the private and public market channels. It examines also the standard operational procedures of private and public marketing networks and estimates mar- keting costs in selected private channels. Chapter 6 analyzes producers' marketing behavior at the village level. It uses the food grain flow statement and marketed-marketable surplus measures to examine the behavior of farmers who are grouped in a sample that is stratified according to ecological zones, farming tech- niques, and farm income. Chapter 6 also examines the patterns and tim- ing of grain sales and purchases, farmers' cash flow situations, and the variability of farm gate prices paid and received. Chapter 7 analyzes grain producers' market linkages and problems. It examines farmers' perceptions of marketing problems and the role of public agencies. It also presents empirical evidence on the issues of grain measurement and marketing credit. Finally, it evaluates the 8 Village Cereal Bank program which provides credit to farmers' groups to buy, store, and sell grain. Chapter 8 reviews and smnmarizes the major findings of the study. Policy recommendations are formulated and future research needs are suggested. CHAPTER 2 APPROACHES IN THE STUDY OF AGRICULTURAL MARKETING IN DEVELOPING COUNTRIES 2.1 Introduction This chapter reviews the literature on agricultural marketing in developing countries in general, and in West Africa, in particular. It focuses on the role of agricultural marketing in the development process and the types of approaches used to study that role. It also serves as a backdrop to the review of the food marketing issues in Upper Volta presented in the next chapter. 2.2 Agricultural Marketing in the Development Process The agricultural sector has long been recognized as a key to eco- nomic development. Within this sector, however, the role of agricultur- al marketing has been often times neglected to the benefit of produc- tion. Part of this neglect may be traced to scholars and economists' long-held belief that only production, or physical transformation, mat- ters and that agricultural marketing passively adapts to economic de- velopment stages (K.D. Harrison, et al., 1974). Starting in the early 1950s, however, economic scholars and plan- ners have reassessed these traditional beliefs and recognized the dynamic role of agricultural marketing in the process of agricultural and economic development. R. Holton (1953) held that if marketing chan- nels were less tortuous, more goods would flow through them and reach 9 w 10 more consumers. J.C. Abbott (1967) stressed the vital importance of integrating the expansion of marketing functions with the expansion of production. N.R. Collins and R.H. Holton (1964) emphasized that dis- tribution can play a vital role in economic development by changing demand and cost functions in ways favorable to development. C.C. Slater, H. Riley, et a1. (1969).conceptualized a market development strategy within the context of a comprehensive approach to development planning. A key to the understanding of the dynamic role of marketing in de- velopment is the view that agricultural marketing should not be limited only to the activities that occur after the products pass through the "farm gate." Rather, it should be viewed as a "primary mechanism for coordinating production, distribution, and consumption activities" or as a "part of the set of activities coordinating various stages in a production-distribution channel such as the food system or a commodity subsystem." Agricultural marketing includes “the exchange activities associated with the transfer of property rights to commodities, the physical handling of products, and the institutional arrangements for facilitating these activities." (K.D. Harrison, et al., 1974). Even though there is no single and unique path to economic growth and development, agricultural marketing plays a central role in all eco- nomic systems. In the development process, agricultural producers rely more and more on the environment outside the farm-family unit to pro- vide food, consumer goods, and farm inputs. In addition, the growth of the urban and other nonfarm rural populations requires more marketing services to coordinate production and consumption activities. Hence, the lack of market access or the high cost of marketing services can cancel out efficiency in production, for producers will not get 11 acceptable cash returns which provide incentives to produce. The back flow of food, goods, and services to producers may serve also as an in- centive to produce more. 2.3 Typology of Approaches in the Study of Marketing Issues Marketing processes extend beyond traditional economic concerns and are embedded in the larger social system.1 Hence, the social makeup and structural relationships are an important element in marketing studies. Marketing activities are also spread over large geographical areas and have spatial features. In sum, agricultural marketing research cuts across many disciplinary boundaries (geography, sociology, anthropology, economics, etc.) and addresses the problems of many different partici- pants or decision makers (G. Johnson, to be published). The approaches to the study of marketing issues in developing countries are varied, but they may be classified in three broad categOr- ies: feasibility studies, descriptive studies, and diagnostic assess— ments (H. Riley and M. Weber, 1979). 2.3.1 Feasibility Studies Feasibility studies concern the evaluation of the economic and/or financial viability of physical infrastructure such as wholesale market centers, food storage facilities, and road construction. Too often, these studies are undertaken in a short time without adequate checks on the validity of the data used. They use optimistic assumptions about 1According to Schmid and Shaffer (1964, p. 13), “The social 5 stem is the aggregate of institutions defining the relationships of any group of individuals. The economic s stem is that particular subset of institutions defining the limits of activity and dependence among in- dividuals in the provision and use of goods and services within a society. 12 the technical and managerial resources in the host countries and the result has been an inappropriate utilization of the facilities. 2.3.2 Descriptive Studies Studies by geographers, sociologists, and anthropologists provide valuable descriptive information on market systems, but they seldom car- ry policy recommendations for market development strategies. Carol A. Smith (1976, Vol. II, p. 6) suggests that "there have been practically no attempts (by anthropologists) to correlate spatial configurations with the formal properties of other aspects of social life." She adds that "geographers...tend to deal with the human component as given," and that "regional science still has far to go to make its models rele- vant to those who are concerned with real-world rather than normative or ideal systems." ‘ On the whole, geographers and regional scientists have encountered many difficulties in the modelling of marketplaces and periodicity in West Africa on the basis of central place theory (see C.A. Smith, 1976; R.T. Smith, 1978; and C. Meillassoux, 1969). This is because in de- veloping countries, the marketplace has social, political, and religious roles in addition to its exchange function. Furthermore, marketing activities may actually take place outside of physical market facili- ties2 and market days, in what has been called "house" or "hidden trade" (P. Hill, 1969; C. Meillassoux, 1969), which weakens the importance of marketplaces in centralplace theory. 2Polly Hill's definition of a marketplace is “an authorized con- course of buyers and sellers of commodities, meeting at a place more or less limited or defined, at an appointed time" (R.H.-T. Smith, 1978, P. 2 . 3. 13 Sociologists and anthropologists have found that many West African markets may have originated from the long-distance trade (in food grain, cola nuts, salt, animals, etc.) that linked in years past Saharan to Coastal states along the Benin Gulf. These researchers stress that be- cause of this long-distance trade in West Africa, millet and sorghum have been traditionally grown not only for home consumption, but also for sale (C. Arditi, 1975; C. Raynaut, 1973). Raynaut (1973, p. 214), for example, contends that for centuries, in the Sahel, millet and sor- ghum have been the main commercial crops in the trade between herders and farmers, and that the flow of grain from rural to urban centers is a relatively new phenomenon. He adds that this commercial nature of mil- let and sorghum has been recently transformed into "forced" or "distress sales" to meet cash needs introduced by the new direction of trade, the increased monetization of the subsiStence economy, and the power of trader, which leads to the unsettling of social structures. 2.3.3 Economic Diagnostic Assessments Diagnostic assessments aim at providing inputs to policy and pro- gram development. A French researcher (C. Arditi, 1975, p. 13) admits that, in the past, economic research on the rural food marketing were carried out more by English-trained researchers than by French-trained ones. It would appear that, in the past, the French have not carried many economic analyses of rural food marketing because some believed that the products exchanged were not "traded goods;“ that is, they don't have any exchange value. This has led them to believe that there was no “counercial” relations in the traditional food marketing (see Morel, 1973). However, this contradicts others' findings about the secular commercial nature of grain in West Africa. 14 Most rural food marketing studies conducted in West Africa have been in English-speaking countries (Nigeria, Sierra Leone, Ghana) and, only recently, in Francophone Africa by English-trained economists: Clark Ross (1979) in Senegal, J. Sherman (1981), and the present study in Upper Volta, etc. (see also the reviews by P. Riley and M. Weber, 1979; B. Harriss, 1978; and J. Sherman and I. Ouedraogo, 1981). The studies by E. Gilbert (1969), W.O. Jones (1972), H.M. Hays (1975), and N. Ejiga (1977) follow the same pattern in the use of a structuralist interpretation of the structure-conduct-performance (S- C-P) approach3 applied to price analysis to determine market competi- tion and evaluate market efficiency. On the basis of paired market price correlation coefficients, seasonal and spatial price differentials, and profit margins they conclude that the traditional food marketing system is performing fairly well and that government active interven- tion is counterproductive. E. Berg4 (1980), in studying marketing poli- cies in the Sahelian States, reaches the same conclusions. The recom- mendations that flow from these studies ask governments to provide only 3Presented by Joe S. Bain (1968) in his "Industrial Organization," - the S-C-P model postulates a predictable relationship between the struc- ture (organizational characteristics) of the industry, the conduct (be- havior) of firms within that industry, and the performance of the in- dustry. According to B. Marion (1976, pp. 4-5), most agricultural mar- keting economists using this approach have followed in the past three schools of thought. The first school of thought, "technological deter- minism," concentrates on the design of systems from a "logistics-produc- tion economics" point of view. The second, "behavioral," focuses on g9! the system functions, with strong emphasis on conduct affecting inter- firm vertical relationships. Finally, the third, "institutional," tends to concentrate on structure-performance relationships. It is the last school of thought and the perfect competition norm which provide the methodological underpinnings to past S-C-P studies of food marketing conducted in West Africa. 4See also cuss (1977). 15 facilitative services (infrastructure, economic incentives, and market information) and leave the market to itself. B. Harriss (1979a,b), a former user of S-C-P herself, "goes a- gainst the grain" of these conclusions. She asserts that the “fetishism of competition" and the "laissez faire" ideology of these researchers have weakened the value of their studies. She contends that price cor- relation is a faulty measure of market integration or competition owing to the facts, among others, that "high coefficients indicate stable margins or stable prices and by themselves could easily indicate monop- oly conditions as perfect competition" (l979b, p. 202), and that "mar- kets may well be integrated and yet have low correlations because market towns are centers of supply, nonfinal demand, and final demand" (l979b, p. 203). She also condemns the use of secondary data on market prices and the simplistic assumptions about price behavior over space and time, and she shows that in many studies there is no logical link between the empirical results and the conclusions drawn. Riley and Weber (1979) have also criticized the same type of S-C-P approach used in previous marketing studies, but on different grounds. They note (in p. 12) that the aggregation bias with respect to partici- pants' behavior and the data problem--stressed by Harriss-—often make the results of such studies inappropriate for policy recommendations to provide services to specific target groups such as smallholders. They add that these studies overlook the dynamic process of coordinating pro- duction, processing, distribution, which takes place even in subsistence economies. In particular, they stress that the fundamental weakness of such S-C-P studies is their overemphasis on structural variables-~and, l6 hence, the exclusion of conduct variables--as determinants of market performance. Another approach to the study of marketing in development is that of marketed and marketable surplus of food grain. The analysis is usu- ally cast in terms of the response to socioeconomic factors of house- holds' food crop production and disposals. Most studies of marketed and/or marketable surplus have been carried out in the Indian sub- continent or the Far East (see, for example, R. Krishna, 1962; J.R. Behrman, 1966 and 1968; Z. Toquero, et al., 1975; K. Bardhan, 1970; and W. Haessel, 1975), and only a few in Africa (see A.I. Medani, 1975; and Ross Clark, 1979; N. Ejiga, 1977; P. Matlon, 1977; V.E. Smith, et al., 1981; and G. Lassiter, 1982, also mention the issue). Many studies on marketed/marketable surplus provide empirical evi- dence of positive marketed/marketable surplus response to prices. Still, a few studies (P. Mathur and H. Ezekiel, 1961; T.N. Krishnan, 1965) show a negatiye response. It appears clearly that different researchers use many different definitions and even though these definitions may be justified on the basis of the characteristics of the system studied, it must be realized that the concepts are fraught with hidden assumptions (M. Newman, 1977). 2.4 A Framework for Studying Rural Food Marketing The review of literature shows the existence of many approaches for studying agricultural marketing in development. They all yield valuable insights in one aspect or another. However, the descriptive studies by geographers, sociologists, and anthropologists do not provide policy recommendations for improvement, feasibility studies do not include social relationships, and finally, many econOmic diagnostic assessments l7 patterned to the structuralist interpretation of the S-C-P approach are too narrow to provide workable recommendations to policy decision makers. To remedy these weaknesses, Harrison, et a1. (1974), J. Shaffer (1970), Slater, Riley, et a1. (1969), on the basis of their experiences from Latin America, have suggested a broad approach to study and formu- late recommendations about food marketing in development. This approach is based on a modified S-C-P framework and emphasizes market processes as economic coordination activities in a food distribution system. It uses a descriptive-diagnostic-prescriptive approach and the focus on marketing system avoids the sterile dichotomy between production and marketing. Market coordination is defined as "the process in an ex- change system5 whereby producers, distributors, and consumers interact to exchange relevant market information, establish conditions or ex- change, and accomplish physical and legal transfer of economic goods" (K.D. Harrison, 1966). A key to the understanding of market coordina- tion is the analysis of market participants' standard operational pro- cedures (SOP). In addition, the approach to market system diagnosis argues that the performance6 of agricultural marketing systems can be improved through a variety of government active interventions owing to the observation that "the most important marketing problems related to 5A. Schmid and J. Shaffer (1964, p. 13) define the exchange system as the "subset of the social system governing transactions between in- dividuals and groups which result in the exchange of property rights of future control of assets." The exchange system (intangible social rela- tionships) and the physical distribution system (tangible physical rela- tionships) make up the marketing system. 6Performance refers to economic consequences such as "efficiency in the use of resources in marketing activities, effectiveness in market coordination, fulfillment of consumer quality preferences, and competi- tive flexibility and willingness of market participants to innovate and progress" (K.D. Harrison, 1974, pp.4-5). 18 achieving the desired structural transformation are in the design and promotion of new technologies and new institutional arrangements which nay be unprofitable or unavailable to individual market participants, but if adopted by all participants, could yield substantial system im- provements" (H. Riley and M. Weber, 1979, p. 13). To facilitate the study, the food-distribution system may be broken down into three levels or entry points: (1) the farm or firm (trading enterprise) level; (2) the marketing channel level; and (3) the system level. In particular, the analysis at the channel level focuses on com- modity subsystems defined as "the entire set of activities performed in the production, assembly, processing, distribution, and consumption of a single product“ (K.D. Harrison, 1974, p. 56), where vertical coordina- tion refers to ways, such as the price mechanism and/or administrative regulated system, of harmonizing the vertical stages of a food-distribu- tion process (K.D. Harrison, 1974, p. 27). The framework of the present study takes from the descriptive diag- nosis approach. However, local conditions and information resources (see Chapter 4) require an adaption of the approach. The study focuses on two staple crops, sorghum and millet. It stresses producers' deci- sion making at the farming system and at the channel level. Standard operating procedures of major participants at the channel level are also analyzed. Problems of participants are diagnosed as information permits and recommendations for improvement are suggested. 2.5 Summary Agricultural marketing plays a dynamic role in the economic develop- nent process. But to study agricultural marketing in development, its subject matter and multidisciplinary nature must be recognized. In l9 particular, one must go beyond narrow approaches which yield unreal- istic recommendations to decision makers. The broad descriptive-diag- nosis-prescriptive approach provides a useful framework for such stud- ies. This study takes from that approach, but should only be considered as an entry point in this framework because resource limitations re- strict its scope. CHAPTER 3 MAJOR TRENDS IN FOOD GRAIN MARKETING IN UPPER VOLTA 3.1 Introduction The regional food marketing system is bounded and influenced by the national economic, political, and institutional environment. An under- standing of this environment is required in order to put into better perspective the understanding of the food grain marketing system in the Eastern 0RD. To do so, this chapter provides an overview of the evolu- tion of the grain marketing issues in Upper Volta over the last decade. The chapter discusses the general trends in economic growth and food availability and the public debate over the food grain marketing issues. 3.2 General Trends in Economic Growth and Food Availability Upper Volta is primarily an agricultural country. More than 80 per- cent of the estimated 6.5 million people are dependent on the agricul- tural sector. In recent years, the role of this sector in the national economy has been declining, but it still provides 39 percent of the Gross Domestic Product with crops contributing 24 percent and livestock 15 percent. The agricultural sector is dominated by two crops, sorghum and millet, which are cultivated in 90 percent of the arable land and accounts for more than 70 percent of the gross value of crop production (World Bank, 1982a). Agriculture provides over 90 percent of recorded foreign exchange earnings. Despite its modest contribution to GDP, 20 21 "livestock alone furnishes from one-third to over two-fifths of ex- ports" (World Bank, 1982b, p. 10). The major export crops are cotton, groundnuts, and sesame. Shea nuts, a gathered product, also play an important role in export. According to World Bank's estimates, the country has experienced a strong growth rate in 1977-79, 4.7 percent in real terms. However, per capita income of $180 is still very low. There is a high population density, especially with respect to the arable land, because of the small size of the country and the small proportion of its arable land (33 percent of the country). The low per capita income and the high population density contribute strongly to migratory flows that were in- stituted by the previous colonial power. The World Bank estimates that the longer-term prospects for economic growth are less encouraging than the 1977-79 performance. Another discouraging trend is that of the food situation. Accord- ing to USDA figures,1 total food production per capita has been lagging behind population growth since the 1969-71 period. Sorghum and millet production per capita picked up for three years after the 1968-73 drought, but since then has slipped behind population growth. (Because of the migration, the World Bank estimates that the growth rate of the total resident population hovers around 1.6 percent). Consequently, food imports, made mostly of cereals and flour, have increased dramati- cally over the same period. From the annual average value of 600 to 800 million FCFA before 1968, food imports have ballooned to an average value of 3 billion FCFA since 1973 (World Bank, 1982a). The fact that ISee USDA's “World Indices of Agricultural and Food Production," Statistical Bulletin Numbers 669, 689, and 697. 22 a main part of the food import is food aid is only testimony to the lack of resources of the country. Nevertheless, many studies point out that Upper Volta has the po- tential to reach food self-sufficiency during normal years (see CILSS, 1977; and World Bank, 1982a). While the Central part of the country is overpopulated with respect to agricultural resources and constitutes the main grain deficit area, the Southwest and, to some extent, the South- east are thought to have large food production potentials if provided with the right mix of development incentives. But the interregional productive differences and the overall variability of food supply in- duced by erratic rainfall put food grain distribution in the forefront of the government strategy of food self-sufficiency: Even if production is increased so as to cover everyone's needs, food still has to be made available at the right time and place to all consumers. The awareness of the importance of the food grain marketing by pub- lic officials coincides more or less with the declining food production per capita and the squeeze on total food availability in different parts of the country put on by the 1968-73 drought. Local urban pressure groups, food donors, and other international development agencies have jointly contributed to an attitude of more and more direct public inter- vention in the food grain marketing in a pattern similar to that of neighboring countries such as Senegal, Mali, and Niger. 3.3 Evolution of the Debate on Food Grain Marketing 3.3.1 The Private Trade For centuries, sorghum and millet have been traded throughout the Sahel, both on a local basis and in the long-distance trade (see 23 Chapter 2). In Upper Volta, however, the rise to preeminence of local merchants has a more recent history. Beginning in 1954, French colonial trade companies, which dominated the entire country since 1945, started to scale down their scope of operations and retreat into the two major cities of Ouagadougou and Bobo Dioulasso in order to reduce logistical costs. This strategy opened up new opportunities in the countryside, first for Lebanese and then for local traders. They were contracted by the French companies to bulk export crops (groundnuts, sesame, shea nuts) and even distribute locally a line of imported consumer goods. The capital accumulated through 1959 to 1966 enabled many merchants to ex- pand their scale of operations in grain marketing (0. Bollinger, 1974). Today, there is a powerful grain merchant union which the public authority has to reckon with. The network of private traders provides assembly, bulking, transportation, and storage of grain on a larger scale than public agencies. Unfortunately, it appears that the public authorities have often confused the mishaps of aggregate performance of the food grain marketing system with the belief that private traders can be replaced altogether. 3.3.2 Historic Perspective of Government Interventions in Food Marketing The first effective attempt by a Voltaic government agency to con- trol agricultural marketing can be traced back to 1964 with the estab- lishment of the Caisse (Caisse de Stabilisation des Prix des Produits Agricoles-—CSPP). The mandate of the Caisse was to: (l) stabilize the producer prices of groundnuts, sesame, cotton, shea nuts (harvested from semiwild trees), and paddy rice; (2) promote the export of these prod- ucts (except rice whose domestic production hardly covers local demand); 24 and (3) increase their productivity. The motivation provided for the establishment of the ngggg was that France, because of its entry into the EEC (European Economic Community), could no longer offer directly advantageous terms of trade for products exported by its former colon- ies. Another motivation was that too much competition, introduced by too many traders vying for the farmer's produce, was alledgedly re- sponsible for the producer price fluctuations. Public intervention was therefore deemed necessary. The Caisse, however, allowed the whole marketing organization (as- sembly, bulking, transportation, and export) to remain in the hands of licensed private and public institutions and concerned itself with export pricing and taxation: On the basis of a minimum producer price and a reasonable profit margin, a reference export price was set.within a range (fourchette) of a ceiling and a floor price based on past and V projected world prices. The price stabilization scheme called for the Cgigég to refund exporters the difference between the reference and the floor price if actual world price fell below the floor price, and the exporters to pay the Cgigge the difference between the world price and the ceiling price if the world price rose above the ceiling price. This pricing arrangement always worked in favor of the ng§§g_un- til 1974 when large funds had to be disbursed to compensate exporters who had paid high prices to producers, but stood to lose money be- cause of unfavorable world price movements. Since then, the ngggg has been positioned as the sole exporter of groundnuts, sesame, and cotton 25 seed.2 The licensed traders must now sell all their products to the Cgiggg (CILSS, 1977; CSPP Annual Reports). In 1966, the Voltaic government started the establishment of Re- gional Development Organizations (0RD for Organisme Régional de Dé- veloppement) in all 11 regional administrative units (Départements) of the country. The 0RD system is a departure from the former one-commod- ity agricultural development approach and signaled a more integrated rural approach at the regional level. The 0RD mandate was to coordinate all development efforts (and implement some) at the regional level through its network of multipurpose extension agents (agents de develgg; pement communeautaire). A particularly interesting aspect of the 0RD status was that each may engage in revenue-generating activities in order to self-finance its development activities. The above legal provision opened up the door to the various market- ing operations the 0RD engaged in. In July 1968, to help the ORDS a- chieve financial autonomy, the government granted them the monopoly rights3 to purchase export crops (groundnuts, sesame, and shea nuts) from producers. These monopoly rights lasted only two buying campaigns and were repealed in 1970 because of the 0RDs' poor performance. Very few ORDs (only Bobo, Banfora, and Bougouriba had sizable purchases) 2The cotton subsector is vertically integrated by CFDT (Com a nie Francaise our 1e Develo ement des Fibres Textiles, a private Frenc corporation). CFDT has now fgrmed a partnership with the Voltaic Government in SOFITEX (Societe Voltai ue des Fibres Textiles) in a pat- tern that isrepeated throughout all former French colonies in Africa. The company has the monopoly for cotton input distribution, cotton gin- ning, and lint export. Public institutions (ORDs), which help organize cotton assembly, are paid a commission. 3Traders, licensed (agréés) by the Caisse, were also granted the same rights. 26 actually participated, either because of lack of interest or lack of financial and managerial capabilities. There was also no help from either the Caisse nor the National Bank of Development (BND). Actually, the monopoly did not hold, for even nonlicensed private traders were al- lowed to collect the produce. Furthermore, the ORDs, their organiza- tional system overextended by the marketing function, were forced to contract with private traders to bulk or move out products. Yet, during the 1968-70 episode, the ORDs were introduced, if only marginally, to food grain marketing, for a number of farmers brought grain along with export crops to the 0RDs' buying points. So far, how- ever, in spite of public statements to the contrary, the government agricultural policy had been biased toward export crops. A testimony to the fact that extension and marketing efforts were better articulated for export crops than food crops is provided by the relatively well- structured operations of the Caisse and CFDT, and by the 1968-70 attempt to monopolize export crop marketing. The drought that plagued Sahelian states from 1968 to 1973 helped bring food crop orientation in govern- ment policy in the forefront. It is only in 1971 that the Voltaic government actually began ac- tive interventions in the grain marketing system by creating a grain marketing board, OFNACER, Office National des Céréales. The mandate given to OFNACER was to stabilize producer and consumer prices by grain purchases in surplus areas and sales in deficit areas. Grain was also to be purchased in bumper crop years to be sold in deficit years in order to smooth out interannual prices. To achieve these goals, OFNACER set up a target of 30,000 to 40,000 tons of grain to be purchased every year. In addition, OFNACER was also to build an emergency food grain 27 reserve. The establishment of OFNACER follows the pattern of food crop agencies in neighboring countries (in Mali, OPAM was created in 1965, in Senegal, ONCAD was created in 1965, and in Niger, OPVN was created in 1970).4 It follows also the need for a central agency that would handle the growing flow of food aid given to relieve drought stricken victims. In fact, in its first few years, OFNACER's actual role was largely to manage the storage and distribution of food aid. Abruptly, in 1974, the government granted the monopoly rights of grain assembly to the 0RDs and traders licensed by OFNACER which, in turn, was given the monopoly of grain distribution to consumers. The objectives of the reform was to: (1) reduce alleged excessive private speculation; (2) promote rural organization by encouraging farmers' par- ticipation in agricultural marketing and introducing better grain meas- urement devices; and (3) provide the 0RDs with a self-financing opera- tion. By virtually all accounts, the 1974-75 campaign was a disaster (see CCDR, 1975; CILSS, 1977, Vol. II; D. Wilcock, 1977, pp. 194-239; and OFNACER reports). First, the decision was taken without knowledge of food grain marketing conditions which are different from those of export crops. Second, the 0RDs, which already lacked the financial, managerial, and logistical means, did not have much lead time to set up their buying networks. Third, the Bank (BND) was also caught by surprise, and re- flecting the disastrous 1968-7O campaigns, did not rush to loan money to the 0RDs Or OFNACER. In fact, for a campaign that was to start in November and end in March, the 0RDs did not get any funds from BND until 4OPAM stands for Office des Products A ricoles du Mali; ONCAD, Of- fice National de Coo eration et d'Assistance our le Develo ement; Sid OPVN. Office des Profiucts Vivriers du Niger. 28 late January, and OFNACER did not receive enough funds to buy from the 0RDs. Fourth, the official prices (22 FCFA/kg. to the producer and 37 FCFA to the consumer) were unrealistic: Because of good harvests, the actual market consumer price was around 30 and the producer price below 22 (CILSS/Club du Sahel, 1977, pp. 22-23). These market conditions prompted OFNACER to buy from private traders and not from the OROs. Fifth, in the meantime, the 0RDs sold at 28 FCFA/kg. to international 5 organizations (OSRO and FAQ) which were attempting to build food re- serves from local sources. In addition, another public institution, the ’6 Sous-comite was distributing grain from food aid at 15 CFA/kg., thus, undercutting OFNACER's efforts. The confusion of the 1974-75 campaign was carried into the next year. Prior to the 1975-76 campaign, the 0RDs tried to unload their carryover stocks to OFNACER. But OFNACER could not buy because its own storage facilities were filled with unsold imported maize. Therefore, the 0RDs were forced to default on their loans extended by BND. Further- more, they had to discontinue their operations, which, coupled with the fact that they intervened late the previous year, was very instrumental in creating the mistrust that producers now show with respect to the 0RDs'marketing capabilities. In fact, it would appear that the only gainers of the 1975's grain monopoly experience was again private trad- ers who stood ready to buy when producers wanted to sell, and who were also contacted by the Ords and OFNACER to provide transport services in many instances. By 1976-77, virtually all 0RDs had ceased to intervene in the grain marketing. SOSRO is a UN/FAO drought relief operation. 6Sous-comité de Lutte Contre 1es Effects de la Secheresse (sub- conmittee ‘t'o comfif Elie effects of the drought). 29 The E-ORD, which had been just officially instituted in 1974, had also very poor marketing campaigns during this episode. It had attempt- ed to involve farmers,7 but its hierarchical and centralized marketing network, which required too many decisions to be made at the central headquarters, was too rigid to be effective. As a result, funds were not delivered on time to buying agents and the situation of the stock (held in local warehouses, in extension agents' and farmers' houses, or left in the open) was very confused. In 1974-75, the E-ORD never a- chieved more than 43 percent of its target in millet and sorghum pur- chases (E-ORD reports). Even worse, by "December 1975, the 0RD had not been paid for over 80 percent of the sorghum and millet it had deliver- ed...to national grain marketing (agencies)“ (Eicher, et al., 1976, p. 24). Consequently, it defaulted on loans extended in 1974 and did not receive funds in time in 1975 so that the 1975-76 campaign was temporar- ily cancelled in December 1975. In 1978, the official grain monopoly rights granted to the 0RDs and OFNACER were repealed.8 OFNACER administration, which up to now was under the Ministry of Commerce, was now brought under the Ministry of Rural Development which oversees also the ORDS. OFNACER was now to collect grain from its own agents, its licensed traders, and village groups (groupement villageois). But, in fact, OFNACER buys also from 7During the 1974-75 campaign, 76 village groups (groupements vil- la eois), 138 village committees (comite's villageois which are set up where there is no village groups),174C CTUbs (youth organizations),and 11 "isolated intermediaries" could act as buying agents for the 0RD in additgon to some of the 0RD own extension agents (Eicher, et al., 1976, p. 27 . 8Donor agencies, which were dissatisfied with the monopoly rights, played some role in this new change of policy by holding or delaying their contributions (0. Wilcock, 1977, p. 201). 3O nonlicensed traders. Thus, a more liberal market view has been adopted. But OFNACER still holds fast to targets of grain purchases of 30,000 tons in order to affect grain price movements and control price fluctua- tions. Furthermore, it has also targets for emergency food reserves of 30,000 tons, a three-month food grain supply reserve. Another development in food marketing has been the creation of Cereal Banks, whereby producer precooperatives engage in grain trans- actions in order to meet their members' food requirements during the hungry season. According to CILSS (1979), more than 30 such Cereal Banks were built between 1975 and 1979. A few more Cereal Banks were also to be built in the Eastern region. Today, OFNACER has clearly emerged as the main instrument of the government food marketing policy. However, OFNACER's performance is still not very satisfactory: (l) OFNACER's local grain purchases have always remained below its target of 30,000 metric tons. Even in the 1978-79 season, regarded by OFNACER as its best ever (as in 1981), it managed only half of its target. In fact, OFNACER's operations are still dominated by its food aid sales in urban centers (see OFNACER's pur- chases and sales in Tables 3.1 and 3.2). Because OFNACER purchases small quantities, its impact on the level of producer prices is rather nonexistant. It may only possibly affect price levels in areas of difficult market access and large producing areas. But OFNACER may have an effect on consumer prices in urban cit- ies because of its concentration on urban cities and the relatively large volume of food aid available for sale. The food reserve built up by OFNACER obviously cannot come from its local purchases. Instead, food aid has been sought to build up the 31 -.ovugoeeou an ozcvnapc> mumpuom any we .emucaomc .»_=o a~\_M\m enacted moan moPama .covuaavcum_u owe» meow m:_u:pu:_v .mEaLmosa mpo vcow Love: mums mucous_u .a—nau p~:_m_go soc» umm_>mm .Azouao>om. =o_aam.a xu.p.a_m:OQmme we» mam: «up; we mmpum ago momagueza .m~-¢~m_ .wucau .ommm_ .xcum u_coz "mugaom .. _~.m— .. mm.~¢o.— .. ~m.mm .. oa.m¢~.m .. mm.—m .. mm.nmm.~ .. ¢~.om 11 mm.mmo oo.op m~.m— mo.¢m om. om.p ep.mp —~.o- e~.m ~¢.mo—.— mm.mom.— -1 mm.mmm .. om.mm¢ oo.e~ mm.m~ mo.—~ po.m~ No.mm mm.e~ ~—.~v—.m mm.cmo.p 11 m~.wo~.— .. ~o.om¢ m~.om po.m~ mo.- m—.oe —m.~m ~—.mN ~o.oom.~ mm.om¢.p u- mo.mmm .. mm.omv sv.om mo.mp mm.m ms.¢~ em.m~ p¢.- o_.pm~ ¢m.mo~ u- op.mom .. Np.op m~.m mm.m pm.op o¢.mp n~.~ on.o mo.o mo_u_uca:c A+ mmmomcucmv mxooum :_ mmmcmgu ——.omm mos—u> mm.om mo_upu:u=c nausea: _auo4 co mm—om pooewu em.~wm mm=_m> no.8m »u_u=a=o..pauop om.¢mm nonpo> m~.mm ma_u.u=a=o umueoge_ _aoeou m~.~m mos—m> cm.~ mo—upucoaa muoxeoz,hmooa co mamasueam pooeou com-mka_ oa~-m~op cunnsm— unumnop nmsimnmp umnncsmp «mumsop munwmm— msa—sm— mm\mmop..muxpmm— .mcowumgmqo mpmwgmu mmu Loan: mo :o_mmm :Lmummm ~.e aezmwu .Awmapv someeaz..< ”mugsom Sam. 8.. 8 ska. ..23: .22. 2.0. .33 cm £62.80 23> 832.. a. 8.3 .823» 8E :28 m A ...E a :35! an 8:8 3.5.8 8.338: .. 2.8.3 5:069:23; is: 11.1 ....... .l 5 E§~ 1; Seam“. \ . 13.3000 .nfl. WIIIII\ m a... ton. . no, 11.5—.02. 1A.; . :smmwxnax .1 £3.85 1 .\. m0_.rw_mw._.0 ban: 8 550m E2mom .38; I I I 12%.- I 2.... 8a A a A a 04 38 Table 4.1 Agroclimatic Characteristics of Surveyed Zones lpprossnate Catt-ated Population Doltnant Lonetero ngtgd by Order 2! lggggtggce; Danetty‘ Ethnlc Major Sou Average lone (persons/kl?) Group Typesb laIntall° NaJor Crops Livestock Types 1. Bogande 27.1d Cour-a leached sandy soIl 690 eIIlet o sorgnue o ease-e. eroundnuts goats. sheep, cattle 2. Rent 17... Cour-a leached sandy soil 610 sorgnu- o seaele. groundnuts sheep. goats J. Ptela 32.1 Cour-a leached sandy sell 750 esllet o sesale. |roundnutl, rice goats. sheep, cattle I. place 3|.) Roast poor laterttlc soIl 680 etllet o coupea. groundnuts, race goats. sheep, cattle 5. Logobou 00.0r Cour-a silty to sandy clay 960 sorehua o coupes. sorghul o ltadl‘, sheep. goats. groundnuts, rice, tubers cattle (taurin) 6. Partlaea 24.2 Gourla hydro-orphic black clay 900 sorenul o coupes. ealxe. rice, sheep. goats, cattle or sandy clay tubers. cotton 7. Yonde 13.0 Nose; black clay overlaying 900 eIllet . sorgnul o coupes, groundnuts goats. cattle. sheep hydro-orpnlc vertlsols 8. Dtapangou 19.? Cour-a clay and sandy clay 910 sorgnul o eIlIet . coupes cattle. sheep, goats 9. Botou 19.9" Source sandy clay and black :Iay 856 sorghul or etllet v coupea o sesele. cattle. goats. snoop (reundnuts, cotton IO. KantcnarI 9.3 Gourea sandy to sandy clay 870 sorghu- . eIllet o coupes. eaiae. cattle, sheep. goats aanIoc. cotton II. Ju‘JrOu 3.6t Gouroa clay 880 sorgnu- . coupes, eaIae cattle. sheep. goats 1:. Pena 2.7 Cour-a hydro-orpnsc black cIAy 1060 sorghu- . coupea.l sorghu- . IIadIl. goats, cattle (taurInI or sandy clay tuners, cotton. rice Source: G. Lassiter, 1982. a1979 estimates taken from Mehretu and Wilcock (1979, Table 3, p. 20). bFrom Bureau de Production Agricole, "Determination des Zones Homogenes en Vue de l'Installation d'Un Reseau d'Essais Multilocaux," 0RD de l'Est, Fada N'Gourma, Upper Volta, August 1977. cFrom J. Weldring, "Synthese sur 1es Amenagements Hydro-Agricoles dans 1'ORD de l'Est: Fada N'Gourma,“ Direction du Fonds Developpement Rural, Ouagadougou, May 1979, pp. 5-6. Weldring took his figures from an un- cited 1974 S.A.E.D. report and thus they probably represent 20-year rainfall estimates extrapolated from a few national rainfall stations from similar latitudes. In cases where Weldring did not present an estimate for a survey village, regional averages were used: Bogande (Bogande + Thion), Mani (Coala), Botou (Bilanga + Yamba), and Diapangou (Fada N'Gourma). dDensity for Thion canton used. eDensity for Coala and Bogande cantons used. fThis is a rough estimate of the effective population density in the survey area. The density of the Gobnangou canton is only 9.8, but the majority of the canton area is nonarable rock ridge or wildlife reserve. gNiadi is a short season, 60-day millet grown only in the wetter regions of the EORD. hDensity for Bilanga canton used. lDensity for Matiacoali canton used. 39 and Southeast (Pama, Arly) is covered by forest formations which host one of the richest wildlife habitats in West Africa. Because this por- tion may also have the highest agricultural potential, there is a poten- tial conflict between the development of wildlife and agricultural re- sources. The Eastern region, while being the largest in size, is also one of the least populated area of the country: it hosts only 7 percent of the nation's population, and its density is eight people per square kilo- meter, whereas the average of the country is close to 25 people per square kilometer. More importantly, however, it has a very uneven popu- lation distribution. Although the Western stretch, from Coalla in the North to the Togolese border, covers only some 25 percent of the area, it contains more than 50 percent of the population (A. Mehretu, 1982). Elsewhere, the population is mostly clustered in centers and what remains has very sparse population settlements (see Figure 4.3). It appears as if the. population settlements have been oblivious to the climatic and soil con- ditions, which results in a very unequal pressure on the land endowed with different potentials. For example, the driest part, in the North, is more populated than the wetter and richer part in the South. This uneven population distribution can be related to the ethnic make-up: Mossi, the dominant ethnic group in Upper Volta, and affili- ated groups are mostly found in the West which borders the predominant- ly Mossi areas of Koupela and Kaya OROs. Fulani are found in the North, and Gourmantche--the dominant ethnic group in the Eastern region--are found elsewhere. Thus, this population distribution may be explained by historical reasons. But another part of this uneven distribution can 4o 6- 1'- i- Eosiem Region of Upper Volta POPULATION DENSI I Y BY CANTON . ' I’d .fi' 1\‘\. N e O O :1 2"‘1’. .60!" /' y . a .K - "or ............ _. 5 I . » . g ,, ' ’5’ /' W ’ .—v' . 9 O I ‘ , “ .l . ) a u .> 2? ./ . . '\ .l _/ WI . I ,/ ( £ ' . 1 fl (. - ”“ —-~._ I ‘\/ ...... . .\." '''' I. ”M \ K m. 9 z 7' .- \ I 1.! . /‘ ' Ira \ . . D . \.\. / . {.1 u \ l, I o ‘, "_ (s.._ __ / .g . . /' “ .>'- ... .\ I w ."‘e .‘. I“... ........ s \ \ eM II. . :\. g L, 1.! ,....,‘ \ ‘fl ~r-val '”“ \ me o ,- f ' 0 HP ° It" mac-mum (oceans/sane) . heritage D o-‘., flLMUmm E] s—s.s £2! rake-13m [3 Io-Iu D Is-Is.e E3”*’ 9.9.1L.9-9.@m m asun- l- 1 Source: A. Mehretu (1982). Figure 4.3 Eastern Region of Upper Volta: Population Density by Canton 41 also be explained by the fact that the South is infested by the simuli fly which causes "river blindness" (onchocerciasis), a deterrent to human settlements. 4.2.2 Relevant Institutional Characteristics The Eastern region is one of the 11 Départments in Upper Volta, and also one of its 11 0RDs (Regional Development Organizations). While the Eastern department is administered by the highest ranking official in the region, the Eggfgt, who is accountable to the Ministry of Interior, the Eastern 0RD (E-ORD) is administered by a director who is accountable to the Ministry of Rural Development. The headquarters of both the Department (the Prefecture) and the 0RD are in Fada, the regional center. The E-ORD, in theory, is the coordinator of, and the main agencw're- sponsible for, all economic development efforts in the region. In par- ticular, it is charged with developing agricultural production, live- stock, agricultural water resources, and forestry and tourism.1 But, in fact, it concentrates almost exclusively on agricultural extension. For that purpose, the region has been divided into sectors (eight), the sec- tor into subsectors (23), which coincide approximately with the cantons_ (traditional administrative unit), and the subsector into extension zones. The zone, in turn, is made up of a group of villages. A certain number of villages are officially structured into village-groups (Groupe ment Villageois), a sort of precooperatives which are being promoted as 1The 0RDs were given mandate to coordinate development efforts in forestry, wildlife, and tourism in 1974 when the Ministry of Rural Development oversaw also Forestry and Tourism. In subsequent minister- ial changes, the ministries of Rural Development and the Ministry of Forestry, Tourism, and Environment were separated and the role of the E-ORD in forestry and.tourism is not practically inexistent. 42 the cornerstone of the government's rural development policy. The over- all extension system operates in a top-down approach. Along with the Prefecture (Eastern department) and the E-ORD, OFNACER (the grain marketing board) may be considered as the third im- 2 operating in the Eastern region. portant development institution OFNACER has its headquarters in the capital city, but in most Depart- ments, it operates a subunit, the CRG (Centre Regional de Gestion). The CRG manages a permanent stock warehouse in Fada, has important out- posts in Diapaga and Namouhou, and sets up every year periodic buying and selling points according to production conditions. Overall, the socioeconomic infrastructure of the E-ORD is very poor, even according to the standards of Upper Volta. For example, the E-ORD has the second lowest rates of school enrollment in the country, 7.8 percent as of 1979. Based on an effective service radius of about- five kilometers from the health care facility, the current health deliv- ry system covers only 4 percent of the area. Domestic water is supplied by individual wells that may dry up part of the year. Finally, assuming an effective use of all roads and motorable trails to be limited within a 10 km range along the roads, the road accessibility index reaches only 37 percent. In fact, only 13 percent of the total road system (2,823 kilometers)3 is made up of all-weather roads (see Figure 4.4). 2Other major institutions are two bank branches of the National Bank of Development and the International Bank of the Voltas (both With government majority participation) a post office, and various religious institutions which provide development aid. 3Figures taken from Mehretu, 1982. 43 L 9i mm Wheel “Tulle Whom Tannerifillepe al.000erlore Eastern Region of Upper Volta SURFACE TRANSPORTATION 13'4 ”"4 Source: A. Eastern Mehretu (1982). Figure 4.4 Region of Upper Volta: Surface Transportation 44 4.3 Research Design and Instruments This study relies on numerous sources of primary data, the most im- portant of which is the 1978-79 MSU/E-ORD farm survey. The other sources used as a complement to the marketing information of the 1978-79 farm survey are the 1979-1980 marketing investigation and the 1980 E-ORD vil- lage census. 4.3.1 The 1978-79 Farm Survey The 1978-79 MSU/E-ORD farm survey4 collected information on the economic activities of 480 households surveyed in 27 villages from 12 major ecological zones from May 1, 1978 to April 30, 1979. The sample was made up of 355 "hoe" cultivating farmers (348 of whom were randomly selected and seven were village chiefs purposely included to assure political support) and 125 animal traction (ANTRAC) cultivating farmers purposely selected (see Figure 4.5 and Table 4.2). The sample was stratified at three levels. First, 12 zones were selected to be representative of ecological conditions, population dis- tribution, animal traction (ANTRAC) uses, and the eight sectors of the 0R0. Second, from a frame list of villages of the zones identified, two _vi11ages were randomly selected in seven zones with predominant hoe farming (traditional) techniques. In the other five zones with ANTRAC use, traditional villages were randomly selected as a control group to match nonrandomly selected ANTRAC villages. Only in two of the five zones was the pairing (one traditional against one ANTRAC village) not _ 4The sampling and data collection procedures, and sample character- istics are very well documented in Lassiter (1982, pp. 31-42), Tapsoba (1981, pp. 30-47), and various MSU team six-month reports. 45 k k - —e¢ Lab eucaoaac 3.22.5.5 cote». cocoa.- 1cac< 393 81.5 ole-In 0 one!) iii . .a: 353 "'I 90985.! “N 30.0501! Emir » p o 8.3....” . . IN. 22.32 32.3 as: 1...... . . . an ago 2355 . A. «main; cud-2.9. do .2: 3...... :1 Eastern ORD--Farm Survey, 1978-79 Figure 4.5 0RD de L' Est-Fada. Map of Sampled Villages: Source: 46 Table 4.2 Distribution of the 480 Samples Agricultural Households by Agroclimatic Zone, Village, and Technology Number of Households . . Population in Each Village or Area Estimates Animal (1975 Zone Village Traditional Traction Census) Gbanlamba 18 -- 1,160 Bogande Komboassi* 18 -- 1,119 Mani Lanyabidi 18a -- 161 Bonbonyenga* 18 -- 604 . Dabesma* 18 -- 234 PM“ Piela -- 13 --C Monkontore* 18 -- 377 . Lantaogo -- 18 --C ”'3” Diabo 1d -- 17 --c Diabo 11d -- 13 «c Namponkore* 18a -- 2,138 Logobou Kindikombou* 18a -- 2,032 Logobou -- 18 __c . Bomondi* 18 -— 1,063 Pam 395' Dubcaal 1 18 -- NA Ouobgo* l7 -- 627 Yonde Kondogo 18a -- 302 . Tilonti* 18 -- 402 Diapangou Diapangou -- 18 --° b Botoub -183 -- 600 BOtOU Ougarou 19a -- 547 . Mantchangou* l7 -- 525 Kantchar1 Moadagou 18 . -- 285 Poniokon1i* 18 -- 315 Ougarou Ougarou -- 18 --C Tindangou* 16 -- 462 Pama Kpajali 16 -- NA Total: 355 125 -- Source: 1978-79 Farm Survey and 1975 Population Census. *Residences of enumerators. aVillage chief included as a nonrandomly selected household head. b North of Fada. c1975 data missing because of confusion over names of villages. d zone, the 13th. In some computerized data files, Diabo I and Diabo II make up another 47 respected: in Diabo, three ANTRAC villages were selected against one traditional village, and in Logobou, two traditional villages were se- lected against one ANTRAC village. Third, in the 20 randomly selected traditional villages, 18 households per village were selected (village chiefs were purposely selected in seven villages), but in all seven non- randomly selected ANTRAC villages, 18 households were purposely selected by extension agents for their successful use of ANTRAC in the village (see Table 4.2).5 The 1977-78 farm survey used a "cost-route" approach of flow data collection. That is, multiple interviews are regularly spaced through- out the year in order to reduce recall errors of flow data. The ques- tionnarie, patterned to Matlon's data collection method in Nigeria (1977), was disaggregated by activities: crap harvesting, sales of farm crops, sales of grain bought for trade or processing, etc. Furthermore, heads of households were asked to recall information on a weekly basis, for it is thought that the weekly recall period may help farmers better focus on past activities. In a few other cases, farmers were interviewed week- 1y to collect labor data, or three to four times during the year, or only once to collect such information as consumption, grain stock, house- hold personal characteristics, etc. This detailed "cost-route" approach and questionnaire design cre- ates a very large quantity of data and requires close supervision. To meet this requirement, one enumerator speaking the local language was assigned to two villages, and supervisors visited the enumerators 5The source of the village and household lists is the 1975 popula- tion census. Because of the low population density of the region, it is assumed that village size is independent of the farming system. 48 regularly to check on the accuracy of the data. Further checking was made at the E-ORD headquarters and at CENATRIN (computer center in Ouagadougou) where the coded data were punched, edited, and recorded on tape. This study uses the MSU-computerized data files which relates to grain marketing (see next section, and Chapter 6 in particular). 4.3.2 The 1979-1980 Food Grain Marketing Investigation The 1979-80 food grain marketing investigation was carried out by the author from July 1979 to July 19806 as a complement to the 1978-79 farm survey. Because there was no basic food grain marketing informa- tion available for the Eastern region as of July 1979, preliminary in- formation on the physical market infrastructure, the identification of the main market participants, and the nature of market processes had to be gathered before any in-depth analysis could be conducted at the vil- lage level. Unfortunately, the size of the Eastern region, the very poor road conditions, and the curtailment of resources did not allow this preliminary investigation to be completed in the short time (a few weeks) it should call for. As a result, the preliminary investigation overlapped with part of the in-depth analysis which took place later during the period. The reconnaissance of the food grain marketing system was conducted through unstructured interviews, all conducted by the author either directly in French or More, or through a translator when the interviewee spoke only the other local languages. 0n the public marketing side, all 6Other lines of duty of the author included the supervision of the 1979-80 farm survey follow-up, the clean up of the 1978-79 farm survey and 1979-80 farm survey follow-up, and the preliminary analysis of the 1978-79 farm survey data. In addition, he helped design and supervise part of the 1980 Village Inventory. 49 23 subsector chiefs and eight out of nine OFNACER buying agents were in- terviewed. 0n the private traders side, merchants and their agents were interviewed at marketplaces and at home, when available. In addition, marketplaces were visited to gain a visual feel of their importance. Finally, all village groups who participated in 1979 and 1980 in the village cereal bank project were visited.7 Food marketing information was also collected at the producer level through structured interviews by enumerators from February to June 1980, once the enumerators had been freed up from the 1978-79 farm survey fol- low-up. Two groups of farmers were interviewed; one group from a sub- sample of the 1978-79 farm survey (from now on referred to as the 1980 farm survey, see Table 4.3) and the other at selected marketplaces (from now on referred to as the 1980 marketing survey). The sample of the 1980 marketing survey at the farm level is dif- ferent from that of the 1978-79 farm survey. The 1978-79 sample was re- duced from 18 to six households per village in order to save resources and a few villages were dropped or added. In addition, given the time constraint and the fact that many farmers had been complaining of re- peated interviewing without tangible improvement of their lot, the "cost- route" approach was abandoned in place of one-shot interview. Data was sought on producer's marketing behavior, in particular: - perception of marketing problems; - measurement of products: and - marketing credit provided by merchants. 7A few other village groups financing their own cereal banks were visited as well. 50 Table 4.3 Location of Sampled Villages of Surveyed Farmers, 1979-80 Farm Survey Number of Animal Households Zone Traditional Traction Selected Bogande . Lanyabidi 6 Mani Bonbonyenga* 5 Piela Dabesma* Piela Area 3 6 . . Lantaogo Area D1ab° Tibga Diabo I Area* 2 Namponkore* 6 Logobou Kindikombou Logobou Area 6 6 Partiaga Foanboanli 6 Ouobgo* ‘ 6 Yonde Kondogo 6 Diapangou Balga Diapangou Area .13 Botou*a 6 BOtOU Ougarou 6 Mantchangou* 6 Kantchari Moadagou 6 Sambalgou 6 Ougarou Egygokonli Ougarou Area 2 6 Kpamkpaga 6 Kpoali 6 Pa“ Soudigui 5 Tindangou* 6 Total: 27 168 Source: 1979-80 Farm Survey. *Denotes residences of enumerators. aNorth of Fada. 51 The 1980 marketing survey was also conducted at marketplaces pur- posely selected on the basis of their importance in the zone or in the entire region (see Table 4.4). Enumerators attempted to interview pro- ducers as they finished their transactions with merchants. A producer who refused to be interviewed was replaced by the next available until 10 were interviewed if market attendance permitted. The purpose of this part of the 1980 marketing survey was to identify eventual differences of marketing behavior patterns between the heads of households of the 1979-80 farm survey and a more heterogenous group of producers (heads of households as well as other members of households) visiting markets at the same periods. To achieve that purpose the same basic questions (with slight modification to account for the fact that enumerators were strangers to market visitors) were asked to both groups. 4.3.3 The 1980 E-ORD Village Inventory Survey This survey, conducted in April-May 1980, was designed to present simple socioeconomic information which could help describe the basic level of economic development of all 644 E-ORD villages listed in the 1975 population census. The questionnaire was structured into five main categories: population and socio and economic infrastructure, water sup- plies education and health, grain-marketing, processing industries, and other industries (see 0. Wilcock, 1982). The type of information was either a physical count of some characteristics (population, trucks, etc.) or the existence/nonexistence of some trait. Questions were asked to a group of knowledgeable villagers (usually the village chief and other leaders) fbrewarned of the arrival of the interviewers (the local extension agents usually accompanied by one or several members of the MSU-team). 52 Table 4.4 Sample of Markets and Producers Surveyed in the 1980 Marketing Survey Number of Number of Sectors of Farmers Villages Farmers of E-ORD Marketplaces Interviewed Came From Diaka 32 7 Manni 32 9 3°9ande Piela 33 10 Diapaga Logoubou 40 5 . Nadiabonli 16 8 Diapaga Namounou 42 15 Naponsiga Comin-Yanga Bousgou 2 2 Fada 2 2 Bilan-Yanga 33 15 Fada Diapangou 36 _ ll- Ougaroua 34 17 Tentiaka 26 7 Kantchari Kantchari 10 7 Matiacoali 27 6 Matiacoali Nassougou 9 2 Boulgou l7 3 Pognoa 5 3 Pama Tindangou 8 4 Total: 463 144b Source: 1980 Marketing Survey. aNorth of Fada. bCombined figures for all markets. 53 The relevant marketing information of this 1980 E-ORD village in- ventory survey concerns: whether the village has a marketplace; whether merchants visited the village to buy millet-sorghum; whether OFNACER had bought grain; how much sorghum and millet (in 100 kg bags) left the village; the price of one 100 kg bag at harvest: 8 whether there existed a colonial silo in the village and if it existed, whether it was used. 4.3.4 Data Limitations The many sources of information referred to above provide a large data base, but in many respects they also present some limitations that will be addressed in detail in the next chapters. The following com- ments, therefore, serve only as an overview. The data in the 1978-79 farm survey is a single-year cross-section set and, thus, presents limitations as to the representativeness of the producer's behavior that can be derived. (The 1980 farm and marketing surveys have been gathered in an even more reduced time period, since it was conducted from February to June.) In addition, the fact that the unit of analysis is the household or farm-family unit introduces some bias. This is because the head of the household, obviously, may have a good recall of what he did, but his recollection of other members' 8Concrete silos of various sizes were built by the French colonial power around WW II to hold the supplies of the Sociétés de Prévogance which purpose was to purchase, store, and sell cereals to rural and urban populations. This local level grain stabilization practiced through forced cereals collection had since been ended and most silos had remained idle. (Personal communication--see also E. Tapsoba, 1981, p. 49; and D. Hilcock, 1977, p. 200). A. 54 activities could be less sharp. (This fact was not an oversight of the survey design. Rather, it was imposed upon the design by the fact that other members of the household, especially women, were reluctant to answer questions without the approval of the head of household.) Another limitation of the data is that the 1978-79 farm survey pro— vides an imperfect coverage of both the marketing seasons of 1977-78 and that of 1978-79, although it covers a full lZ-month production season. Such partial coverage of the marketing season prevents one from follow- ing the disposals of farm crops from one harvest to the next and com- plicates further the analysis. (A detailed discussion of the 1978-79 farm survey coverage and its implications are provided in Chapter 6.) A further shortcoming of the data used in this study is that house- holds may have a better recall of the value of transactions compared to the recall of the volume of transactions. The fact that containers, in. which volume of transactions were reported, are not standardized intro- duces an added difficulty. In the absence of actual measurement of vol- ume, the kg conversion used serves only as an approximation. As for private traders in the E-ORD, most are illiterate and none keep regular records of transactions. Besides, their mistrust of pub- lic officials, justified by repeated government attempts to reduce their activities, is not conducive to the sharing of precise information about their transactions. But the fact that the marketing investigation at- tempted to cover the entire E-ORD, instead of zeroing in on a few_mar- ketplaces and merchants, was also responsible for the lack of more de- tailed information on traders. In fairness to the design, however, the .fOcus of the marketing survey was still put upon farmers, as was the 1978-79 farm survey. 55 The above limitations are not atypical of many other studies con- ducted in developing countries. E. Berg (1977, p. 2) reminds us that wherever possible, numbers generated under these conditions "should not be used as point estimates--simple numbers--but rather as range esti- mates, however inconvenient that is.... And conclusions about reality which rest on these data should be expressed in the most tentative of terms." This study pays heed to Berg's warning. 4.4 Summary The E-ORD is a large geographic area with contrasted climatic and population distributions. The 0RD is given an extensive economic de- velopment mandate, but concentrates mainly on agricultural extension. OFNACER regional office, the CR6, is the other main relevant institu- tion. The Eastern region has been neglected for years and its poor socioeconomic infrastructure (e.g., roads) is reflective of this fact. The information on which this study is based has several sources: (1) the 1978-79 farm survey; (2) the 1980 marketing information; and (3) the 1980 village inventory survey. The data set is rich because of its large base and detail, especially in the 1978-79 farm survey. It still suffers, however, many limitations, for it is only a single-pro- duction year cross-section data set, the unit of analysis is different from the unit of inquiry, values and volumes of transactions might have recorded with different levels of accuracy, and in many aspects, the information on grain merchants is not as detailed as that on farmers. CHAPTER 5 SORGHUM‘MILLET MARKETING CHANNELS AND PROCESSES 5.1 Introduction The focus of this chapter is on the identification and character- ization of grain marketing intermediaries, the description of the market- ing channels and standard operating procedures, and the diagnosis of some marketing problems at the overall system level. Whereas economists' reference to the market is with respect primar- ily to a pattern of exchange and then the physical location where this exchange takes place, geographers and regional scientists have long sought to gain understanding of the development process by emphasizing a central place theory approach which focuses attention on the spatial characteristics of marketplaces. With regard to developing countries, many regional scientists have hypothesized that markets and their even- tual periodicities are set up in order to minimize travel costs of itin- erant traders (see R.H.T. Smith's review, 1978, pp. 11-25). In turn, neoclassical structure-conduct-performance studies measure correlation coefficients between market prices to make inferences about market in- tegration. The underlying assumption of both regional economists and structuralist S-C-P researchers is that all transactions take place in the physical marketplace, or that prices reported in these mar- ketplaces are representative of transactions taking place elsewhere. 56 57 But all transactions are not conducted in marketplaces. In house trading (P. Hill, 1969), food grain is purchased or bartered directly between village consumers and village producers. In another arrangement, sometimes called "periodic trade without periodic marketplaces" (R.H.T. Smith, 1978, p. 18), food grain is sold by producers to middlemen at the farm gate, and exported from there to other regions. What happens then is that a very small volume may actually reach the marketplaces of the exporting region, which, coupled with the imperfect knowledge of participants, may lead to "thin" market conditions where prices are more volatile than otherwise because they are no longer representative of the overall supply and demand conditions. Another consequence of the smaller volume passing through marketplaces is that checkpoints established at markets to measure region-wide grain flows will yield underestimated re- sults. These alternative market channels may present other special char- acteristics and problems that should be investigated along with those of the more traditional marketplace-oriented trade. Thus, even though this study focuses major attention on grain pro- ducers in the E-ORD, a treatment of the grain marketing channels and processes is an essential backdrop to the understanding of the marketing behavior of grain producers in the E-ORD. Furthermore, recommendations to correct malfunctions of the marketing system should be based on all relevant information about the present system. These recommendations are essentially geared at shaping the behavior of the market participants in ways consistent with overall system objectives. Thus, it is important to try to uncover the major participants' standard operating procedures (SOP) and assess how these SOPs might be related to some of the market- ing problems. 58 The sources of information used are the 1980 marketing investiga- tion (see Chapter 4). These sources are varied, but an important part is made up of the open-ended interviews with knowledgeable merchants and agents at Namounou, Piela, Fada, Bassieri, Pama, and Bilan-Yanga. The major limitations of the information gathered are the lack of market price series and volumes of transactions. These limitations, which are explained by the difficulties of data gathering as well as by the market- ing characteristics of the region, prevent us from testing hypotheses related to market price volatility and addressing a broad set of system- wide problems. Instead, the balance of the chapter addresses the following, re- stricted, but still important points: (1) characteristics of marketplaces in E-DRD; (2) typology of major grain marketing participants; (3) major private and public marketing channels; (4) standard operating procedures of participants; (5) illustration of private traders' costs. 5.2 Marketplaces in Eastern Upper Volta1 5.2.1 Number and Periodicity The 1980 village inventory survey identified 1782 markets among 635 3 villages; that is, on the whole, two markets for seven villages. It 1Section 5.2 relies heavily on A. Mehretu's report based on the 1980 village inventory survey (1982, pp. 73-83). The author helped designime marketing component of this 1980 village inventory survey. 2A previous count by the Ministry of the Interior (December 1976) found only 103 markets. _ 3The 1975 census identified 644 villages, a few of which were tem- porary rainy season settlements (cam ements de culture). A few other Villages were overlooked by the pfififigatTfifi'EEHSUETT‘T' 59 would appear that the existence of a market is strongly related to the population and size of the village and that rural market distribution follows the scattered pattern of the population distribution. Pama and Bogande, however, present two exceptional cases. In these administra- tive centers, there is no agricultural produce market even though a few hawkers sell a small line of consumer goods (soaps, imported or locally manufactured packaged food products, etc.). Markets in Upper Volta follow four types of periodicity (see the Ministry of Interior, 1976, Report, "Les Marches de Haute-Volta): (1) daily; (2) once every three days: (3) once every five days; and (4) weekly. In Eastern Upper Volta, markets are of types 1, 2, or 4, but daily markets are not very common (see Table 5.1). These daily mar- kets occur only in important centers and usually still have a longer periodicity (three days or weekly) on which days market activities are more important. The empirical data presented in Table 5.1 and the population density in Map 4.3 seem to support the hypothesis that the denser the population and shorter the distance between centers, the shorter the periodicity of markets. For example, the periodicity of markets is of the three-day type in the densely populated areas of the West (Comin-Yanga, Diabo, and Fada), whereas, the periodicity is of the once a week type in the areas of low population and village density of the East (Matiacoali, Diapaga). An alternative explanation of this pattern (or at least part of it) is that market periodicity is related to the social traditions rather than the population density in Eastern Upper Volta. Market services are mostly performed by nonGourtmantches because, traditionally, the 60 .mgocgm mgwncaoc we mmsmumn cop on a: can we: cc m—muou meow "muoz .ua_cwmma.u ooze .zmuxcm>m mpauppo>m mums ppwum mew mmcmz new .uoow ummmxumn .muoom Lusam -cou saw a mumxcms uwco_cma umgmp =_ .x~m>wumcgmup< .zuwu_uowcma cmmcoF a ammx Ppwum mamxcme appmu “mozm .Nmmp .ypmgzmz .< .»m>gam agoucm>cm mmmppw> ommF "mueaom xwmz o.m~ o.mm N.mm ~._m o.m o.o m.em nm.¢ _.m a maze mama 9.2 ii m.mm Wm mdm 0.2: o.o m.om 98 3...: .055 w.~ o.o o.o o.o m.m o.o m.m w.e e.m mzmuzgw>m pouch can; Ppmoomvumz penguucmx ascaow.z-momm mmcm>1:PEoo mmoqm_a onm_o mucumom xucmzcmcm omo mgouomm omo Haggai. Logaam comm cw apwupuowtaa metata> ;a_3 mumxgaz mmmF__> mo ucmugma "hawu_nopgmg pmxgmz p.m mpnmp 61 Table 5.2 Rural Markets in the Eastern Region That Serve More Than Five Villages Mean Market Mean Market Location . Number‘ Radius Radius b a of Village of User (Crude) (Weighted) c Rank Market Population Villages km km Rank 1 Fada 13,067 190 41.5 43.9 1 2 Kohogo 3,041 10 38.9 21.7 3 Diapaga 5,617 23 35.2 30.1 4 Namounou 5,048 35 31.3 17.1 10 5 Botou 1,839 16 27.3 21.7 5 6 Bilanga Yanga 1,573 32 23.3 23.5 4 7 Bogande 4,351 47 23.0 18.2 8 Namoungou 837 10 22.3 15.8 13 9 Kantchari 2,883 22 21.7 17.1 10 10 Pama 2,265 4 19.7 9.9 18 ll Dzembende 1,472 23 18.7 15.7 14 12 Boulgou 1,116 8 16.1 26.2 3 l3 Matiacoali 2,683 12 15.6 16.1 11 14 . Comin Yanga 3,603 9 15.0 7.7 20 15 Piela 3,974 53 14.6 17.6 9 16 Manni 2,212 43 12.9 12.0 15 17 Tibga 3,004 21 12.9 11.3 16 18 Bassieri 1,324 7 _ 12.8 19.0 7 19 Ougarou 547 6 11.3 8.6 19 20 Tiantiaka 940 13 11.0 6.8 23 21 Yamba 1,399 12 10.9 15.9 12 22 Boussirabougou 695 17 10.1 7.2 21 23 Diabo 1,277 54 14.9 20.1 6 24 Karkouri 105 7 8.0 6.9 22 25 Diapangou 1,249 18 7.6 10.6 17 26 Diaka 1,488 17 6.2 4.6 25 27 Pori Gourma 1,201 8 6.1 6.3 24 Source: 1980 Village Inventory Survey, A. Mehretu (1982). aBased on crude mean of market radius in km (Eucledian distance in km from market to user villages). bComputed as follows: - £[d. - P.) d. 1.1 J 1Tb;— where, a . weighted mean market radius for market i: d - Eucledian distance from market 1 to user village j: ij - population of user village j. p 5 “The weighted market radius (distance to user village) is considered a better and more reliable estimate of market shed because it takes into consideration the sizes of the user villages." (A. Mehretu, 1982, p. 78). cBased on (b) above. 62 Table 5.3 Distribution of Village Markets in the Eastern Region by Mean Radius of Market Area Number of Number of Mean Markets Markets Total Market Based on Based on Number a Radiusc Crude d Weighted of User Type of Market km Distance Distanced Villages 1 . Regional Center 40+ 1 l 190 2 . Regional Markets 30-39 3 l 68 3 . Subregional Markets 20-29 6 5 131 4. Area Markets 10-19 17 13 278 5 . Local Markets <10 35 42 121 6 . Village Marketsb -- 116 115 116 Total: -- 178 178 -- Source: 1980 Village Inventory Survey, A. Mehretu (1982). a . . . . . The first four categories of market serv1ce five or more user Villages. Local markets service between two and four villages. t)These are village markets which do not serve other villages besides their own. SEucledian distance in km from market to user village. dSee computation in Table 5.2. 63 LOCATION AND SIZE OF MARKETS IN THE EASTERN REGION (mean market radius. weighted distance) UPPER VOLTA 9 Regional Center * Regional Market as Subregional Market . Area Market - Local Market .____1 ' Village Source: A. Mehretu (1982). Figure 5.1 Location and Size of Markets in the Eastern Region 64 information into the location and size of markets in the Eastern region. It appears that more than 65 percent of the rural marketplaces do not serve villages other than their own. The lingering impression from an observer of the physical appearance of these local markets and some area markets is their small dimension. In many of these markets, the stalls are simply made up of a straw roof supported by a few wooden poles, open from all sides. Actually, a great number of these village markets are almost completely shut down physically during the rainy season.8 The latter observation is supported by the fact that the staples millet and sorghum are available during the hungry season when 9 of all markets (A. Mehretu, 1982). stocks are low in only 33.1 percent Thus, the fact that 28 percent of the villages have markets (178 mar- kets for 635 villages) is a misleading indicator of the importance of markets in the region. The assessment of the importance of rural markets in Tables 5.2, 5.3, and Map 5.1 is on the basis of villagers' attraction to markets. Marketplaces attract local residents because of the exchange in agri- cultural products, but also because of other central place character- istics of these markets. To illustrate, in addition to agricultural products, manufactured goods (consumer goods, hardware, kitchenware, 8Market facilities are shut down in the rainy season because of transportation difficulties and also because time for market visits com- petes with agricultural activities. Exchange then takes place among households. However, in his compound, the local trader may still sell some grain to local residents. ‘ 9This figure is computed as the percent of the number of markets where sorghum and millet are available in the rainy season to the total number of markets (178) in the E-ORD. 65 4 Two other most dominant Gourtmantche have despised these activities. groups perform market services in the region: The Mossi are numerous in the West, and the Haoussa are numerous in the East at the border with Niger where many come from. It would appear that the Mossi have imposed their three-day market frequency in the populated area of the West, whereas, the Haoussa have imposed their weekly market frequency in the sparsely populated parts of the Center and East where their influence is mostly felt.5 5.2.2 Location and Size of Marketplaces 0n the basis of the results of the 1980 village inventory survey, Table 5.2 identifies only 27 markets which serve more than five vil- lages,6 7 and ranks them according to the distance from the marketplace to the user villages. In addition, Table 5.3 stratifies all the 178 markets in six categories on the basis of the number of villages served and the mean market radius. Finally, Figure 5.1 translates this 4Many actual Gourtmantche traders trace their origins to non- Gourtmanche background, from Mali notably (Jean Dahni, personal com- munication). 5As the legend goes, the Mossi Emperor's wives, who alledgedly set up the first regular market meetings in Ouagadougou, instituted the three-day market periodicity to coincide with the three-day brewing cycle of the popular sorghum beer (dolo). In the literature, it has been suggested that weekly market periodicity is often related to strong Muslim traditions. (Nearly all Haoussa in the East are Muslims, where- as, the religious beliefs of Mossi are more diversified and more tol- erant to alcohol.) 6It was not determined for which particular reasons(to buy and/or sell agricultural or other products) residents in (user) villages at- tend the markets identified. 7Distance is mean Euclidian distance in kilometers from the market to user villages. In Table 5.2 this distance is referred to as market radius. 66 agricultural implements, clothes) are also exchanged and services (bicy- cle and auto repairs, grain milling, traditional and modern medicine services) are also provided in the regional center, regional and sub- regional markets. Also, Manni and Piela markets attract numerous market participants because market-goers can take advantage of available health care facilities provided almost gratis by religious institutions. Furthermore, many markets in categories 1, 2, and 3 (see Table 5.3) are also administrative centers (e.g., the E-ORD sectors centers), and only a few markets, for example Namoumou and Piela, have deve10ped outside government-provided services. The importance of rural markets with respect to grain marketing may not coincide with the assessment based on the more encompassing central place characteristics. Knowledgeable market participants and the au- thor's own observations suggest a slightly different ranking of rural markets with respect to agricultural marketing (see Table 5.4). For example, Namounou is regarded by all participants as the most important produce market in the E-ORD on the basis of the volume of transactions and the number of participants attracted to the market. The fact that it is also the major market of smuggled manufactured products in the region adds to this importance. Another category of rural markets is the frontier markets at the borders of Togo and Benin (Pognoa, Zembede) which have their counterparts (with the same names) on the other side of 10 the border. Further information on the role of rural markets is pro- vided later on in this chapter. 10The ranking of Table 5.4 should not be viewed as inconsistent with that of Table 5.3 simply because the two differ. This difference is es- sentially due to the fact that the ranking of Table 5.3 is based im- plicitly on a broader criterion than grain marketing, the subject of the present study. 67 Table 5.4 Major Agricultural Produce Markets in Eastern Upper Volta 1. Regional Centera Namounou 2. Regional Markets Fada, Piela, Diapaga 3. Subregional Logobou, Nadiabondi, Mani, Diaka, Bilanyanga, Nassougou, Bassieri, Tansarga 4. Frontier Markets Pognoa, Zembede Source: 1980 marketing investigation. aPouytenga, located outside the E-ORD and 5 kilometers off the Fada- 0uagadougou highway, is an even larger rural market, and plays an even more important role than Namounou. Actually, Pouytenga is regarded as the largest rural market in Upper Volta. It happens that it is also the largest smuggling market in the country. 5.3 Typology of Major Market Participants There are many ways to categorize rural market participants. In many cases in Africa, sharply delineated definitions such as "whole- salers" and "retailers" are not operational research concepts because many intermediaries perform all of these and sometimes other functions (see also, H.M. Hays, 1975: and N. Ejiga, 1977). Here, we categorize the major middlemen on the basis of the following criteria: (1) Title to the product or ownership of capital. The inter- mediary may own the grain he buys or he may do so on behalf of a third person. This category distinguishes between the independent trader (merchant middleman) and the dependent trader (agent middleman). 68 (2) Remuneration of the agent middleman. The dependent trader may work for an agreed upon commission, or he may be an employee or a sort of apprentice whose ultimate re- muneration is a share in the patron's enterprise or a grant to start his own business. (3) Trading range. Merchants may trade at only the village level, the canton level, the 0RD level, or across several 0RDs. 0n the basis of these criteria, we distinguish six main categories of middlemen operating in grain marketing channels in the Eastern re- gion.H 5.3.1 The National Wholesaler-Retailer (NWR) This is a merchant middleman with a large capital base, located outside the 0RD in the large national market centers (Pouytenga,' Ouagadougou, Koupela, Boulsa, and Kaya) and who trades across several 0RDs. Most of the NWRs are Mossi. Many of those large traders OWn their own trucks and transport purchased products which have been col- lected by a network of agents or bought from smaller traders in regional markets. They buy large quantities in the region, but may also sell large quantities, depending on supply and price conditions, at whole- sale and/or retail in the markets they operate. nThis typology, implicitly, reflects the fact that market channels are mostly set up to move products outside the E-0RD. As will be seen later, many of these participants have a much more reduced role in the backflow of grain in the E-ORD. 69 5.3.2 The Regional Wholesaler-Retailer (RWR) This merchant middleman is similar to the NWR except that he is located within the 0RD, at one of the larger regional rural markets or centers (Namounou, Fada, Piela, Diapaga) and has a trade basis in the region. Most RWRs are Gourtmantche or Haoussa and only a few RWRs own their own trucks. Furthermore, their network of agent middlemen differs from that of the NWR as will be seen later. The RWRs collect truck- loads of grain that they sell outside the 0RD and sell smaller quanti- ties at retail in their principle business locations. NWRs and RWRs are also active in the marketing of crops other than grain. They are often licensed (aggégs) by the GOUV to buy export crops (peanuts, sesame, shea nuts) and sell them to the export crop marketing board (Caisse de Stabilisation) and CITEC (nationalized oilseed plant). Furthermore, many of the RWRs have made substantial profits in cattle trading from the 0RD to coastal countries (Togo, Benin, and Nigeria). Finally, some own other businesses such as grain mills, consumer goods stores, and bars. 5.3.3 The Local Independent Trader (LIT) This is a merchant middleman with a small trading capital base who operates over a small range of a few markets and villages. He is often based in the larger markets of the 0RD, but many come also from outside the region (adjacent 0RDs of Kaya in the North, and Koupela in the West). These traders do not own trucks, but some may possess donkey carts, and almost all make use of bicycles or motorbikes when it comes to transporting small quantities of grain. The average LIT is also en- gaged in some consumer goods retailing, but, and more so than most RWRs, may also be a farmer who produces grain for his own needs. 70 5.3.4 The Trader's Apprentice (TA) The TA is generally not a salaried employee, but rather an appren- tice in the merchant middleman's enterprise. Oftentimes, the two are related and may live in the same compound. These relationships are re- flected in the fact that the apprentice's remuneration is usually not a straightforward commission based on volume, but something which includes living expenses, a reward based on the success of a given trading period, and ultimately anticipated rewards such as a partnership or the capital needed to start his own business. NWR, RWR, and LIT have apprentices who buy grain in outlying villages on their behalf during the marketing season, and return to help in retailing grain and consumer goods in the market centers. 