\ - be 2 ABSTRACT { AN ECCNOMIC ANALYSIS OF comm A. PRODJCI‘ICN , MARKETING AND PROCESSING IN NORTHERN ZAIRE By Eric Frans Tol lens Presently, cotton production in northern Zaire is 1&8 than half the 1960 level when Zaire gained independence from Belgium- Productim has declined because of a reduction in the number of cotton farmers. cotton acreage per farm and reduced yields per hectare. The Government of Zaire is now committed to develop northern Zaire based on agricultural and industrial development of the third grcwth pole centered in Kisangani. Cotton production will play an important role in a comprehensive strategy for developing the rural areas as emlogical conditions favor the production of cotton and domestic and EXPO“- demand is favorable for cotton lint. The Purpose of this study is to describe the cotton subsector in “0me Zaire , identify barriers to expanding cotton production. evaluate alternative strategies for increasing cotton production and fonnulate POliCy prescriptions for expanding cotton production and for 1%me the PErformance of the cotton marketing system. This stUdY Utilizes the concept of a subsector study and focuses on cotton producr tlon, marketjng and processing in northern Zaire, including bOth the vertical . and horizontal relationships for this subsector . Eric Frans Tollens The major features of the agricultural economy in Zaire were reviewed aria the essential economic characteristics of the cotton subsector in Vrthern Zaire and of its environment were described in order to mderstand how the subsector operates. A farm business survey of 160 randomly selected cotton farmers was conducted in northern Zaire over the 1972-1973 cotton production year. , Enumerators visited each farm twice a year, once before harvest and once after harvest. The survey generated information on the characteristics of cotton fanners, cotton production practices, the "crcp imposition system," the agricultural extension service, farm size, yields, incomes, marketing of seed cotton, and the major constraints on expanding cotton production. The survey of 160 cotton farmers revealed that seed cotton yields were low, 399 kg. per hectare, cotton acreage per farm was 0.36 hectares: and the gross income per farm from cotton production was 8. 79 Zaires (17-53 dGuam) . There was a large variation in cotton acreages and YiEIdS bGtween farms. Generally, farmers were not following simple but momnended agroncmic practices such as time of planting, spacing, etc. m o - addition. there were no farmers in the survey using chemical fertil- izers, Pesticides , tractor mechanization or agricultural credit. W13: Sfme cotton fields were treated with insecticides by the agriolltural service of ONAFITEX, the national cotton marketing office- An analysis of the returns to labor for cotton and for other sele‘ied cmmodities in northern Zaire revealed that the returns for 0°th were OGIparable with those for the major food crops bUt were below . . the legal minimum wage of unskilled workers in rural areas. The niturns. - for perennial crops, cocoa, coffee and oil palm, were more than Eric Frans 'Ibllens three times higher than for cotton and were well above the legal minimum wage. A multiple linear regression model was used to explain the variation in cotton yields and in cotton acreages between farms. Dummy variables representing the major differences in ecological conditions and in agronomic practices between the four subregions in the North were among the most important variables explaining the differences in cotton production between farms in northern Zaire . Alternative strategies for increasing cotton production in the North were analyzed and evaluated. These strategies must be examined in the cmtext of increasing regional commodity specialization within northern Zaire and in a political economy framework of developing the North. In the immediate future, the most promising avenues for expanding cotton production are a revision