5.3.5 The Commission Agent (CA) Similar to the trader's aide, the commission agent does not take title to grain, but purchases for others at the village level. What distinguishes the CA from the TA is that he does business on behalf of a merchant middleman for fixed commissions. Furthermore, the CA‘s in- teraction with his employer lasts a shorter time than that of the TA (particularly if the employer is outside the 0RD, in Ouagadougou for example) and occurs during the buying season. The CA is often a farmer himself, and/or he may also be a craftsman in one of the villages in a producing area. Another difference between the two is that the CA works at the village level, whereas the TA works at both the village level and in regional marketplaces. Finally, the CA tends to work almost exclu- sively with outside traders (NWRs). The CA is active in his home village, but also in surrounding vil- lages. He may also operate in distant villages in which he relies on 71 village resident buyers (VRB). Both the CA and the TA perform the im- portant function of gathering information for their employers. 5.3.6 The Village Resident Buyer (VRB) The VRB is also an agent middleman who buys grain on behalf of a third party and is paid on commission. His distinguishing feature is that he operates in one village only. The most common picture of the VRB is a village chief, or some other prominent figure in the village hierarchy to whom farmers are willing to sell their produce once he an- nounces that he is buying. In effect, the VRB lends his social influ- ence to an outside middleman and gets remunerated. But the VRB is pro- vided with less capital than the other commission agents, and he tends to work more with grain merchants located in the E-ORD. In a village, there may be households that provide accommOdations for outside traders who want to stay a few days in order to conduct busi- ness. The room given to the trader also serves as a storage room for the grain bought. Although the villager host (called a lgggur_or a land- lord) may influence producers to come and sell grain to the trader, he is not paid a commission. Rather, he is given gifts, oftentimes in kind, as a reward. OFNACER also uses this system, but agents pay the lggggr for providing and/or watching the storage facility. The preceding paragraphs have described the major middlemen in the traditional grain marketing system (see Table 5.5). There are other categories of participants in the total grain system that we should also mention. One is OFNACER which buys grain through its buying agents, has modern warehouses in Fada, and sells grain mostly to civil servants in Fada and other cities. Another is the 0RD village group with a village cereal bank which is analyzed later. Futhrermore, food processors, 72 .cwmcm co zmcos m—ucm; p.=mmou w; .co_p_ucm :H .cowmmwaeou a upmn yo: m_ was .cmmzumnuom m we zpco muum umc_w may .cm>m3o: .mm> on» on ce—_E_m m_ Aczmmopv whopucmp mews .cowumu_umm>:_ mcwumxcms camp "mucaom .mme—__> mco .mmmp copaeu_wpuecw co Acwaam cu a_u_p ozv Amm>v ecm>3m Lo mace; uaau_cummm -_P> ca ago .m =_;u_z mad =o_mm_esou Pau_aau oz ceme_mam ama___> .mmmmppw> .mumxces co mmmmp mod Acwmcm on wpa_a ozv A a? mac .u cvspwz cowmmweeou pmu_amu oz ucmm< :owmmweeou .mmmmp . -~_> new mumxcme .mumxcms :_ oao .m mcmguo ace Acwocm on m—u_u ozv Aceme= pma_aeu oz muwucmcga< m.cmuecp .mmmep 1~P> new mumxems .mpmxcae cw ago .m Ahfiev gonna» 3mw m we emcee spasm muwmgao Lo :_;HP3 “Pecan mmmm P_msm ucmucmamuca _mqu .mmumpg _mcacmu .omo \maaxaae amen, Amzmv ca_capa¢ .m cm>o emcee mac; cw omo .m cwguwz uwmoca mmmm magma lampmmmpogz pmcowmmm .mamo chm>mm Amzzv cmpweumm mmocom emcee ago; .amo .u mu_mu:o uwmocm mmmm omen; leopammpozz .m:o_uuz mmcom m=_vach cowpmuoe cowuecmczsmm :_mgu op :msmpuc_z co mmaxw opuwh\Pmu_aeu . swampee_z Lowe: co mmnxp new =o_uu~wcommaeo co mwcmu_cu m.m mpnwp 73 mostly women,who buy grain for the purpose of preparing and selling traditional beer and food,play an important role, particularly in local grain transactions. Finally, there are transporters, such as truck, donkey, and donkey cart owners. As already noted, many of the NWRs and some RWRs own their trucks (commercant-transporteurs) which transport passengers as well as grain, but there are also private truck transporters who have regular lines or provide their services on demand. Besides these options, many traders make use of government-owned trucks with or without the knowledge of the state organizations. There also are a few donkey owners in the Namounou area who engage in transporting grain, but most donkey cart owners are primarily in the ubsiness of transporting firewood which pays more than transporting grain for a third party. (Revenues from donkey carts can be a relatively important source of income for ANTRAC farmers. See Barrett, et al., 1981.) 5.4 Grain Marketing Channels in E-ORD Marketing channels may be viewed in terms of the many alternative combinations of market intermediaries who participate in, or facilitate, the change in title of products from producers to consumers. They may also be viewed in terms of the physical movement of grain (assembly and bulking centers, transport routes, storage). The description of the major marketing channels that follows is mainly in terms of the coordina- tion among the major marketing institutions. For lack of quantitative estimates, the treatment of the physical distribution will be less de- tailed. There are two major types of institutional grain distribution sys- tems in the Eastern region. One type involves the public institutions, 74 while the other does not and is referred to as the "traditional" system. The traditional marketing system involves private traders' participation, but also has direct house trading among village households. As will be seen later, there are some links, if only minimal, between private and public intermediaries. Another interesting feature of the E-ORD grain marketing system is that transactions involving public and private in- termediaries may be a market-oriented trade or a farm gate-oriented 12 trade. In addition to these institutions, there is the village cereal bank which is operated by the precooperative village groups (groupements villageois) and the ORD. The importance of the village cereal banks will be analyzed in Chapter 7. 5.4.1 Traditional Grain Marketing System 5.4.1.1 Marketplace-Oriented Trade Figure 5.2 is a summary flow chart showing possible distribution channels with marketplace-oriented trade in the traditional system. Pro- ducers may also be retailers or consumers: for some go to these markets to sell, others to retail, and still others to buy grain. The middle- men are either located in these markets or come from other centers. The double-headed arrows linking various participants denote possible reverse flows of grain which will be discussed later. The main distribution channels in the Eastern region are the follow- ing: (1) Producer/Retailer-Consumer. This direct exchange is a local trade which redistributes local grain surpluses in small lzlt is worthy of note that although house trading occurs also out- side marketplaces, it involves mainly transactions between producers and final consumers located in the same or neighboring villages. Trader's Apprentice 75 Producer/Retailer/Consumer g - - - - -> Local Independent Trader - - - - 9 National Wholesaler-Retailer, Regional Wholesaler-Retaileri: -—-> -5 e——-> 11 Women Retailers 11 Within Eastern-ORD/Outside E-ORD Denotes one-directional flow. Consumers Other Than Producers Denotes one-directional flow and internal flow to merchant middleman's business. Denotes two-directional flow. Figure 5.2 Traditional Grain Distribution Channels in Marketplace-Oriented Trade 76 quantities. In general, both participants are female, and it is often difficult to determine whether the seller is re- tailing her own (or her husband's) grain, or if she is re- tailing grain previously acquired from other participants. In the latter case, these other participants are producers or merchants, but consequently the channel is no longer direct. (2) Producer-LIT-Consumer. In this channel, the local in- dependent trader assembles grain for retail in other mar- kets or in the same market later in the year during the “hungry season." (3) Producer - LIT - RWR or NWR - Consumer. In this channel, grain collected by the smaller merchants (LITs) is sold in the same or next larger market to larger merchants (RWRs and NWRs). The RWR retails the grain in his business loca¢ tion which is within the E-ORD, while the NWR located out- side the E-ORD retails the grain to consumers outside the E-ORD. This is an important channel linking rural and urban areas. (4) Producer - LIT - RWR - NWR - Women Retailer - Consumer. This represents the longest major marketing channel which moves the grain outside the E—ORD to consumers in other parts of Upper Volta. 5.4.1.2 Farm Gate-Oriented Trade In the Eastern region, grain assembly by private intermediaries oc- curs not only at market sites, but also at the farm gate. Figure 5.3 illustrates this farm gate-oriented grain assembly while also showing 77 [ (Neighboring) Village Households J 4} n n ap_____—+¥ Producer . }---€> 8 (———-[ Village Resident Buyer 1 :3 , e e c a) :i — i 2‘ 2' -——)L Local Independent Trader k S t” ~r \ '5 s. I fl :1 - t ' E E —-—)Eegional Wholesaler-Retailer' ; .8 , : I _ L it -—--4>| National Wholesaler-Retailer Denotes possible transactions between merchants in marketplaces. Figure 5.3 Traditional Marketing System Showing House Trading, Assembly at the Farm Gate, and Merchants' Transactions in Marketplaces 78 the house trade and possible transactions which may later take place in market centers between merchant middlemen. For clarity, the transactions in marketplaces, between middlemen, and consumers are not shown. The following are the major types of transactional channels: (1) (3) (4) Producer-Village Households. This direct channel has farmers selling to consumers in the village. These con- sumers may be other producers, craftsmen, or fulani herd- ers. This is part of the "hidden trade" or "household trade" as reported in marketing literature. In addition, exchange involving producers and other participants may include barter transactions such as gifts, credit, etc. Producer-Commission Agents-NWR. Commission agents are used mostly by the large outside traders (NNWRs) who do' not have a reliable network of village resident buyers. Even though we have not shown a link between commission agent and village resident buyers, this link may exist, but a CA will use VRBs only if the volume requested by the NWR is very large. Producer-—LIT. Here we note that the local independent trader who goes directly to producers in the village to buy crops. He may also go through a village resident buyer (VRB). Producer-Trader's Apprentice-Merchant Middlemen (LIT or RWR or NWR). Trader's aides here are doing the buying just as the buying just as the LIT, but they do so on behalf of a merchant middleman who may happen to be a LIT, a RWR, or a NWR. 79 (5) Producer-VRB-TA-Merchant Middlemen. Merchant middlemen often send their apprentices to VRBs when the area over which they want to collect the product is large, and/or when they do not have enough apprentices, but they have confidence in several VRBs. 5.4.2 State Marketing Channel In this section, we describe the structure of OFNACER (CRG of E-ORD) buying organization. OFNACER headquarter in Ouagadougou every year decides on the number of buying agents to send to the Eastern re- gion who will be directed by the OFNACER regional chief (Chef de CRG). In the 1979-80 buying campaign, OFNACER employed nine buying agents in the Eastern region to operate in the marketplaces and also directly at the village level. They buy products (millet, sorghum, cowpeas, and paddy rice) directly from producers at marketplaces and at the farm gate, and they also buy from traders. In view of our categorization of the participants, the marketing structure of OFNACER is as follows (see also Figure 5.4): (1) At the village level, OFNACER agents buy directly from pro- ducers at preannounced times and places. The buying points are generally the compounds of the village chiefs in one of which the agent usually resides. The village chief plays the role of a landlord (199235) who is paid a rent for the agent's housing and an amount of money based on the number of bags and days of storage. (2) When the AFNACER agent operates in the marketplace, he usually purchases large quantities from the merchant mid- dlemen who are willing to assemble and bag the small 80 Producer _ 1 L . . I 7 1 Landlord (logeur) 1 I I 71/ F——>[ OFNACER Merchant Middleman (LIT or RWR) 1 (Mostly Urban) l Consumers ---> At the marketplace. ---> At the village level. Figure 5.4 OFNACER Marketing Channel 81 quantities sold by numerous producers. In most instances, the trader buys the grain at the market, bags it using OFNACER's bags, and sells it right away to OFNACER. In other cases, the trader may rely on his connections in villages to organize grain assembly and transportation to the OFNACER market buying point. In contrast, OFNACER pur- chases from producers at the marketplaces are not very important. 5.4.3 "Back Flow" of Grain The preceding presentation of the grain market channels is typical of the unidirectional presentation of the flow of agricultural produce from producers to rural and urban consumers. Added to this picture should be the sequence of middlemen who interact to sell back food grain to rural producers/consumers. In particular, we are interested in the inflow (If grain to rural areas through the channels under a normal production year' and/or' during the rainy or "hungry" season. RWRs and LITs located in markets close enough to outside markets such as Pouytenga and Ouagadougou are known to go to the NWRs outside E-ORD to buy food crops. The RWRs may, in turn, sell the products to other LITs or retail it themselves in the large rural markets where they are located. However, the local independent trader (LIT) is the only intermediary with large volumes that comes in close contact with pro- ducers/consumers and other rural consumers in the\village. [These pos- sibilities are illustrated in Figures 5.2 and 5.3 by reverse (up) arrows.] The other intermediary is the small retailer whose trading range is severely limited in the rainy season. 82 The small proportion (33.1 percent) of rural markets where grain food is available during the hungry season indicates that many rural consumers are not well served by the network of rural markets during the rainy season. Unfortunately, the network of agent middlemen (TA, CA, and VRB) which teamed up to move out products from farmers is not oper- ating to serve back producers/consumers. This is because once the buy- ing campaign is over, TAs rejoin their respective traders in the larger marketplace, and CAs and VRBs go back to their own business of farming. As for OFNACER, its distribution network usually covers urban cit- ies only as indicated in Chapter 3. (In E-ORD, OFNACER's main food dis- tribution outlet is in Fada.) Only in some instances of severe drought does OFNACER organize interant selling points of food imports. Unfor- tunately, these selling points are set up in the main regional centers (where most civil servants are reached) and food is sold directly to consumers. Thus, whereas there are important market channels that remove grain from the region through markets as well as through exchange arrangements which operate outside marketplaces, only in important rural markets in the rainy season is grain available to rural consumers through market intermediaries. In all other instances, producers/consumers and other rural residents must have accumulated grain stocks or buy grain from fellow grain producers. (In a few cases when the market is shut down during the rainy season, a local trader may still sell from his com- pound.) But if grain harvest happens to be disastrous for all producers in the area, food shortages may be severe. 83 5.4.4 Major Grain Physical Movements in E-ORD Even though quantitative estimates of the total grain handled by market participants are lacking, the following account still provides a good summary of the physical grain movement in the E-ORO based on our present knowledge of the grain marketing system. Four main movements that take grain out of the E-ORD have been identified: (1) The movement which extracts grain from the Namounou-Diapaga- Logobou area to the large market and consumer centers of Pouytenga, Ouagadougou, and Fada, through the main Diapaga- Kantchari-Fada Highway, is regarded by many knowledgeable merchants as the most important. This flow is fueled by added grain picked up along the road as trucks head back to Ouagadougou or Pouytenga on Sunday nights or Monday mornings following the Sunday market of Namounou. (2) Along the West stretch of the E-ORD, grain flows from centers like Coalla, Manni, Diaka, Bila-Yanga, and Piela to the mar- ket and consumer centers of neighboring 0RDs, Yalgo and Boulsa (Kaya 0RD), Pouytenga and Koupela (Koupela 0RD), It appears that such movement is more important than the flow along the second road network of the region (Bogande-Piela- Fada). (3) An important part of the grain also flows to the neighbor- ing countries of Togo, Benin, and Niger. Historically, the E-ORD had been considered a "food shed" for the neighboring countries of Niger, Benin, and Togo. Today, there is still a great deal of produce being exported to these countries (especially millet and cowpeas). Togo and Benin draw their 84 their supplies mainly through markets along the borders (particularly Zembende and Pognoa on the Togo border). These markets often have their counterparts or "twins" on the other side of the border, and transactions will involve Voltaic producers, foreign consumers, and small traders from both sides. The visible volume traded.and exported may not be impressive in any market day, particularly if no large capacity trucks are present, but it may be quite important over the entire trading season. On the other hand, Niger draws its supplies deeper in the region because it has a longer border and better roads with the region than T090 and Benin. (4) Finally, the direction of flow of grain from the center of the region is less distinct than in other parts because of the very poor road network. They branch either to the Piela- Pouytenga route or to the Namounou-Fada route. In contrast with the numerous alternative grain export flows from the region, there is but one major route along which grain flows back into the E-ORD. This flow originates from Ouagadougou and Pouytenga, the very centers which imported grain from the E-ORD during the com- mercialization campaign. The path follows the main highway in the re- gion and reaches Fada and Diapaga. 5.5 Standard Operating Procedures This section examines key standard operating procedures involving agent middlemen, merchants, and producers. 85 5.5.1 Private Agent Network At the large marketplaces (Namounou, Piela, Logobou, etc.) there is a keen competition, on the one hand, among small independent traders, traders' aides, and women restaurateurs, and on the other hand, between national and regional wholesaler-retailers. At these markets, some as-. semblers may be known well enough to attract a clientele of producers, but most assemblers pay children to guide uncommitted producers to them. This practice of touting producers is taking a peculiar twist in the Fada area. In Fada, women restaurateurs do not wait for producers to come to the market. Instead, they meet them at key crossroads a few kilometers outside the marketplace. In addition, they also send their children on the smaller trails to intercept farmers on bicycle. The child usually attaches a distinctive piece of cloth to the bicycle to indicate to which woman the farmer will have to sell his/her grain. These transactions outside the Fada market have become so prevalent that almost no producer goes to the market of Fada for the purpose of selling grain (those who do, usually women, intend to retail ratherthan sell to middlemen). Women restaurateurs have come to rely on this scheme in order to avoid purchasing grain from merchants or paying double taxes (one for buying grain, another for selling food) when they buy grain at the market. Farmers, for their part, are content to avoid the inconven- ience of the town (notably the police which may harass them for various reasons: nonpayment of head taxes, lack of identification, lack of proper fixtures on their bicycles, etc.). As a result of this practice, an assembly grain market may be developing outside the Fada market. Buying grain at the marketplace is a time consuming task which yields disproportionally small results for any single assembler because 86 of the small quantities sold by a score of farmers and the small units of measurement used. In the small assembly markets, large merchants (RWRs and NWRs) may rely on their apprentices to collect grain, but in larger markets, the number of aides may be too small to take advantage of the available supply. Consequently, in the latter case, larger mer- chants contract with smaller ones (LITs) to supplement their needs. (To hire more aides is riskier because of potential cheating.) An important feature of these contracts is that large merchants re- luctantly provide credit to LITs to buy grain. Rather, they wait until the LITs have assembled a large enough volume to buy (usually at the end of the market day, but sometimes as the market activities go on). It has been reported that RWRs and NWRs fear that LITs (and occasionally the apprentices) may use their capital to buy grain and sell it to other merchants at more profitable prices to them. (When this happens, LITs and 1A5 usually pretend that market supply conditions were too poor.)13 When the LITs do not receive credit from their larger contractors, they are free to strike a deal with anyone willing to buy grain, but at the same time, it substantially reduces their access to working capital.14 In the farm gate oriented trade, the village resident buyer (VRBs) and landlords (logeurs) play a central role. The VRB tries to convince other farmers to sell their grain to him, and thus saves the trader (RRW, LIT) the time it would have otherwise required to go from household to household. As previously mentioned, the VRB who takes an active part in 13In the past, the E-ORD had also experienced such misfortunes with some village groups. 14Timely access to working capital is a major barrier to greater par- ticipation in grain marketing. As we will see, the easy access to large amounts of capital explains the appeal of the village cereal bank scheme. 87 handling grain and cash, is paid a commission, but the landlord who acts only as a go-between receives only some gratifications (usually gifts in kind). Sometimes, this handling of grain and cash by VRBs creates prob- lems. The VRBs may not measure the grain the way the merchant might have, or the VRBs may cheat the trader out of his money the same way the NWRs fear the LITs do. The standard operating procedures described above have emphasized the fact that cheating may exist to show how it may effect the system operation. Controlling the buying agent (bet it a VRB, a CA, or a TA) appears to be an overriding concern of the merchant. There is a trade- off between the cost of control (time and money costs of physical con- trol) and the benefits of reduced money loss (because of cheating and inexperience). Merchants operating within their supply area can afford to give their agents small amounts of working capital (one-fourth or a fifth of the total at one time) and make more frequent control visits. In contrast, menchants located outside their supply area in the E-ORD, for example in Ouagadougou, have to provide their agent with larger operating funds and make control visits only a few times (no more than two). As for the evolution of agricultural marketing in E-ORD, it was suggested by knowledgeable traders that more and more Gourmantches are entering the business and are challenging traders of other ethnic groups (Mossi and Haoussa) who have been dominant so far. The change may be due to inroads of Muslim and Protestant beliefs more tolerant to trade than traditional Gourmantche religious beliefs (most large and wealthy Gourmantche traders are either Muslim or Protestant). This rise of Gourmantche traders forces particularly Mossi traders from Ouagadougou 88 to rely more and more on LITs or commission agents who are part of the local communities in the E-ORD. (It should be noted that there is a large number of Mossi permanently established in the region and who are no longer considered strangers.) 5.5.2 OFNACER Agents'Operating Procedures Table 5.6 shows the locations and characteristics of OFNACER's agents who purchased grain in the E-ORD during the 1979-1980 campaign. Most hardly understand the predominant local language (Gourmantche) and almost all have been in position late in the buying season. (In Bogande the agent was positioned in April to sell grain.) As for purchases, ap- proximately half of the agents did their major purchases in villages (by- passing marketplaces) and half in the marketplaces of the village in which they were located. Because the relevant performance criterion on the basis of which agents receive a bonus is the total volume collected, and because most agents are handicapped either by the language or their knowledge of local trade conditions, the agents rely on intermediaries to bulk the grain in order to reduce transaction costs when they are given the op- portunity. At the marketplaces, the intermediary is a merchant15 (li- censed as well as nonlicensed), and at the village level, it is gener- ally the village chief who plays the role of landlord (see Figure 5.4). In the latter case, producers are given empty 100 kg bags to fill and 15Some agents provide capital to merchants to collect the grain; these merchants, who thus receive the credit that many LITs don't have access to, deliver the volume based on the official price and keep their profit. When market prices rise above official prices, they stop ac- cepting the deal. 89 mucmu_mmc .macoma on» on age: mcmuzuoca .uamumcH .mpnm_.e>a no: u .<.zm .__am as .uoxcas mom—PP> ecu :_ camagocza ma: :_acm o: .Pcaoncaz cu .mumxgas mzu mu_mp=o umuuaucou Ppwum mam mmmagucsa “mos «an .mmwuwpaoo— amen» cw mumxcms mam wcmchu .mucamom =_ pagans xpnsmmma cameo o: mp wedged .mco mmewumeom .mgucos go» a apco m_ .cm>mzo; .uowcma newewacu _azaum make .ucaumcmuqa mucaa_uwucaa poxcms Acme ;u_;z .mcoz ape» a umam— an xamam .amcamem» use A.<.zv sogmmaao cw uamuxm .mucmma —_< .m50pcaEL3oc cow vene; m? xawp_nmaau amazmcagm .=o_umm_amm>:H pagan: owlmump “muesom we mm\-\—F mcam> m cw>o Loom Loom ummma___> _geonuaz mm m~\_\NP mcam> m cm>o Loom Loom ummma_~_> mmcmmcah em mm\p\mp memo» m cm>o coca ucmppmuxm umxaaz nonomoe AN m~\om\__ mcam> m Lm>o Loom ucmp_muxu pagan: Puconawumz mm m~\_\mp meme» m gm>o coca Loom pmxcaz aoczosaz mm m~\m\mp meow» m cm>o voou . Loom «oxen: umaaa_o m.<.z m.<.z m.<.z m.<.z m.<.z .6332; 39.8.5 mu m~\p\mp mgucoz o coon Looh. “oxen: exawo om m~\e Lam> mco coco Loom ummmaFFP> mvcamom Amcem>v ucom< mco_amcmao mo nacmm< mmu omumpzocx ommsmcma Lona: mucauwmmm omm—nmnmp :P xgozamz .mucmm< amuwm mam cowuauzaeou co m_veumo n302 .cwmxm we was mx1oo~ a oco.¢ =m m:_ao¥caz omm. "mugzom aauwea coac_m -_=cu< co & ea - o.mm - m.mp 1- N.om 1- ~.N - m.m~ pmou mcpumxca: o.oop ¢._N o.oo~ N.“ o.oop m.op o.oop _._ o.oo~ m.m "umou papa» m.m F.N o.m a. m.mm m~.m 1- - n.~ N.o mmmcoam o.no m.¢_ m.¢n m.m - - - - ¢.om m.~ cowumucoamcech m.o_ m.m m.o~ o._ m.~_ mm.p o.oo_ _.F m.oF o._ m:_~u:a: N.PP om.m 11 - -1 - - 1- - - m:_»:m “moo mx\mca An cmcczucH mumou mcwumxcmz m>wuacumzppa m.m mpnmh 100 even though we try to account for regional differences, cost figures, apart from transportation, are normalized for the entire year. Indeed, handling and use of facility in a few areas can be higher (50 CFA for loading/unloading, 50 FCFA-for use of facility, for storage) than the rates used here. Fifth, taxes were unaccounted for due to the lack of information on annual volume. Hence, for all preceding reasons, the cost figures should be viewed as conservative; in other words, as mini- mum costs. Despite all the preceding qualifications, it is possible to make a number of general comments. Firstly, the costs of transportation (when it has to be undertaken) is the highest and may account for up to 80 percent of total marketing costs. A look at the rates charged for dif- ferent distances show that the transport cost is a step function with very extended plateaus. Transport rates are quoted between important centers without any discount for secondary centers in between (except when they are very close, by less than 10 km, to one of the extreme points). Examples of the stepwise behavior of transport are shown in Table 5.8. The overall conclusion is that transport costs are very large for medium-range distances and between centers off main lines. In contrast, transportation over long distances (to Ouaga for example) is the least expensive on a kg per km basis. In effect, the transport cost structure favors the grain export from the region, and within the Eastern region, export from the producing area. Secondly, storage facilities do not seem to be of much concern for these traders. Only large traders build a facility for the express pur- pose of holding grain, but most merchants make use of whatever space 101 Table 5.8 Grain Transport Costs Along Selected Transport Routes Distance Transport Route Mode Mn FCFA/kg FCFA/kg/km Namounou-Fada Truck 228 500 21.9 Kantchari-Fada Truck 150 500 33.3 Fada-Ouaga Truck 230 500 21.7 Namounou-Ouaga Truck 458 750 ' 16.4 Kodjari-Namounoua Truck 37 500 135.1 Tansarga-Namounoua Truck 22 400 181.8 Piela-Pouytenga Truck . 63 500 .79.4 Logobou-Namounou Donkey Cart 50 750 150.0 Naponsiga-Namounou Donkey Cart 35 500 142.9 Basseri Area Donkey Cart 15 150 100.0 Source: Marketing Investigation. aOff main transport lines. 102 happened to be available. Physical losses, provided the grain is not kept in storage for more than a year, do not present much concern for the trader. This is more so because most of the damage is passed on to the consumer when the physical appearance is not very much affected. This being the case, lower figures for storage loss from the trader's perspective are misleading as far as consumers are concerned. There may be warranted action to try and lower physical and quality damage to grain in storage. Thirdly, total costs of physical distribution vary according to whether the trader is a full service merchant or performs a few functions, but also according to whether he operates in the marketplace or mainly outside the marketplace. The costs of operating outside the market presented here are certainly more conservative than the costs in- curred in the marketplace, for we did not account for all expenses (travel, control, food, gifts for goodwill building) attached to operat- ing at the village level. Despite these higher costs, large wholesalers are still willing to operate there because of the larger volume secured in a short time. On the other hand, traders incur less cost when they operate at the marketplace because producers are performing some of the marketing functions. 5.7 Summary This chapter has described the characteristics of rural markets in the E-ORD, identified the major market intermediaries, described the major private and public grain market channels, analyzed key operating procedures of major participants, and attempted to illustrate the out- of-pocket costs of marketing services (physical distribution) performed by private traders. 103 The region's poor road system, uneven population, and socioethnic traditions are key determinants of the rural market network and period- icity. There is but a small number of major assembly markets and the marketing channels show how private traders attempt to adapt to these limitations by developing a sophisticated network which sometimes by- passes the marketplaces. The coordination of this marketing network has shortcomings caused by short supply of working capital, mistrust, and the demanding physical control that must be applied in order to reduce losses. OFNACER inter- vention in the marketing system tries to emulate private traders and in- troduces some competitive pressure that benefits producers to whom OFNACERs go to in their villages to buy grain. The marketing channels are geared to move products out of the re- gion, and in spite of their shortcomings, are more effective in doing so than coordinating the back-flow of grain to consumers. The highly dis- persed population and the deterioration of the already poor road infra- structure during the rainy season partially accounts for an ineffective back-flow. These conditions may also explain why physical distribution costs incurred by private traders are dominated by transport costs when merchants' time costs are left out. The lack of working capital and the low concentrated effective demand accounts for the small storage costs shown in the illustrative cost figures. This chapter constitutes a backdrop against which the marketing behavior of grain producers is analyzed in Chapters 6 and 7. CHAPTER 6 VILLAGE AND FARM-HOUSEHOLD LEVEL ANALYSIS OF GRAIN MARKETING BEHAVIOR 6.1 Introduction The present chapter analyzes food grain producers' marketing behav- ior at the farm-household level. A number of studies by anthropologists and economists have attempted similar analysis in other countries. They have described market transactions, estimated grain disposals, and mar- keted or marketable surplus, but few have generally examined the behav- ior in the light of the whole farming system. Economic studies, in par- ticular, even though they point to the complex nature of crops produced for home consumption, sale, and other nonmonetary exchange, have often sacrificed the understanding of this complexity to the needs to general- ize for the purpose of mathematical modelling. Too often not enough at- tention has been paid to the "back flow" of grain to the farm-family unit through market as well as nonmonetary exchanges. In addition, the inflow and outflow of grain has not been related to the farm-family cash flow situation. This chapter will attempt to answer the following questions: (1) In the make up of the grain flow, what is the relative im- portance of monetary and nonmonetary transactions? (2) Are the marketed or marketable surplus, if any, related to the farming system (hoe and animal traction)? 104 105 (3) How are grain purchases financed given the cash flow situa- tion? (4) What is the timing of sales and purchases of grain, and does it relate to price variability? (5) Do farmers sell and buy back grain, or are selling house- holds different from buying ones? The following specific points will be addressed: (1) the concepts and operational definitions of grain flow statements, marketed and mar- ketable surplus; (2) the limitations of the data and the consequent estimation procedures; (3) the empirical results of the yearly and seasonal grain flow statements; (4) the empirical estimates of the year- ly and seasonal marketed and marketable surplus; (5) the yearly and seasonal cash flow statements: (6) the timing of sales-purchases of grain and related farm gate price behavior; and (7) the empirical evi- dence of distress sales and repurchases of grain. 6.2 Concepts and Operational Definitions of the Grain Flow Statement, Marketed and Marketable Surplus 6.2.1 Grain Flow Statement The grain flow statement stems from three related concepts: (1) the conservation of flow principle; (2) the cash flow statement generated in farm management analysis; and (3) the food balance popularized by FAO publications.1 The conservation of flow principle states that what flows into the farm-household system either flows out or adds up to stock (see Figure 6.1). In other words: 1Other definitions are food budget, food matrix (Smith, et al., 1981): Etc. . 106 Stock Change > Input Flow ' Output Flow Figure 6.1 Conservation of Flow Principle Output - Input = Change in Stock (6.1) where output includes all possible household and farm uses, and change in stock is defined here as beginning-ending stock. The concept of food grain flow is also closely related to that of the cash flow of the farm-family unit and serves similar purposes. The annual grain flow examines the total availability and use of food grain by the household unit, and the seasonal grain flow examines how the vari- ability of the sources and uses of food grain might lead to eventual seasonal food shortages. Thus, both annual and seasonal food grain statements help identify problems that might not have been apparent otherwise. Finally, the grain flow statement at the farm-family level may be thought of as the micro level version of the food balance sheets estab- lished for entire countries or regions of the world by FAQ. The purpose of the food balance sheet is to estimate food supply available for con- sumption by accounting for export-import, industrial use, and waste. A similar measure can be derived in micro level grain flow statement. Following the pattern of a typical cash flow statement, the funda- mental identity of the grain flow statement reads as follows: Total Sources of Grain = Total Uses of Grain '(6.2) 107 One important dimension of the marketing component of the 1978-79 farm survey was the identification of all the possible sources and uses of sorghum-millet in Eastern Upper Volta. The sources of grain to the farm-family in Eastern Upper Volta include: (1) (2) (3) (4) (5) (6) (7) Beginning stock from the previous year crop, purchases and gifts, etc. Production--the 1978 crop. Purchases of grain: (a) for farm-family uses; that is, home consumption, animal feed and seed; (b) for the purpose of later resale; (c) for processing and sale of prepared foodstuffs. Gifts received. Grain borrowing. Grain received as payment for loans extended (in cash or in kind). Grain received as payment for work and services. On the other hand, the uses of grain include: (1) (2) (3) (4) (5) (6) Household consumption. Seed. Sales: (a) of own stock and production; (b) of grain previously bought for resale: (c) of processed grain (in particular, gglg, sorghum beer). Gifts given. Loans extended. Grain given as payment for borrowing (in cash or in kind). 108 (7) Grain given to hired workers. (8) Eventual losses. (9) Ending stock as of April 30, 1979. In order to simplify the presentation of the grain flow and related statements, let us classify the sources and uses of grain as identified above into four broad categories: production, exchange, farm and house- hold disappearance, and stock. Production is self-explanatory. Ex- change, for convenience, is defined to include all-purpose purchases and sales, gift, credit, and labor transactions. Farm and household disap- pearance, again for convenience, includes farm uses (seed and feed), household consumption, and losses. Finally, stock refers to beginning and ending inventories. To simplify further, let us call barter transactions those made up of gifts, credit, and labor transactions: and let us call fanm-household uses those for seed, feed, and household consumption. As a result, we inay define the following: Exchange Inflow = Purchases + Barter Inflow (6.3) Exchange Outflow = Sales + Barter Outflow (6.4) Net Exchange Outflow = Exchange Outflow - (6 5) Exchange Inflow Let us also define change in stock as: Change in Stock = Ending Stock - Beginning Stock (6.6) The use of the variables identified above allows an explicit yet compact representation of the fundamental identity of the grain flow statement. Identity (6.2) may be rewritten as follows: Beginning Stock + Production + Exchange Inflow = Consumption + Farm Use + Losses + (6.7) Exchange Outflow + Ending Stock 109 From the above expression, identity (6.6) which defines the change in stock may be rewritten as: Change in Stock = Production + Exchange Inflow - Exchange (6 8) Inflow - Exchange Outflow - Farm-Family Use - Losses ‘ Throughout the literature, measures such as food availability, grain disposals, marketed and marketable surplus are used to evaluate the be- havior and the performance of farmers with respect to food crops. These measures can be derived from the grain flow statement. Yet, this simple derivation shows that pitfalls and hidden assumptions, which may obscure the meaning of the measures identified above, go often underreported in the literature. Examples are shown below. 6.2.2 Food Available to the Farm-Family Unit For example, from identities (6.7) and (6.8), we see that: Consumption = Production - Net Exchange Outflow - (6 9) Farm Use - Losses - Change in Stock ' In food balance sheets such as those constructed by FAO, food supply available for consumption is computed as: Food Available = Production - Net Exchange Outflow - (5 10) Farm Use - Losses . Also, in many studies and consultant reports, one finds estimates of production available for consumption derived on the basis of one .year's data and computed as "raw production minus 4 percent seed and 5 Percent for losses" (CILSS, 1977, appendices). Obviously, since change in stock is not accounted for, the under- lying assumption is that change in stock is zero. But this assumption Inay hold more for the long-run that the short-run. That is, it may be safely assumed that in the long-run (using long time series data), stock fluctuations over particular productive or destructive years even out so 110 that, on average, beginning stock will equal ending stock. However, in the case of one particular year (using one year's data), there may be a difficult ground to assume identical beginning and ending stocks2 (see E. Simmons and T.T. Poleman, 1974). Given the riskiness of rain-fed agricultural production in semi- arid tropical conditions, food stock levels and changes play an impor- tant part in farmers' strategy for survival. This consideration makes the question of stock change an empirical issue. Also, given the focus of the present study and the national trend toward the promotion of vil- lage grain reserves to alleviate local seasonal shortages, an attempt should be made to estimate the magnitude and behavior of grain stock change. 6.2.3 Grain DispOsals The literature also abounds with grain disposal studies where at- tempts are made to measure the magnitude of uses from farm grain produc- tion. Written in terms of the above grain flow statement, grain dis- posed of by the farm-family is: Disposals = Sales + Barter Outflow + Consumption + (6 11) Farm Uses + Losses ° and stock residual is derived as: Ending Stock = Production - Disposals (6.12) 2One may still assume identical beginning and ending stock, pro— vided there is reason to believe that the year being studied has followed a long series of consecutive normal years so that it can be argued that producers have no particular reason to add to, or decrease their level of stock. Alternatively, one may assume that over a good representative cross-section of farmers, with grain production following a normal dis- tribution, some producers will increase their stock over the year, but others will decrease theirs so that one can argue that the overall mean will show no stock change. 111 Conceptually, as shown in the grain flow statement, such studies of grain disposed of provide only a partial picture of the grain movement within the farm-household context. First, the backflow of grain through monetary and barter transactions is ignored. Second, the eventual begin- ning stock is assumed away. Yet, to the extent that there is any grain in stock at the end of the marketing year, there will be a carryover at the beginning of the next year, which then indicates that there might have been a stock at the start of the present year. If producers do car- ry grain stock from one year to the next for security reasons, the end- ing inventory derived above may provide only a partial understanding of producers' behavior. Again, we are faced with the empirical question of stock change. 6.2.4 Grain Marketed and Marketable Surplus The concept of marketed and/or marketable surplus of food crop has received a great deal of attention thanks to the pioneering work of Raj Krishna (1962) and the many other studies that followed with added re- finements (see for example, J.R. Behrman, 1968). The concern over mar- keted-marketable surplus arises from the fact that estimates of price elasticity of supply, which are an important potential ingredient in agricultural policy design, may not be reliable if they are derived through econometric supply functions which treat food cr0ps like export or industrial crops: owning to the income elasticity of producers' own consumption out of production, the total quantity harvested and its portion made available to other consumers (particularly urban consumers who purchase from the market) may respond differently to prices and other relevant variables. 112 The main emphasis of the studies mentioned above has remained the derivation of regression coefficients which would relate some measure of marketed or marketable surplus to production, consumption, price, in- come, and other relevant variables. In the literature, the ongoing de- bate centers on: (1) how to approach marketed-marketable surplus (i.e., through direct observation or as a residual): and (2) the sign and magni- tude of these coefficients. The debate over the sign of the coefficients is an empirical issue (J.R. Behrman, 1968), but it also hides some pri- mary concerns. The fact that the concepts are operationally ill-defined because of hidden assumptions (discussed later on) obscures the debate. It is also apparent that producers are considered as grain exporters who are satis- fying their consumption needs.. But what happens if producers are net or absolute importers? .The complexity introduced by such a situation is handled with more or less difficulty by current econometric techniques.3 But more importantly, the results of such techniques are very difficult to interpret in a policy context given the generally poor quality of data used. Thus, before one adds new ammunition to the debate over 3Standard econometric techniques can handle a dependent variable (marketed surplus) with negative, but nonzero values. A difficulty a- rises over which price series (prices received by producers, prices paid by producers, or a composite market equilibrium price) to use. Current techniques used only one price (retail or wholesale market) in single or simultaneous equation systems. Another and more serious difficulty stems from the fact that the dependent variable (marketed surplus) may assume negative, positive, and zero values because, as will be shown later, there may exist four categories of producers: (1) those without sales nor purchases: (2) those with purchases only; (3) those with sales only: and (4) those with both sales and purchases. Tobit analysis, which has been used to model only sales (see Garcia, 1978; and Sands,; 1984) can handle dependent variables with zero values, but without negative values. 113 regression coefficients, it is worthwhile to ponder the meanings of the concepts themselves in a low income economy. From the ongoing analysis of the Eastern Upper Volta marketing in- formation, it is clear that a distinction can be made between the mone- tary transactions involving sales and purchases and the nonmonetary transactions, here loosely defined as barter transactions, which include in-kind gifts exchange, credit transactions, and labor service payments. In addition, it must be realized that even in a predominantly rural eco- nomy, food grain is not only sold, but is also purchased. Consequently, marketed surplus to truly represent what actually flows out of the farm- family unit or to serve as an indication of the degree of monetariza- tion, should take account of the back flow of food grain. Marketed surplus is therefore best defined as net market sales; that is: Marketed Surplus = Net Market Sales = (6 13) Total Sales - Total Purchases ° Here, total sales include sales of produced grain, sales of grain béught' for trade, and sale of processed foodstuffs from grain. Similarly, total purchases include purchases of grain for farm-family use (seed, feed, and human consumption), purchases of grain for trading purposes, and purchases for processing. Since grain is also exchanged through nonmonetarytransactions, the net barter position may be defined to include net gifts given, net credit outflow (loans plus debt payment less borrowing less payments for loans), 114 and net labor payments (wages of hired labor plus cost of labor "invita- tions"4 less wages received). Consequently, a better indicator of the total grain available to consumers outside the farm-family unit include not only marketed surplus, but also the net outflow from the barter transactions. Let us call it, for convenience, net exchange outflow: Net Exchange Outflow = Marketed Surplus + (6 14) Net Barter Outflow ' ° Identities (6.13) and (6.14) provide an excellent basis to examine the common alternative definitions of marketable surplus found in dif- ferent studies. The concept of marketable surplus relates to the total food crop potential available for sale. Some authors use marketed and marketable surplus interchangeably; that is, sales net of purchases (identity 6.13). For others, marketable surplus is the net total ex- change (identity 6.14) available to participants outside the farm-family unit. Still others add a stock component to either net market sales on net exchange outflow to make up marketable surplus. The stock component, often included, is accounted in two different ways. First, some authors consider that the residual stockg after'an ex- plicit level of security stock is subtracting from the ending stock,will be available for sale. For example, Lassiter (1982, p. 202) defines marketable surplus as "the total value of crops either sold or available for net sale net of requirement for subsistence consumption and food security stocks." The difficulty, of course, is to determine both 4Labor "invitations" refer to traditional work parties of a recipro- cal nature in which the "invited" are rewarded in kind (food, sorghum beer, and cola nuts). 115 reserved consumption and security stock out of ending stock. It should also be noted that the "surplus" stock will include potential sales as well as other potential uses. A second group considers the change in stock as potentially avail- able for sale. The hidden assumption in such a definition is that the level of the beginning stock is deemed satisfactory or planned, so that any stock above this level at the close of the year might be sold off (or exchanged) once the farmer realizes it. But change in stock may be either positive, when stocks are being built up, or negative, when stocks are being depleted. The interpretation of marketable surplus, which is straightforward when change in stock is positive, becomes less obvious when change in stock is negative. A negative marketable surplus may be saying that the farm-family unit has given up or reduced stocks poten- tially more than it should or wanted (again, assuming that the level of beginning inventory was planned): Marketable Surplus = Net Exchange Outflow + (6 15) Change in Stock ° Thus, for this second group, marketable surplus represents not only a measure of the food cr0p potentially available for sale, or better for exchange, but also a measure of the potential erosion/strengthening of the food balance position of the farm-family unit. However, even though this marketable surplus accounts for change in stock, it does not explain the decision making of households as to the build up or depletion of stocks. The two different ways of estimating marketable surplus (assuming an explicit security stock level and accounting for change in stock) yield different results unless beginning stock is identical to ending stock and equals the explicit security stock level. A marketable surplus that 116 uses an explicit end of year stock security level will never drop below the net exchange outflow (or marketed surplus) and will be negative only if net exchange outflow (or marketed surplus) is negative and large. On the other hand, a marketable surplus estimate that includes change in stock may be smaller than net exchange outflow (or marketed surplus) and will be subject to a great deal of variability. Marketable surplus defined as in (6.15), however, is consistent with the distinction of the direct and indirect approaches of marketable sur- plus derivation encountered in the literature of supply analysis. The identification of marketable surplus in terms of its components as in (6.15) is referred to as the direct approach. The indirect approach, on the other hand, identifies marketable surplus as a residual on the basis of the grain flow statement. From identity: Change in Stock = Production - Net Exchange Outflow- (6 8) Consumption - Farm Use - Losses ’ we can derive also marketable surplus (change in stock plus net exchange outflow) as: Marketable Surplus = Production - Consumption - 1 Farm Use - Losses (5' 5) Thus, implicitely, the indirect approach also assumes that beginning stock is planned or desired stock. As to the controversy of which ap- proach is best (see Medani, 1975; and G. Gemmill, 1978), it appears that the indirect approach has fewer components to estimate, but the degree of difficulty of either approach hinges on the availability and accuracy of the data. To sum up, the grain flow statement is a useful tool to describe and diagnose producers' behavior and performance with respect to grain 117 marketing. An important element of this statement is the year-to-year change in stock which affects producers' ability to feed themselves properly. Unfortunately, in the literature, the change in stock is often conceptually assumed away rather than being treated as an empirical question. From the grain flow statement some of the conceptual limitations of alternative measures of producers' behavior, such as grain disposals and marketed-marketable surplus, are examined. A majority of grain disposal studies have provided only a partial picture of the grain flow within the farm household. This section also clarified the concepts of marketed and marketable surplus to show that part of the debate over marketable sur- plus can be clarified if it is realized that the different alternative definitions may yield different empirical estimates. By themselves, the nagnitude of these estimates can dispell some myths about producers' grain "export" or "import." Ultimately, however, the difficulty of estimating directly a marketable surplus hinges on the availability of data. The next section examines openly the limitations of the data in the current study and the consequent estimation procedures advocated to de- rive the grain flow and other related statements. 6.3 Sources and Limitations of the Data and Estimation Procedures The grain flow and other related statements are accounting identi- ties, and thus, may be marred by accounting errors. In similar cases, accounting errors are usually checked by matching the observed figure of a "balancing item“ against its estimated figure derived as a residual. In cash flow statements, for example, ending cash balance serves as a 118 "balancing item" to provide an internal consistency check. Unfortunate- ly, the limitations of the 1978-79 farm survey data will not allow a "balancing item" to serve the purpose of internal consistency checking.5 Herein, we address these limitations and the consequent estimation pro- cedures adopted. But before addressing these issues, a few words about the sample selected for the present analysis are in order. 6.3.1 Sample Selection In order to permit comparison between animal traction and hoe (tradi- tional) farmers while reducing the larger 0RD-MSU disaggregated data base to a manageable size, the present grain flow analysis is focused on zones where animal traction (ANTRAC) practices are used. In those zone, pur- posely selected ANTRAC households in one village are matched against randomly selected traditional households in another village (for detail, see Chapter 4).. However, for the purpose of this analysis, only one ANTRAC village (in place of the three) was selected in the Diabo Zone in order not to overrepresent the ANTRAC farming system vis-a-vis the hoe system (represented by one traditional village). In addition, one household in Ougarou Village was judged "atypical“ and dropped, because of its grain trade dealings being larger than that of all the total sample combined. Also, one household that was dr0pped in Kindi-Kombou Village because its credit transactions were unusually large, obviously 51n the cash flow statement developed on the basis of the farm survey data there is no true "balancing item" either. Partly be design and partly because of difficulties in data collection, beginning and ending cash balances were not measured. Instead, a net cash flow bal- ance is computed and forced into the identity: Net Cash Flow = Sources - Uses. 119 a misrecording.6 The total sample size of 13 villages and 234 house- holds was consequently reduced to 11 villages, with 107 traditional households and 89 ANTRAC households (see Table 6.1). In the sample thus selected, in addition to the breakdown between ANTRAC and traditional farming systems, it is possible to identify vil- lages with known important grain market activities such as Piela and Logobou; villages with easy access to the main road such as Diapangou and Ougarou; villages off the main road such as Diabo (only slightly), Logobou (of difficult access). For purpose of the grain flow analysis, the sample data is also 7 The disaggregated by season of the year into three broad periods. first, from May 1, 1978 to September 3, 1978, corresponds to the pre- harvest and rainy season; the second, from September 4, 1978 to December 31, 1978, corresponds to the harvest season: and the third, from January 1, 1979 to April 30, 1979, corresponds to the after-harvest and dry season. This breakdown was performed primarily to accommodate the fact that more detailed monthly observations were not available for many 61" Village 23, one household had purchased 164,073.6 kgs. for trade purposes and sold back 13l,562.8 kgs. The total volume for all other households of the whole farm survey sample are respectively 25,940.7 kgs. and 18,628.4 kgs. In Village 10, it was misrecorded that one household had paid back 400,320.1 kgs. The total volume for all other households of the whole farm survey sample is 3,833.3 kgs. 7Lassiter (1982, pp. 141-142) broke down the year into four seasons: (1) Workin Season (May 29-August 30); (2) Hungry Season (August 21- October 15); (3) Harvest Season (October l6-January 7); and (4) Dry Season (January 8-May 28). However, the farm survey started May 1, 1978 and ended April 30, 1979. 120 .a~-- .a. .a .aa ._mm_ .aaa_ma.s sac. gas.“ a. «3.3..4. «3.2 use” .mamau u..a> «a. co “.maa «as can .uo_>osoa upau.q»ueu no onaauoa aw. «map—_s e. tonnage a. opoeomaog «coo .ucoo— eo_: a_nacomaocca co «manage o_‘ ome___> cw emanate a. u—ozomaog econ .ouae.umo cuuuan co gun. to» “no; a“ RAN new .m~ .mv moma___> m we "_a cos can: «en a. coon u<¢hz< ecu usn .ocon «.2» :. aoumAm newscay u<¢hz< “camacaoL-co>o ea you gotta c. tonnage «to Augusta» u<¢pz~ .xo>e=m Scam o~1m~¢p "oucaom c_~.n—— coo.mo_ _ho.mo~ sop.oo~ mmm._m moo.wn~ on~.on— «no.mc— n—n.na ono.~o_ . _mo.wm _mm.__ Ax: ace goo:w mocoowwcoo ucoogoo o>ww-auocwzo .Loo» Loo mspcos NP wo mwmon on» :o meow» oucw ooocm>eoo mo: mgucoe cw wswwo .A532mcom oco wow—way moose wwo Low ogamome ouwmoosou o mw ogsmww mew .ocooeoo omm.F mw cowooe on» one Ammou _v ocoocoo mm on Amomoo mv ocoN sogw mw ooze; onwo .Aqu—we apcoov woow: cow moo: .Eocmgom Low RN oco .uoP—ws cow momcoomog mm .Acocuo one go oco cow woos mo: cowaocwamwo ocv powwwe-535mgom cow mmmcoomoe an ego: meocwo .»m>gzm Econ mw-mwmw ”mogoom mmo.m o» mmo.~ Nem.~ omm.~ co eoo.P mNo.~ new. ‘ o, 4N o.H.o Homoeoo mm com: .H.u ucmoeoo mm coo: momou momowww> mo_o;mm=ox o wo Looszz oopoo_om oouoopom opcoocoo cw ouom omgao> cw wo conga: wo coosoz mmoo oaocoom woocc< omocoum wo osww Amouoewomu .mowogomzoz omuoopomv opom moo; omacoum woacc< Eogmcom-uoppw: N.m mpnoh 129 estimating minimum grain consumption and deriving change in stock as a residual. However, this change in stock is easily interpreted if we as- sume that storage losses are negligible; that is, within the margin of statistical error. We will choose this interpretation on the basis of the empirical estimates of farmers' judgment of millet-sorghum losses. Despite all the mentioned shortcomings, it is felt that the pursuit of the grain flow analysis is worthwhile. First, the 1978-79 farm survey is the best data bank set available on Eastern Upper Volta. Second, the grain flow analysis provides a simple, but sound approach to analyzing . food grain marketing behavior. Third, from such an analysis of the farm survey data, many shortcomings can be exposed that should be taken into account in future studies, notably the oncoming proposed food grain mar- keting study by CRED and USAID in Upper Volta. 6.4 Grain Flow Statement 6.4.1 Annual Grain Flow Statement . Table 6.3 presents the annual grain flow statement for the average animal traction and traditional households in the five zones being stud- ied. Overall, the relatively low levels of grain production and exchange flows convey very well the modest level of agricultural development of the average farmer in Eastern Upper Volta. Only a few hundred kilo- grams are dealt with on the average, and though the ANTRAC households deal with more volume than hoe (traditional) counterparts, we must keep in mind that ANTRAC households also have larger families (see again Table 6.1). Grain production, as expected, makes the better part (74 and 84 percent) of the total annual grain inflow into ANTRAC and traditional 130 Table 6.3 Annual Grain Flow and Performance Measures by Zones and Technology (Selected ANTRAC Zones) rig. Ito-s Piela lone . tantaooo zone Looobou lone Diapaanu Zone Ougarou lone Total Sa-ple (Mr-Isa t9 ----—— by household) 481880 1840 ANTRAC 1880 ANTRAC TRAO ANTRAC TRAO ANTRAC TRAD AulkAC TRAO Sources of Grain. Household Purchases 330.3 115.7 132.4 97.3 502.5 257.3 288.0 116.8 920.4 235.5 429.3 179.3 Trade Purchases 0.0 2.1 1.4 0.0 108.6 61.0 639.9 0.0 276.4 54.4 204.5 29.5 Processino Purchases 0.0 0.3 40.7 11.9 82.6 51.0 69.2 21.2 33.8 24.8 45.8 26.2 Total Purchases 330.3 118.1 174.5 109.2 693.7 369.3 997.1 138.0 1,230.6 314.7 679.6 235.0 Gifts Received 2.1 0.0 6.5 1.8 23.8 19.7 .9 4.4 34.8 56.1 13.4 16.9 lorreuinq 0.0 5.6 5.6 9.0 7.4 2.9 0.0 25.1 0.0 5.3 2.6 8.5 Loan Repay-ant 0.0 0.0 0.0 0.0 2.1 102.1 0.0 0.0 75.9 45.2 14.9 41.0 lages , , 0.0 0.0 0.0 0.0 0.0 0.0 0,0 0.0 0.0 2.9 0.0 0.5 Total Barter Inflowb 2.1 5.6 12.1 10.8 33.3 124.7 0.9 29.5 110.7 109.5 30.9 66.9 Total Exchaneo Inflow 332.4 123.7 186.6 120.0 727.0 494.0 998.0 167.5 1,341.3 424.2 710.5 301.9 Production 781.4 597.2 1,493.4 751 3 1.735.3 1,964.7 1,486.8 563.5 4,890.9 4,129.6 2,045.9 1,659.0 70541 500'903 11113.8 720.9 1,680.0 871 3 2,462.3 2,458.7 2,484.8 731.0 6,232.2 4,553.8 2,756.4 1,960.9 eeeteef . Fare Sales 0.4 0.0 27.3 9.4 16.2 144.4 71.2 20.7 150.4 424.0 52.0 123.6 Trade Sales 0.0 0.0 0.0 0.0 107.2 89.9 393.9 0.0 160.9 54.4 132.1 38.5 'PDCISSI99 $414! 0.0 0.3 40.7 11.9 82.6 51.0 69.2 21.2 33.8 24.8 45.8 26.2 Total Sales 0.4 0.3 68.0 21.3 206.0 285.3 534.3 41.9 345 1 503.2 229.9 188.3 Gifts Given 1.0 1.0 49.3 17.2 69.0 102.6 40.0 91.0 123.4 83.8 55.7 66.0 loans Extended 0.0 0.0 0.6 0.0 18.0 10.4 8.7 0.0 15.3 1.1 8.4 3.6 Debt Revlywtn¢$ 0.0 0.0 0.6 9.0 7.4 2.9 0.0 1.1 0.0 48.7 1.6 10.8 Hired Labor . 0.0 0.0 1.1 0.0 0.0 0.2 2.1 8.8 1.8 0.0 0.9 1.5 Hort ‘lnvimioM' ' 0.0 0.0 2.8 0.0 8.7 3.9 3.1 9.7 22.6 15.9 7.3 5.6 Total Barter Outflow 9 1.0 1.0 54.4 26.2 103.1 120.0 53. 110.6 163.1 149.5 73.9 67.5 10541 51¢04094 Outflow 1.4 1.3 122.4 47.5 309.1 405.3 588.2 152.5 508.2 652.7 303.8 275.8 Seed 23,7 30.6 47.3 22.2 42.3 42.2 53.4 29.6 49.0 33.2 43.1 33.2 Feed 6.3 3.3 22.8 13.1 180.4 33.4 125.0 78. 173.1 17.1 100.7 29.7 15ta1 Fare Use" 30.0 33.9 70.1 35.3 222.7 75.6 178.4 107.6 222.1 50.3 143.8 62.9 Consumption (estimated) 1,686.2 1,095.1 2,131.9 1}182.3 2,122.3 1,978.5 2,131.9 1,395.5 1,826.5 1,308,2 1,901.5 1,485.1 ParIpFa-ily Use 13716.2 1,129.0 2,202.0 1,217.6 2,345.0 2,054.1 2,310.3 1,503.1 2,048.6 1,358.5 2,125.3 1,548.0 Total Uses ‘ 1,717.6 1,130.3 2,324.4 1,265.1 2,654.1 2,459.4 2,898.5 1,655.6 2,556.8 2,011.2 2,429.1 1,823.8 5¢°¢k C0409. (residual) -603.8 -409.4 -644.4 -393.8 -191.8 -.7 -413.7 -924.6 3,675.4 2,542.6 327.3 137.1 Performance Phasure (I) ‘ Para Sales/Household ‘ ' ' Purchases .1 0.0 20.6 9.7 3.2 56.1 24.7 17.7 16.3 180.0 12.1 68.9 Total Sales/Total . Purchases 0.1 .2 39.0 19.5 29.7 77.2 53.6 30.4 28.0 159.9 33.8 00.1 larter Inflow/Household _ Purchases 0.6 4,8 9.1 11.0 6.6 48.5 0.3 25.2 12.0 46.5 7.2 37.3 Barter Inflow/Total - rce 0.2 .8 .7 1.2 1.4 5.1 0.0 4.0 1.8 2.4 1.1 3.4 Exchange Inflow/Total Source 29.8 17.2 11.1 13.8 29.5 20.1 40.2 22.9 21.5 9.3 25.8 15.4 Peru Sales/Production 0.1 0.0 1.8 1.2 .9 7.3 4.8 3,7 3.1 10.3 2.5 7.4 Fare-Family Use/ ' Production 219.6 189.0 147.4 162.1 135.1 104.6 155.4 266.7 41.9 32.3 103.9 93.3 larter OutflouVFare . Sales 250.0 - 199.3 278.7 636.4 83.1 75.7 534.3 108.4 35.3 142.1 70.8 Total Sales/Exchange tflew 28.6 23.1 55.6 44.8 66.6 70.4 90.8 27.5 67.9 77.1 75.7 68.3 Exchange Outflow/Total . Uses .1 .1 5.3 3.8 11.6 16.5 20.3 9.2 19.9 32.4 12.5 15.1 Source: 1978-79 Fare Survey. ‘Ieginning stock is not observed. ‘smar inflou - gifts received 1 berrouing . 161-1- repay-ht 1 mos (11- 1111-111. cExchange outflow - total barter inflow 0 tdtal purchases. ‘Endino stock is not observed. 'Ieier to traditional work parties of a reciprocal nature in which the “invited' are rewarded in food, sorghu- beer, and cola nuts. 'See a and 6. 9See b and c. “Total m- use 1 seed 1 feed. 'Stock change - total source . total uses. It is cuuivalent to ending stock - buuinning stock. 131 (hoe) households. (Beginning inventory, however, was not estimated) Similarly, grain consumption is the most important component (81 percent) of the uses of grain by ANTRAC and traditional households. (Here also, ending inventory was not estimated). Conversely, the exchange of grain, acquired or disposed of, plays a relatively smaller role for the average ANTRAC or traditional household (see Figures 6.3 and 6.4). Table 6.3 shows that for the average ANTRAC and hoe farmers barter plays a very small part in grain acquisition compared to household pur- chases let alone total purchases (see also Figure 6.3). But grain given out in similar barter form plays a relatively larger part in grain dis- posals compared to total sales, let alone farm sales (see also Figure 6.4). For example, the average ANTRAC household gives up more grain in barter form (gifts, credit, and labor payments) than through farm sales. For the average traditional farmer, total grain bartered out represents 70 percent of the farm sales. It is worthy of note, as Table 6.3 shows, that the amount of grain used by households may be larger than the year's total grain production. This is because the extra grain is acquired through some exchange (mone- tary or barter) and/or comes from stock carryovers. As already noted, grain carryovers were not observed, but the change in stock can be esti- mated on the basis of the grain flow statement. In the zones being studies, the results of the grain flow analysis 14 over the shows that the average household would have accumulated grain year if consumption were to be kept at the minimum of 175 kg. per capita per year. In particular, the ANTRAC household would have accumulated 14It should be kept in mind that stock changes include also storage losses (here assumed negligible). 132 mcoEcou mo: oco oo>c=m Econ mn-mmmp "mocoom Aoooswocw «oz mxooum co>ozccoov , Aoooz—o=H uoz mxuoum co>oaccouv newscod mo: cow powpwz one Eozmcom wo moocoom mcoscod uc:m stow mN-mwmp Hoocoom zooospoem ooz meooom ocwocuv zooospo=H ooz meooom ocwocmv meoELod mo: zn powpwz oco Ezzmcom wo mom: mcoacod ue~= cuppa xuoun m:_u:o umuumnocan .Aacmsouuuu zap» c—ugm oomv «mango xuoum Auv m? “mm>cn: ogovua xuoum acwcc—mog .umo>ca; axe: o» Lops: nouo_aou on o» vessmma ago mxuOHm oucvmn .»m>g:m scam m~-mnm— “mugaom 73? 5.82. odmm- .63.. N43 mama- 933. 932 «Juno- mdmm- 92o <80. #8? N6; 98m- baccafi 3on 9.33” fimouc o :3. «.50 o 93¢ 93:. o TN: 92o o mg: Nd; o savouuonoct . . gooom we've“ fimome o mama «.8... o mag 93: o 1&8 920 o 7,8 Nd; o 993 283 9:538 My... oco~ accomac acoN :oa:aaa.o ucoN nonomoa ozoN avenues; acoN a_o_g mama" A.mxv pmm>caz “cocczm op cowca gooum oemN we co_uaszmm< any Lows: Amm Anzv xuoum mzwwcu cmuuwwocm m.o mpnmp 143 the zones being studied. 0n the basis of a sensitivity analysis using alternative grain consumption requirements and the changes in sign of the stock change and the marketable surplus as stopping points, we tenta- tively set the actual annual grain consumption in the zones studied between 165 kg. and l90 kg. (see Appendix H). This transaltes into a daily grain consumption per capita between .45 kg. and .52 kg., which is low. It would not then come as a surprise that there is hardly any mar- keted or marketable surplus as presented below. 6.5 Marketed and Marketable Surplus Grain marketed and marketable surplus are key components of the grain flow statement rearranged in a different way to highlight the ability of producers to supply outside consumers while accounting for the farm household grain requirements. Marketed surplus is defined here as total sales less total purchases, while marketable surplus is the sum of marketed surplus, net barter, loss, and change in stock (ending less beginning inventory). In Table 6.6 the annual marketed and marketable surplus for the average ANTRAC household is compared to that of the traditional house- hold in the selected ANTRAC zones. The picture that emerges throughout the zones shows that both ANTRAC and traditional households were mar- keted and marketable deficit producers in l978-79, except in a few vil- lages. The average traditional household in Ougarou is the only example of marketed surplus producer in the study zones. The stock build ups of households in the Ougarou Zone and the traditional villages in Logobou help make them marketable surplus producers. But the marketable surplus is only significant in Ougarou. I44 .Auv + Au. n Avv + any + An“ u ma—nsan o—aouuxgmxo .mom: - mougaom u agouco>cv u:.=:—moa u xgoucm>=v mcfiuco - «mango xuOHmu .Aav + on u zapmcy macasuxm _ouoa - zo—muao amcasuxu _ouou u zapuuao mocaguxo umzu .3opmcv caucan page» . xopmuao copcua _ouou u sopwuao caucus umzn .mommgueaa _auou - ma_am pauou u mm_um um: n m=_ae:m nouoxgmxm .xo>g:m econ m~-m~m_ “oucaom o.___ e.mh- _.ps~.~ m.~¢m.m o.mmm- m.m~m- ¢.mm-. ~.moo- «.moo- o.mo~- m.me- m.cmm- mmapqcam apnouoxcu: _.~m_ m.-m m.~em.~ «.mno.m o.c~m- ~.m~e- 5.- .m._mp- m.mmm- c.¢co- «.moe- m.moo- . vomcagu xuobm —.o~- ~.mo¢- m.m- p.mmm- onmp- «.mce- ~.mm4 a.~_¢- m.-- ~.¢o- e.-_- o._mm- axe—maze mmcmzoxu uuz o.o~ o.m¢ c.oe e.~m p.—m . o.mm ~.o- m.mo e.m_ «.me o.¢- _._- pro—uuao smegma uoz ~.me- u.oeeu m.mm_ m.mmm- ..om- «.mce- o.em- o.~me- m.~m- m.mop- m.~_p- m.m~m- omapqeam umumxcmx mo: u<¢hz< mo: u<¢hz< mo: u<¢hz< . we: ucmm -no mo gmnE=c F_m2m mg» .gm>o3o: .emoo. u wucmuww_cm_m .vmmw.~_ u mgmzcmuwgu .mcow¢m>gmmno amp co ummmm .mmom. u mucmuwmvcmwm .MFNm. u m m>o=< .mcowum>gmmno amp co ummmmM .mmmm. u mucmu_$wcmwm .momp. u m m>oc< .mcowum>gmmao cop co commmc .mooo. u mucmuwmwcm_m .opoo.m u u m>oc< .mcovpm>cmmno «mp co ummmmu .oooo. u mucmuwywcmwm .opmm.w u m m>o=< .mcowum>cmmao cap co ummamn .mmmoggzg m:_mmmuoca can mung» cow mummgucsq ucm mopmm amp—Fe use Eszmcom muspuxmm .>m>c:m Scum msumnmp ”mucaom m_ «N c . N.o~_._o_ ,.o_ m.mmm o.onN mammzueza u_o;am=o; ucm mmFMm Exam spam m m N m.~mm.om. N.m ¢.NNN N.oo~N spec mm_mm scam m. me am. m.mmo.m__ .w.m m.mm¢ o.amm_ »_=o mammgueaa v.0;amzoz m mp o. 7 m.mmo.o._ N.m . N.mmm m.oom_ m=o_uumm=mep oz mprbcmza msoucu «_Pbcmma mamauav capo; Aa;\.mxv A.muxv vgm mpuvwz um— mEoucH -mmao: uvpmw> ocowguau mmccmuuom upogmmaoz cma mgma Ezsmcom -ocm Essa :o_uumm=mch wmmmmmpu m5ou=~ n_o;mm:oz mmmco>< name mo vcm -Lom new cpuco pmauu< z. mc_o;am=o= do canaaz canszz ba.__z ba___z mmmcm>< mmmgm>< moccm>< mccmuuma :owuummzmch :ngw ow mcwcgouu< m50u=~ u_o;mm:ozuscmm new u—mw> :Pmcw .cowuuauocm :Pmcu we copusnwcum_o m.o m—nwh 151 .mmmcoammc .mu—osmmzo; om. co ummmmc .uxmu as» cw mmpnmu cacao cw mmcamw» ;u_3 xucmamgumwu mg» com uczouum mcoaca m=_c::omu .mmzpm> op m>wmemc .mmx =_ momma m:_mmwe An umuzmmcpcv mp .cm>mzo; .xocmamLUmwu u:m_pm.< memccm oumcmc .wwc may :_ muposmmzo; mo conga: msmm mg» mw>wm Agwucmzc mo ummum:_ mwzpm> mo mwmmn as» so cowpmu_wwumcum mshn .mmmoacaa mcwmmmuoca ucm mung“ Low mommcucza ucm mmme umppws can aszmgom muapuxum .»m>c:m Exam mm-mnmp ”mugzom m.mmp m.m~¢ m.mmp o.Nm ~o_ mm u_muo» “.mPN ¢.omm m.m¢m N.mom mm Pm mammgucaa vpozmmso; new mmpmm Scam guom o o m.mom o.ow m. e x—co mmpmm Scam N.omm _.N~m o o Ne mm >_:o mommgugaa upocmmsoz o o o o o_ N_ . mco_uummcmgu oz mo: u< can «Pom mmmcm>< a do cansaz mccmpumm cowpummcmgh :_mcu vac amopocsuop mcwscmm new: cowumpmm :P apnea yo mmmmgugza uposmmao: ucm mopmm ELM; m.o anmh 152 Added to the fact that the average number of members does not vary significantly across the four groups, these results lead one to believe that farmers without grain transactions were not self-sufficient as might have been thought. But at the same time, there is no indication that the poorest and the smallest grain producing farmers are the ones that are forced into sales and then repurchases. Table 6.9 lends support to the previous observation that on aver- age ANTRAC households bought more grain for farm-family use and sold less farm grain than their hoe counterparts. But an attempt to charac- terize in the same way the sale and purchase behavior on the basis of income distribution is more difficult given the small number of observa- tions in many cases (see Table 6.10). The results of this attempt, how- ever, suggest that across the transaction groups the volume of grain sold or bought increases with the household income bracket. This lends support to the hypothesis that as total household increases, farm sales increase (or that farm sales increase constributes to total income) and household purchases increase too. Looking particularly at the group of households who sold and re- purchased grain, it can be seen in Table 6.11 that overall almost 80 percent of the households had repurchased grain in excess of their farm sales. But, on average, hoe households sold more grain than they re- purchased contrary to their ANTRAC counterpart. On the other hand, 153 .mmmcoammc .mcFogmmaog amp co vmmmma .mmmoaeza mcwmmmuoca use mumc5 c05 mommcucaa use mmpmm Ezgmcom new 5m—pws mvzpuxmm .>m>c:m Esme mnuwnmp "mucsom P.m¢¢ N.NmN m.No, «.moN N.m¢ m.__ Ne mm Ne a_moo» ¢.NN¢ o.NmN m.P__ m.on N..o_ m.mm my «N m mamasuesa u_o;am=o; use mm_am scam guom o o o ©.NmN m.mm ¢.¢o_ N a N »_=o ma_mm seam _.mmN m.mmm ..NqN o o o m_ me mN N_=o mamagueza upocmmaoz o o o o o o m m_ o_ m=o_puam=aep oz 25.55 2.85 m _ .5330 353:3 9:85 m _ 5.330 m F 5.53 9:85 3.525 gem mpuu_z gem ucm m_uu_z pm_ ucm byway: pm. amecapuma :o_5umm:mch :_ upocmmzoz emu A.mxv :5 uposmmao: cog A.mxv mummmpu msoucH uposmmao: :_mcu Fm=5u< «monogaa apnea mmmcm>< m_mm :wmgo mmmcm>< an? mn_o;mm=oz mo smasaz macm55mm :owuummcmc» :wocu can co_5=awg5mpo msoucH uponmmaoz :55: cow5m—mm cw :wmco yo mmmmgugzm upogmmzoz new mmpmm scam o—.o anMH 154, Table 6.ll Relationship Between Farming Technology and Sale/Purchase Volume Differential for Households with Grain Sales and Repurchases . Number of Average Volume Ratio of F . . Sales to ngge- Householdsa Differential (kgs.)d h°1d P”r°“ases ANTRAC TRAD ANTRAC TRAD Greater than 1 6 l4 (+)182.l (+)506.6 Less than 1 l4 l5 (-)559.8 (-)2l2.6 Total 21” 29 (-)321.2c (+)l34.6 Source: l978-79 Farm Survey. aBased on 50 households' responses in that particular transaction group. bOne household had sold and repurchased the same amount. cBased on the total number of ANTRAC households (2l). . dVolume of grain sales exceed that of purchases if (+). 155 Table 6.12 shows that in this same group only farmers in the third in- come quantile had farm sales in excess of their household purchases.22 These results combined with the large number of exclusive buyers, pro- vides further insight in the grain deficit position of the farmers over the selected ANTRAC zones. Not only did producers buy more grain than they sold as shown by the grain flow situation, but also more producers bought or had larger purchases than producers who sold grain. 6.6.2 Timing of Farm Sales and Household Purchases The timing of farm sales-household purchases are readily interpret- 23 ed by reading off the cumulative percentage figures plotted against the time axis of 13 four-week periods ("months"). Evidence of large sales at harvest (distress sales) will be shown by cumulative percent 24 sales larger in period 5 through 9 than in other periods, and the 22It should be stressed that the results in Table 6.12 are based on a sma1ler number of households' responses than that of Table 6.11. Be- cause of the small sample size, a problem of degrees of freedom arises which does not permit a meaningful analysis which would be used in con- junction with farming technology (ANTRAC-hoe households), income classes, and grain transaction patterns as independent variables. As already shown in Table 6.1, on average, ANTRAC households have larger total in- come than traditional households, but preliminary cross-tabulation re- sults (Chi-square==l.5267 and significance =.4661) do not suggest any significant relationship between farming technology and the distribution of households in income classes. 23The cumulative percent figure is with respect to the total year's transactions and not production. Such a measure avoids the pitfall of assuming that the observed pre-1978 harvest sales are a proxy for the post-l978 harvest sales that were not observed. Yet, such a measure has a merit of its own, for it avoids the pitfall of considering that all farm sales are from the current year's crop. As already indicated, in one marketing year, farm sales may also generate from stock carryovers. 24In fact, actual harvest (of early millet) starts about the end of August or early September. By including the whole of period 5 (weeks August 21 to September 17), we increase the likelihood of harvest dis- tress sales in the interpretation of the timing of sales. 156 .mmmggueza wo 5gcu gmmuxm mmwgm wo mE=_o> ww A+vm .m_w5egac ms5 cw on mg_o;mm:o; wo emaazc Pguou co gmmgmg .opaguw—aag 5ozu .5caosg msgm mc5 gmmgzoeaame gcg gpom gm; gpozmmao; anon .agoem cowuugmcgeu eg_:Uw5egq 5gg5 cw mmmcogmoe .mgwogmmzo; mg co gomgmg .xm>e:m seam mmuwnmp "muesom m.mge+v m.gmee-v gg.eme-v m_ gN gg _gggw g.ggge-v N.Ngme-v g.mge-e m g_ m _ egeg wage m.gmge+v m.me_e+v g- o_ o. o F egeg egggaeg mpw5egzo «sou:_ mpw5egao mpwuegzo mEou:H mpw5egzo gem «_gng gm. gem g_gng 5m_ mgmgegega g_ge -mmaoz o5 mmpgm gmmmmgpu mEoo:H upozmmzo: see; we owuga we.ge v Fgwgegegwweg gag_g> aggegeg cw mgFosmmzo: wo emnsaz mmmggueaamm gcg mmwgm :wgem 55w: mgpogmmzo: eow —gw5:wemwwwo N_.© epoch manpo> magnueam\m_gm ucg cowuanweumwo mEoo:H g_o;mm:o: :mmzumm awsmcow5ngm 157 curve would be s-shaped. Conversely, evidence of concentrated pur- chases in the hungry season would be shown by cumulative percent pur- chases larger in periods 1 through 5 than in other periods, and the curve would likely show a plateau in periods 5 through 9 indicating no purchases in the harvest period. But if the curve were to lie on the 45-degree line, the timing of transactions would be interpreted as steady and regular throughout the year. When we compare and contrast the behavior of ANTRAC and hoe house- holds, the evidence in Table 6.13 does not suggest any harvest distress sales for those farmers who sold and repurchased; only 22 and 35 percent of the annual farm grain was sold during harvest (periods 5 through 9), while 63 and 57 percent of the grain was sold after harvest. For house- holds who had had sales only, the pattern of timing of hoe households (cannot be meaningfully compared to that of ANTRAC because of the small number of observations for the latter. (Traditional households had heavy sales, 45 percent of the annual total, in the harvest period, but also important sales, 32 percent, in the preharvest period.) As for pur- chases, the results show that traditional households had heavy purchases, 61 to 68 percent in the hungry season, while ANTRAC had 41 to 54 per- cent in the same period. In sum, when the timing of grain transactions of ANTRAC households can be meaningfully compared to that of hoe house- holds, Table 6.13 and also Figure 6.625 show that the major significant difference is with respect to the pattern of purchases. Otherwise, there is no strong evidence of harvest distress sales. 25In order to plot the cumulative percent curve, it will be assumed that all transactions are made at the end, rather than midway or at the beginning of the period. 158 .momoeueoega e. oeu ..meueoeev ago.eaq mewveoemoeeouo o.e2gm umea.eu mg vouaeeou mouwee augewumo o» vogoue no: on.“ opeEgm omen. new .o—esom xu>eam Eegw m~-mwm. o.oex we» eo gomgm .Amangeueae e.gem g.oeou:oev ezg .Amo.gm e.gem seowv mwu ..seueoeev go.eoe xomz-e eugu eow meageo>~ a .noeou ueem ecow o~-m~m. "uueeom m..m . N.mg N..g g.mg o.mn ..N. ..om ..om a... ..Nm o.gg N.Nm m.gm g.az. g.ge mog.ee g... ..og ..NN ..mn g.gn ..NN N.NN ...g ..nm m.nm N.og ..mm ..Ng g.me. gge.oug¢ mgu.ee amouwee cage gnaw N.g g.m g.m o.N n.o ..N N.N g.m o.n. e.N «.0. N... a... gee. 05. g; .5 .6 N6 a.. Na ea ma .6 .g g; mdN 92:2 Nmeo manoeueaa «mac: m.m N.. ..m N.N N.g. ..g o.o a... m.e o.o e.. a... N.g. gge. ...... o... .3 ea c... o... o... 9.. a... a... 0... ”.3 g... 22:2 .».ec mo.gm eeow g.n N.. m.N ..o ..g o.g g.. N." m.g a.. m.mN ..N. g.g. gee. g... .... 3 n... SN 92 ....N o... «N .6 mg. 92 ..N. USE: momgeueem v.0eomeo: N.ON N.gN o.m N... N... m.¢ m.. ..N n.o a." N.e g.g g.. gae. g.o g.gm g.m g.N. g.g m.m m.e n.g g.N. ..N o.e n.m g.g g<¢.z< no.am seam mongeueem g_oemmeox geg mwpgm seem eaom u.... .N.. .... .o.. .m. .g. ... .g. .m. .g. .m. .N. ... a:eu...e mNm. a... NNN. ..N. ..N. gem. gem. NNN. meg. gem. gem. gem. gem. no.56gmgge. cm ee< — ee< v ea: c new 5 egg 0. uoo ~— >oz m. ago N. qum ow oe< n~ A.33 cw oeaw mm Ag: ewgeu pgeuue -N ea< -m ege -m age -e egg -.. gag -N. eoz -g. egg -g. .aam -.N gag -.N ..gg -gN «egg ..N e.: -. age Aux ewe woumm mmuwee m5go Seem mmgem>< neg .Amse.o> —g=ee< wo omguemoeme ».e5eozv xeommugu eowuugmegew .g:5u< an mmmgeueee g.oemm=o: gem mm.gm Eege ewgeo wo mews.» m_.m mwnmh 159 meemuuge eowuogmegew ewgeu geg xmopoeeumw an mmmgeueee gpoemmeo: gee mmpgm see; we mewsww g.g gegg.e $328.33. new mmpgm ewgeo 5.2.. mupoemmeozv mummeueee e.oewmeo: gem mmwmm see; we mewaww go.m meemwe 950— ....m .2le I «I; _._ ».,.L "_naa_z..7w «..mxlz:cm_ -.0..—292. I hmw>mu_' 160 meemuuge eowuogmegew ewmew gee zooFoeeumw a; mmmgeueee epoemmeo: gee mm.gm gene we mewsww g.g gegg.e Ax.eo mmmgeueee ewgew gpoemmeo: epwz weasegev mwmgeueee epoemmeo: wo meweww gg.g gegg.e arm. .cm :5: 1 D21.— wzn "...aa—zua r._mx-m:a: ..0..—23:; - .meeege - m. N. .. o. a g e e m g m N . a _ i. . . . . _ . _ . Tr». .. a 1 ON .x. I O. .. - om -.:.:.. megeeuege we: ..... x on ..x. memeeeeee gee.ze In: oo— u:z::..au:r—~>uu LUMUUZP 161 meew55ge eow5ugmegew ewgeu gee zmo.oeeumw an mammeoeee epoemmeoz veg mmpgm seem wo mewsww c.o meemwe AAFeo mopgm ewgeu seem euwz mgpoemmeozv mmpgm see; we mewsww uo.o weemww as... .3” fixe: 3......— »c: Luce—z... ...:2. 13:. 3.320;; - .meeege - m. e. .. g. m g g e a g. m N . o ---=;_--.1 _ _ _ . e. . ...-au...|..-.-:.| w . a..- a eN e. - eg .... ....... - cg neseg eee ..... 1 eg meeem Nee.ze.n|. Do. UDI::-0C’—-—>U LMXUMJZI— 162 The timing of sales and purchases can then be interpreted with reference only to grain transaction patterns ignoring farm technology (and gaining more degrees of freedom). (See Figure 6.7.) There is hardly any difference between the timing of sales of the households who sold only and the timing of purchases of households who purchased only. Both groups tend to have a regular pattern of transactions throughout the year. It is only for households with both sales and purchases that the data suggest heavy purchases in the hungry season, but for the same group, farm sales are again mainly concentrated after harvest. when we contrast the behavior of households in the three income classes (discarding the grain transaction patterns), the evidence in Table 6.14 and Figure 6.8 show a somewhat regular pattern of household purchases for all three classes of income throughout the year. For ex- ample, by mid-October households in all income classes have purchased about 55 percent of the total amount they bought that year. As for sales, there is only a slight indication that middle income households had heavy sales (42 percent of the year's total) in the harvest period, but even then, an equal amount of grain was sold at a regular pace after (harvest. Here again, there is no strong indication of harvest distress sales. 6.7 Farm Gate Prices The next question that comes to mind is about the impact of market prices on the timing of farm sales and household purchases. Unfortunate- ly, there is no reliable market price series for the Eastern Region for the time coverage of the 1978-79 farm survey. An attempt made in 1978- 79 to secure monthly market prices at important marketplaces throughout 163 mmweemmugo eowuugmegew ewgeu >5 mmmgeueee neg mmpgm e.gew mmmgeueee ewgew gpoemmzo: wo mewsww n~.o wezmww as. .an :5: - 92.... t... 2.253.... 5.3.2.5.. ......za... ..N._.e.eepeegm~.. e.. . .115. r elipllei .. 29.5.2: ...... ......» ...... 2:5... :2... 335...... ...... 2:5...“ l.l... nnnnnn a ..‘V 9' a: . 3:. ..g gegg.. U::=JC—->U manna—'8'- mm.gm e.gem Eege wo mews.w gm.o mezmwe .an ._¢w¢ . ope... we: ..m==_¢.k £-ux-xaab_ .03—862. aha. ...N...e...g.ggg.~.e — p — p p p — _ - p p p .. oooooooo O ..... -..N .... \\I am— -8 335...: ...... 2...... ...... 2.5.... ..... l\ 2.... 2.3 .....S ...... were... II . .III.||1I.11|.-.I.IIII.. . 3.: u:8=a¢~-’U uuxuuxh .moaofieee... e. 2... $322.... «37.8 niecefiteeeg .oeae a. ee.uuemegeu .oeuea we mo.eeuouau xa exeexgeea ee .momeu we emcee: ..a:« use ea me.xeu .momeeemee .me.eeomeee mm. ee eomome .meeeeo me.ge=ee e5 me.zo ee. e. an egg uee age ...ewg .xe>e=m snow mu-mmo. "oueaem 1(54 ..a ... N." ..N e... e... ... ... g.g e.. e... ..N ..gN e...e.=g gen N.. ..g. ... ... N.o. e.e . e.. ..N e.. ... e... ..N. g... usage. e.gg.e a. e. 9. t. e. 1N QN. . fig a. a. 9.. me e. ezeaaees umomeeeeeu e—eeomee: ..N. e..N ..N g... m... ... e. ... o.N ..N ... ... ... o...e.ga gen g... ..e. N.. g... m.g. ..o ..N ..N o... ..N g.N g.e .. usage. e.gg.e e.g. e.. g.o .... ... g.o. N.N .. g.. o.o e.g N.e. g.