of cotton pricing policies , diffusion of improved agronomic practices , an overhaul of the cotton extension service, a pilot cotton development project in Bas-Uélé , and accelerated research on new varieties . Although the government has raised the producer price of seed cotton a number of times since independence, the real purchasing power of seed cotton in December, 1973 was only about half the June 1960 level. Local fiber prices are kept low to protect the Zairean textile industry and Zairean consumers. A substantial increase in the price from 6.5 Makuta to 10.0 Makuta per kg. for the first grade seed cotton is needed to provide adequate incentives for farmers to expand cotton production. An increase in the domestic lint price of 15 to 20 K. per kg. would generate enough revenue to enable ONAFITEX to raise the producer price. Eric Frans Tollens A price increase of 15 to 28K. per kg. of cotton lint for dorestic spinners would raise the price of domestic lint to about 60 to 65K. per kg. of lint, still well below the price they would have to pay for imports of the same quality and grade. Higher producer prices would, in turn, reduce per unit cotton marketing and processing costs and will provide (NAFITEX with substantial export earnings. When more favorable producer pricing policies are adopted, the cotton "imposition system" can be removed and the cotton extension service can gradually shift from its traditional regulatory role to an educational level. There is a large opportunity for increasing cotton yields in northern Zaire by introducing improved agronomic practices . However, a major overhaul of the cotton extension service is needed in order to create the administrative structure and leadership for effective small farmer extension programs. Extension reforms include retraining and upgrading of extension workers, creation of regional cotton extension committees and the establishment of an extension department at the National University of Zaire and at the National Institute for Agrmomic Studies and Research. Extension agents should focus on assisting farmers in adopting simple agronomic practices such as timely planting and harvesting, correct spacing, frequent thinning and weeding, etc. We also recomend that the extension service encourage farmers to purchase locally made hand powered Sprayers or dusters . Farmers should be taught ho» to protect their crops from cotton pests instead of relying on the cotton extensim service to apply pesticides to their fields. There is a strmg case for establishing a pilot cotton development project in Bas-Uélé, centered in Bambesa. Such a project, focusing on f“! Eric Frans 'Ibllens an extension of improved agronomic practices, would have a great potential for increasing cotton production. Moreover, specialization of productim in this area would substantially reduce per unit assembly, transport and ginning costs . Researdm on new varieties for the North should also be accelerated and improved varieties should be imported from neighboring countries. Chemmal fertilizers, tractor mechanization and agricultural credit are less promising strategies for increasing cotton production in northern Zaire in the immediate future. More technical and economic research is needed at the farm level to determine the profitability of alternative rates of fertilizer application, selective mechanization and agricultural credit. When the cotton producer' 5 price is raised and when the extension service is upgraded, these strategies may becore feasible. AN ECONOMIC ANALYSIS OF COKE!!! PRODUCTION, MARKETING.AND PROCESSING IN NORTHERN ZAIRE By Eric Frans Tollens .A DISSEREATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR.OF PHILOSOPHY Department of Agricultural Economics 1975 © Copyright by Eric Frans Tollens 1975 ii ‘V x' Many people have provided encouragement and advice during my gradu- ate studies, though it is not possible to give adequate tribute to all who have assisted