° a...em=a ... ewe—em see. g.... .N.. .... .g.. .0. .e. ... ... ... ... .n. .N. ... a... a... a... a... a... a... g... a... e... a... a... a... a... mg....e.ge usage. ea g. ee< . ee< . ea: . ea. . gag g. ego N. ,0: m. .eg .. be». ON one NN ..gg .N eegg eN .e: . “engeeege gee «e... -N eeg -m ea: -m .e. -e gag -.. gee -.. eoz -.. .60 -e. ..em -.N as. -.N e.=g -gN «egg -.N .g: -. ea: ememme.u eseee. e.eeem:e= xe memeeeeee ePeeemee: gee me.gm eweee Seem we 5eeeeee >.e5eez ¢—.o m—eeh 165 memme.u eseee. an memeeeeee eeg mm.gm e.eee we mews.. m.e eeemwe .xe>e=m see. munmwm. .eeeeem memmepu eseee. an memeeeeee ewgeo epeeemeez memme.u maeee. xe we.em eweew See. we mews.» ae.e gegg.. ge.g geeg.. arm. .on :52 u arm. u. 2.: Inna—mum Kau3-¢q5m. 9:29... a. Na .— a. m h m w v m ~ _ 0 er... .... .....ee - e2... .. ......r “39:“... .....-ge: ....zee. o . ; . - . . . . 7 L . L . . . e e .. .... .. ...... .. e . e .. g ... r . e w . h— L. 1 .e x 8 x. _ . 1 ea 6.. \. 3:53 ”:82. 8:59... . .8. £8... 3829.6 8 .. .8\ 3:53.. $.82. 55:1. .....u . ....eeee 8.5....» I A “=82. 38.55.. 8 u._.uc:o wm¢~uclla U::DJ¢b->U fiUKUUZh ‘ I 93==JCh-’U ..UCUURP- 166 the E-ORD (Mercuriales des prix) was not successful because of the poor performance of E-ORD extension agents and the difficulty of gathering accurate prices without actually measuring the weight of the nonstandard- ized devices used by participants. Farm gate prices, however, can be derived from the 1978-79 farm survey. Because the values of transactions entered in computerized files are net of marketing costs (handling and transportation), the unit prices obtained are farm gate prices. Given that there are but a few instances of cash outlays to pay for marketing services by producers, these farm gate prices may approximate market prices in most cases. One difficulty arising from the derivation of the farm gate prices is the possible measurement error in the kilogram conversion for millet and sorghum. For this reason, the farm gate price may reasonably be ac- curate on an annual basis (previous analysis umxia benchmark of annual price received of FCFA 45-5 per kg.), but not so much so on a more dis- aggregated monthly or zonal basis. In particular, the smaller the num- ber of transactions, the stronger the possible bias of price variability due to the volume conversion. To circumvent this difficulty, monthly farm gate prices are estimated as trimmed sample averages26 (see Table 6.13). In addition, in Table 6.13, trimmed sample averages were computed on the basis of grain transactions for the entire 1978-79 farm survey sample (i.e., 480 households) in order to add more degrees of freedom to 26This measure "trims off" all observations below the first quartile and all observations above the third quartile. The trimming makes the measure less sensitive to the existence of a few unusually large or small observations (see G. Bhattacharyya and R. Johnson, 1977, p. 31). 167 the estimates. (Trimmed sample averages on the basis of the reduced sample of 196 households show outliers caused by the smaller number of cases which serve for the computation, see Appendix K). In Figure 6.9, the plot of farm gate prices received is compared to that of farm gate prices paid. It can be seen that average farm gate prices received in 1978-79 were at a level below the official price of FCFA 40 per kg. from mid October to early April. In the preharvest season, however, the level of 27 The lowest level is reached in October- prices was quite high. November, but Table 6.13 shows that producers are not marketing large quantities then, and that grain is mostly sold after January and some- times a sizeable proportion in the preharvest period. Thus, some pro- ducers seem to be able to avoid the very low prices associated with the harvest. However, this price pattern works against the group of households with both sales and purchases since they mostly bought before harvest and sold after harvest. Overall, prices paid are contained in a narrow band (FCFA 35 to FCFA 60) , and thus, shows somewhat less variability than prices received. Based on Table 6.13 the coefficient of variation of prices paid is only 12.6 percent, whereas that of prices received is 21.7 percent. This may be due to the localized nature of grain purchases, but also the fact that food grain provided by aid agencies was sold at low nominal prices and contributed to dampen the variability of prices paid. 27This general pattern follows that of market prices in other parts of Upper Volta reported by J. Sherman (1981) and ICRISAT (A. Bonkian, 1982: see Appendix L). AeeEEwewv meuwee eueu Seem peppwz eee seemeem m.e eeemwe .xe>eem see. mmummmF .eeeeem men... .om .chmzmhm. .. era: ..moonmm :guznmno.. mrwzoze . #323: u .m a; m m N. m m. w m N w . . _ . _ _ _ _ . _ 0'1 Cd c-o _- 168 . .om LULCE LINK XO geeezeeee eeeeemeee ..... meeee eme..lla . .-:.: om 169 However, despite the overall higher level of prices paid relative to prices received, the two series of prices appear to move together. In particular, prices paid are not significantly different from prices received in the preharvest season. This suggests that prices paid may bear some relationship to prices received, which is consistent with the picture presented by Figure 6.7 above in which, for a number of pro- ducers at least, the timing of farm sales may be similar to that of household purchases. 6.8 Cash Flow The lack of marketed and marketable surplus in most of the zones being studied, and also the large number of households who bought food grain without selling any in l978-79 prompts the question as to how the households finance their food and other cash needs, given the fact that cash income actually contributes less than 20 percent to total household income.28 To shed some light on this matter, we analyze the cash flow in relation to the level of technology, the zonal difference, and the grouping of households with respect to actual grain transaction patterns (in farm grain sales and household grain purchases). The cash flow statements presented are shown in a light consistent with the derivation of the grain flow statement. As in the grain flow statement, the cash flow statement does not estimate any beginning and ending balances; instead, the net cash is estimated as a residual just as is the change in stock. Crop sales refer to sale of sorghum-millet and all other crops (e.g., peanuts, bambara nuts, cotton, etc.) grown 28Preliminary analysis shows that based on 189 households' responses, cash income accounts for 14 percent of total household income on average (see also Barrett, et al., 1981). 170 by the household. In addition, net cropping revenue is defined as the (sum of all crop sales and cropping related revenues (custom plowing, for example) less the sum of crop input costs and related service expendi- tures (costs of custom plowing, for example). And net cropping cash surplus, defined as net cropping cash revenues less foodstuff expendi- tures, measures the capacity of the household to finance its additional cash food and other expenditures out of cropping revenues. Table 6.15 presents the annual cash flow for households stratified with respect to technology, zone, and grain transaction patterns in the zones being studied. Sources of cash are derived mainly from grain and other crop sales, livestock, credit. and nonfarm revenues (off-farm em— ployment, artisan enterprises, trade in consumer goods, and in a few cases military pension or heritage). It would appear that a large part of total crop sales in the source of cash is not associated with large sources of cash or positive net cash flow. In fact, Table 6.15 sug- gests that regardless of how one stratifies households in the zones be- ing studied, crop sales are not the first contributor to cash revenues (except only marginally in Ougarou Zone). The average household relies more on livestock and nonfarm enterprises to generate cash income. The above results help explain why the average households in the largest and the smallest grain producing zones are the only ones with negative net cash flow, and why they have also the lowest level of sources of cash. In fact, as shown by the negative net cropping cash surplus, cropping revenues are not adequate to provide cash to purchase food in most instances. The above results show also that animal trac- tion technology is net yet well developed to pay for itself through cropping revenues. 171 .nomeeemae .ue.eeoaeoe co. co venues .aa>eem aeaw oscoso— "oueeem w.m— ..N ¢.m. ... o.o~ o.~. o.m~ ~.o «.m— o.m m.mo omexeee. ..eN w.- m.om o.nn o.o~ o.mv o..~ m.nm o.m~ m.~n o.v. oueeemxeeeo>m¢ _ge.e< m.n~. ..o. ¢.v m.m «.mp o.m e.m~ o.~ o.m— m.n e.g oueeemxo.om an. meeemaaz uuegaeeeeem mom.— o«~.m no~.op- me- onm.—- ooe.o—- ~.~.~—. ~mm.-—u poo- m—_.ou .e..n.- ue—eeem ae.eeeeu uez. mmm.~a pn~.m moo.v~ oee.om vuo.~— om..a~ ¢——.o—. —m~.mo .mu.m ~v~.ov ~me.e.- 3a.. game 502 mac... Ne..e~ mme.m omn.~— vmm.n mvm.e~ ~.m.~— mmm.m~ e.e.v— e-.n a. eeeeo me~.op 0.... omm.m— .me.— pm..m mmm.—~ nme.m~ mmm.m. oom.v— meo.m «mm... «weeemeee. nee .oe.n ec~.. ewe mew ~eo.. e~ ..m.. ao~.~ me... e.. .emse.eew o.o.~. Nom.m o.~.o. noo.m so... mc~.e~ emm.o mom... mem.m. ~.m.a_ ..~ ..eeeu ooo.on meo.o. eee.mc .ee.m~ c.e.—— mm..~.. .mm..n ooo..~— .m...~ omo.e~ ee~.. xeeeme>.. ~oe.n m.e.~ .mm.c ~me.. ~ov.— mee.m .v... m~... ~om.m m.~.~ o.o oe.eeeuae eee mewmmeueee .o< m...¢— ~ae..~ .m~.m e.e.o— eom.o omn.o— ..o.~ vm~.mm m~o.m— ~.o emm.. me.eaep —aeeu.eu.em< omn.c mwo.v vmm.n eee.~ ~.m.~ ..m.¢ o.m.n ..m.o emm.m m.o.n .m... eeeu. woo.~o— moo.¢m ..o..o_ e.v.~e ame.—v mm~.me. eee..o mvo.nm~ mgo.vm omm.mm cmo.e. emgu we mew: nuo.e. n.o..~ .m..on www.cv v...v— no.._m omm.v. m¢~.~m «Na... .o..wo co. moueeem eeeeo pm~.o. mem.v em..~. om~.n moo.c mom.m— .c..v v.o.m— .mo.~— ~.a.n— .mm a.eoeu .eN.eN new..N am...e me..." ng.o. oe..ee N.m.eN eme.Ne. e...oN ee..nn Nee.. eggeme>.. sum.m mnm.o mm~.v .e..~ oes.~ moe.o .me.o mm... pa..o mom.n mom oe.eee5ee eea oe.mmeeeee .me vo..~. e.m.e~ mmv.m mwm.m eNm.n mmm.¢_ .me.o mm..v~ m.~.m. mm. —. me.eoew —oee..eu.em< o.m ..m .ec on o owe o- men «am mm. o nee.>eem see. qme.n~ o.~.m_ vmm.m mme.m men.o— mmo.~— a....~ me..m Nem.m. nmm.n mmN mo.om eeeu ovo.oe— ano.va -e.nnp m.n.no «no.vm n.e.om— e.m.m. m.~..on ne..me noo.~o— nmn.o gene we neueeem xem eeo x_eo » eo mee_.eomeoew g. . eeegm: someone :eeooe. eoeeee a e. meme. ze_. ence ..em ..em em oz e w u empeeee .>e>eem Eeeu mulomm. "meeeem «.me am. m.g. mm. mepem Feswe< ooo. No.o~u o.o mm moo. . N—.m m.mN um mmpem peewe< oz ¢.pw mo o.o~ oo mopem eweeo Noo. no.~ m.mm NNP ooo. o¢.mu p.~w mm— mm—em eweeo oz exew—weeeeee ee=Fe> eeez memeoetg e>5PFweeeeee eeepe> eeez memeu we meeeuuee .wew-eeo up eeesez .wewueeo -w eeeEez eewpeemeeew —eswe< manna. ewuu HeHeH au zefluefl emuu HuueH eee eweeo mepem .eswe< we 5eeeeee e5 mepem eeeo we peeoeee mee.5eemeeew .eswe< eee eweeo we meeeepee eee me.em .eewe< eee meeeo emeeeew eewpeeeeeo emeo eee35em ewemeewuepem o—.o m—eew 174 .meueswume eeeewee> empeeee .ae>e=m see. mulwmm. .eeeeem m.om ov— .mmmeeueze eweeo ooo. o¢.mu ¢.¢ me memeeuezm eweeo oz exuwpweeeeee eeepe>lw eeez memeo we meeeuuee .wewneeo eeeE:z eew5eemeeew eweeo zepwueo emeo .euew e5 mmeeuwgeeexm eee. we peeeeee meeweeemeeew eweeo we meeeeeee eee mwweumeee. ee meeepweeeexm emeo eeezuem ewemeewuepem n—.o mweeh 175 largest net cash flow. The large positive cash flow may have resulted from underreporting, or be explained by the fact that households may have postponed payment, of sensitive cash outlays such as animals or credit (see Tapsoba, l9Bl). The monthly cash flow statements (see Appendix M) show that in spite of his positive annual net cash flow, the average ANTRAC farmer spends the early part of the growing season with a negative cumulative cash flow. This pattern, however, may be more representative of farm- ers with grain sales and repurchases than for ANTRAC. If such pattern was to be repeated every year, it may present some liquidity constraint to this group of farmers. The monthly cash flow statements do not clearly show an inverse re- lationship between crop sales and animal sales which would suggest that at one time households may raise relatively more cash from animals and at another time raise relatively more cash from crops. This may be due to the fact that animal sales are averaged over all households even though not all households have effectively sold animals. To correct this impression, we consider only households which had actually sold both grain and animals and we compute zero and higher order (control- ling for zone, technology level, and period of the year) correlation co- efficients between the percent contribution of crop sales and animal sales to cash revenues. These correlation coefficients indicate that crops and animal sales may indeed by inversely related significantly, even if the strength of the relationship is weaker than expected (Table 6.18). Overall, the evidence from the cash flow analysis shows that crop sales are a relative minor part of the sources of the farm-family cash 176 Table 6.18 Relationships Between Crops and Animal Sales Contribution to Cash Inflow Third-Order Correlation Zero-Order (Controlling for Zone, Correlation Technology and Time) ‘ Correlation Coefficient -.602 -.576 Number of Cases 127 looa Significance Level .001 .001 Source: 1978-79 Farm Survey. aMissing data (and list wise deletion) accounts for the small number of cases. revenues. Crop sales, at most, makes 40 percent of cash revenues in November-December for households who sell and repurchase grain, and less than 5 percent in December for households without grain transac- tions. These figures convey the low monetization level of farmers with respect to crops, but not necessarily with respect to other enterprises. They also show the diverse nature of the farm-family economy even at this low level of development. In particular, they show the critical importance of livestock revenues and also that taken together, revenues from agricultural trading, processing, and gathered products (e.g., shea nuts) are larger than revenues from farm crop sales. Once again, we are shown that food marketing at the village level must not only be concerned with crop sales. 177 The figures also show that cash outlays for food are relatively more important for traditional households than for ANTRAC households. Yet, in absolute terms, ANTRAC households still spend in cash more on foodstuffs than traditional households (partly because of larger family size), but cash outlays for cropping enterprises, livestock raising, and equipment are even higher. Thus, they are being forced into more cash outlays while the cash generating power of the cropping enterprise remains very small. 6.9 Summary This chapter has analyzed producers' behavior with respect to grain marketing at the farm-household level by examining the grain flow statement, marketed and marketable surplus, the timing of farm sales and household purchases, farm gate prices received and paid, and the monthly cash flow statement. The overall conclusion is that grain mar- keting covers both purchases and sales by producers even in a low-in- come economy. Conceptually, the grain flow statement, rather than grain disposals, provides a better approach to the understanding of producers' behavior; and a key to the understanding of this behavior in the short-run is the annual change in stock. Marketed and marketable surplus which are de- rived from the grain flow statement have hidden assumptions with re- spect to the change in stock, and thus, may yield significantly dif- ferent results depending on the treatment of stock. In addition, by establishing the grain flow statement, it can be shown that short-run cross-section studies of food availability at the farm-household level 178 may yield inaccurate results because they often assume implicitely that the change in stock over the year is zero. The grain flow and marketed-marketable surplus statements were developed for a selected subsample under the assumption of a minimum consumption requirement. Furthermore, the statements are restricted to three seasons that span a growing agricultural year rather than a mar- keting year. Despite these limitations, they show clearly that the average ANTRAC and traditional households were both marketed and mar- ketable deficit producers with the exception of a few households in only a few villages. The results suggest that many may even have had grain consumption below the assumed minimum requirement. This situa- tion worsens particularly for the ANTRAC households because of their larger family size. The pattern of farm sales and household purchases are slightly biased when they are examined as if all households buy and sell. When households are stratified into four categories of actual transactions, the results show that the largest group of households buy grain only. Also, ANTRAC households tended to be more buyers than sellers. There is but a small difference between ANTRAC and traditional households when it comes to the pattern of sale and purchases, except for the fact that some ANTRAC households have had large purchases in the harvest season to accommodate social events. Also, the distribu- tion of farm income does not affect significantly the timing of trans- actions. Rather, it appears that the difference may be related to the types of actual transactions. Households which sell only, or which purchased only, seem to trade more evenly year-round than households who sold and bought. For the latter group, the pattern of transactions 179 show heavier sales after harvest and heavier purchases in the pre- harvest season than in any other period. Thus, for the last group of households, the evidence suggests that price variability might further worsen their situation. Farm gate prices received tend to move together with farm gate prices paid. The high level of prices paid relative to prices received may affect more the relatively few farmers who bought and sold. For others who tended to have sales and purchases nearly better balanced throughout the year, the price variability will have less impact. The importance of crop sales in the cash flow is not as dominant as might be expected. To generate cash producers are engaged instead in other farm and nonfarm related activities such as livestock rais- ing, trade, and employment off-farm. In 1978-79, the average household, in the zones being studied, was not able to cover its food purchases out of its net cropping revenues. Introduction of ANTRAC puts con- sequently added cash constraints on farmers. CHAPTER 7 GRAIN PRODUCERS' MARKETING LINKAGES AND PROBLEMS 7.1 Introduction This chapter focuses attention on the identification and diagnosis of grain producers' marketing problems. It is based on information about the structure of the E-ORD grain marketing system (see Chapter 5), producers' patterns of grain marketing behavior (see Chapter 6), and other empirical data from the 1978-79 farm survey and the 1980 market- ing investigation. Public officials perceive that producers are at a disadvantage in (their linkages with private traders. One such area of concern relates to the measurement of grain when producers sell to private intermedi- aries. Another area relates to marketing credit which private traders allegedly provide it at usuary terms to producers in order to gain a low cost access to grain supplies at harvest. OFNACER (see Chapter 3) was instituted partly to minimize these alleged negative effects, and aims at reducing private traders' market shares and promoting better measuring techniques. Further, in the E-ORD, "Village Cereal Banks" have been set up in village groups (groupements villageois) to help - producers cope with local seasonal food grain shortages and the con- sequent credit need referred to above. The design of effective grain marketing policies requires empirical evidence to analyze the issues mentioned above. It also requires an 180 181 evaluation of the Village Cereal Bank scheme. Does the scheme fulfill its intended objectives and farmers' real needs? Policy design requires also an analysis of producers' perceptions of problems and, if any, suggested solutions for improvement. 7.2 Marketing Problems Felt by Producers We are concerned with first, producers' awareness of marketing problems and then solutions to these problems which producers may think of. Also, of particular interest to public officials is the perception of producers towards the E-ORD and OFNACER after these agencies'previous lackluster marketing campaigns (see Chapter 3 for details). 7.2.1 Producers' Awareness of Marketing Problems At the close of the 1978-79 farm survey, producers were asked first to recall the major topics discussed with the extension agent, and then to note important topics which might have been overlooked by the exten- sion agent, but should require attention by the E-ORD. Obviously, the fact that E-ORD management had no grain marketing plans in 1978-79 means that very few, if any marketing themes would likely be on the extension agent's agenda when they meet with farmers. In effect, only a handful of farmers' responses (12 out of 253 valid responses) directly identi- fied marketing issues: Regular purchases of produce by the 0RD (five responses), building of warehouse and rural general store in the context of Village Cereal Banks (four responses), problems of crop storage loss- es (two responses), and food aid (one response). Clearly, a better indication of the importance of marketing is provided when producers are allowed to express their concerns in an open-ended way. Table 7.1 shows how marketing issues compare with one 182 .meweewee eeweu we meemmw eeuwu me.eeem=ee emu .ze.ee.ee eeeeem we meemmw eeewe mepeeemeee mppe .»5.eewee ume.w we meemmw eeewe mepeeemeee mmme .epesem eeweee xe>e=m see. emuw.m. “eueeem o.oo_ No en.g. eom e.~ ”_e5eh e.e. N. m.~ .. N.. ewe eeee ¢.~ N u w. e. meeseew e5 xeee ..em m.m . T . w.N meewee eeeegeee emceeee. ~..m Ne o.m e.. . N.N. meeee we memeeeeee ee.emem ¢.~ N - - m. meewee eeeew eezee ..eN .. m.m e.m g.m .eeeeeoeee. geoe memeeemem 5eeeeee we eeeEez .memeeemmm .eeew we peeeeee mw5e.em —.~ mpeew 183 another. Despite the seemingly small number of issues related to mar- keting, it should be noted that not shown here is the fact that taken individually as a first priority producers request that the E-ORD buy regularly farm crops everywhere and anytime after harvest (or that the E-ORD keep its promise of doing so) ranks only second to the demand of more extension agents. (Better servicing of agricultural implements comes third and water supply fourth, all requested by at least seven percent of the producers.) The previous questions asked to the heads of households were center- ed on the E-ORD and tried to identify the whole range of problems of interest to producers. Even though the results in Table 7.1 show clearly a number of problems, not all marketing problems have been iden- tified. Hence, an attempt was made to identify the problems and also the suggestions of solutions by producers themselves. But before we analyze the suggested solutions, we will examine farmers' perceptions toward the E-ORD's past marketing campaigns. 7.2.2 Producers' Attitudes Toward Public Intervention in Grain Marketing From Table 7.1 it appears that one major concern of producers is the availability of regular market outlets. Also from the overview of marketing issues in Upper Volta (Chapter 3), we are aware that public marketing operations have not been very successful. Hence, it is impor- tant to get producers' assessments of past and present public interven- tions in the grain market system. Questions were asked to determine whether producers knew of and/or participated in previous E-ORD grain marketing operations (mainly in 1974-75 and 1975-76), and whether they would be willing to sell to the 184 E-ORD if the E-ORD were buying grain in 1978-79 (that is in spite of the 0RD's disappointing results in the past). Table 7.2 summarizes producers' answers to these questions. A sizeable number (380 or 80.5 percent) know of the E-ORD previous marketing campaigns. Few (26.1 percent) of these producers, however, actually sold to the ORD, and still a smaller percentage (26.1) of these producers said they had crops to sell to the E-ORD if it were buying. Conceivably, producers who would not sell to the E-ORD, even if it were buying, might not have had much to sell in 1978-79 anyway (see the patterns of transactions in Chapter 6). However, the large magnitude of the negative responses to the questions shows that it is unlikely. Therefore, these numbers should be of concern to the E-ORD management should it decide (or be allowed) to trade grain again because they suggest that although producers might like to see more public grain. purchases, they have lost confidence in the E-ORD capabilities. It is important that OFNACER, now the only public agency buying grain in the E-ORD, understands why some producers do not sell to its agents, if it wants to better serve producers. Looking at the pattern of answers (by two subsamples of farmers) to this question, the following broad picture emerges from farmers' perceptions: First, OFNACER has limited access to producers; second, many producers think that a somewhat large quantity is required if one wants to sell to OFNACER; and third, a sizeable number of producers had never heard of OFNACER before (see . Table 7.3). OFNACER's limited direct access to producers is well known since the agency has only a handful of buying agents (see Chapter 3). But the fact that the information about the agency's grain purchases is not 185 .xe>eem see. mmiomm. Heueeem Apmv .ommo .ommv .mmeo Amemee ee—e>o o.oo. o.oo. o.oo. o.oo. 5eeeeee .epew o.oo. m.m. N.ee m.m. ez o.o ..em m.mm m.ow we» Aueeeeeev .5eeeeeeo A5eeeeeev .5eeeeeev Nzee excum we. Nzee omo-m e5 meeee ..em e5 meeee ..em News. omOTM eeu e5 Nmemweesee meewumeeo e5 mew. :ex e.=e3 pee e5 mew. no» epeez ..em eex ewe aee mewuexees e5 meezme< .memeeeeee emee emolm .memeeeeee 5mee .memeeeeee .mee umee excum we we zeex 5ee ewe 3e» omo-m we 3eex ee> omolu we zeex ee> zeex eex ea meweexeez e.eeo omoim we meewueeeeee N.m m—eeh 186 Tab1e 7.3 Major Reasons Why Producers Did Not Sell to OFNACER in 1980 Evidence From Evidence From Reasons Given by Producers 1980 Farm Survey 1980 Market Survey Number Percent Number Percent Too small volume to sell to OFNACER 34 25.4 115 34.3 OFNACER did not come in the area 81 60.4 153 45.7 Never heard of OFNACER 18 12.3 55 16.4- Official prices are too low 1 .7 2 .6 Had not found empty bags -b -b l .3 OFNACER doesn't buy my cropsa —b -b 4 1.2 ’ I was away when OFNACER's b b .agents came - - 5 1.5 Total 134C 100.0 335d 100.0 Source: 1980 Marketing Investigation. aOther crops than millet-sorghum, maize, and cowpeas. bReasons not stressed by the heads of households of the 1980 Farm Survey. cFrom 157 households'valid answers, 146 had not sold to OFNACER. How- ever, 134 provided valid reasons why they did not sell. dFrom 409 households'valid answers, 352 households had not sold to OFNACER. However, 335 provided valid reasons why they had not sold. 187 that widely circulated in the area shows that OFNACER impact is very limited. In addition, the latter observation suggests that other mar- keting information, on prices, supply and demand conditions, may be poorly circulated as well. Producers also believe that OFNACER prefers to buy the larger quantities grain merchants have bulked for resale rather than the smaller quantities individual producers bring to the marketplaces (see Chapter 5). The latter proposition could have been the object of more empirical testing had the 1978-79 farm survey: (1) distinguished among the loca- tion (market/elsewhere) of transactions; and (2) made a clear distinc- tion among the marketing participants on the basis of the typology of Chapter 5. As it turned out both features were overlooked by the sur- vey, which makes the comparison between OFNACER and merchants' market shares very difficult.1 1One may still want to have an idea about OFNACER's market sharelas compared to all other categories of first handlers (including merchants, their agents, and rural residents). OFNACER market share in the E-ORD may be estimated in one of two ways. First, an aggregate figure for all zones of the E-ORD, regardless of whether OFNACER intervened, is of primary concern to macro policy. But second, a figure computed only for the areas and periods where and when OFNACER operates is a better in- dicator of OFNACER's performance in direct competition with other handl- ers. (OFNACER intervenes only in selected zones and the 1978-79 farm survey was not designed to be representative of OFNACER's activities.) To compute the seconf figure, we selected a subsample in which only the periods (from October 15 to April 1) and the villages (from which at least one household had sold to OFNACER) when and where OFNACER would have intervened. The purpose of such an ad hoc and ex-post quasi experimental design is to try to pair households who sold to OFNACER against a control group of farmers who had not, and to help through this pairing control for the area heterogeneity. The results of such computations show that OFNACER's share of pro- ducers' farm grain sales (which hovers around 20 percent) is very small. But OFNACER's aggregate share of farmers'purchases which hovers around 3 percent is even smaller (see Appendix N on market shares). 188 7.2.3 Producers' Suggested Solutions From the reading of some of the problems identified, one may infer some of farmers' perceived solutions. Unfortunately, not all such pro- blems contain implicitly their solutions. Furthermore producers surveyed in the farm surveys have somehow acquired a better awareness of the E-ORD (through the numerous interviews) and thus may have a different apprecia- tion of the role of the ORD. In addition, these producers are heads of households and may have different concerns than other members. For these reasons, questions regarding producers' suggestions for the im- provement of the market performance were asked to two groups of farmers. One is made up of the heads of households of the 1980 farm survey, the other is made up of a more diversified group of producers interviewed at key marketplaces. Table 7.4 summarizes the answers of these two groups of producers to the question of marketing improvement. The responses show the same main problems identified previously (bad road system and inconsistent purchase policy). They suggest also that the heads of households and farmers interviewed at the marketplaces have different priorities. Of particular interest in this difference are the suggestions regarding the standardization of measures and the need to delay crop sales until prices rise, both suggestions favored more by producers inverviewed at nmrketplaces than heads of households. This difference may be at- tributed to the level of responsibility and the kind of experience of the two groups. On the average, the heads of households have more re- sponsibility than other groups of producers, and in addition they may have more grain to sell which could indicate a different approach to the issues of standardized measures. 189 Table 7.4 Producers' Suggestions for the Improvement of The Performance of the Marketing System 1980 Farm Survey 1980 Market Survey Answers Percent Answers Percent Improve roads 36 37.1 6 3.4 Provide official buying agents 20 20.6 25 14.2 Impose standardized units 3 3.1 47 26.7 Increase official prices 17 17.5 31 17.6 Provide advice and information 7 7.2 8 4.5 Provide credit to buy grain 2 2.1 6 3.4 Tighten control on merchants 2 2.1 3 1.7 Attract more merchants in area 5 5.1 4 2.3 Improve market facilities 2 2.1 - - Need to store grain and wait for higher prices 3 3.1 46 26.1 97 100.0 176 99.9a Don't know 39 - 211 - Source: 1980 Farm and Market Surveys. aTotal percents do not add up to 100 because of rounding errors. 190 These interpretations must be regarded with caution, however. Given the large number of producers who had no suggestions (a much larger number than that of respondents in the case of producers inter- viewed at marketplaces) they are only tentative. 7.2.4 Problem Diagnosis In summing up the analysis of producers' marketing problems, we need to determine, to the extent possible, whether producers' percep- tions are supported by the available empirical evidence. It appears clearly from the farm survey results, and it is of com- mon knowledge, that there is very little emphasis on marketing issues by the extension agents. The farm survey results also show that pro- ducers do not trust anymore the marketing capabilities of the E-ORD management, which is a reflection of E-ORD past failures. Yet, there seems to be a demand for official buying agents. This demand may be due to two related factors. First, many farmers interviewed by (what they perceived as) public officials may tend to emphasize the help that pub- 1ic agencies may provide. Second, many farmers are either still unaware of OFNACER's existence or see that OFNACER agents prefer to buy from merchants when possible, which is supported by the fact that OFNACER intervenes only in selected areas, at selected times, and is able to buy directly only some 20 percent of farm grain sales in the areas mentioned. More generally it is clear that farmers are expressing a genuine need for the access to secure markets to sell or purchase products, through road improvement and attraction of more buyers (official or pri- vate) to their region. It does appear that a great deal (60 percent) of the farm grain is sold and the grain for domestic use is bought at 191 distances of 10 kilometers or more from the farm compound (see Appendix 0: Sales and Purchases at Distances from the Village). Unfortunately, unambiguous estimates of first and last handlers'participation in these sales and purchases cannot be provided. The interpretation of producers' perceptions of official prices is also difficult. On one hand, very few producers who did not sell to OFNACER in 1980 mentioned that official prices are lower than merchants' prices. On the other hand, relatively more producers thought that low official prices is an important dimension of the performance of the marketing system. There is no available market prices series to resolve this issue. But on the basis of farm gate prices (see Chapter 6) it would not appear that at the time OFNACER intervenes official prices are lower than prices received by producers. Thus, farmers who men- tioned that official prices are low may have been mistaken, which sug- gests again that they lack accurate market information. Other problems referred to throughout the analysis of producers' perceptions and suggestions relate to credit in order to buy and store grain, the building of warehouse and general store, and grain measure- ment. These problems are also of particular interest to public offi- cials because they had set in motion Village Cereal Banks to provide credit and food security. Public officials are also interested in pro- moting better grades and measures. These issues are the subjects of the next sections. 192 7.3 Measurement of Farm Crops 7.3.1 Issues and Hypotheses "The standardization of market weights, qualities, and practices is so widespread (in developed countries) that it is taken for granted, and the role of standardization in the market process is sometimes un- appreciated." (Kohls and Uhl, 1980, p. 360). Grading and standardiza- 2 facilitate price discovery and benefit producers, traders, and con- tion sumers in several ways. First, the use of grades and standards enable the buyers to obtain the particular qualities of produces they desire. Second, it encourages the production and marketing of a better quality product. Third, it decreases transaction costs by eliminating the need for personal inspection. Fourth, it reduces logistical costs by sort- ing high quality products suitable for storage and by facilitating con- solidation shipments of homogeneous products. Fifth, it imporves equity to individual producers, sellers, and consumers by reducing the poten- tial cheating in trade. Finally, it facilitates the exchange of market information and helps maintain effective competition by providing a common language for buyers, sellers, and market reporters. Grades and standards are not well articulated in Upper Volta in general and in E-ORD in particular. A great number of varieties of millet and sorghum are distinguished throughout the region on the basis of consumers' preferences and agronomic qualities. Yet there is a good deal of pooling locally. This is of great significance to the cost 2Grading is the sorting of products into homogeneous lots, accord- ing to characteristics or grade specification that might include physi- cal as well as subjective factors. Standardization is the process es- tablishing one set of grades among buyers and sellers, and it involves defining weights and measures and indication of quality. 193 of personal inspection when the volume exchanged is large, as it is at the wholesale level. However, the demand for grain grading is not strongly articulated and public officials have not provided any grading system except in vague terms for rice.3 There is also a great deal of diversity in grain measurement. For example, the 1978-79 farm survey recorded some 50 different containers used by producers to store and/or measure grain. Some of the devices are officially accepted: The so-called 0RD "tine" is a calibrated 20- liter container sturdily constructed to prevent losing shape and thus volume tampering (a half unit is also available), the jute bag is usu- ally of approximately 100 kg., but is also available in 50 kg. or even 25 kg. versions, and there is also the scale. But a larger number of these devices are not standardized. The so-called merchant's (or local) ftjflgfl_is a nonstandardized unit made often from used oil drumsoof very light construction, and thus easily reshaped in somewhat smaller or larger volumes. Market participants also use a whole array of plates, the most popular being the "Yoruba plate," so-called because it is sold by Yoruba women from Nigeria or Benin, as many kitchenwares are. Final- ly, market participants may also use baskets or calabash, the most popu- lar being the small lgughg_(calabash), which volume varies according to participants (merchant or producer) and season, and yet is always quoted for the same face value of five FCFA. The 0RDs and OFNACER have sought to promote the ORD tine in order to: (1) make the handling of produces during transaction more efficient; , 3For example, millet and sorghum are treated by OFNACER as if they were one product. This thesis does the same, but only because, in most instances, the data on the basis of which the analysis is performed do not distinguish between millet and sorghum. Furthermore, such treatment is consistent with previous analysis performed upon the same data set. 194 (2) help prevent possible wrongdoing by grain assemblers and retailers who may be tempted to buy from one local ting, but sell from a smaller one just by reshaping or changing the device; and (3) make the collec- tion, dissemination, and monitoring of prices meaningful and easier. The concern about grain assemblers' wrongdoing stems from the way food grain is measured. When the tjgg_is used, it is filled up above the rim until a cone of grain (called the "hat") is formed and stabil- ized on top (that is, until the cone cannot retain any more grain). This "hat" is considered a volume discount. The 0RDs and OFNACER want to get rid of this practice; unfortunately, sometimes their agents themselves-- as the author observed--use the same practice. (The I'hat" of grain, would say the agent, is for his chicken.) This has become one of the ways OFNACER and 0RD agents take advantage of both their agencies and the producers, just as licensed traders allegedly do. When the small calabash (lgughg) is used, the assembler holds the device with his thumb and forefinger and attempts to take more grain in the craddle of his palm as he scoops the grain from the producer's container into his. Yet, even though they may be in a position to take advantage of the measurement, many traders would rather use a larger device like the ting to speed up the transaction and move quickly to the next patron. They claim that in the Namounou and Diapaga areas, where the lgughg is pre- dominantly used, producers refuse any other device when they sell their produce. The analysis focuses particular attention on producers and attempts to determine whether producers usually measure (using devices similar to grain assemblers) the produce on the farm when they plan on selling at the market, why they think they have to do so or not, what possible 195 effect measuring on the farm has on the perceived fairness of the mar- ket transaction, and which devices they would like to see and why. On the basis of the available information the analysis attempts to infer whether the authorities' campaign to promote the ORD tine is taking hold among producers, and it also attempts to estimate the price spread that results from the different devices assemblers and producers use. The analysis is based on the results of the 1980 farm and market surveys. In general, farmers of the first group (heads of farm house- holds) interviewed at the farm, were asked what they usually did, while farmers in the second group (heads and other members of farm household), interviewed at selected markets, were asked what they did on that parti- cular market day. Despite the ability of the well-trained enumerators, many farmers of the second group who have never been exposed to inter- views before (contrary to the heads of households in the first group), lost their patience or got afraid of the many questions asked and did not respond to all questions. This results in numerous missing cases, the bias of which cannot be determined, but the sampling design has emphasized a better knowledge of farmers' marketing problems rather than statistical rigour. 7.3.2 Measurement on the Farm Most heads of households (1980 farm survey) said they usually mea- sure the grain at the farm4 when they plan at selling at the market, and only 20 percent said they did not. Producers primarily measure on the farm to determine precisely the quantity which helps them estimate 4Of course, when producers sell on the farm some measure is per- formed. This measure, however, is not planned except possibly when the producers are forewarned on the buyer coming to the village. 196 the sale value to be expected on the basis of some price expectation. A secondary reason is specifically to guard against any measurement tricks of the assembler. The reasons for not measuring include the lack of the proper measurement device, the fact that they trust the family member they send to the market to sell the product on their behalf, the fact that assemblers impose their measures at the market, and also the fact that very small quantities don't require any measurement (see Table 7.5). If most heads of households have stated that they usually measure grain on the farm before market sales, almost half (47.4 percent) of producers interviewed at key nerketplaces (and providing valid answers) had not actually measured their products at the farm prior to the market visit.5 In particular, only 36.8 percent of producers who brought sorghum-millet for sale had actually measured the grain on the farm. If anything, the evidence from the 1980 market survey raises the possi- bility that a small proportion of the diversified group of farmers at- tending markets actually measure grain on the farm, contrary to heads of households. But is there any relationship between grain measurement on farm and perceived market fairness? 7.3.3 Measurement by Producers on Famn and Perceived Market Fairness Two proxies for market fairness are used; one is the percentage of farmers who said they usually received at the market the value ex- pected for their crops, and the other is the difference between the value expected and that actually received for farmers who had sold all the grain they brought for sale. 5152 farmers, among 296 millet-sorghum sellers, provided valid answers. 197 Table 7.5 Frequency and Reasons for Grain Measurement 7.5a Do Producers Measure Grain at the Farm Before Market Sale? Answer Number Percent Yes 105 79.5 No 27 20.5 Totala 131 100.0 Source: 1980 Farm Survey. aMissing cases: 14. Table 7.5b Reasons Why Households Usually Do Not Measure at the Farm . . - Reasons they do Percent Reasons they don't Percent To estimate sale value 78.6 Do not possess a tine 37.5 To guard against tricksa 20.4 Trust in parentb 31.3 Traders require it 1.0 Traders impose their 100 0 measure 25.0 Quantity for sale is too small 6.2 100.0 Number of cases 103 Number of cases 16 Source: 1980 Farm Survey. aGrain assemblers' measurement tricks. bParent or member of household sent to sell grain on behalf of household. 198 The results of the 1980 farm survey suggest that heads of households have a favorable perception of the market fairness which does not depend on whether they usually measure on the farm or not. Also, the heads of households see only one of two reasons why they may not get the value expected at the market and the test results do not suggest any signifi- cant impact due to measure of product on the farm (see Table 7.6). Thus, on the basis of the responses from the heads of households, only a minor- ity of farmers find the market (and assemblers) unfair as far as expec- tation about the value of products brought to market is concerned. The evidence from farmers interviewed at marketplaces is not that clear cut, however. Millet and sorghum sellers who had measured on the farm expected a significantly higher value for their sales than farmers who had not. But the perceived gap between the expected and actual values is only marginally (.01 significance level) significant for the two subgroups. This is partly because, although measuring farmers had significant higher expectations than nonmeasuring ones, both subgroups. of farmers did not receive that significantly different actual value for their products (see Table 7.7). Table 7.7 also suggests that non- measuring farmers, on average, received more than expected, while those who had measured thought they were short of FCFA 217. Hence, on the basis of responses of farmers interviewed at. marketplaces one may put forward the hypothesis that measuring on the farm actually raised expec- tations as to market value anticipated which possibly translates into' perception of market unfairness. One needs, however, to put forward such a hypothesis very cautious- ly. It contradicts the perceptions of the headsof households and we lack information about various farmers' price expectations (see Table 199 Table 7.6 Measure of Produce on the Farm and Perceived Market Fairness (Chi-Square Tests) Perceived Measure Don't Measure Total Sample Market Fairness and Causes Percent Percent Percent Cases Market Fairnessa Usually get at least what expected 80.5 65.0 78.0 96 Usually get less than expected 19 5 35 O 22 O 27 100.0 100.0 100.0 123 Reasons for gettingb less than expected Lower prices than expected 45.0 57.1 48.1 13 Assemblers cheat 55.0 42.9 51.9 14 100.0 100.0 100.0 27 S aChi-square is 2.7907 with 2 degrees of freedom and significance level b ource: 1980 Farm Survey. of .2477 Chi-square is .01298 with 2 degrees of freedom and significance level of .9093. 200 .eeeeeexe eme5 wees ee>weeee meeseew .me. me5ee.eew e=.e> e>w5emee «ewe ......eaeaee ..a.-aeen .eeeewee> empeee ee eepe>uhe .>e>e:m eexeez owe. ”eeeeem A.88. name. mace. .eee. aeeee...=m.m eNN.N em... eoe.N we..m.egem-. em.em.-. e...N.+. ..N... 0..... . m...o. e.em.. .maeee. «ego gee: ee me ee me oe mm momma ea eeee=z 5oz em: eeeemeez ea: 5oz em: eeeemeez em: .ez em: eeeemeez em: eez meeaeee eee .a..=eee.e.o a=.g> ea ewe.-. eez meeEeee eee mee.e> .e35e< ee ewewlw eez meeseee ee. me=.e> eeeeeexm ee anew-» ..eweeeeewwwo eee .ee>weeem appe=5e< .eeeueexm me:_e> eweeo we memewuwv mmeeewe. 5exeez ee>weeeee eee see. wee ee eeemeez ..N m—eek 201 7.6). It may simply be that farmers interviewed at the marketplace had a poorer knowledge of actual market prices than heads of households, which explains why some farmers would think they got more than expected. 7.3.4 Farmers' Preferences for Measuring Devices Equity in grain measurement is but one dimension of grain measure- ment. The 1980 surveys also sought to determine which measuring devices producers would like to see in use and why. Not surprisingly, all mea- suring devices chosen were those already in use and tested by partici- pants. That would not mean that new devices would not be accepted; these alternative measuring devices would have to respond to producers' needs, however. Table 7.8 shows the reasons why heads of households (1980 farm sur- vey) and farmers interviewed at marketplaces choose measuring devices. The most striking difference between these two groups of farmers is that the heads of households show a much stronger preference for the ORD ting than households interviewed at marketplaces. Thus, it ap- pears the heads of households have paid heed to the promotional cam- paigns of OFNACER and the E-ORD. As for the producers interviewed at marketplaces, their choice of measuring devices may reflect the fact that these devices are the ones used by merchants. (No trader uses the ORD tjgg_at markets and heads of households may be measuring at home at relatively higher numbers than producers attending markets.) Both subsamples of producers are more in agreement with the rea- sons of preferences. They chose one measuring device for three main reasons: (1) they trust its fairness;(2) they value its convenience; and (3)they appreciate its wide usage in their areas. ENJZ .memeu me.mmws we emeeuee —e5e5 eeeem me. e. e: eee pee ee m—eueueem «ewe .xe>eem Eeew oom— .eueeem m .n m. .am en N eN. .N ma amenag e.ee. e.ee. e.ee. e.ee. e.ee. e.ee. e.ee. e.ee. e.ee. .eeeeee .eea. - - ... - m.eN - N.e ewe ..N .a.e..ea a... e.ee. e.NN m.em e... a... e.em. e.em m.Ne m... ease. - ..e. - ... - - ..e m.m ..e we... .eeae - m.e e.mm e.eN - - m... n... ..N. euee.eu>eae - - - e.N m.e. - N.e e.o ..m oea mesa eeeea. - m.em - m.e. n.m e.em N.e. e.eN N.e. an: eg.z-eee< 9:093.— ucmogmm acougmm ucmugon— H.593... ucuuwmm ucmugom 0..—09.0.5 0:095; «.aum human” .mwewe. m%h%_ WWW” emaea.ge wumfiuw .wnum~wue eeuuuez wwemnwwwmn meuw>eo an »m m.:z hm m.:: an A>e>eem See. ooo.. e_eeemeez we meme: em.w e.eew meew>ee peeeeeemeez eew eeeeeeweee we meemeem o.m «peep 203 .uomau me.mu.e we «menace .eueu eeeem one o» e: eeo nee ee mpgueueem mews .>e>e=m poxeee ooo. .uueeem m av cu no F oe sow oo —o ememeu ew~e> o.oo— o.oo— o.oop o.oop. o.oo_ o.oo— o.oo— o.oop o.oo— ueeueoe peach u n u u u I n a n pewuwwwe m.5~ o.oo_ m.o~ m.~o s.mm o.oop o.o¢ ~.¢e o.o¢ m.mm awash u v.o~ u ~.m u m.u~ ~.~— m.e— —.—p mmpem hope; a v.o~ m.~— m.nm a u o.¢p «.mp o.e~ oueowcu>eeu . e.~ - m... . .m.~ ~.. - e.e. eee meze easee. n o.ev u m.o n o.o~ o.p~ v.- m.op mm: oewzueoe< ueoeeee .eoueee .eeeeee 5eeeeee .eeueoe neeueee ueeeeee «eeueee .eeeeoe 28,. been”... ewe“... .eaeg..a.. ance—e. .....eflews ee..a.......eem geeeewzaez gwneagfi mwwemummmn mee.>oe eewze—pem we. xm huwrhzmw emee. an meee.e5exeez eeuuepem 5e eezew>ee5e. meeEee. em.. eweew eeeewueeo o.~ mpneh 204 Only a few heads of households (8.3 percent) chose a measuring de- vice because it is the official one,while a sizeable proportion of pro- ducers interviewed at market (12.7 percent) would like a device to handle their small-scale transactions. Again, the fact that heads of households think of official measuring devices is indicative of the fact they are aware of the 0RD's campaign. In contrast, producers who attend markets often bring small quantities for which a small-scale measuring‘ device is very appreciated (see Table 7.8). The fact that more than one—fourth of those who chose the 0RD tine did so because that was the official unit is very encouraging for the 0RD promotion campaign to spread the use of that unit; very few producers, however, chose the scale, and even though they did so because of trust, we are aware that most farmers simply do not know how to operate it. Tab1e 7.8.and other previous empirical evidence suggest that the more diversed group of farmers who visit. marketplaces may have a different pattern of behavior with respect to measurement devices than the group of heads of households. 7.3.5 Price and Volume Differential Related to Measuring Devices There is no doubt that the nonuse of standardized measurement de- vices leads to imprecisions regarding the volume being bought and sold, and may encourage some form of cheating by persons doing the measure- ment. For example, grain traders who sell back to OFNACER complain that a "100 kg.“ bag filled to the limit as requested by OFNACER, who does not weight it in the presence of the merchant, may weigh some 18 to 20 percent in excess of 100 kg. They content the OFNACER agent later repackages the bag into 100 kg. units and saves the difference 205 for personal use. An estimate of the quantity of price differential, however, is difficult to determine because of thw unwillingness of most merchants to allow any measure of the capacity of the devices they use. At the Namounou Marketplace we attempted to estimate the quantity differential between the merchants' calabash (lgughg) and the producerst Even though Namounou is the region's largestrural market, male and fe- male producers insist on selling by the lgughg (small calabash). They would compare their own lggghg_with the merchants', and if not satis- fied, may go to the next grain assembler or attempt to retail their product. In the latter case, they do the measurement. It was an esti- mate of such a volume differential between the traders' and the pro- ducers' lgggththat we attempted to measure in conducting an experiment at the Namounou Market on February 17, 1980. A sample of 12 grain assemblers and 12 producers willing to parti- cipate was taken. One hundred FCFA worth of grain was weighed for both groups. (Even though the lgughg_is quoted for FCFA 5, FCFA lOO worth of product is made up of 27 louches and not 25 because of a volume "discount.") The results indicate that an average FCFA lOO worth of sorghum measured by the traders weigh 34 percent heavier (at 3.07 kg.) than the same retailed by producers-retailers (at 2.25 kg.). Assemblers on the average bought sorghum at FCFA 33.22 per kg., while farmers, who decided to sell it themselves to consumers, were selling it at FCFA 44.44 per kg. Of course, some portion of these differentials is necessary to cover different marketing costs involved in wholesale and retail trade. .At a minimum, the differentials do represent the opportunity cost of 206 producers' time spent in trying to retail their products rather than sell- ing it more quickly to traders. When assmblers sell back the grain they have purchased from farmers, they must charge higher prices to cover _ their operational costs and normal profit levels. Unfortunately, in our analysis we have no way of determining the degree to which the price differentials are justified on the basis of the differential marketing costs. 7.3.6 Synthesis The empirical evidence presented shows that heads of households with more responsibility may have a different pattern of behavior with respect to measurement as compared to a more diversed group of farmers (some retailing their products) visiting marketplaces. The heads of households tend to measure their grain before taking it to market, while it seems a smaller percentage of producers interviewed on market actually did. While heads of households, regardless of whether they measured on farm, thought they were treated fairly at marketplaces, average producers interviewed at markets had higher expectations about the value of their products when they had measured at home than other- wise. The effects of possible different price expectations could not be determined, however. Heads of households who have been having contact with 0RD agents, seem to be more aware of the official 0RD tine than producers interview- ed at markets. Both groups would prefer a measuring device which is trustworthy, convenient, and in wide use. At Namounou it was found that the 193911; merchants use to buy from producers is significantly larger than the louche producers would prefer to use. This differential 207 partially represents producers' time opportunity spent to retail own product and accounts for traders' operating costs and normal pro- fits. In conclusion, it appears clear that producers are aware of the benefits of standardized measures. However, the promotion of the 939 tigg_and other standardized units, with the possible exception of the 100 kg. bag, is not taking hold fast enough. ORD and OFNACER should step up their promotion campaign by: (l) holding demonstrations showing the usefulness and superiority of the 0RD tine over its nonofficial a1- ternative;(2) providing new containers which respond to the expecta- tions of producers (e.g., subunits of tige);and (3) devising ways to en- force the use of standard measures by all participants. 7.4 Marketing Credit and Advanced Grain Sales 7.4.1 The Issues Credit transactions cover many dimensions6 in the rural economy of the Eastern Region but those between producers and merchants are of particular interest because of the public perception of producers being exploited. Producers may borrow money, or buy fbodgnain on credit, be- fore harvest against the promise of payment in grain at harvesttime. _ In both cases, the quantities of the future grain7 deliveries are 6 7Repayment in kind may involve other crops than millet-sorghum (in some cases even animals). The pattern follows regional differences: for example, in the peanut growing area of Bogande, the primary crop sold in advance is peanuts. We have evidence that in the Pama area yam producers commonly sell their crops before they mature or even before they plant them. The contract, which may affect a portion or the whole field, calls for the producer to care for the yams until harvesttime. (The latter is a genuine case of "green sales"). See Tapsoba, 1981. 208 determined on the basis of the amount of money or food borrowed and a preharvest food grain price is in effect set up. The pledging of part of future harvest grain as collateral for a loan is not a "green sale," but an advanced sale8 and may be conceived as a preharvest contractual arrangement between producers and merchants. But, it is thought that this type of arrangement favors only merchants who are provided with an opportunity to secure, some three to six months before harvest, grain supplies at prices even lower than the depressed harvest prices, while locking producers into a vicious circle of indebtedness. The questions confronting us are: (1) How widespread the practice is in the E-ORD? (2) What is its magnitude for the producers who are concerned? (3) What ranges of prices and quantities of grain in repay- ment are agreed upon? And (4) What annual nominal interest rates these prices and repayments in kind imply? These questions are difficult to investigate in the Eastern Region because first, money or food grain bor- rowing is socially perceived as an inability on the part of the borrow- er to feed his family on his own, so that farmers may hide or underre- port the extent of their borrowings. Second, because they have been depicted as userers, merchants don't easily admit that they buy crops in advance, while some even say that the practice goes against their re- ligious beliefs. Finally, it is often difficult to determine who the lender is; the merchant, the merchant's agents,and village resident buy- ers working on behalf of the merchant, or even some farmers no longer feeling a bond to help their fellow farmers the way they formerly did and requesting an interest payment when fellow farmers pay back a loan. 8A "green sales" arrangement assumes that the future production of an entire field has been sold. Here we use the term for convenience. 209 Because of these difficulties, the analysis relies on different sources of information: 1) the 1978-79 farm survey has recorded the credit transactions of heads of households over the year-u-a comprehen- sive analysis of this data is provided by Tapsoba9 and this analysis makes use of some of his results already published; (2) the 1980 farm and market surveys focused on the pattern of advanced sales by heads of households and a more diversified group of farmers interviewed at the markets;(3) the comprehensive 1980 village inventory survey focused on the magnitude of "green sales" on a village basis;(4) finally, we will also rely on personal communications. 7.4.2 E-ORD Areas with Advanced Grain Sales The results of the 1980 village inventory survey show that in 12 percent of the villages visited regularly by grain merchants, part of the grain is bought in advance or "in green" (see Table 7.9). These villages are located in all sectors of the 0RD except Diabo, Comin Yanga, and Matiacoali. The evidence from the 1980 market survey (in selected marketplaces) shows also that producers who sold in "green" that year came from 19 out of the 126 villages (15.0 percent), and that 63.2 per- cent of these villages were located in the Bogande sector, 21.0 percent in Fada, 10.5 percent in Pama, and 5.3 percent in Diapaga. As for the 1980 farm survey, producers acknowledged that they knew of the practice taking place at 19 out of the 27 villages surveyed, which is similar to 9Tapsoba has covered this point in his analysis of the "commercial segment of the informal credit system" (1981. Pp. 193-210). 210 Table 7.9 Village Locations Where Merchants Usually Buy Millet-Sorghum in Green Villages visited Villages where part by merchants of the grain was 0RD Secteurs brought in'greed' # # Percent l. Bogande 69 10 14 2. Diabo 2 0 O 3. Diapaga 21 2 10 4. Comin Yanga 3 O 0 5. Fada 26 3 12 6. Kantchari 5 l 20 7. Matiacoali . ll , O . O 8. Pama ' 6 1 17 142 17 12 Source: 1980 village Inventory Survey. 211 the results of the 1978-79 survey. Thus, the general impression convey- ed by the various sources of information is that advanced sales of some sort takes place all across the E-ORD. 7.4.3 The Magnitude of Advanced Grain Sales The question now is how important are advanced sales for producers and lenders? The results of the 1980 farm survey identified only 13 heads of households who had sold part of their grain in "green" that year, and a total of 21 heads of households who had done so at least once the past three years. Only 35 (7.6 percent of) such farmers were identified in the 1980 market survey (13.3 percent of all farmers did not answer the question). Thus, it would seem that there is a small number (possibly due to underreporting) who sell in "green." The results of the 1978-79 credit analysis support this conclusion with the same qualifications (see Tapsoba, 1981, p. 205). Only sixteen (16) loans which called for repayment in grain with interest were record- ed, with an average of 67.7 kg. of grain paid back. From the lender's- side, however, 37 loans paid back in grain (an average of 102.6 kgs.) ‘0 Also from the record of advanced were extended by heads of households. sale transactions provided by one grain merchant in Piela, 12 farmers were interested in the scheme which involved an average of 4.83 local tings (or approximately 91.77 kg.) of grain to be delivered at harvest by producers. (For the grain merchant, that we classify as a regional wholesaler or a medium to large trader, the total quantity to be 10We have assumed here that these "commercial loans" have been ex- tended mainly by grain merchants (some being grain producers themselves), or by their agents. 212 collected,approximately 1,102 kg., represented only 5 percent of his grain transactions.) Finally, the grain flow analysis (See Chapter 6) shows that the amount of grain used to pay back loans is smaller than that given away as gifts. 7.4.4 Some Patterns Related to Advanced Sales Most heads of households with advanced sales interviewed in the 1980 farm survey usually deal with one trader at a time, while Only a few (25 percent) with two traders, and only 30 percent of the households would label this merchant their "regular" or exclusive partner. Also, very few (10.5 percent) producers go to another village to borrow money in exchange of future grain deliveries at harvest (see Table 7.10). We infer first that this kind of transaction is confined to the premises of the village, and second, from the producer's point of view, the grain merchant has not secured an exclusive source of supply because of his , money-lending. Sometimes producers may not get the entire amount of money or food grain they wanted to borrow. A third or so of the heads of households in the 1980 farm survey have faced such a possibility. This happens when the merchant says he cannot fulfill the credit demand or, reason the heads of households, the merchant is afraid that the borrower may not pay back the entire loan (see Table 7.10). Thus, it appears that the merchants are cautious not to overlend money, or more accurately, not to attempt to buy too much of farmers' harvest, for fear farmers W0U1d have difficulties to pay back. (We will see later that this possibility still occurs.) 213 .ae>e=m see. ooo. "eeeeem .8 .2 .N. .N. E ... $3.8 8.3.... m. o— m— a. ow om memee we eeesez e.oe_ e.ee. e.ee. e.ee. .o.ee. e.ee. .eeew a a 191...... mm: a a oz e... ..ge «.8 ...... 98 e... we. 5eeueee 5eeueee 5eeeeee 5eeeeee .eeeeee peeeeee muwwm .Immmmml eee_ emee ewe—_.> eezee eeeeem eee memzme< e>weeee\.em ..e 5em ..e 50m eewmeee ee.=mee e »_ee seew a..eem= appeemo x—.eem= seeeem aegeeees m. zeeeem eweeo we mePem eeeeu e5 eeee.em meeeeuee eewaeemeeew .meeeeeeee o_.m m—eeh 214 For most heads of households (70 percent), the period of July- August is the time when they engage in "green" sales; for 25 percent it is May-June; and for 5 percent it is September-October. Money is bor- rowed during these periods to buy food--from either another trader or a fellow farmer, (6l.9 percent of the answers), to pay for celebrations (9.5 percent), or to buy consumer goods (28.6 percent). As for food grain purchased on credit, all heads of households do it in July-August.H As already explained, the amount of grain to be delivered at har- vesttime is calculated on the basis of the loan or food grain borrowed during the hungry season. In the case of a loan, a price (by local ting) is agreed upon, and in a case of food grain borrowed, the exchange rate is in terms of local tjgg§_to be reimbursed for one borrowed. These prices and the resulting interest rates are the main concerns of public officials. Next section summarizes the results of the 1978-79 credit data, and the insights that the 1980 farm survey provide. 7.4.5 Grain Prices in Advanced Sales The heads of households in the l980 farm survey were asked to quote prices per local ting_that merchants usually set for the loan in May- June, July-August, and September-October. In addition, they were asked the number of local tjg§§_of grain they usually pay back for 6 local tines borrowed in the same periods. These quotes are provided by 3712 HBased on answers of 21 households who borrowed from merchants at least once in the last three years, 1980 farm survey data. 12Actually, the total number of "knowledgeable" heads of households is 56 in the l980 farm survey, among whom 2l had green sales at least once in the last three years. Only 37, however, provided valid answers to the questions above. 215 heads of households who acknowledged that they know of the practice of green sales, among whom only four had actually sold grain that way at least once in the last three years. Table 7.11 shows the price quotes aggregated across periods and sectors. FCFA 250 per ting (3 l3.l6 FCFA/kg.) is the most quoted price while the 95 percent confidence interval on the mean (FCFA 323.98) is FCFA 300 to FCFA 350. The analysis of variance, to the extent that it is warranted owing to the number of missing cases, suggests that these prices do not vary significantly over the periods, but vary across the four sectors. This pattern may suggest that prices may be somewhat institutionalized in the different areas the practice takes place. The average of the price quote (l7.l FCFA/kg.) is very low compar- 13 but it is closer to the low- ed to the official price of 40 FCFA/kg., est farm gate price received at harvest of 27.8 FCFA/kg. (see Chapter 6). The records of advanced sales transactions of one grain trader re- ports a more generous price of FCFA 500 per ting (26.32 FCFA/kg.). Also, the fact that among farmers quoting prices only four farmers had actually had advanced sales introduces some potential for underestima- tion due to merchant bias. ' Thus, when compared to the official harvest price, advance sale prices appear very low, hence, justifying public conern. However, these advance sale prices are not that low when compared to actual harvest prices in the region. The price quoted in advance sale is discounted 13The average harvest price computed from the quotes of the same producers is FCFA 521.25 a "tine," that is, approximately 27.3 FCFA/ kg. It can also be seen the average actual harvest price is less than the official price, but almost identical to the lowest farm gate price derived in Chapter 6. 216 Table 7.11 Price Quotes for Advanced Sales of Millet-Sorghum Aggregated Across Zones and Periods FCFA FCFA Frequency Cumulative per tine per kg.a (Percent)b Frequency 200 10.53 3.1 3.1 250 13.16 35.7 38.8 300 15.79 18.4 57.1 350 18.42 22.4 79.6 400 21.05 8.2 87.8 500 26.32 11.2 99.0 600 31.58 1.0 100.0 Source: 1980 Farm Survey. aA local tine with “hat" weighs approximately 19 kgs. b Number of cases 98. 217 by the merchant and includes an interest and risk charge for the amount of money loaned. When everything is netted out, this price may not ap- pear as low. As for the exchange rate in kind, 80 percent of the heads of house- holds quoted 12 tings for six tings, that is a two for one ratio (see Table 7.12). However, a few producers still maintain that no interest is charged by merchants; that is, six tings borrowed are paid back by six tings. Also many knowledgeable participants we talked to referred to nine tings for six, that is, a three to two ratio, as a more common exchange rate than Table 7.12 would indicate. The tests on the vari- ability of the quantity quotes show no effect of periods or sector. Therefore, we are tempted to infer that the two to one ratio is an accepted benchmark ratio of repayment in kind in the region. 7.4.6 Inputed Nominal Interest Rate of Marketing Credit The prices and exchange rates are only one dimension of the issue of advanced sales; the nominal interest rate that results from these prices is a second one. Estimates of the nominal annual interest rates actually charged by traders in relation to advanced sales are difficult to generate because of: (l) the imprecision as to the length of time involved; (2) the imprecision regarding prices in the preharvest season and at harvest; and (3) whether we consider the borrower or the merchant side in light of the fact that producers may default on their loans. Table 7.13 shows the inputed interest rates on actual cash borrowing and loan transactions recorded in the 1978-79 farm survey (entire samfle) Tab1e 7.13a represents the producer-borrower side while Table 7.13b re- presents the producer-lender side. It appears that the inputed nominal Exchange Rate Quotes for Repayment of Food Grain Bought on Credit in the Hungry Season 218 Table 7.12 Number of tines paid back for six tines bought Items Total 6 7 8 9 12 Number of quotes 3 9 5 6 80 103 Percent 2.9 8.7 4.9 5.8 77.7 100.0 Source: 1980 Farm Survey: aMissing cases: 8. 219 Table 7.13 Advanced Sale Prices and Nominal Annual Interest Rates Tab1e 7.13a Actual Borrowing Sum Grain Green sale Harvest Value of Repayment Annual Zones borrowed paid back price rice re a nt period interest (FCFA) (kgs.) (FCFA/k9) (réinxkgid [183?) (days) rate (2) (l) (2) (3) (4)a (5) (5)b (7) (8)c Bogande 1,000 74 13.5 40 2,960 42 1680.0 1.000 37 27.0 40 1,480 49 352.6 Mani 1,000 38 26.3 40 1,520 161 116.4 2,500 74 33.8 40 2,960 . 182 36.0 Logobou 500 19 26.3 32 608 161 48.0 Partiaga 2.250 101 22.3 36 3.636 154 144.0 2,000 101 19.8 31 3,131 91 223.2 4,000 202 19.8 35 7.070 35 788.4 Ougarou 1.500 95 15.8 26 2,470 112 208.8 1,000 38 26.3 26 988 203 -2.4 600 38 15.8. 51 1.938 154 520.0 300 19‘ 15.8 26 494 140 164.4 300 19 15.8 26 ' 494 119 193.2 1.200 76 15.8 26. 1,776 168 136.8 500 38 13.2 26 988 91 385.2 Pane 3,000 114 26.3 27 3,078 91 10.8 Source: adapted from Tapsoba (1981): 1978-79 farm survey. Nzw(n Wards C (cg-*2) , 330 , 100 220 Table 7.13 Continued Tab1e 7.13b Actual Lending Sum lent Grain Green sale Harvest Value of Repayment Annual Zones (FCFA) received price price repayment period interest (kgs.) (FCFA/kg) (FCFA/kg) (FCFA) (days) rate (2) Bogande 1.500 57.0 26.3 40.0 2.280 147 127.2 2.500 95.5 26.2 40.0 3.820 112 169.2 1.500 57.0 26.3 40.0 2.280 77 243.1 600 19.0 31.6 40.0 760 91 105.5 1.000 38.0 26.3 40.0 1.520 98 191.0 2.000 55.5 36.0 40.0 2.220 . 42 94.8 2.500 191.0 13.1 40.0 7.640 112 661.2 1.000 76.0 13.2 40.0 3.040 91 806.4 2.000 76.0 26.3 40.0 3.040 153 122.4 1.500 76.0 19.7 13.0 988 98 -125.3 Logobou 1.500 101.0 14.8 32.0 3.232 245 169.2 1.000 101.0 9.9 32.0 3.232 - 182 440.4 1.500 101.0 14.8 32.0 3.232 189 219.6 1.500 101.0 14.8 40.0 4.040 145 395.8 Partiaga 5.000 202.0 24.8 36.0 7.272 77 212.4 2.500 101.0 24.8 36.0 3.636 77 212.4 2.500 101.0 24.8 36.0 3.636 77 212.4 5.000 202.0 24.8 31.0 6.262 105 86.5 14,000 606.0 23.1 42.0 25.452 112 262.9 2.000 101.0 19.8 42.0 4.242 126 320.3 2.000 101.0 19.8 42.0 4.242 105 384.3 2.000 101.0 19.8 42.0 4.242 182 221.7 2.000 101.0 19.8 42.0- 4.242 91 443.5 1.500 101.0 14.8 31.0 3.131 77 508.4 Kantchari 1.000 76.0 13.2 34.0 2.584 105 543.1 750 38.0 19.7 34.0 1.292 112 232.2 Ougarou 1.250 57.0 21.9 40.0 2.280 182 162.9 1.250 57.0 21.9 40.0 2.280 182 162.9 3.000 152.8 19.6 26.0 3.973 126 92.7 2.000‘ 76.0 26.3 26.0 1.976 133 -3.6 2.000 152.8 13.1 51.0 7.793 175 595.8 1.000 76.0 13.2 51.0 3.876 168 616.2 500 38.0 13.2 51.0 1.938 168 616.2 250 38.0 6.6 51.0 1.938 168 616.2 250 19.0 13.2 40.0 760 154 476.8 1.000 76.0 13.2 37.0 2.812 189 345.1 1.000 76.0 13.2 37.0 2.812 112 582.4 Source: adapted from Tapsoba (1981): 1978-79 farm survey. 221 interest rates are very high even if, here and there, some become nega- tive. But it is also interesting to note that the level of interest rates borne by borrowers is lower than that accruing to lenders, mean- ing that we need to look at both sides. The analysis of one merchant's transaction record book confirms that the two sides may indeed be different. The record shows that in June 1979, the merchant lent a total of 29,000 CFA Francs to 12 farmers at an agreed upon price of 500 CFA Francs for a local _t_i_rl§_ (approxi- mately 26.3 CFA Francs/kg.) of grain to be delivered at harvest, thus, expecting 58 ting§_(approximately 1,102 kgs. of millet) to be delivered sometime in November. 50, on average, a producer borrowed 2,416.7’Francs to be reimbursed by 4.83 tigg§_(9l.77 kgs.) of grain. At harvesttime, however, only six farmers fulfilled their commitment and delivered 39 tjgg§_(74l kg.) leaving the merchant short of 19 tings (361 kg.).' One year after he had lent the money and thus seven months afterifie grain had been delivered, the merchant sold the grain the the neighboring market of Pouytenga at 42,250 CFA Francs (net of transport cost, but in- cluding storage cost). Table 7.14 shows the gross interest rates com- puted from both the lender's and the borrower's sides. The merchant's accrued interest rate is low compared to the estimates in Table 7.13 and may be even lower if storage costs are accounted for. (But had the merchant sold the grain right off to OFNACER at the official price in- stead of speculating, his gross profit margin would have been higher.) 0n the producers' side, those who paid back were charged a moderate in- terest rate compared to estimates of Table 7.13, while those who de- faulted that year were subsidized, and overall as one group producers benefited from a negative interest rate. Of course, the magnitude of 222 Table 7.14 Inputed Interest Rate Charged to Producers Nho Paid Back and Received By One Selected Grain Merchant Merchant's Side Producer's Side Loan (CFA) 29,000 19,500 Quantity of grain repaid (kg.) 741 74] Value of sale/repayment ((CFA) 42,250a 20,475a Time lapse (months) 12 5 Inputed annual interest rate 42.7% 12%c Source: personal communication. aValue of grain sales net of transport costs, including loss due to de- fault. An estimate of this loss is: 361 kgs. * (42,250 FCFA/741 kgs.) = 20,583.3 FCFA. b= 39 tines * 525 Francs. The harvest price used is the merchant's quote. CIf one were to account also for producers who defaulted on the loan, the resulting interest for all producers would be negative. 223 the inputed interest is highly dependent on the price and loan period estimates, but the exercise just conducted illustrates the point that a high interest rate accruing to merchants needs not be associated with a high interest rate charged to producers. The inputed annual interest rate on food grain bought on credit should not be computed on the basis of the exchange ratio alone. Rathen, it should also consider the opportunity cost of the grain borrowed in the hungry season, harvest price, and the loan period. The estimated interest is highly dependent on the assumed spread between the hungry season and harvest prices. Ceteris paribus, the larger this price spread, the smaller the interest rate. 7.4.7 Synthesis In summary, the various sources of information available suggest that millet and sorghum advanced sales do occur, but the proportion of producers involved seems small and each has promised to deliver only a small amount (110 kgs.) of millet-sorghum. Prices paid by merchants for these advance sales are low relative to harvest prices, but it is mainly the very short repayment period that makes the resulting annual interest rates very high in most cases. But from the lender's point of view, if only 50 percent of the farmers ever pay back, then high im- plicit interest rates do not yield high profits, but help to cover high costs. The insights gained from these results and some personal communica- tions suggest that the relations between producers and grain merchants are evolving. It would seem that many producers have learned to play tricks with grain merchants by: (1) delaying the repayment; (2) repay- ing only the capital of the loan in cash; or (3) defaulting on the loan 224 altogether. The public campaigns may have encouraged these tactics and also the fact that some merchants who provided the funds did not live in the area. Thus, the advance sale arrangement has become a somewhat costly and risky way for merchants to secure grain supplies, and a high cost source of credit for the producers who do actually repay on time. Money lenders have found ways to enforce cash payment of the loans (see Tapsoba, 1981). In case of credit transactions which involved pay- ment in kind it appears that merchants may be relying more and more on village resident buyers who are given the funds to lend to their fellow farmers and collect the grain at harvest. Being members of the com- munity, these agents are in a better position than merchants to screen borrowers and enforce the contracts. _These contractual arrangements which lead to advance sales, no matter how unpopular they may be, fulfill producers' genuine needs for cash or food during the "hungry season." To do without them, alterna- tive sources of credit in cash and/or food must be found. In Eastern Upper Volta a new such institution to allow producers access to cash and food grain, the "Village Cereal Bank" has been attempted. The fol- lowing section analyzes the performance of this scheme. 7.5 An Evaluation of the Village Cereal Bank Program 7.5.1 Objectives of the Village Cereal Bank The Village Cereal Bank program started in 1977 as a pilot project in three selected producers' precooperatives (village groups) financed by USAID. Its justification was the perception that producers were being forced into distress sales at harvest in order to fulfill their cash needs, and into borrowing to buy food in the hungry season when 225 grain may only be available outside the village. The original idea, as promoted by the MSU's credit expert, looked like a nonrecourse loan program.14 In subsequent reformulations, however, emphasis on commun- ity development was added with the additional expectation of developing a food reserve, and provision of social and other services by the vil- lage group to its members. Three broad objectives were stated in the final project document: (1) To timely provide cash to producers so that they would not be forced into distress sales at harvest for depressed prices. (2) To make food grain available to rural consumers at the vil- lage level at reasonable prices during the rainy season. (3) To increase revenues of the village group through grain sales in order to allow it to self-finance agricultural in- puts, such as improved seeds and fertilizer, and other services or businesses such as a rural pharmacy, a dis- pensary, and a general store. The program administered by the E-ORD provides loans at 8 percent annual interest rate to village groups (selected for their progressive- ness) to allow them to buy food grain at harvest, store it in the ware- house built by the village group, and sell it back to members and other rural consumers in the rainy season. At that time, the poor transporta- tion network isolates many villages and often make food grain unavailable 14A nonrecourse loan, as operated by the U.S. Commodity Credit Cor- poration, provides loan to farmers who put their crops as collateral on terms determined by the government price support level (loan rate). Farmers repay the loan (plus storagecost) and take possession of their crops if the market price moves above the support price, or keep the loan if the market price stays below the support price. 226 to the community. The E-ORD recommends that the purchase price to mem- bers and other producers be higher, and the sale price to the same parti- cipants be lower, than the respective prices offered by grain merchants. It is assumed that the gross margin would still be adequate to yield a profit after paying for the cost of capital, handling, transportation, and storage. These revenues would finance social and commercial ser- vices and their proceeds, in turn, would make the precooperative financi- ally independent in a matter of six years so that the loans could be phased out. The first year, the village group receives a loan of FCFA 600,000, the subsequent years, FCFA 480,000, FCFA 360,000, FCFA 240,000, and in the last and fifth year, FCFA 120,000. 7.5.2 Performance Dimensions From the Village Cereal Bank objectives and operational set-up, it can be seen that the scheme is a multi-facet project. It operates not only as a consumption credit institution, but also as a grain reserve and food security program set up at the village level, a marketing cooperative, and an agent of change to improve storage conditions and foster community development spirit. An evaluation of such an institu- tion is difficult because of the many alternative performance criteria related to the various goals and because of the lack of recorded in- formation about the operations of the Village Cereal Banks. There are, however, three overriding and interrelated questions of interest to farmers, the E-ORD management, and donor agencies: (1) Do the Village Cereal Banks achieve their stated objectives? (2) Do they fulfill farmers' needs (which we now understand better as a result of the analy- sis of the grain marketing system in Chapter 5, farmers'food grain mar- keting behavior in Chapter 6, and farmers' marketing problems? And 227 (3) What specific problems fact the Village Cereal Banks and how can they be resolved? Whether the Village Cereal Banks achieve their stated objectives can be examined by looking into the evolution of their membership, the record of procurement and sales of grain, and the financial viability of the system. The other questions can be examined through the diagnosis of the Village Cereal Banks' problems. 7.5.3 Evolution of the Village Cereal Bank Membership The number of Village Cereal Banks which received some financial support from the E-ORD at one point or another has gone from three in 1977-78, 15 in 1978-79, and on last count (June 1980) five in 1979-80 (see Table 7.15). Two out of the three initial Village Cereal Banks were dropped the next year because of inadequate performance, but also because of inappropriate selection in the first place. Donors' agen- cies, however, seemed to have been convinced by the usefulness of the program, for in that next year three more donors joined USAID and the number of Village Cereal Banks was expanded to 15. In 1979-80, however, only one new Village Cereal Bank was financed and all previous Village Cereal Banks were dropped except one. Thus, the extent to which donor agencies have supported the Village Cereal Bank proved to be critical. Likewise, the E-ORD team had a successful campaign promoting the concept of the Village Cereal Bank to producers. In 1978-79, 30 vil- lage groups started their own Village Cereal Bank financed by their own reserves accumulated over the years (FCFA 50,000 or less). Even though many of these villages might have started the Village Cereal Bank in order to attract future loans from the E-ORD, their decision demonstrates 228 Table 7.15 Number of Active VCBs (1977-78 to 1979-80) Year Village Cereal Banks (Village Groups) 1977-78 Tapoa, Logobou, Piela 1978-79 Tapoa, Bourgou, Kikideni, Ougarou, Samou, Boudieri, Boulel, Diagorgou, Gayeri, Koulwokou, Sakpani, Tampoudin, Tindangou, Toussiegou, Yenkoali 1979-80 Bourgou,a Kikideni,a Ougarou,a Samou Zembede Source: 1980 Village Cereal Banks Survey. aOperated that year without the E-ORD loan. that producers in E-ORO were receptive to a certain idea of the Village Cereal Bank. . Yet, it does not appear that the village groups who received the financial help to set up Village Cereal Banks attracted new members because of it. Table 7.16 shows that, in most instances, the member- ship after the creation of the Village Cereal Bank is not larger than what it was before. We must keep in mind, however, that comparison of "before" and "after" is a crude proxy to capture the impact of a proj- ect. In the few cases (four) where membership has increased, some other aspects than the Village Cereal Bank may have been added to the village group. (For example, at Tapoa, a lowland rice project attracted new members in the village group.) The figures also show that the amount of the loan extended was not in line with the village group membership when .2352. 2 bee «..0: 9.05 J2» 9.80m .38qu 9.8-3.. «5 ho 02.33 82 5 2.8 2328. :oza .3..an ago.“ £3.23 .38 2.28 ounZ; 82 33.53 229 m._~o.a_ _._._.cvn ~_.¢_ - m.“— - - can: - ooc.-_.o AN" - _~n - - ..QOF o.coo.o~ coc.oom A. o~a_ m. asap «caBEaN .._Nm.m~ ooc.oc~ m mem— e 828. sooeau=_» game e.coo.oo_ oo°.oom . osm— m -m_ ..aoxca» c.oom.~o 888.com o aka. a New. socaaso o.ooo.om ooo.oom 8 «Na. N_ Nem— peahau ...ou~_uax m.mem.~_ cm..em. ~p use. ~_ New. somo_mm=o» c.ooo.m~ ooc.co~ m. «em. a new. .eaaxam o.coo.o~ ooo.oo~ o. «Am. a. -a_ someoma_a m.mmm.o ooo.oo a wee» «N nea— _co.u=om o.ooo.o~ ooo.co. a mem— o 258— .o_=om _caeuueau o.m~m.o ooo.oo~ e. 85w. _N «Na. .eac_x_g some o.om~.a_ ooo.ooo .~m mam. ~m neo— coach o.ooo.oo ooo.ooo c. “No. o_ 828. 3080004 ameaawa a.ooo.on ooc.oo~ o. «Na. o. “No. =_e=oa5ah o.ooo.om ooo.ooo NM meo— ~_ -m_ axo;_=ox oa~.a c.coo.o8 coo.ooo a. sea. o. “Rafi a.a.a . o.oow.~_ ooc.om_ o. mem— ~_ 258— ageam m.m~m._ omm.~m_ co use. oo— a~a_ somcsom oucauou emfiwwwu eoSQoucuuv an; cow... ””2. 352833 333.5 005.333 96.5 2:393 W?“ as v3.3.6 u> ..0 pamwwvumum; atwm>couamax waofluflu 03:; 96 can ence ucaoe< _aoop a.gmcans~: cam. op “he, 2022 mo> ago to awcmcaasaz ago co eopozposm e_.~ «_nae 230 the Village Cereal Bank was set up, for some cases loan per member was as high as the total annual farm income and in other cases it was only a few thousand FCFA. 7.5.4 Procurement and Sale of Grain Table 7.17 shows that the Village Cereal Bank did buy at higher prices and sold at lower prices than grain merchants. This, however, says nothing about whether prices offered had a relationship to costs of operation and long-term financial viability of the Village Cereal Banks. In most instances, the prices offered were strictly imposed by the donor agencies or by the E-ORD. In other instances, when purchases were made very late in the season, the Village Cereal Bank had to buy at the pre- vailing market prices. Table 7.18 indicates that indeed many purchases were made late in the season because of the poor scheduling of the loan installments. The scheduling of the loan installment has handicapped many Village Cereal Banks, but all managed somehow to buy food grain. The next ques- tion then becomes how much grain was sold in the hungry season. Table 7.17 shows that only one Village Cereal Bank (Tindangou) had not sold its stock in the preharvest season. But for the others, some of the sales were made at credit and had not been paid for as of June 1980. This situation has forced some four Village Cereal Banks to default on their loans and the others to delay repayment. It was learned that in some cases, for example in Gayeri and Yenkoali, most of the credit was made to the E-ORD extension agents. (No Village Cereal Bank would turn down such a request by agents who represent the authority in the mind of producers.) 2231 .e.8co on» 808888888.“ 883 88>.28 888.8 a» =o>.m 8888 meow .8.88t8 88 8.8. 8.8. n tap—803.0 an U—Ou mxuam _.—. 088 88 88.588 .asonaaa mu> an nouoaa no... .88 co. - >< «v u was.» n 8058888 9: ..><. 88.8» 8880.88< .08 .8 on.» .88 888.8 88: 88.88 88:: .9882. .xu>cam scam .ooeou amap.8> camp .uueaom 88 888.8 888.8 8 888.8 8.888.8 88 888.888 8888888.. 888. 8.. 888.8 888.. 8. 888.8 8.888.8 88 888.888 .888.8.8 888. 88.. 8. 888.8 888.8 88 888.8 8.888.8 8.. .. 888.888 ..88888. . - 888.8 888.8 88. 888.8 8.888.8 88. 888.888 88.8888 8.888 888.. 888.. 88 888.8 8.888.8 .8 888.888 ..8.88 ..888888: 8.. 888.8 888.8 .8 888.8 8..8..8 88 88..88. 8888...:8. 888 888.8 88..8 88 888.8 8.888.8 88 888.888 .88Q888 88 888.8 88..8 8. 88..8 8.88..8 8. 888.88 ..8.8888 - - - - - - .8 888.888 888.888.8 88 888.. 888.8 .8 888.8 8.888.8 88 888.88. .8.888 .2888.888 - 888.. 888.8 88 888.8 8.88..8 888 888.888 88.8. 88888.8 - 888.. 888.. .. 888.8 8.888.8 88. 888.888 8.888388. 8.8 888.. 888.. .8 888.8 8.888.8 88. 888.888 8.88.888 888.8 - 888.8 888.8 88 888.8 8.888.8 88 888.88. 88888 - 888.. 888.8 8.. 88 888.8 8.88..8 8.. 88 888..8. 8888.888 8888888 “Name?“ 8......8moa8. .mwdmag 88888888.”... .. “8%. 38.84.8888 memoms E8. .8... .383... .8883... . 3 .8 .8858: 8.88888. 888Lo>< $8 .0838: no... 8.8m ou~go>< 888582 8.-8.8. 8. 888. .8 8.88 888 888888 8888.88-88...z m... m—nmh 232 Table 7.18 Number and Scheduling of Loans to Village Cereal Banks Number Year of Loan VCB Dates of Loan Installments Installments 1977 l Piela ll/5/77 2 Tapoa ll/28/77 NA 2 Logobou 12/8/77 NA 1978 l Bourgou 2/26/79 1 Samou 2/28/79 2 Koulwoko 2/2/79 3/2/79 2 Tampoudin l2/2/78 2/9/79 2 Tapoa l/2/79 2/3/79 2 Boulel l/2/79 2/3/79 3 Diagorgou l2/l9/79 l/3l/79 3/26/79 l Boudieri 12/16/78 3 Sakpani l2/l5/78 l/l2/79 2/22/79 3 Toussiegou lZ/l6/78 2/3/79 3/2/79 3 Gayeri l2/18/78 l/l9/79 2/8/79 . 2 Ougarou 12/18/78 l/l9/79 - 2 Yenkoali l/lZ/79 2/2/79 l Kikideni 3/9/79 2 Tindangou l/24/79 March 79 Source: l980 Village Cereal Bank Survey. NA: Not available. 233 Another dimension of the procurement and sale of grain is the extent to which members of the Village Cereal Bank sell to and buy from their cooperative. Interviews of the members reveal that these sales and purchases were not very important. The small volume of sales back to Village Cereal Bank members, obviously, is due to the small member- ship of the Village Cereal Bank relative to the volume of grain the loan can buy. As for the sales, some members were not aware that they could sell their grain to themselves and buy it back later. 7.5.5 Viability of the Village Cereal Bank System To be viable, the Village Cereal Bank should be able to cover costs, make a normal profit, and pay back the loan on time. The evalua- tion of the profitability of the scheme requires a financial analysis of the project. Unfortunately, neither the 0RD nor the Village Cereal Bank members kept any records of the transactions. The E-ORD criterion of evaluation was not the Village Cereal Bank's profitability, but rather its ability to repay the loan. Table 7.19 shows the loan repayment rate for the Village Cereal Bank operating in 1977-78 and l978-79. Ignoring the_fact that Logobou paid back more than two-thirds of its loan after the due date, the re- payment rate in 1977-78 was very impressive (93.3 percent). The fol- lowing year, however, the loan repayment rate had dropped to 47.7 per- cent. In fact, only three Village Cereal Banks had repaid back the en- tire loan plus interest, while four Village Cereal Banks had not repaid anything, up to 1980. Thus, on the basis of the repayment rate, the Village Cereal Bank had not performed very well and this has contributed to the weak support of the Village Cereal Banks by donors in 1979-80. 2294 Table 7.19 Loan Repayment Rate of VCBs b Total Amount Not Yet Annual m (32:2, ‘33:? 35:21.23 ...... (FCFA) (FCFA) (Percent) 1977-7a Piela 600.000 33.539 633.539 - Tapoa 600.000 32.153 632.153 - Logobou 600.000 40.194 512.329a l27.865 162.1 1,300,000 105.886 1,778,021 127,865 93.3d 1978-79 Bourgou l57.850 6.435 164.285 - Samoa 150.000 8.055 78.904 79.151 Koulwoko 600.000 37.392 150.000 487.392 Tanpoudin 300.000 8.091 308.091 - Tapoa 480.000 15.474 495.474 - 060131 130.000 10.283 ‘ 162.000 28.288 Diargorgou 200,000 l2.879 52.460 160.419 Boudieri 30.000 5.593 ' - 85.493 Sakpanic 200.000 9.420 207.800 1.620 7603313960 154.140 9.326 100.000 63.466 Gayeri 600.000 38.970 480.000 158.970 Ougarou 500.000 32.636 - 632.636 13.118311 500.000 32.932 58.000 474,932 1112108311 200.000 10.345 - 210.345 Tindangou 200.000 12.274 - 212,274 18:31 4.501.990 250.110 2.257.014 2.495.086 47.4e a353,587 was repaid after the I978 due date. in l979. bThe due date is March 3l. every year. cSakpani acknowledge having paid only 100,000. which is consistent with the income from the total number of bags (20) they sold for cash. dComputed as: 100 . 1.773.021/1,905.386. eComputed as: 100 ~ 2,257.014/4,752.100. 235 7.5.6 Problem Diagnosis The poor performance of the Village Cereal Bank stems from the organizers' misunderstandings of farmers' real needs. In essence. there are conflicting objectives set for the Village Cereal Banks. This poor performance is related also to the fund's availability and timing of delivery. the managerial capability of the Village Cereal Bank. and the role of the Village Cereal Bank in relation with other institutions in the Eastern Region. Misunderstanding and Conflicting Objectives The results presented in previous chapters suggest that an impor- tant number of producers are not producing enough grain to be self-suf- ficient. However, there was no strong indication of either harvest distress sales or important advanced sales as far as the number of par- ticipants and volume dealt with are concerned. Rather, given the export orientation and other characteristics of the grain marketing system. there is a problem of the reverse flow of grain to producer-consumers through market intermediaries. But one conflict arises because the availability of grain in the village during the hungry season is valued differently by Village Cereal Bank members of different age groups. Elder members who are heads of households and obligated to provide food to the households value the availability of food grain more than the younger members who don't have this responsibility. For these young members. the profit motive is stronger than food security and, given the large size of the Village Cereal Bank loans relative to village membership, many have attempted to invest in other and more profitable commodities such as rice, peanuts. and shea nuts (see Table 7.l6). 236 Many members were not also clear on how the financial reserves would be built in view of the fact that some activities of the business (pharmacy and general store) may be risky. Indeed, there may be another conflict between the build-up of an independent financial basis and the provision of services that do yield little, if any, monetary bene- fits (dispensary). Some members thought that the credit would be avail- able on demand and not compulsory as the phase out suggested. In another area still, the goal of price stabilization may conflict with the profit motive. I Funds' Availability and Timing of Delivery The availability of funds is a major problem in view of the limited resources of the E-ORD. At the same time. one cannot justify providing funds to Village Cereal Banks regardless of its membership unless it is expected that by so doing the precooperative will attract more members. As we have seen, no such effect took place. The numerous (30) village groups that have established their own Village Cereal Bank may be an indication that producers are willing to pay for the scheme. What may be needed is a meaningful contribution of each village group. The untimely delivery of the loan installments has had an adverse effect on the performance, as it did on previous E-ORD marketing cam- paigns. For example, the last Village Cereal Bank created (Zembede) got its second installment in May when no grain was available at a price that could have provided an opportunity for resale and payment of all costs. Although there is no strong indication of harvest distress sales, late Village Cereal Bank purchases caused by late loan delivery reduce the likelihood that members of the village group would be able to supply the Village Cereal Bank. This prevents the local village economy from 237 capturing all the benefits of the scheme which entails purchasing grain from members at higher prices than traders. Managerial Capabilities of the Village Cereal Banks At the outset. the sheer volume of the loan confused the members of the village group. many of whom had never seen so much cash at once. The money was allocated to two to four members to purchase the grain. When the loan came late. they had to move farther away from their vil- lage to find available supplies. and some (from Tampoudin) even went to Ouagadougou. And it also happened that appointed buyers embezzled the funds. The Village Cereal Bank bought the grain at many sources and usu- ally not from their village. let alone from their members. (Exceptions are the Village Cereal Bank of Ougarou which had only to go to the next big market, Nassougou; and the Village Cereal Bank of Gayeri which had to go to the next village, Bassieri.) Therefore, they had great logistical difficulties in trying to bring the grain to one place. Only a few Village Cereal Banks (Bourgou, Samou. Ougarou, and Gayeri) were able to keep the grain under one roof. while the others were forced to use two to four different locations to store the grain. Later on, the business of selling the grain was complicated because, again, too many sellers and too many selling points were involved. making the control difficult. if not impossible. The grain is usually stored in bags in one member's house. Some Village Cereal Banks. however, have built their own storage facilities as promoted by the program. Others have used available public facilities such as the 0RD warehouses, school warehbuse. and in one occasion 238 (Bourgou) a concrete silo built during colonial times. Except for the concrete silo, farmers felt that the storage facility used would be adequate only if the grain is stored for a short period and sold right on time in July-August. They estimated that they would lose the grain if it stayed two to three seasons in the members' houses. (In effect, nearly all the stock of the Village Cereal Bank of Tindangou, which was not sold after two years, was rotten.) If the grain is stored just for a few months, the storage loss is low coming only from humidity and the domestic animals, rather than from termites.‘ (Usually, only three doses of insecticide--HCH--are used.) The Village Cereal Banks who had their own storage facilities had built them just like dwelling units with tin tiles. The E-ORD manage- ment sometimes helps in the transportation of the tiles and sometimes donates the cement, but there is not much emphasis on how to help im- prove the storage conditions on farm. No one farmer would think of building the same facility to store the household grain for an extended period of time. Finally, the grain is often sold at credit and there does not seem to be much pressure on the members to pay back. They tend to think that the grain is not so much theirs as it is the E-ORD's. In many in- stances. they also feel somewhat safe in not paying back since the exten- sion agent has also taken the grain at credit and not repaid Market Policy Coordination A grain reserve scheme operated for the needs of rural producers and consumers is very risky because of the precarious logistical condi- tions. the lack of transportation. and the lack of effective storage facilities once the grain has to be held the entire rainy season. 239 There are financial risks as well due to the variability of prices. In the mind of many members of the Village Cereal Bank, since the 0RD is the promoter of the idea. it should also help coordinate the difficult instances when the hungry season is passing by without the grain being sold. But the role of the Village Cereal Bank in relation to OFNACER and the private sector was not thought through. Forced by the flaws of the loan installments or the lack of local supply. the Village Cereal Bank buys outside the village. Likewise. it sells to other villages because the grain bought often exceeds its members' demand. But when there is not much demand from its members and other consumers in neighboring villages (because of good harvests and consequent stock build-up). the access to far away markets becomes important. Logobou, in l977 and Tindangou. in 1978-79 faced these prospects. Members of the Village Cereal Bank of Tindangou had re- quested in vain a truck from the 0RD to move the grain to Fada or other deficit areas. The grain that stayed too wet seasons in inappropriate storage facilities was lost. 7.5.7 Synthesis The Village Cereal Bank program is a multifacet project of which the marketing cooperative and consumption credit dimensions have been adopted by producers in the E-ORD. The poor performance of the Village Cereal Bank financed by the E-ORD can be traced to the poor definition and implementation of the concept. Producers have not clearly under- stood what their roles and responsibilities should be. and lack the managerial ability to run a reserve system when the lax control by the E-ORD management compounds the logical problems and encourages embezzle- ments. 240 All these factors show the important need for further research and understanding of what roles private and well-run public firms should play in order to promote a financially viable operation in the face of the tremendous uncertainty and logistical problems. OFNACER can play an important role in identifying useful marketing services and new mar- keting practices which need to be supported by market participants. In turn, the Village Cereal Bank must reduce its ambitious objectives, tighten its control, and provide members with the managerial tools to conduct the operations themselves. 7.7 Summary Grain producers in E-ORD are aware of a large spectrum of market- ing problems. It appears, however, that they are not given enough op- portunity to discuss them with the E-ORD extension agents. Overall, their attitude towards the E-ORD marketing campaigns is negative because of organizational failure. 'In addition, a great number of producers are not even aware of OFNACER's existence or think that the agency is not interested in their crops. Given the poor infrastructural support (roads and markets). it is not surprising that market access ranks high in producers' concerns. In spite of the fact that some producers do not know of OFNACER, the agency's market share in the areas and periods it intervenes looks some- what higher. at 20 percent. than the ball park figure of lO-lS percent. Unfortunately. a comparison between OFNACER's market share and that of grain merchants is not possible because the questionnaire design did not account for the merchant's agent network. One major concern of public authorities concerns the shortcomings caused by nonstandardized measures. Grain producers are also aware of 241 these shortcomings. Though the empirical evidence shows mixed results as to the producers' perception of market unfairness related to grain measurement, it indicates that producers show preferences for widely used and convenient measures they can trust. Indeed. there is a dis- tinct differential between a small measuring device (louche) used by assemblers and producers-retailers. However, part of this differential represents producers' time opportunity cost and wholesale-retail price margin. At any rate. assemblers themselves would rather use larger measures. Thus. there seems to be a genuine opportunity to promote more effectively standardized measures that meet the needs of market par- ticipants. Another major concern of public authorities is the alleged prac— tice of "green sales," actually advanced sale arrangements. Advanced sale prices quoted for harvest deliveries may be low when compared to actual harvest prices, but the high imputed nominal interest rate of credit derived is_mostly due to the very short repayment period. How- ever, these figures should be put in perspective because of the fact that the magnitude of advanced sales among producers is rather limited in terms of number of participants and volume of grain. The possible. high default rate does not encourage merchants to practice the arrange- ment on a large scale. The credit arrangement between producers and traders or among pro- ducers themselves (mostly without interest) takes place because they fulfill a genuine need for farmers often faced with low food availabil- ity in the "hungry season." For this reason. the concept of the Village Cereal Bank has caught on with many producers in the region. The 242 Village Cereal Bank project. however. is itself ill-conceived and ill- managed. and runs the risk of deepening the mistrust of producers toward the E-ORD. All of these results show that a well-run public agency can play an important role by identifying useful marketing services and new marketing practices which need to be supported by the main groups of market participants. There is a real need for more research to help discover ways to help village groups. private traders, and OFNACER improve food availability and security in the E-ORD. CHAPTER 8 SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS 8.1 Study Objectives The overall objective of the study was to improve the understanding of the processes of the rural food grain marketing system and of pro- ducers' grain marketing behavior in Eastern Upper Volta. The specific research objectives were the following: (1) to review the major food grain marketing issues in Upper Volta from the country's independence (1960) to the late 19705; (2) to describe and analyze the rural food grain market channels and processes in terms of the organization and standard operating procedures of the major market participants; (3) to analyze producers' grain marketing behavior at the village level; (4) to provide a diagnostic analysis of farmers' market linkages and problems; and (5) to suggest policy recommendations to improve the performance of the system and to identify future research needs. The study was part of an applied microeconomic research effort con- ducted by MSU in the context of an E-ORD Integrated Rural Development Project financed by USAID. A farm survey of 480 farmers conducted from May 1978 to April 1979, and a market survey conducted from July l979 to June l980 provided the main sources of information (see Chapter 4). 8.2 Major Research Findings The major research findings can be grouped into six broad areas of interest: (l) historical perspective of food grain marketing issues in 243 244 Upper Volta (see Chapter 3); (2) characteristics of the E-ORD's food grain marketing channels (see Chapter 5); (3) farm-households' food grain flows and transaction patterns (see Chapter 5); (4) timing and farm gate prices of farm grain sales and purchases (see Chapter 6); (5) farm-house- holds' cash flows (see Chapter 6); and (6) selected food grain market- ing problems and issues in the E-ORD (see Chapter 7). 8.2.1 Historical Perspective of Food Grain Marketing Issues in Upper Volta Active participation by Voltaic governments in agricultural market- ing concerns can be traced to the early years of the country's in- dependence with the establishment in 1964 of a stabilization fund for export crops. But direct public interventions in the buying and selling of grains can only be traced to the early 19705 with the establishment in 197l of the National Cereal Board, OFNACER. In the mid lQSOs, private traders gained an access to the assembly of export crops and the capital thus accumulated, through the early 19605, enabled many to expand dramatically their scale of operations in grain marketing. Monopoly rights were granted in l968 to the Regional Development Organizations (0RDs) for the assembly of export crops, and in 1974 to the ORDs and OFNACER for the assembly and distribution of food grain. These government attempts to monopolize agricultural market- ing failed because of poor logistic planning. the lack of coordination among numerous public agencies. and strong competition from the private sector. Through each of these attempts, the importance of private mer- chants grew stronger so that in the late l970s government policy has become more liberal toward the private sector. There is. however, strong evidence to suggest that government active participation and sometimes 245 direct intervention in grain marketing will continue. Better informa- tion about the actual workings of grain markets needs to be provided to make this intervention more effective and efficient than in the past. 8.2.2 Characteristics of the E-ORD's Food Grain Marketing Channels The major participants in grain marketing in Eastern Upper Volta performs several functions. Farmers are grain producers. but they are also known to retail their crops. to play the role of middlemen. and to buy back food grain for consumption. Private traders may engage both in wholesaling and retailing and are thus best categorized on the basis of the title to the grain (merchant and agent middlemen). the renumeration of the agent middleman (trader's apprentice. commission agent, village resident buyer, and landlord), and the trading range (national whole- saler-retailer. regional wholesaler-retailer, and local independent trader). The government cereal marketing board (OFNACER) is active in grain marketing. but the E-ORD management is not. Truck transporters also play an important role in grain marketing. The grain exchange system in the E-ORD is a mixture of house trad- ing among village households. farm gate oriented grain assembly, and marketplace oriented trading. The farm gate-oriented grain assembly is particularly adapted to the E-ORD local conditions. The region has a sparsely and unevenly distributed population, a very poor road accessi- bility. and a rainfall-induced instability of grain production which all translate into small rural markets (l78 markets for 635 villages) most of which are periodic. or even more temporary, and (65 percent) serve but one village. Under these conditions. quite a few transactions must take place outside the marketplaces whereby grain merchants set up a 246 network of agents to collect grain (and other agricultural products) at the farm gate. The description of the food grain marketing system reveals that market coordination in the private channels is hampered by the lack of working capital and the mistrust among participants which translate into costly control procedures of the agent network. OFNACER intervention in the system at the farm gate is intended to directly benefit grain pro- ducers. At the marketplaces OFNACER purchases from small local independ- ent traders put some competitive pressures on national wholesaler-re- tailers and creates indirectly some effective demand for farmers' prod- uce. OFNACER purchases from merchants cost very little in terms of logistical set up, while allowing private traders to reach producers in more isolated areas. Both private and public marketing networks are geared at;exporting grain outside the E-ORD and are less effectively structured at serving rural consumers in the hungry season and/or in isolated areas. For example, food grain is available in only 33 percent of the rural markets in the hungry (rainy) season. and there are few agent networks set up by grain merchants to distribute grain to rural consumers at the farm gate. Thus, rural consumers in the hungry season and/or in isolated areas must rely on their own stocks or the house trade which distributes local grain surplus among village households. For illustrative purposes. the study constructed physical distribu- tion costs (buying, handling. transportation, and storage) incurred by private traders. Transportation costs account for up to 80 percent of total marketing costs. when merchants' own time costs are left out. In most cases. the costs of storage are small because traders emphasize a 247 rapid turnover of the stocks to cope with the lack of working capital. and/or limited stocks to cope with the low concentrated effective demand of the sparse and poor rural populations. 8.2.3 Farm-households' Food Grain Flows and Transaction Patterns The grain flow statement was used as a practical tool to gain a better understanding of producers' behavior with respect to grain mar- keting (see Chapter 6). The grain flow statement compares all sources of grain (production. purchases, gifts, credit, and wages) to all uses of grain (sales. gifts, credit. labor payments, farm uses, and consump- tion) and derives a change in stock as a residual under the assumption of a minimum annual consumption requirement of 175 kg. per capita. Per- formance measures. and marketed and marketable surplus are estimated for households located in five ANTRAC zones of the l978-79 farm survey. Despite some data limitations, this analysis helps bridge some gaps in the understanding of producers' marketing behavior in the E-ORD. Overall, a few hundred kilograms are dealt with and these rela- tively low levels of production and exchange flows convey very well the modest level of agricultural development of the average farmer in Eastern Upper Volta. Although ANTRAC households seem to deal with more volume than hoe (traditional) counterparts, we must keep in mind that such a difference is largely due to the larger family sizes of ANTRAC households. For the most part, the grain flow analysis shows a zonal heterogeneity in which the large producing zone of Ougarou contrasts with the drought- stricken and low-producing zone of Piela. In three out of the five zones selected for the study, grain stock by the end of the year has been decreased relative to the previous year's level. Because the change in stock is derived under the assumption of a 248 minimum consumption requirement. when the size of the stock depletion is large relative to the year's production, we infer that the household may have actually consumed less than the annual minimum requirement of 175 kg. per capita. This possibility is strongest for the average household in Piela Zone. In all the zones studied, the results of a sensitivity analysis tentatively set the estimated average annual grain consumption between 165 kg. and 190 kg. per capita. An important result of the study is that in all 11 villages except one, the average producer bought more grain than was sold in 1978-79. That is. in only one village does the average producer show a marketed surplus. As for marketable surplus (which include a stock build-up potentially available for sale) in only two villages does the average household show a surplus. Unfavorable weather conditions may explain the low production. But these heavy purchases should dispel the myth that grain marketing in E-ORD can only be viewed as extracting grain from producers of the region. Another important result from the grain flow analysis is that the amount of grain given out through barter may be more important than farm sales. Thus, a narrow definition of marketable surplus is likely to provide meaningless results. The analysis also shows that overall, the amount of grain given out to pay for past debts (which include loans taken from traders) is less than that given out as gifts. When the data on farm sales (out of farmers' production) and house- hold purchases (for farm use and home consumption) are disaggregated on a household basis. the results show four patterns of grain transactions: In the five ANTRAC zones studied. 14.3 percent of farmers have had no grain transactions at all. 50.5 percent have bought, but did not sell. 249 9.7 percent have sold. but did not buy, and 25.5 percent have bought and sold grain in 1978-79. In the last category. almost 80 percent of the households had repurchased grain in excess of their farm sales. Thus, not only have producers bought more grain than they sold as shown by the grain flow situation, but also more producers have bought or have had larger purchases than producers who have sold grain. The analysis also indicates that households who did not sell any grain were the ones with the poorer grain harvests in 1978-79. These results help explain why grain transaction patterns vary significantly across ecological zones. The results suggest that farm grain sales in- crease with the household income bracket and show inconclusive evidence on the impact of market location on the pattern of transaction, but these results were presented with caution because of some data limita- tions. When compared and contrasted. the major significant difference in the behavior of ANTRAC and traditional households is in the pattern of purchases: Traditional households had heavy grain purchases (61 to 68 percent) in the hungry season, as would be expected. while ANTRAC house- holds had only 41 to 54 percent of their purchases in the same period. But for both households. the evidence does not suggest any grain distress sales at harvest: Only 22 and 35 percent of the total grain was sold during the harvest period, while 63 percent and 57 percent of the grain was sold in the postharvest period. 8.2.4 Timing and Farm Gate Prices of Farm Grain Sales and Purchases The evidence on the timing of sales and purchases also indicates that the groups of households with only farm grain sales and those with 250 only grain purchases tend to have a regular pattern of transactions throughout the year. For the group of households with both sales and purchases. the timing of transactions shows heavy purchases in the hungry season. but farm sales are still mainly concentrated in the postharvest. The timing of sales and purchases further indicates that in all three income classes. grain purchases are made in a somewhat regular pattern throughout the year. There is a slight indication that middle income farmers have heavy sales (42 percent) in the harvest period. but even then an equal amount is sold at a regular pace in the postharvest period, thus showing no strong indication of harvest distress sales. As expected. farm gate prices are lowest at harvesttime and highest at the hungry season. For the period studied, the average farm gate price is lower than the official producer price. Overall, farm gate prices paid are higher than farm gate prices received, but the two are not significantly different in the hungry season. In other periods, the two series tend to move together. which is consistent with the fact that for numerous households the timing of sales and purchases are somewhat similar throughout the year. Because the timing Of sales analysis shows that the average house- hold does not sell disproportionately more grain at harvest than in other periods (that is, there are no indications of harvest distress sales) the negative effect of low producer prices at harvesttime on the average farm income is likely to be dampened. 8.2.5 Farm-households' Cash Flows The evidence from the cash flow analysis indicates that sales of crops (food and cash crops combined) are a relatively minor part of the sources of the farm family cash revenues. At most. crop sales make up 251 40 percent of cash revenues in November-December (harvest) for house- holds who have no transactions in grain. (Grain still remains the most important farm crop sold by the average farmer in E-ORD.) In 1978-79. as indicated by the negative cropping cash surplus, the average house- hold in the zones studied was not able to cover its food purchases out of its net cropping revenues. Hence. the introduction of animal trac- tion which had not had any dramatic impact on yield put added cash con- straints on farmers.1 The small contribution of crop sales to cash revenues convey farmers' low monetization level with respect to crops. but not necessarily with respect to other activities. In most instances. the most important source of cash is livestock. But other activities such as trade in agri- cultural products. agricultural processing and nonfarm and related labor sales type activities also contributed significantly to cash income. The cash flow analysis also shows that cash outlays for foodstuffs are relatively more important for traditional households than for ANTRAC households. In absolute terms. however, ANTRAC households still spend more cash on food than traditional households because ANTRAC households have larger family sizes. But for these ANTRAC households even more cash is spent on cropping enterprises, livestock raising, and ANTRAC equipment. Thus, ANTRAC households are forced into more cash outlays because of ANTRAC adoption while the cash generating power of the crop- ping enterprises remains very small. 1See Lassiter (1982). 252 8.2.6 Diagnosis of Food Grain Marketing Problems and Issues in' the E-ORD The diagnostic analysis of farmers' market linkages and problems examines farmers' opinions on marketing problems, the issues of grain measurement. the issues of marketing credit and advanced sales, and the performance of the Village Cereal Bank program (see Chapter 7). The analysis finds that producers are not given enough Opportunity to discuss marketing problems with extension agents. This. and the ORD's past failures and unkept promises can be suggested as the reason why very few producers (26 percent) would still want to sell to the E-ORD. Some 12 to 16 percent of producers who have not sold to the OFNACER have never heard of the agency before. some 25 to 34 percent think that they have too small a volume to sell, and some 60 to 45 percent wanted to sell. but could not gain access to the agency. The survey reveals than an access to a secure market [through road improvement or the presence of public (or private) buying agents] is an important dimension of the marketing problems perceived by producers. The ORDs and OFNACER have long sought to promote standard units of .measurement in order to make the handling of products during transactions more efficient, to help prevent possible wrongdoing by those who measure grain during transactions. and to make the dissemination of market in- formation easier. The analysis reveals that 80 percent of heads of households inter- viewed on the farm usually measured grain before taking it to the market. In addition, regardless of whether they have measured grain on the farm. these heads of households think that they usually receive at the market the grain value expected. But another group of farmers (made up of heads and other members of households) interviewed at the marketplaces indicates 253 that less than half of them had actually measured grain on farm prior to the market visit. In addition, those farmers who have measured grain on the farm tend to have higher expectations about the value of market sales than those who have not. The study hypothesizes that this dif- ference is due to the fact that heads of households have more responsi- bility. more volume of grain to sell. and possibly better market informa- tion than the average farmer interviewed at the markets. Producers at marketplaces have the opportunity to retail their products themselves if they do not like traders' measuring device and/ or price offers. An experiment conducted at Namounou shows that the non- standard devices (louches) used by merchants and producers differ sige nificantly. It was not possible. however, to determine whether the wholesaler-retailer price spread in the same market on the same day for similar products justified traders' marketing costs and normal profit return. It appears clearly that producers are aware of the benefits of standardized measures and traders will welcome larger units than the small calebash (louche) used in Namounou. However, the promotion of the official ZO-liter ting and the scale is not taking hold. Producers' preferences are for a trustworthy standardized measure which would be convenient to use and would have large acceptance in the area. Credit transactions cover many dimensions in the rural economy of the Eastern Region. but the ones between producers and merchants are of particular interest because of the public perception of producers being exploited. The analysis helps put a better perspective on the alledged traders' exploitative behavior through an analysis of the importance of advanced sales, which are often inappropriately termed "green sales." 254 The results show some arrangements are used whereby traders lend money or sell grain on credit to producers for payment in kind at harvest. The imputed interest rates on these arrangements may be quite high mostly because of the very short repayment period. But the magnitude of the advanced sales itself is very small in terms of the number of producers involved (from 8 to 14 percent) and the volume involved (from 70 to 100 kg.). There is also evidence to suggest that such arrange- ments are very risky because in one instance up to 50 percent of the producers defaulted on or postponed the payment in kind. Obviously these risks reduce the likelihood that the practice will spread beyond its present limited use. Advanced sales arrangements (between producers and traders) and other informal credit arrangements are used by some producers to gain access to food, especially in the "hungry season." Thus, the idea of a Village Cereal Bank scheme which would provide credit to farmers to buy grain and reduce food shortages in the hungry season is accepted by many village groups. But the performance of the Village Cereal Bank scheme as supported by the E-ORD through a loan program financed by USAID and other donors is very disappointing. There is no indication of increased membership of the precooperatives because of the scheme. the repayment rate is very low. and members are confused and disenchanted with the project. This performance was caused by the rather poor definition and implementation of the concept: The many and sometimes conflicting objec- tives of the Village Cereal Bank are confusing. loans are not delivered timely. and cooperative members have no managerial talent to keep opera- tional costs low. The whole scheme runs the risk of being rejected by producers even before it has the chance to be adequately tested. 255 8.3 Recommendations Policy actions are recommended in broad areas. on the basis of an improved understanding of the E-ORD marketing system and under the as- sumption of continued public involvement in grain marketing. 8.3.1 Improved Sorghum-Millet Varieties and Farming Techniques An effective marketing system articulates consumers'effective demand and provides economic incentives to producers to increase food produc- tion. However, the low level of commercialization in the E-ORD is clearly related to the relatively low level of production. The extent to which food production can be increased is limited among other things by the biological potential of food crop varieties. To provide abundant food supplies in E-ORD. the food grain biological potential and farming techniques must be improved along with the improvement of the marketing system. 8.3.2 Physical Infrastructure Given the poor rural infrastructure of the E-ORD, there is no doubt that a better road system will increase the performance of the grain mar- keting system in the E-ORD by providing access to markets and by reducing transport costs. In that respect, the USAID financed road project in the Namounou area will increase farmers' access to the large rural mar- kets of Namounou and Diapaga. But much more still needs to be done (for example in the Pama area) to increase road accessibility in the E-ORD. 8.3.3 The Marketing Role of the E-ORD Management Past experiences indicated that the E-ORD management lacked the financial. managerial, and logistical capabilities to set up a sound food grain buying and selling network region-wide. In the foreseeable 256 future these capabilities are likely to remain a constraint. To obtain maximum effectiveness of the limited human resources, we recommend that the marketing role of the E-ORD be focused more on the following areas. First, do more to provide marketing advice to producers. At present. there is almost no marketing information exchange between producers and extension agents. Extension agents can help producers better plan their sales and purchases by providing market intelligence on prices and supply conditions in important centers of the E-ORD and neighboring regions. Second, facilitate grain transactions between OFNACER and farmers precooperatives if OFNACER continue 1x1 buy grain. What will be often needed will be for the extension agent to inform farmers of OFNACER buy- ing campaigns, and help OFNACER agents get access to potential sellers. Such an arrangement (which takes place in Madjoari for example) helps OFNACER agents. often ignorant of the local conditions. reduce logistical costs. Third, the ORD should consider actual buying operations only if these purchases are meant to promote a new variety or crop. The E-ORD must then stand ready to subsidize part of the marketing cost of the new crop until stable (private) market channels develop. If there should be no proven market access in sight. then such a new crop should not be introduced. For example, soybeans were introduced in the past without a proven market. The first year the E-ORD bought the crop for a good price to producers. But the next year, the price was cut in half and the third, the E-ORD ceased to buy the crop. Such an experience contributes only to develop further producers' mistrust toward the E-ORD. 257 8.3.4 The Role of OFNACER Given that in the E-ORD the back-flow of grain is not well struc- tured and that there is no strong evidence of distress sales at harvest, OFNACER may want to reconsider its concentrated purchases at harvest for export outside the region. For example, OFNACER can emphasize more than in the past to reach back to rural consumers in the E-ORD. At present, OFNACER sales to these consumers are limited only to times of crisis and take place only at important centers. To reach consumers in remote areas, OFNACER may play the role of wholesalers and rely on private re- tailers who have closer contacts with these consumers. Second. OFNACER can play an important role by researching areas of marketing improvement and by assisting private traders in the adoption of these improved techniques. possibly in the areas of bookkeeping and logistical cost control practices. In effect. OFNACER would be extend- ing services to private traders the way the E-ORD extension agents are supposed to assist farmers. At present, unfortunately, OFNACER itself lacks the managerial know-how to pass it to traders. A training program along these lines for OFNACER agents could be undertaken. There is need for better information to estimate the full costs and benefits to all concerned participants from OFNACER actual purchases in order to recommend in which areas at the farm gate or at marketplaces. OFNACER should concentrate its purchases. In most areas, OFNACER farm gate grain purchases are welcomed by producers because the official price is higher than market prices at harvest. But OFNACER operational costs are likely to be higher than private traders' agent network operations. Many traders benefit also from OFNACER marketplace purchases. These traders gain access to working capital and thus expand their scale of 258 Operations in remote areas to providing better opportunities to producers. In these marketplace purchases. OFNACER's logistical costs are reduced, but producers receive somewhat less than the official in order for trad- ers to cover marketing costs. 8.3.5 The Village Cereal Bank Program In some areas producers face food shortages in the hungry season. The objective of the Village Cereal Bank should be more sharply defined to help fulfill this need of food availability. But since different members value this need differently. proper screening of the Village Cereal Bank membership will be needed to increase adherence to that single objective. Also, no Village Cereal Bank should be established if the members have no willingness to pay for part of the scheme. This will insure that members will regard the grain stored as theirs rather than the ORD's. Furthermore, if a loan is to be provided. it should be given on the basis of the village group membership. OFNACER and the E-ORD assistance will be crucial to insure a viable Village Cereal Bank. What is needed most is better management and con- trol. awareness of overall logistical costs, and a mechanism for selling stored products in reliable markets outside the village should no effec- tive demand materialize from local rural residents. The E-ORD management could help Village Cereal Bank members set criteria that would indicate (preferably before the rainy season) whether there would be enough local demand for the grain of the Village Cereal Bank. OFNACER or private traders may be contacted then to buy off the stock in order to avoid potential heavy storage losses when the grain is stored over the rainy season. Alternatively. storage rooms could be 259 built so that the grain can be kept for two full years with minimum damage should the early warning signal fail and roads be made impas- sable in the rainy season. However. this Option would be much more ex- pensive and require ample justification. 8.3.6 Standardization of Measurement Standardization of measurement can be promoted by OFNACER and the 0RD in the E-ORD through the following measures: First. public of- ficials should provide standardized volume measures with appropriate sub- units. For example. 10-liter, two-liter, and one-liter tings should be made available to replace "Yoruba plates" and small calebash. Second. promotional campaign should be conducted in key marketplaces and days to demonstrate the superior reliability and trustworthiness of the standardized units over the nonstandardized ones. Third, steps should be taken to enforce the use of standardized measures. Traders should be required to use the standardized measure. Deliverance of license or registration of trader could be made subject to the possession of the devices. Also, the traders' union may be requested to help enforce the measure. The recommended actions command a rethinking of the relationship between public officials and private traders. In many instances, public officials cooperate, if only reluctantly. with the private sector. This cooperation should be more emphasized by public Officials. They should seize the opportunity to work with private traders to help find ways to police the trade and enforce contracts. OFNACER and the E-ORD manage- ment, in their respective domains, should consider taking the lead to do this. 260 8.4 Further Research Needs This study has increased the understanding of the food grain mar- keting system in E-ORD. It is hoped that this understanding will be a useful comparative ground for the USAID/CRED food grain marketing study currently underway in Upper Volta. But this understanding remains only a part of the overall system picture referred to in Chapter 1; hence. more extensive understanding of the whole system is needed. For example, we need better estimates of the flows of grain through the various dis- tribution channels and an assessment of the degree of coordination and efficiency in these channels. Further detailed information is needed in the area of storage at the farm level and by private traders. Storage at the farm level has obvious implications for the Village Cereal Bank scheme. For how long, under which storage conditions, and how much can farmers keep grain in storage as a security against future food shortages? In areas where farm level storage is adequate, the Village Cereal Bank will not be needed as a reserve scheme. Storage conditions at the private merchant level have also to be examined in connection with the Village Cereal Bank and the back-flow Of grain to rural consumers. Further detailed information is needed also in the area of market- ing costs incurred by private traders and public agencies. Such informa- tion is needed to determine the best combination of OFNACER intervention in the E-ORD, and to determine acceptable marketing margins charged by OFNACER and private traders. These studies should be accompanied by an in-depth analysis of public and private enterprise management in the E-ORD to identify cost-saving practices. 261 Finally, there is a need for long-term gathering of key important statistics in the E-ORD. Producer and consumer prices in key producing and consuming areas are still sorely lacking in E-ORD. Also. the ef- forts of the 1980 village inventory census should be pursued to update the socioeconomic information it provided. The design of further marketing studies to be conducted in E-ORD should be built around the following major points emphasized by this study. It is important to properly identify merchants' agents in the village studied. Figures of volume exchanged cannot be accurately esti- mated by relying only on transactions that take place at marketplaces. In order to follow producers' disposals of grain, the research should extend over 15 to 18 months to cover one production and one marketing season. And an attempt should be made to measure beginning and ending stock. It will not be possible to abandon totally the cost-route ap- proach, except in measuring production, seed used, and storage. The overall approach should.emphasize the identification and diagnosis of unexploited economic opportunities and barriers, and the identification and analysis of behavioral and institutional changes to improve the per- formance of the marketing system. At the farm level, the behavior of producers should be studied in the context of the overall farming system. APPENDIX A APPENDIX A PHYSICAL DISTRIBUTION COSTS OF GRAIN MERCHANTS OPERATING IN EASTERN UPPER VOLTA: AN ILLUSTRATION A.1 Local Independent Trader (LIT) The physical distribution costs are illustrated for actual LITs located in Namounou. But by looking into three different cases we can cover reasonably well the major roles of LITs in the Eastern ORD. In case one, the LIT buys in other markets and sells to larger traders at Namounou. In case two, he buys at Namounou and sells there to larger traders. In case three, he buys grain at Namounou and stores it for re- sale later in the rainy season. A.l.l Case 1 The representative market outside Namounou is taken to be Logobou, a weekly market open on Wednesdays. located behind the Gobnagou hill range at 45 kilometers or so from Namounou at the other end of the road Namounou-Logobou that USAID was then helping to upgrade. At Logobou, the LIT is helped by teenage boys who bring him sellers and bag the grain in 100 kg. bags for FCFA 50 a bag. The merchant buys grain until dawn and hauls it to a friend or the transporter's compound for the donkey cart owner to carry it to Namounou for the Sunday market day. Hauling the grain from the market stall to the warehouse costs FCFA 25- per 100 kg. bag. Overseeing the grain costs another FCFA 25 per 100 kg. bag regardless of the period over which the grain is stored. There 262 263 are no specific charged for loading onto the cart or unloading Off it and one may assume that they are included in the transport charges. The out-of-pocket costs for such an Operation are presented in Table A.1. A.l.2 Case 2 The representative LIT located at Namounou Market is well known enough to attract grain sellers without the help of teenage boys. Since the grain is sold immediately at Namounou to larger traders, out-of- pocket costs involve only bagging (FCFA 25 per 100 kg. bag) and the user cost of the bag (FCFA 83 per 100 kg. bag) for a total of FCFA 108 per 100 kg. bag. A.1.3 Case 3 The LIT buys grain at Namounou as in Case 2, but stores a few bags for resale during the rainy season (only a few bags are set aside be- cause the LIT cannot afford tO tie up more capital. We will assume that 10 bags will have been set aside by November, that the stock is kept from November to May, and that retail sales are uniform from June to October. Under these assumptions, 10 bags are held in storage for seven months, an average of five bags are held for five months, and for the year. the (weighted) average stock is eight (100 kg.) bags. Alter- natively, a bag stays an average of eight-tenths of 12 months in stor- age. The costs for such an Operation are presented in Table A.2. A.2 Regional Wholesaler-Retailer (RWR) The representative RWR considered is a small merchant who has thus far gained access to grain trading through a USAID-sponsored loan program which aims at promoting local entrepreneurs. He buys grain at Namounou from LITs, transports it on regular line trucks to Fada, and 264 Table A.1 Estimated Physical Distribution Costs (Small Merchant Operating at Namounou and Logobou) Cost Total Cost (FCFA per (FCFA per Percent of Cost Components 100 kg. Bag) 100 kg. Bag) Total Cost Handling 158 16.9 Hauling to warehouse- 25 Bagginga 50 Use of bagb 83 . Transportation 750 80.4 Grainc 750 Storage . 25 - 2.7 Use of facility 25 Lossesd -- Cost of capitald -- Total Cost: 933 100.0 Source: 1980 Marketing Investigation. aTeenage boys are paid for bagging and helping bring sellers to LIT. bA 100 kg. bag costs FCFA 250 and has been assumed to be reusable three times 0 ‘ cDonkey cart charge is on a bag basis in the Namounou area. They include loading and unloading dThere are only four days between Logobou's market day (Wednesday) and Namounou's (Sunday);cost of capital and value of storage losses are very insignificant during this period. 265 Table A.2 Estimated Physical Distribution Costs (Small Merchant With Storage at Namounou) Cost Total Cost (FCFA per (FCFA per Percent of Cost Components 100 kg. Bag) 100 kg. Bag) Total Cost Handling 133 12.7 Bagging 25 Hauling to warehouse 25 Use Of bag 83 Storage - 913 87.3 Use of facilitya 20 Cost of capitalb 397 Lossc 496 Total Cost: 1,046 100.0 Source: 1980 Marketing Investigation. aPrevailing storage space rent is FCFA 25 per bag, but average stock is eight-tenths bAssumed opportunity cost of capital is 12 percent, assumed purchase price is FCFA 4,000 per bag, and the resulting total value of one bag in stock is FCFA 4.133. Therefore, capital cost for the average stock is: 4,133* .12*8/lO = FCFA 397. cInventory costing uses "historical" rather than market (sale) prices. Here. however, the "historical" price is the purchase price plus costsoi= related services required to bring the grain to the warehouse. This is the same value upon which capital cost is computed. Hence, .15 * 8/10 = FCFA 496. 4,313 * dBecause total annual volume handled by the merchant is not known. tax cost per bag cannot be assessed and be included in the total cost. 266 sells it to urban consumers and women restaurateurs. His strategy is to achieve a high turnover thus taking advantage of regional price differential rather than speculating on seasonal price differential. The RWR does not incur any buying costs since he buys from an assembler (LIT). Handling services include hauling. loading/unloading, and use of the bag. Transportation charges are FCFA 500 per bag from Namounou to Fada but in addition the merchant has to pay FCFA 1,000 round-trip passenger fare. (A. nontrader would have paid FCFA 1,500 for the same service.) Grain transport charge is spread over 20 bags, a volume the trader secures on average on a trip. In association with two other merchants, the NWR has rented a tin-roofed lOO bag-capacity mud house at Fada market to store the grain. Sometimes for a couple of days a few bags are left outside the warehouse for lack of space. and also in the‘rainy season humidity may affect bags on the top layer (when the roof leaks) and bags in contact with the noncemented floor. The merchant estimates, however, that storage loss is insignificant be- cause of the high turnover (20 bags over a 2-week period). The costs of such an Operation are presented in Table A.3. A.3 National Wholesaler-Retailer (NWR) The representative NWR considered uses an agent network to collect grain at the farm gate in the Bassieri area where we have accounts of one commission agent who bought grain for a total value of three million FCFA on behalf of a large merchant from Ouagadougou. Here, however. only the costs related to the NWR's operations in the E-ORD, up to un- loading in Ouagadougou will be estimated. 267 Table A.3 Estimated Physical Distribution Costs (Small Regional Wholesaler-Retailer in Fada) Cost Total Cost (FCFA per (FCFA per Percent of Cost Components 100-kg. Bag) lOO-kg. Bag) Total Cost Handling 158 20.5 Hauling to truck 25 Loading onto truck 25 Unloading plus stacking 25 Use of baga 83 Transportation 575 74.5 Grain 500 Passengerb 75 Storage 39 5.0 Use of facilityc 17 Cost of capitald 22 Total Cost:e 772 100 0 Source: 1980 Marketing Investigation. aSee Tab1e A.1. bFCFA 1.000 spread over 20 bags on each trip. CMonthly rent of 1/3 * FCFA 2,000 per merchant is spread over 20 bags held in storage for two weeks on average. d price of FCFA 4,000 per bag; 52 weeks make up a year. eSee Table A.2 Total value of one bag in stock is FCFA 4,733 assuming a purchase 268 The costs incurred by the NWR include the commission agent's fees (FCFA 250 per bag). They also include handling in which loading in Ouagadougou may cost four times as much as in the E-ORD. The transport charges include that for moving the grain from the farm gate to the bulking point by donkey cart, and that from moving the grain by truck from Bassieri (bulking point) to Ouagadougou via Piela. For lack of a better estimate, we will approximate the truck transport cost by the cost professional truckers charge for grain transport between Bassieri- Piela and Piela-Ouagadougou, even though the NWR uses his own truck. (It is important to note that large merchants buy trucks not only to attempt to save on transport costs but mostly to lower overall distribu- tion costs by being on hand to take advantage of lower prices at harvest, by minimizing the risk of heavy physical losses that occur when rains fall early on the grain left in the open at the farm gate, and by making speedy delivery to take advantage of spatial price differential. That is, the availability and speed of transportation are as importantmas,if not more important than, the out-Of-pocket cost of transportation). We will assume also that the NWR faces some losses in storage and possibly monetary losses. The costs of such an operation are presented in Table A.4. 269 Table A.4 Estimated Physical Distribution Costs (Large National Wholesaler-Retailer) Cost Total Cost (FCFA per (FCFA per Percent of Cost Components 100 kg. Bag) 100 kg. Bag) Total Cost Buying 250 11.7 Commission fee 250 Handling 233 10.9 Bagging 25 Loading at village 25 Unloading at Ouagadougou 100 Use of bag 83 Transportation 1,450 67.6 Assembly to bulk pointa 150 Trucka 1.300 Storage _ 210 9.8 In village storage 25 Capital costC 131 Lossesd 55 Total Cost: 2,144 100.0 Source: 1980 Marketing Investigation. aDonkey cart cost per bag over a distance Of less than or equal to 15 kilometers. bProfessional truckers charge FCFA 300 per bag from Bassieri to Piela and FCFA 1,000 per bag from Piela to Ouagadougou. cThe total value of grain in storage at the village includes the purchase price (FCFA 4,000), bagging (FCFA 25). the user-cost of the bag (FCFA 83), and the commission fee. Loading and unloading take place when the grain is leaving or has left the village. dSee inventory costing in Table A.2. 4,358 * .05 * 3/12 e FCFA 55. eSee Table A.2. APPENDIX B 270 APPENDIX B KG CONVERSION FACTORS FOR SELECTED MILLET-SORGHUM MEASUREMENT DEVICES Measurement Mean Kgs For Mean Kgs For Devices Unthreshed Grain Threshed Grain Large straw basket 10.14 46.50 Small straw basket 7.75 31.75 Large wooden basket 7.50 40.35 Small wooden basket 6.30 30.10 0RD tigg with "hat" 5.15 18.50 ORD tjgg_without "hat" 4.70 16.90 Trader's tine 5.30 19.10 100 kg-bag 29.4 101.30 50 kg-bag 14.10 ° 50.60 25 kg-bag 7.0 25.30 Large plate 8.0 27.65 "Yoruba" plate .70 2.53 Large calebash 3.30 11.20 Medium calebash 1.60 5.60 Small calebash (louche) .30 1.10 Large pail - 12.80 Small pail 1.20 4.40 Scale (one kg) .75 1.0 Source: 1978-79 farm survey. subroutines. Derived from kg conversion computer APPENDIX C 271 NO0.00 OO0.00 ONH OH OO O NOH OH ONN ON0.0 O NH OON Ovm Om NH NON OOO HOH OH ON HO HO OH NOO NHO.H HOH OH OHH OHH O OH OOO.H NOO.H mHH OH OON NNN Owe OH NHH.H OHO.H mOm OH «Om O OON OH NNm.OH OOH.HH OOO NH OOO ONO.H OON.H NH HOO.N ONO.H Om HH mON OOO OOO HH .a.: .m.: .m.= OH OHN ONO ON OH OO0.0 ON0.0 NOO O OON OOO OO O OOO.N ONN.O OO O OON OON OO O OHO.e «ON.O Om N ON OHO O N mNN.O OOO.H O O .e.c .m.: .e.: O OHO.N NO0.0 HOH O OOH.H O NON O OOO.HH OOO «ON.¢H O ON OOO ON e NO0.0 OOH.Om O O O HO HO O HOH.OH ONN.OH mom N HOH O.NNO O N NHO.H OOO OHO.H H OOm O 0.000 H mev mev mev Huxv mev mev Nxm>eam Oo Numm>cmz xuoam me>gam NO Hamm>cmz xuoum eemv xeoem eaoe . a_eeeHeN emeHHH> mmO ONNOOONO zH NOHOHZ oz< zzzumom mo mmhOHO NLLOOV umHsz Ocm EOOOLOO No mmugaom OH.O OLOOHO OH.O mesa HO N~.~nO~.N an: ace. sense. wanna. aua‘. ac—o— uao xu—cco depo— w...~»...u i1 aguczuxaa H¢.O_ q... he... . a. uu—rcc depo- N..¢Pu~.rh 20-U=Occs No.0.uro.ho 89-.s89azou 275 OOON omomucmd OH HOHHNz Ocm Ezcmcom No mmmO use mmuczom N.O OLOOHO Heee=HOeH “oz axeeem meeeewv HOHHH: new EOOOLOO Oo mam: ON.O OLOOHO ”..Nnno.~ . u . ya: see. our.“ «9". “flaw—N...“- t3. uua¢298=s ac—o- 8- cause. 48.9— nn.~oucn.ua sc-pscauxou .Na>g=m Seam ONiONOH "mueaom AOOOOHOOH uoz mxooum sm>O chmOv HOHHHE use Eagmeom No meccaom ON.O «LOOHO u—--... I No.6uu.9 nonuoo za—Ouaabus 276 OOON OOOOOOO OH OOHHHz Ocm EOOOLOO mo mmmO use mmueaom m.m ee=m_m HepeeHueH “oz mxeepm meHeeOv HOHHHz use EOOOLOO mo mum: em.“ seemem "mum -.e_u~.o. . madam Scene HMO.HM1u use «whee: Hana» mm.ooum.oo ZOHNOcnmzou Nm.m_um.oH mumctuxag ache» sn.nuo.n ... 5...... 2:2 , .No>g=m Egan ONiONOH "musaom HOOOOHOOH poz mxuoum LO>O NLLOOV HOHHHZ use Ezzmeom No mmugzom ON.O OLOOHO uu.onns.mp :o_hu=aorc 277 pa."- andco I HeeeeHeeH poz axeoem eeeeemv — OOON OOOOOOOHO :H HOHHHz Ocm EOOOLOO No mmmO Ocm mmucsom e.u pc=a_a Hm>e=m scam ONiONOH ”mugaom NOOOOHOOH uoz mxuoum cm>O Hegmuv pmHHHz new EOOOLOO No mam: HOHHHE can Ezgmcom No mouezom ea.“ ecsmHa ea.“ mange; NN.¢nn.o we: cues w..nuo.n . Nae cu—ICO acne— ~N.mnum.un numczuxa; dupe- no.mcu¢.nc ZOH—uaaccs No.enn.c 8. Bu—zco dc—o— no.0huo.hp -1- zo_—s::mtou 278 OOON :ogeOOO OH use EOOOLOO No mmmO m.m aa=m_a AOOOOHOOH uoz mxoogm OOHOOOV HOHHHz Ocm EOOOLOO No mmmO em.m aeemwa No.3ho.a ~N.¢."N.o. .m: are. mu.¢m Se.o. 1111111: wo.cna.c :5 8:53 2:9- \ to. ...-.5955 uoHHHz use mwuezom .Nw>gzm Esau ONiONOH ”mugaom NOOOOHOOH aoz mngpm cm>O NLLOOV HOHHHE use EOOOLOO No mmueaom mm.u OEOOHO -...u~... -.1111 mumczwgs ...:O. / -hn 1111 x. ee.¢cn a¢.o. .\ \ / \ .\ . .\., Ne.ncuc.no 2 .\1. an: _vbaact APPENDIX F 279 xuOum .OOHuOEOmcou .cougun .mOHam a mum: e .m .N .— .mmma mm muHca HmuoH mama :H coHuuauoLO Oahu 4 O ape: HuuoH :. mum: u x I .8 C+Elz Emumzm OOHELON gun new ocon emu mvHozomao; me 50:33: c .1 .H a u N E Hump—HE + EOOOLOm .eszmcom .umHHHEO mango e i- .H u a «tug: s. a H». Haw 4 OO— 4 .m m. tau: sou acousoao 283.; O c .H—HauuL cannon No new. Hm>g=m Egan ONiONmH "uuuaom 0.00H 0.00H 0.00H 0.00H 0.00— 0.00H 0.00H 0.00H 0.00H 0.00H 0.00H 0.00— a: nun: Hench N.Om n.mc m.cO m.ON O.He H.Hm N.mN 0.0N O.Nm N.ev N.Om O.Nm aeoucu>cH OOHOOO N.Nm 0.0e O.mN O.nN H.Nm H.O¢ «.mo 0.00 o.NN O.mm m.OO 0.00 :oHua53mcou m.O H.N O.m O.n 0.0 0.0 m.o m.N 0.0H m.OH 0.0 0.0 umcmguxu couenm H.m N.O 0.0 O.N O.H 0.0 N.m ~.O N.m 0.0 0.0 0.0 muHom O<¢h u<¢wz< O<¢N u<¢hz< Oo>c3m use. ON.ONOH "ousaom H m. m NH n.o. 0.0H m.nN 0.9. ..O N.O may: ...oHNse...=O ooeozuuu n.OO N.mN a... n... n.o. o.nn 0.0o o.~u so...=c «usage-uNnu..m ...a— 0.0N H.~v. ..NH o.cv O.oNH N.OON u..¢ O...N .u_.m ec-uNxepuuao cues-O n.Nn N.N n.~n ..N 0.0mH a... o.o O.n nuuegocsa O_e:ou:o:NonNe_ coat- H.OO O.nn O.N.H 0.00 N.mm v.nH 0.0H O.nN “nuggets; ...oNNao..m ...a— o.mo ..N. o.mmH H..N O...— H.O O.HN m.o non-cues“ OHogunaozNao.-m scam Nucoucom a.“ «wean-u: cue-550N503 ...: :2 :3- «.23. 9.2.4... NE... .63. 93.- .03... e. .935 u.n~n.. H.ON..N N.Non H.ono m.OOm v.oom 0.0nN o.noo «OH-n .age— 0.0.m.H n.m~..~ n.0o- H.OOO u.NOn H.nnn 0.0oo ..NHO on: N_.:-..lc-. H.mO..H m.HOO.H n.OOc m.HmO n.an ..OOO m.NNO o.OnO Huo.nu...uH co..e-:neou o.~o o.n.H 0.0 o.ON 0.. N.nn n.o~ m.HO on: Ic-N .age— N.ON N.OOH O.o 0.0N 0.. N.nn H.nH v.on no»; N.nn H.n. 0.0 0.0 0.0 0.0 N.nn H.nv ovum O.mNN O.nOn ..OO 0.0m_ ..N—H n.nu O.Ho m.HO 36...:O oocosuuu ..ueH m.NO O.nN H.O o.nH o.om m.o~ a... «.an onNuao to“... ...o» o.m n.N 0.0 0.0 n.— n.— n.o H.m ..eo..au.se_. ace: m.H ..O 0.0 n.O m.H v.O 0.0 N.O con-H 902.: 0.0H O.H O.n 0.0 O.N O.H 0.0 0.0 «cu-Noaoa anon o.n ..O N.O H.O o.~ N.N «.0 H.o sane...» ucaoH 0.0o N.mm 9.. ..NH n.o. O..N a... H.OH cos.u ...—u n.OO_ 0.0NN n.OO H.mvH 0.0m O.nn O.~c O.Hm un..m ..uo— N.ON o.mc O.N O.NH O.NH 0.0H N.N n.u ...-m oe.u.oooca m.On H.NnH .... 0.0o N.NH o.n ..HH N.O~ ...-a «wasp o.nNH O.Nm o.o¢ m.ON n.nn O... ..nN m.NH u._-m Icon :25” yo no“: o.Ooo._ ..omN.~ O.NN N.Oc~ v.nON.H o.oo~.~ o.nN_ ..NNN nausea» H-HON O.omu.H u.va.N 0.0 O.o O.omo.H o.m.O.N 0.0 0.0 co.uusoecg O.HOn m.O_N O.nN N.O.N v.vOH N.N.N O.nNH H.N~N xOHHeH woe-gunu ...a— O.oo a.On O.HH m.~ O.n. N.N~ ~.o N.» :OHNeH can... _uuo— ....o 0.0 O... 0.0 o6 O... 0.0 0.0 39a: O... a... m.o v.O m.Hn m.... 0.0 0.0 ace-Nunez good m.» o.~ ..N 0.0 ..N 0.0 O.n O.N O‘ancce- 0.0H o.nH O.H ..N N.O N.O a.» .. H.n v.5.ouos a...o O.mn~ 0.0NO N.om N.OnN o.ou 0.0NN N.OHH ..HNN nun-coca; ...a— ~.ON O.m~ O.N O.NH O.NH a... N.N n.o «caucuses oe...ouocg m.ON m..ON N.o n.~OH v.o. 0.00 o.O ..HN ecu-guts; ou-cp n.aN_ n.o~. O.n. n.OHH N.ON n.ONH ..NOH N.OOH. ...-cocaa o.oeuuao: 5.3 No nousaom 35 ”22.2 O5: 355 as: USE: Os: 95:5 HoHogouao: A. o. ooosos<~ ulna. 16.5 2 .8 :57. 3. 2 .2 :22-— .e:. 3 ..n .08.. Jean 2 .n .88-. N... H-HOH NHL-o» «ness.: LouNO uno>caz unu>coz ocoNon .322 25.2 35:: 38582. #3.»...5— 9... O :2 ..E .352! 3.3.8:: 92 Eva—Sm 3 ...—<8 #33... u n—Otuanc APPENDIX H 281 APPENDIX H CHANGES IN STOCK AND MARKETED SURPLUS UNDER ALTERNATIVE GRAIN CONSUMPTION REQUIREMENTS ANTRAC Households HOE Households Annual Grain Daily Grain Consumption Consumption Stock Marketable Stock Marketable Per Capita Per Capita Chan e Surplus Chan e Surplus (kg) (kg) (kgi 1kg) 1kg (kg) 1656 .452 434.2 27.8 217.2 191.3 170 .466 377.4 -29.0 174.7 148.7 175b .479 327.1 -79.4 136.9 111.0 180 .493 263.9 -142.6 89.5 63.6 185 .507 207.1 -199.4 47.0 21.0 190C .520 150.3 -256.2 4.4 -21.6 195 .534 93.5 -313.0 -38.2 -64.1 200 .548 36.7 -369.8 -80.7 -106.7 Source: 1978-79 Farm Survey. a Turning point of ANTRAC households' marketable surplus. b Base run in grain flow analysis. c Turning point of HOE households' stock change and marketable surplus. APPENDIX I 282 .>m>L:m Egon mNiONmH 0.8 eHeeN aeme.e.e.e.e ”mugzom O.HHH ¢.ON1 m.Om¢i H.0OO1 H.ONN.H O.NHm.H 0.000: H.NHO1 mmzHngm mHnOumxgmz H.NOH m.NNm H.mHmi e.Ommi O.NON.H N.NOO.H N.OOO1 m.OONi ummcmso xuoum H.ON1 N.Oovi O.NN N.HO1 N.OH «.mNHi H.NO1 O.m¢H1 oonwuzo mmcmcuxm umz 0.0N 0.00 O.m1 ¢.HH O.mH 0.0 0.0H 0.0N Ozopmuzo megmm umz N.O¢i N.O¢vi 0.0N H.MO: O.Ni N.OOH1 N.NN1 e.ONH1 Own—agam uwumxgmz O Hench Hmm>emz Hmoa Hmm>emx pmm>gmz meg H xHoszO< HOOOINOOOI mum ox mwOOHozxomN >O OOHOOOO m40g=m seam ON1ONOH "muczom OOH OO OO NOH Om Om OO Om OO NOH OO muHocmmzo: mo gmnszz 0.00 O.Nm 0.00 O.NO 0.00 0.0H 0.0m O.¢H O.N 0.00 O.NO mmmmsugza OHocmmzo: Ocm mmHmm Egan :uom N.OO 0.0m O.NO H.NO 0.0H 0.0m H.NO 0.0 0.0 0.0N H.HN HHOO mmHmm Egan 0.00 N.OH 0.00 0.00 H.NH H.HH 0.0N N.OH 0.0N O.NO O.NO HHOO mammcugza OHocmmsoz N.OO 0.0m 0.00 0.0e O.m N.OO H.N H.Nm ¢.HN H.NO O.Ne meowuummcmeh oz Omom Omom umxgmz so; OOOOOO son OO umxgmz 1 1 1 1 OHOHO OOOOOz:OON OO OO>OO Oz< OZOHNOme zmou OcHOOouu «a: «— maHasam Oman OONOOOLO «a: O Oo>c=m luau ON1ONOH uuugaom O.NH N.OH O.HH N.O O.HH 0.0H O.MH H.O 0.0 0.0 0.0 N.OH 0.0N O.NM Nucougoav gang Oo . OmmszoOogugaa 0005 0.00 0.00 0.0N M.M 0.0 H.NN O.NO 0.0N 0.00 O.NO N.OH O.NO O.NO 0.0H Hueuueogv smog No Ouuuaomxmoaco>o¢ HuBHc< 0.0 0.0 O.NN O.N O.NH N.O O.N O.N O.H N.O N.O O.N 0.0 O.MH Hueuueugv smug Oo OOOLOOONOOHOO aogu OOO.OH1 NOM.H1 OOM.H NOO- NOM- OHO1 OON.N1 NOM1 OOO1 OH- OOO.H1 ONO.M1 OOH.H1 NMO.O1 HOuqu OOHOLOO “OOO OOHOOOLO Ho: OON.MN ON0.0 MOO OOH.O1 ONO.N1 OH0.0 HNO.M NOO.N1 ONN.OH OOO.M OOO.M1 ONN.H NO0.0 OMM.N1 3oHu amen no: OO0.0N OON.H MOO ONN.N OMO OHH OMO.H HOO.M OHO ONO.H NO0.0 MOO.H NNN ONO.H mum: Losuo MOM.HN MNO.H NMM.H OOM.H OHO.H HOO.H HMO.N OOO HON OMO NOO OMO.N MOO.H NOM.O mOsaumuoou NOO.H OMM OM O H OO O HH OO NOO NHH NOM OOM NOH Hu¢¢N2<. HOOEOHOOO OON.ON OON.M NOM.H OOO.H NNH.N ONH.H NOO.N ONO.H OON ONO. HHN OMO.N OMO OOO.H Hmucuexaaoa . can mcooHH HHOocu HOH.NN ONO.M OOM.O OOO.N MOM.M HON.O OON.O ON0.0 NO0.0 ONH.O OH0.0 NOH.O MMN.M OOO.M chnH-O au¢umm>HH OO0.0 OON OOO OOO NHO OOM OOO NON MOO MHO NNN ONO. OOO ONN echguOO van Ochuouosa HueauHaquO< 3.3: O8; 83 am 23 H: 33 am; So. So; a... :O 2: NS 9.69; Hts—3:2 HH0.0 OHN HO OO OOH ON ONH OMN OHM OON NMO OOO NOO MOO.H HuouH>LOO new nuaneH. OeHnOocu OON.OOH MHO.NH OHH.NH ONM.OH OOH.O OOM.O NM0.0H HO0.0H NO0.0H OM0.0H HOO.HH MNO.MH HOO.N ONO.MH gmau Oo mum: MOH.HO NON.M OOO.M OOH.O MOO.N NOO.N NNN.O OON.M OH0.0 OO0.0 OO0.0 HOO.N OOO.M OOO.N muucaoO Luguo OOM.OH OON.H MNO MON OHO OMO OOO MON.N HON.O MON NNN OOH.H HNO OOM muOOEOOLOOsHOO we. :oncuoov uHoocu OO0.00 OM0.0H ONN.N MOM OOO OOO.NH OOM.OH NOO.N OO0.0H ON0.0H OOO OOH.O OO0.0 NOO.H OeHmHau xuoumosHH OO0.0 NOO.H HOO.H NON NHO MOO OOO NOO HHM OMO NOH NOM NOO NOH OOHLOOHOO can OcHumuuocg Hogan—OOHLO< OM0.0H OON.M MOO.H OOO.H OHO.H NH OHH.N OOO NOO OHN OOM NNO OMN.H OOM OcHOOEN Hausa—OOHLOO ONO ON ON ON ON NNH NNH MM OO NM OO NOH OO HN mouH>LOO Esau OOO.NH OOO MNO.N OOO OOO HNH.H OOO OOO ONO OOO OOO ONM NNM.H OON OOH-O nocu MN0.00H HOM.HN NOO.MH NON.O ONN.O ONO.NH OOO.NN OOM.HH OO0.0N OMN.OH HHH.O HON.OH NO0.0H NMN.O Haas; muscu>o¢~ zmou Oo mmueaom ON1ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH tau» OM LOO H LO< O cucoz O.OON N OOO OH UmO NH >oz OH qu NH Hnom ON O=< MN N.HzH. ON ucaa ON Ne: menuH 36H; ngu 1N LOO 1O cute: 1O can 1O.:¢O 1HH OOO 1MH >oz 1OH qu 1OH uaom 1HN O:< 1ON OHOO 1ON ucaa 1ON Oa: 1H N»: OOHocmmao: u<¢Nz< LON mmcamumx aucascoNcum use 36H; zmuu OHOHco: H... 2...: 288 .moeaqucoaxu ONOOOOOOON mmoH moacm>we gmou OOHOOOLO Hm: OH OOHOLOO gmnu OOHOOOLO Ho: O Nuts» E: 2.22 "oueaom 0.0N 0.0H 0.0N O.NH 0.0H 0.0H 0.0 N.NN M.OM O.MN O.MM 0.0N N.ON Haeoueumv snag .9 mumaNnomagueam OOOO 0.0N O.NN O.MH O.HN M.ON O.N O.NO O.NN H.NN 0.0N M.ON H.MO 0.0N Hucoucuav smog No moucaomxmoaeo>o¢ HOEHOO 0.0H H.OH 0.0H 0.0H O.HM 0.0M O.NH N.OH O.N O.NH H.OH 0.0 H.ON HHOOOLOO. smog Oo mouesomNmoHaO Ooeu OOM.H- OHH ONM.H MNH OOO NOO OcH ONO1 OOO- NOM1 OON.H- HOH.H1 OOO- HOOOz. OOHOLOO mmnu OeHOOoeu «oz ONO.NH. OOO.H ONO.N MOO OOO- N1 OOO- O1 NOH- ONO.H OOO OOH ONO :OHO gmuu Ha: OOO.M OMH MOH HMO MHO OOH MHM OOH OOM ONM NMO ON OMM mom: L.ofiO HON.O HOO HOM ONM OOO NON ONM OON OOO.H OMO OMO.H OOO NMH.H anaamuooN 8O o o o a o O: O o o 2 ON o 65:5: 22:35 OOH.N OMO.H ON ONO OON OOM NOO NMM MNM OOM HOO OON ONO Hmueusxeaoz be. OOOOH. OHOOLO OHO.HH OON NOO NOO OON OOM MNN.H HOO NON OON NOO.H OOO HOO.H OchHox xuoHuosHH NOO.H NN OO OOH MO HNN NNH OOH NO OO ONH OHH OO OeHcogqu use Ochmouosa Hausa—OOHLO< OOM.O OOO OH MOM OOO.H ONN.H OON.H NNM OO ONN OH OON OO OOHOOLN Hausa—OOHEOO NNO.N ONH OO OO HOH MO OHH OO OOH ONN OHO OMO NNN HnouH>c0O OOO OHOOOHH OOHOOOLO OOO.HO OHO.M OOM.H OOO.N NO0.0 NOO.N NON.O OOO.N OON.N MON.N MM0.0 ONO.M OMO.M OOOO mo mum: OHN.OH MON.H OON.O OOO OOO OOO HON ONO OOO OOO.H NOO.H NOO MOO muueaoO LogHO OO0.0 OOO.H ONO OMM NOO OOH OOO ONM MON OOO OMN OON OHH OHOOEOOLOOEHOO uc- :H36550OH quosu NM0.0H NOH.H OON.H OON NOO MON NON.N OON ONO HMH.H MNN.H OOO.H OMH.H OOWOHOO xuoumusHH OON.N OON NHM ONO OON OHH OOH OHM OOH. OOH OO NON OMH eHcoguaO use . OOHOOOOOLO Hagan—OOHHOO ONO.N OOM NOO.H NON MON OHN MNH OHN HMO OOO NMH.H OOO MOO.H OOHOOLN HocauHauHcO< O O O O O O O O O O O. O O mouH>coO seem OOM.OH OOO MNO.H HOO OON.H MOO NOO OOO MMN ONO HON OMM HOO.H mo.~O Ooeu NM0.00 MO0.0 ONN.O NON.M OHO.M OOO.N OON.O ONO.N OMO.N ONN.M NM0.0 ONO.M OOM.O chcu mo==o>omH gmou so moueaom ON1ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH coo» OM LO< H Lq< O coca: N :OO OH 08 NH >6: OH qu NH HOOO ON O=< MN OHOO ON.o=:O ON xv: manuH zeHu gang -N .2 -O 55. -O e: .2 08 -2 so: -2 ta 1OH :3 -2. m2 .3. :2. 1ON 25.. 1%. 3. -H 5. OOHogmmaoz wax to» Oceans»: OOOOELOOLOO OOO onu gmuu Ongco: N.: OHOON 289 .moeauHecoaxo ONOOHOOOON mmoH moneosoc gmau OOHOOONO Om: OH «OHONOO Omnu OOHOOONO «oz O OO>HOO ICON ON1ONOH "uueaom N.H O N.O N.O H.O O.H O O.N H.O 0.0 O.M O.N O O HucuucosH ngO No mumONmousgoeam Oocu O.MM 0.0H O.NO 0.0 0.0N O.M H.ON 0.00 H.OH N.H 0.0H 0.0 O.H H.HH HOOOOLONH Omau Oo mouuaomxmuaco>m¢ HOIHO< 0.0 M.O O.H O O.N N.N M. N.O O.N O.H O.MH O.NH 0.0N M.O Hueoueoav sumo Oo moueaomNmoHOO Ooeu OO1 ON OOO- OOH- OO NO1 HO- OHH NHH1 OMH- OOH MON1 OOO.H MMO- HOOOzH mHHHHHHOO _ mmuu OOHOOONO Huz OO0.0M. ONO.H1 OON.N1 OOO.M- NMO.H OOO.N NO0.0N NO0.0 OMO NNO.M1 HOH.O HOM.H NO0.0 NMN 3oHO Omau Ho: OOM.NH OON OON MNM.H HON.N OMH OON OON.M O OMO.H OOO ONN OMO OMH.H mum: LOOHO HOO.H O OOM OH O MN O OMO O NM NOH OO O O «OwaumoooO O8 O c o O - c O o NNN o o NON NO 85.55 “5833 MO0.0 OOM NOO ON NOH OMO NNH ONN MOH HHO OH OOO HHO HNO Hmueosxaqox OOO OOOOHH HHOuLO HO0.0N NOO.N HHO.N HH0.0 MON OMN NNO OHO MON.M ONN.N OMM NOO HNM OOO OeHnHOO sucumo>HH NNO.H O OO OMH OOO OOM OON NN OON O O O NMM OM Ochogqu OOO OchmouocO HONONHOOHNO< OH0.0H OON.N HN O O OHM OOO.H ONO OO ONH.N OOH.H NOO.H MOH ON OOHOONN HagauHauHLOO OOO.N HOH HNH OO O N OM HNH NO OOH MOO NOM OOO MOM HmouHseuO use OOOOOHH OOHOOocO OOO.NO OO0.0 OO0.0 HHO.HH OOH.M OOO.H OON.M OO0.0 HON.M OOO.N OOO.N OON.M NON.N OON.M smou No mom: ON0.00 ONO.N HMH.M OOH.N HOO.N MOO.N OOO.M NNN.M OOO.N OoO.N OOO.M NOO.N MON.M ONO.N OOONOOO NOOHO OMN.M HNH MOH OO NOH OMN NOO ON OMO OOO ONN O ONM OM magneOONOOeHua Oc- OHIONNOOH OHOONO OOM.HM ONO OOH.M OOO ONM.H OMH OM0.0H MN0.0 HOO ON OMM.H OON MOH OOM OeHmHog HuaumusHH NON.N O OO ONM ONO ONO NHO OOH HHH ON ON HH OOO HO OeHNOOHOO u:- Ochmuuoea HONOOHOOHNO< MN0.0 O ON o o o NOO OHN o HNM NMN ONO OHN; HOH O52: HEBHOEO OO O O O O O O O O O O .OM OH O mouH>NOO as.“ OO0.0 OOH NNH O ONH OO MO ONO OO HN MOO MNO HNM.N OHN moHOO Oocu ONM.MO OOO.M OM0.0 HO0.0 NO0.0 NOO.M MON.MN NO0.0H HOM.O. ON0.0 HMH.N NO0.0 OON.O NOO.M Neommlmmace>umH Omou No muucaoO ON1ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH E; on 2.2 H :2 O 5:: O 8... N :2. OH 8O NH .62 OH 3O NH 3% ON 21 MN :2. ON 2.2. ON 3. en: .3: 53 -N Lz... -O 5...: -O a: -O :2. -HH 03 .2 so: .2 3O .2 “New -HN O2 -ON :2. 1ON 22. 1ON 3. -H E. mecHHuomcagN OHOLO oz ONH: OOHozamaoz co» OOLOOOO: mucosLoNNua OOO 30H; smog Ongcax m.z uHaOp 290 .mmNOHHOOOOxo ONNOHOOOOO mmoH moacm>vg Omou OOHOOOLO HO: OH OOHOLOO :mou OcHOOoeu an: a xo>g=O Esau ON1ONOH "muuaom 0.0H N.ON H.OM O.MH 0.0H 0.0H O.NH O.N O.NN O.HH M.MH 0.0N H.NN N.MM Hucuugumv guuu Oa OOOONOOOOOOLOO OOO; M.OO O.MO N.O N.NH 0.0N M.ON O.NO 0.0H 0.00 O.HN M.OH 0.00 0.00 0.0N Hueoucumv smou No mousaoONmmaeo>uO HaeHOO 0.0 H.O 0.0 N.N N.M 0.0 O.N N.M O.H O.H 0.0 O.N O.M- O.NH Hueoueomv OOOO No . mouNOOONOOHOO Oocu OHO.OH1 NOO.H1 OHH- ONO- ONO.H1 NOO.H NNO.H- NON- OOH.H1 NOH.H1 NON.H1 NNN.N- HNM.M1 NMH.O- HOquv mHHHNHNHHO mmuu OOHaneO uoz OO0.0N MNM.O HH0.0 OOO.O1 NMO.N1 OO0.0 OOM OON.M1 . OOH.OH MOO.M NOO- NOO.M OHO.M OH0.0- 30H; zmou uoz OM0.0 OOH.H ONN OOO.H NMN OOH NNN NNO.H .NON ONO OOO NNO ONM OOO mum: Noguo OM0.0H OOO.H ONM.H NON.H HHH.H HOH.H ONO.H OMN OOH.H OOH.H ONO OOH.N OOO.H OHO.M ONOOHOOOON OON.H OOM MM o o O. .. o o HO OOO HOH OHH ON 3 85:25 29533 OHN.OH ONN.N NOO NOO.H OMH.N NOO.H OOH.N NHO.H HNO MON HNO OOO.H NMO OOH.H HmucueNnnoz use OcooHH HHOONO OO0.00 ONN.H NOO.H NO0.0 OMN.N MOO.M OOM.O NO0.0 MOH.N NO0.0 NNO.M HOO.M OOO.N HON.M OeHmHna HuoHOO>HH NOM.O OOH NOO NNO ONM OOH OMN NHN MOO OOO MOH OOO OOM OON OOHNOOHOO Oz. . Ochmuuoem HONOHHOOHLOO HON.O NHH.H O OOO NNO NMO.H NMH.N OOO.H OOO MNM MON NO OHM MOO OOHOOLN HONOHHOOHNOO OMM.M MNH OO ON OO HO OHH OOH HOH OHH OOO MHO HHO OOO HmuuH>HOO new OOOOOHH OOHOOOLO NNO.NOH OHM.O OM0.0 HHN.O NOO.N HOO.N OHN.HH OH0.0H OOM.O OMH.OH OON.N OO0.0 OMN.O OON.OH gmmu No man: NOH.OM OON.N OO0.0 OOO.N OMO.N ONN.H MNN.N OOO.N NMN.H NMM.N NMO.M OMO.N OON.N NNH.N moocaoO ensue OOH.NH OOO OHO OOO NNH HOO MOO OOO.N MNO.M OMO OHN NOO.H ONO OHM ONOOEOONOOEHOO OOO :H3OLHOOH OHOONO OOH.NO OON.O MOO ONN OOO NHO.NH ONN.N NOM.H OO0.0 OH0.0 ONO.H NO0.0 NO0.0 HNO.H OchHaz Huoumo>HH OON.O OOM NHH NMO MNN OOO OOM HON NOO OOM NON NNH ONO NOH OOHNOONOO tea Ochmuuon HONONHOOHNOO OO0.0 OOO OMO.N OOO OOO OH OOO NON ONN NOO MHM HON MOH.H HOO OOHOONN Hausa—OOHNO< HOO OO OM ON NN NNH HO O OM O ON OO ON O nouH>ceO luau OO0.0 OON OON OHO OOH OHO.H OON NMN . NHN HHN ONM ONM NOM ONO moHaO OONO NNO.MMH NO0.0H NO0.0 OON.O OM0.0 OO0.0H HOO.HH OON.N OO0.0H NNN.MH OOM.O OMO.NH OO0.0H OON.O Naoeu mo=c~>omH 1 zmou mo mouesom ON-ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH coo> OM cO< H LO< O coeax O OOO N :OO OH OOO NH >oz OH qu NH OOOO ON OOO MN HHOO. ON ocaO ON Ne: meoHH onu :mou 1N ..2 -O 5.5.. 1O a: -O :2. -HH 68 .2 so: .2 to 1OH 23 -HN 5. -ON :2. 1ON 25.. 1ON 3.. -H be OHOO mmmuzueaa :HOLO OHH: OOHozmmao: HON mugamuuz oucofisomsom van 30H“ :mau x-suco: O.z OHOON 291 .OogauHocoOxc OOOOOOOOOO OmuH moaco>og Omou OOHOOOLu «on m. OOHOLOO snag OOOOOOLO an: a xu>sam luau ON-ONOH "ouLOOO H.N O.H O.N O.H O O. O. M.N M.O O.M 0.0 M.H M.H M.O Hucousuav OOOO Oo OOOONOOOOOOOOO OooO O.NN 0.00 0.0 0.0 H.O O.NM N.O 0.00 . 0.0H 0.0 N.OM N.N 0.00 N.O HucuuguOO OOOO mo mousaomxmuaco>om HOaH=< H.OH 0.0H H.MH O.HH O.HN O.NM N.O M.O O.HN O.N O.N 0.0 H.ON 0.00 Hucougogv OOOO Oo mouLOOONOoHOO OOLO OON.O OOO MOO MNO.H ONM.H HOO.N OOM MOM MHO NO MON- ONO- OOM.H NOO.H HOOOzO OOHOOOO mung OOOOOOLO no: HMN.O OOO.M OOO HN0.0 HO0.0 NOH.O- OHH.M- HOH.N- OON- OOO OOO.H OOO NNO.M ONO.N- xOHO OOOO «Oz NON.ON OHN NNM.M OOO.H O MON.O NHO.H HHN NOO.H OOO.N NOO.N OO0.0 O OOO.H «on: L«OOO ONN.H OM NOH NM O NN OO OON ONO OON ONN ONH OM ON OOOOOOOOOO NOO.M O O O O O O OOM.N O O O OOO O MO HO<¢N2LOO - OOO OOOOOHO OOHOOOOO OO0.00 NON.N OMM.O MNN.M OOO.H OHN.HH NMN.OH OOO.HH MOO.N HM0.0 NON.O OO0.0 OOO.N MO0.0 gmau NO mum: MHO.NN HOO.H. NOO.M OOO.N OON.H NOO NON.H NMN ONO.N OM0.0 OHO.N ONO.M HOO NNO nouOOOO OogOO MO0.0 HHH OHH NMO.H OOM OON OOH NOO ONO OHO OHN OHM OHN ON MmucoaongaasHom OOO c—IOLLOOH OHOOLO OOO.HN NON.N OMO NMO OOO OOO.N OH NON.O OOH.H MHO ONO.N NHN NN0.0 HOH OOOOHOO xuoumuOHO OM0.0 OOO OOO ONN OOO OMM OHH ONO HNO OOH OMN HHN OOH OOH OOHLOOOOO OO- OOHmnquOO HOLOOHOuOLO< BOON O OHN HHN. O OOO.H OOO OO0.0 OOO HNO O HNO.H OO1... O 2: 9:2: 53:57.2 HNO O O ON O OO O O . OO . OOH ON ON. O OH muuH>gom luau OHN.OH ONO NMO HNO.H HOO.H .NOO.N OOO NON ONO.H HNO OOH NOO MHO.H OOH.N OoHOO Oogu OM0.00 NO0.0 OMM.O OON.O NO0.0 HOO.N HHO.N NN0.0 NON.N OOO.N OO0.0 HOH.OH HO0.0 OMO.M HEOOO Ouacu>uOO OOOO Oo muugnoO ON-ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH Lao» OM LO< H LO< O Ouguz O OOO N OOO, OH OOO NH >oz OH qu NH “OOO ON OOO MN NHHOH. ON «OOO ON Oa: msuuH 30.; OOOO a~ LO< -O gugax nO awn -O OOO aHH goo -MH >0: -OH uuo -OH uuom oHN ua< ION h—OO -ON OOOO -ON has -H An: OHOO OoHOO OHOLO OOH: OOHOOOan: so» Oosamumx oucuagcwgoa OOO zOHO OOOO OHOOOO: O.x «NOON 292 .OOLOOHOOoaxo OONOOOOOOO Ono. OOOOO>OL :muu OOHOOogu OOO OH OOHOLOO Oman OOHOOOLO an: a 3:3 E: 2.22 "ousaom 0.0H O.MH N.N 0.0H N.O 0.0N M.O 0.0 O.M O.NH M.NH N.NM 0.00 0.00 HacuuLuO. snag No OuOONOoOOOULOO OooO HH.ON N.OH 0.0N M.N N.HN N.O H.ON H.HO 0.00 O.M N.HH 0.00 0.0N O.NN HuOouLOOO OOOO Oo . OOOLOOONOOOOO>OO Has—OO O.MN 0.0H N.OM 0.0N N.OO 0.00 O.NM M.NN 0.0 0.0N 0.0H 0.0H N.N O.NN Hucuugomv OOOO mo OuuLOOONuuHOO OOLO OOM.H OHO NN0.0 ONN OOM.N MOO- NNO OON O- OON HOH.H- MOO.N- OMM.N- NMO.M- HOOOzH OOHOLOO mmou OOHOOOOO um: NOM.N- HHO.N OO0.0 ONN- OOM.M OON.H- OO0.0- OM0.0- ON0.0 OON HMH.O- OON.O- OHO- OOM.H- :OHO OOOO uaz OOO.HH O ONN MO OM OO ONN.N OMO.H NO NOO.H OON.O OMO HOO OO «on: guzuO OON.OH OOO.H OHN HNO OOM HOO.H HOH.H OHO NOM OOO ONO.H ONH.M NOO.H MOM.M OOOOOOOOOO MOO O O OH N N O ON N OMH O ON OOM NH HOHO ONOO.M OO NMH OMO OOM MNN OOO OOH OOH NOO NMH NN HH OOH OOHOOOOOO OOO , OOHOOouOOO HOOOOHaquO< MHH.OH OON OO0.0 ONO OOO.H OOO.H OMM.H OMO OHN MMO OOO OHH OO OOH OOOOOLN HOLOOHOuHOO< OO0.0 OMH OO HN OMH ONN MOH OON OHM HHN NOO MOO OOO OON.H HmuuH>soO OOO OOOOOHO OOHOOOLO NOO.NOH ONH.O OO0.0 OOM.O OMN.M OH0.0 HMN.NH OOM.OH HON.O ON0.0 NN0.0 NHN.O OO0.0 OOO.N OOOO OO mom: MN0.0H OOO.H OOO.H OOO NNO NNO OMN.N OMO OOO.M HOO.H OON.H NOO.H OON.H ONO Ouugaom gosuo HON.OH HON.M HNO.H OHN NHN OMN ONO MOO NON OHM OOO NOO MOO OON mucosumganevoa . OOO ONIOLLOOH HHOOLO HON.ON OOO.H HO0.0 OO HHO.H NON MOO.H OHO.N NON.O OON OHO OHO.N OMO HON.H OOHOOOO xuOuOo>HO NN0.0 NHH.N MNO.N OOO OMH NON OOM OHM. OOH ONO NOH HON ONM OHN OOOLOOOOO OOO . OOOOOquLO HasauHauHLO< OOH.NH HM0.0 OOO.H OOO OON.H ONO O MHH OON OOO.H OM OO OOM OOO.N OOHOOON Hagan—OuHOO< ONO NO OO O O MN NO OO OH OH OH OO.. NOH ONH OouH>LOO ago“ OOO.MN OOO.H OMH.N ONH.H MOO.N HOO.H NOO.N OOO.H NNO OMO.H NOO NNO MON OON.H OoHOO OOOO OOO.OO~ 5.2 ONO.NH OHH; OMH.N O36 ONO.N ON0.0 83: «$4. HOO; NNO; ONO.O OO0.0 Oat 33:2..5 . . OOOO No OuuLOOO ON-ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH ONOH gnu» OM LO< H LOO O nose: O OOO N OOO OH 08 NH >oz OH qu NH «OOO ON Oa< MN N.OO ON acne ON NO: OaauH :OHN OOOO -N OO< -O sugar -O OOO -O OOO -HH UoO -MH >oz -OH OOO -OH HOoO -HN OO< -ON OHOO -ON ocaO -ON Oct -H NO: Oumncugaq OOO OoHOO OHOLO OOOO ONH: OOHOOOOOO: LOO OOOOOOO: oucoELoOLuO OOO IOHO gmau OHgacoz O.x o—Oap APPENDIX N ‘ MARKET SHARES 293 .OOOOHO pmeOE ON OOOHOOOOO OOOOOOOOO HHOEO on One OONOOOOOOO Oz» OOOOO OH OOOOONOOOOOO OEOO O .muOO>OO OH OONONOOOO OOO OOOOO OOO OOONOOO OOOO o: OOOOOOO OOOOOO>OOOOO O.OOO<2OO we OOOOO OOO OOOOLOO OOO OOOoqu on OOO ou Omucmggmz OOO ON OH O OO>OOO Egan ON-ONOH "OOLOOO HOO . NOOO OOO Hmmmau O=_mm_zO O.OOH OOOONOO O.OOH O.OOHOO O.OO OOO HOOON N.OO OOOOOON H.OO N.OOOOO N.OO OOO OmOmOOo HHO O. OOOOO O.M O.ONNO O.O OO OO_HOO OOOO O.M OHHOOH N.O O.NNOO N.O OO OOOOZOO O.O OOOOON O.M 0.0mmm O.O Om mOOmOOOOO mO_mO=O O. OOOOH N. 0.000 m. m OOOOOOOOO OOOH__> O.Om OONONOH O.Om O.NOOOO H.OM NNO OOOOHH_> OOO Oo mgmaemz ”Emu-Hun— OO_OOOOO OOONOOOOOOON OOOOOOLOO EOOOLOO OOO OOHHPz .OOOOOOOOO OO OOOOOO OmeOz .OOOHOOO: OOOO Oo Owgzmmmz OOOOOOOOO< H.z OHOOH 294 .Ocowuummcmgu HOOOO mo OOOUOOO Ow HOmo-O OOO OO mgouumm HHO va OOLOOO OOO<2OO mo OOOOOwO OOOOOLOO< OHOOHwO>O OOO OOO OmuOO OOO“ Omowuumammg OOO OOOHO> OOO .Omwawucmzc O .OOOOOOOOE OwOOO mo OHOOOO Oo OwOOO OOOOOOOO on: H=OOOOOO HHO: OONOOO OOOOHHw> OOO OO OLOOEOE= OEOO OHOwOOOOV OOOOOO OOOOOOw Ow OOOOOOO uomggoucw Ow OOOHOOOO Owgp .OOOOOwOOw OOOOOHO m< u .OOowOOO OOOOLO OOO Ow OOHOO mmmsu OOwOOuOHHO mo NO; O: was OOOOO OOOOOOO OOOOOOO OOO: O .HOOOOOO mama OOO va wHOOqu OOO OOOOOOOwN OOO Hgouumm wHOOOOwOO: OOO va OOOOOOO .wOOoxowOOO .Hgouumm wOOOOOOOx OOO va OOOOOOOOz .OOOOOOOOOOz .Hgouumm OOOOOwO OOO va wHOOUOOO .wOOOsOO .OOOOOOO ”OHOEOO ELOO ON-ONOH OOO we OOOOHHw> chzowwom O OOO OOO» mama OOO<2OO ON OHOO on: OLOOOOOOO O OO>OOO Egon ON-ONOH "OOLOOO 0.00H OOOOON 0.00 O.MONOH 0.00H OMN HOOON m.OO OOOOOO O.NO 0.000HH 0.00 OMH Ogmsuo HH< O.NH OOONNH 0.0H 0.00Nm H.MN OO Oamu<2OO N.H OONO O.H 0.00M M.H M OOOOOOLOE mOwOOOO - - - - - - OOOOOOOOE OOOHHw> 0.0N OMNOOH O.NN 0.00MO O.HN OO mOOHHw> OOO yo OLOOEOz OOOOLOO OquOOOO mOowOOOOOOON OowOOO>OOOOH OOO Ow OOOOOO umxgmz .OOOHOOO: OOLwO we Omgammmz N.z ONOON APPENDIX 0 295 .HOOOOO Omemws MO OOO O.ONO.NO mo OEOHO> HOOOO O Oo OOOOO u .HOOOOO OOwOOwO HHO OOO N.OOO.NO OO OOOHO> HOOOO O Oo OOOOO O .OOOmOou OOHOOOOOOO OOOOO OOOOOOOO OOO SOLO OOO OOOO OOO OO OONHO>OOO OOOOOOOwO OOO .OOOOOOOO Oo OOOHHw> OOO OOOO mgmumsowwx OOO O OOwOO OOOOOOH OO OOO Oosz HOOOOHOO mu OOOOEHOOOOV.OHOwO OOO Ow OmwOOOOOO OwOOu SOLO OwOOO OHOO O>OO OOOEOOO OOO: OOwOOOHmwO OO NOO OOOHHw> OOO SOLO OOuOOquO O .OHOOOO mOwOOm OO>OOO OOO; ON-ONOH ”OOOOOO 0.0 N.H N.O N.OO O.Nm OOOOOOOLOO OHOOOOOOI 0.0 m.OH O.NH 0.0N 0.0m OOONOO OwOLO ELOO .Osx OO A .OEx OOIHN .OEx ONuHH .OEx OH-H .Ex O mo OOOOOOO OOOOHHw> OOOO OOOOOquO mommz OOoszOOI mom zHOO oz< zH