me. However, I wish to express sincere appreciation to Dr. Vernon L. Sorenson, my major professor, and to Dr. Carl K. Eider, my thesis supervisor, for their constant intellectual stimulation and guidance. I wish to thank the members of my guidance and thesis committee, Dr. Glenn L. Johnson, Dr. James D. Shaffer, Dr. James H. Stapleton, Dr. Victor E. Smith and Dr. Jan Kmenta. I wish to express appreciation to those who provided financial assistance for my studies and research. To Dr. Dale E. Hathaway and Dr. Harold M. Riley, successive Chairmen of the Department of Agricul- tural Economics, I am thankful for financial support and for the opportunity to pursue graduate studies at Michigan State University. I am grateful to the Belgian American Educational Foundation for pro- viding me with a C.R.B. fellowship for the academic year 1968-1969. I am grateful to the Belgian Aid Mission for a five-year contract and the qaportunity to work at the Faculty of Agronomy of the National University of Zaire from which this study was conducted. The Ford Foundation funded the field research phase of this study and therefore, I wish to express gratitude to Dr. Haldore Hanson and Dr. William K. Gamble, successive Representatives of the Lagos , Nigeria office . iii I am indebted to many people in Zaire who helped me. In particular, I am grateful to Citoyen Wendi Mbuyi Tshingore, Director-General of the Department of Agriculture, and to Citoyen Tshivuadi-Kalorbo, Director-General of CNAFITEDC, the national cotton marketing office, for their helpful advice and support. I am also indebted to Mr. Michel Vandewalle of (NAFITEX for his patient and skillful attention to many questions and requests over the 1971-1974 period and to Mr. J. C. Thevenin of the C.F.D.T. Mission in Ubangi-angala for his suggestims and ideas. A special note of thanks is given to NE. Linda Sanders for the many hours spent typing seemingly endless manuscripts . I am indebted to Dr. Gerard R. Boddez of the Catholic University of Louvain for encouraging me to pursue a career in agricultural ecmonics. Not least of all, to my wife 'I‘l'érése, I am appreciative for her many sacrifices, patience and help in compiling field data. iv TABLE OFCCNI'ENI‘S Chapter Page LIST OF WEIGTI‘S, MEASURES, CURRENCY EQUIVALENTS, ABBREVIATICNS AND ADMINISTRATIVE OKSANIZATICNS OF ZAIRE. ......... ............. I. IN’I'RODLCTION . ........ . . . . ...... . . 1 Problem Setting and Importance of the Study . . . . 1 Research Approach and Scope of the Study . . . . . 7 Objectives of the Study ....... . . . . . . 9 Conceptual Framework for Studying Cotton Produc- tion, Marketing and Processing in Northern Zaire . 9 II. AGRICULTURE IN THE ECONGVIY OF ZAIRE ...... . . . ll Pecent Performance of the Agricultural Sector . . . 11 Population . . . . . . . . ............ 11 land and Climate .............. . 12 Gross Domestic Product, Sources, Growth Rates and Agricultural Value Added. ......... 15 Structure of Agricultmtal Production and Marketing....... ........ ....16 Agricultural Exports and Imports . . . . . . . . 20 The Ible of Government in Agricultural Develcptent. 21 Agriculture in the Government Budget ..... . 23 Pricing Policies, Wages and Taxation . . . . . . 27 Agricultural Development Projects and External Agricultural Marketing and Transportation . . . 32 Major Problems Facing the Agricultural Sector . . . 35 Achieving Self-Sufficiency in Basic Food Products35 Expansion of Smallholder Export Crop Production. 37 V Chapter The Transportation System . . . ........ Improving the Performance of the Marketing System . . . . . . . ............. III. COI‘I‘CN PHDDIETION IN NORTHERN ZAIRE ......... History of Cotton Production in Zaire ...... Introduction of Cotton ............ Growth of Cotton Production .......... Ecological Conditions Governing Cotton Production .................. Principal Cotton Production Regions ...... Human Factors in Cotton Production ...... Cotton Research ................ Paysannats and Cotton Production ....... Agricultural Extension ............ The Agricultural Service of ONAFITEX ..... Government Policies Affecting Cotton Production .................. One Channel Marketing Office ........ Uniform Fixed Producer Prices ....... Cotton Production Development Projects . . . ThefE-CAKProjectinEIasternKasai . . 'Ihe C.F.D.T. Mission to Relaunch Cotton ProductionintheEquatorRegion . . . . The FIWA—BILI Project of the Belgian Aid Mission in Northern Ubangi ..... Cotton Production in Northern Zaire ....... Climatic Conditions .............. Soils ..................... PopulationandDemogramy........... Infrastructure ....... Land Tenure , Crop Rotation and Cultural Practices....... ............ Role of Cotton Production in Northern Zaire . . History of the Cotton Industry in Zaire . . . . . vi Page 39 40 41 43 43 43 44 46 48 50 S4 59 61 62 63 63 64 67 69 7]. 73 74 74 75 76 77 80 83 85 89 89 Chapter Development of the Cotton Industry ...... . The Fords de Remploi: 1924—1935 ...... . . The Cotton Reserve Fund: 1936-1942 . . . . . . CIXHQKJD: 1943 . . . . ............. The Cotton Decree of June 18,1947 . ...... The Convention of May 30,1950 ...... . . The Cotton Seed Convention of April 19,1951 . . The Convention of January 31,1955 ....... The COVENGD-Convention of June 2, 1962 . . . . . Page 89 91 91 92 93 96 99 100 101 The Convention of June 18, 1962 . . . . . . . . 101 The Decree-Law (Décret - loi) of August 13, 1965 ...... .............105 The Private Cotton Companies . . . . . 105 The Creation of ONAFI'I'EX on August 12,1971 . . 106 Structural Organization of ONAFITEX . . . . . . 110 RoleOfCNAFITEX......... ....... 111 Cotton Marketing and Processing in Northern Zaire . 112 Grading and Collection of Cotton . . ...... 112 Number, Size and location of Ginning Plants . . 114 Cotton Marketing Costs . . . . . . ...... 116 The Evaluation of the Textile Industry. . . 129 Sale of Cotton Lint . . . . . . . ...... 132 CottonBy-Prcducts...............14l Distribution of Cotton Seed to Farmers . . . 141 Cotton Seed Crushing ............ 143 ExportsofCottonSeed...........149 Summary ............. . ..... 150 THE ORGANIZATION (F A FARM BUSINESS SURVEY OF COI‘I‘ON FARMERS IN NORTHERN ZAIRE ...... . . . ..... 152 PurposecftheSurvey...............152 Description of Survey Area ............ 152 Survey Method ......... . . ...... . . 154 Determination of Sample Size . . . . ....... 154 Selection of a Sampling Frame . . . . ...... . 155 SamplingProcedureW ....... 157 Sampling Plan ................... 158 Selection, Training and Supervision of Field Enurerators...... ........ ......l69 locating Farmers . . . . ........... . 162 Questionnaire Design, Tabulation and Handling of the Data . . . . . . ............... 162 Summary ........ ..... 163 vii Chapter VI ANALYSIS OF THE FARM BUSINESS SURVEY DATA ...... Characteristics of the Survey Sample . . ..... Nonresponse and Drop-Out ........... Forest and Savannah Areas ........... Farmers in Paysannats ............. Farmers Possessing a Cotton Farm Book . . . . . FarmersintheCongoRebellion . . . . . . . . Characteristics of Cotton Famers ........ General Characteristics . . ......... . Employment History .............. Family Composition ........... . . . Food Consumption Habits .......... ,. . Land Tenure System in the North ...... . . . Land Occupation Rights ............ Renting of Land and Land Utilization Charges . Selection of New Fields ............ Communal Crops and Fields ........... Crop Imposition ................ Labor Utilization . ............... Types and Units of Labor Used on Cotton Farms . Man-Weeks of Labor Input in Cotton Production . Arducusness of Labor and Labor Bottlenecks on Cotton Farms ................ Road Maintenance ............... Work Performed by Farmers for the Head of the Village of Clan ................ Off-Farm Paid Labor and Legal Minimum Wages . . Farm Tools and Equipment ............. Fertilizer and Pesticide Use ........... Cotton Seed . . ................. Delivery of Cotton Seed ............ Planting of Cotton Seed ............ Crop Rotations and Duration of Fal low ...... Tne Cotton Field ...... . . . . ..... . . The Location of Cotton Fields ......... Number and Size of Cotton Fields ....... Termite Hills . . ............... viii Page 164 164 164 164 165 165 166 166 166 169 169 170 171 171 172 173 174 174 176 176 177 179 180 180 181 181 183 184 184 184 185 186 186 187 Chapter VII. Plant Densities ...... . ......... Field Maintenance and Harmdng ....... Cotton Diseases ................ Number and Size of Other Fields ......... Farm Sizes .................... Cotton Yields . . . ............... Cotton Prices . . ................ Knowledge of Cotton Prices . ...... . . . Importance and Effect of Cotton Prices . . . . The Risk of Growing Cotton .......... Sale of Seed Cotton ............... Time and Conditions of Sale .......... Willingness to Join a Sales Cooperative . . . . The Agricultural Extension Service ....... . Livestock and Poultry .............. Farmer's Revenues, Expenditures and Savings . . . SourcesofFarmRevenLe............ Taxes Paid by Farmers . . ........... Exchange of Money and Products with Family Members in Cities ............... Savings, Lending, Borrowing and Expendittnes. . Visits toLocal Markets ...... . . . . . . Dowry ..................... Farmer's Possessions ............... Farmer ' S Attitudes Regarding Cotton Producticn BasedonOpen-EndedQuestions ...... DoPeopleLiketoGrowCotton? Minon'tYolProduceMoreCotton? WhyDoSore FarmersAbandonCotton Prochtion?. What Should Be Done to Encourage More CottonPrcduction? What Kind of Help Do You Like to Receive From t1eGoverment,ONAFITED(orOtherstoProduce NbreandBetterCctton?............ Summary ........ ANANAIXSISCF'IHEREIUINS'IOIABORFORWAND OIHERSEIEEI‘EDWITIESINNORI‘HEWZAIRE . . . . Introduction................... ix 209 209 210 210 211 211 212 215 215 Chapter Page .Methcdolcgy . . ....... . . . . . . . . . . 215 Returns to Land ....... . . . . . . . . 215 Returns to Capital .............. 216 Returns to Labor . . ..... . . . . . . . . 217 Labor Assumptions . . . . . . . . . . 217 Determination of Standard Labor Inputs for Different Crops . ... . . . ...... 218 Estimates of Labor Requirements for Cotton Production . ..... . . . . ...... 219 Determdnation of Yields . . . . . . . . . . . 219 Perennial Crops . . . . . . . . . . . . . . . 220 Price Data . . . . . . . . . . . . . . . . . . 223 Method of Analysis . ...... . . . . . . . 223 Results . . . . . . . . . . . . . . . . . . . . . 224 Analysis of the Results ........... . 224 The Real Purchasing Power of Seed Cotton Since 1960 . . . ...... . . . . . . . . . . . . . 225 Summemy'. . . . ...... . . . . . . . . . . . 227 VIII. ANALYSIS OF THE MAIN DETERMINANTS OF COTTON PRODUCTION IN NORTHERN ZAIRE . . . . . . . . . . . . 228 Yields . . . . . . . . . . . . . . . . . . . . . 229 Method.of Analysis . . . . . . . . . . . . . 230 Results of the Statistical Analysis . . . . . 234 Results for UbangieMongala . . . . . . . . 234 Results for Bas-Uélé . . . . . . . . . . . 236 Results for HauteUélé . . . ...... . . 239 Results for Bas-Uélé and Haut-Uélé . . . . 241 Results for the NOrthern Cotton Belt, Including Ubangi, Mongala, Bas-Uélé and Haut-Uélé . . . . . ......... . . . 243 Acreages . . . . . . . . . . . . . . . . . . . . 245 The Price Elasticity of Supply (Acreage Planted) . . . . . . . . . . . . . ...... 245 Results of the Mhltiple Regression Statistical Analysis . . . . . . . . . . . . . 247 Results for UbangieMcngala . . . . . . . . 248 Results for Bas-Uélé . . . . . . . . . . . 254 Results for Haut-Ue’le’ . . . ........ 254 X Chapter Results for Bas-Uélé and Haut-Uélé . . . . . Results for the Northern Cotton Belt Including Ubangi , Mongala, Bas-Uele andHaut-Ue’lé ........ ...... Corparison of Cotton Production Determinants Between 1958-1959 and 1970-1971 . . . . . . . . Cotparisom Between the Official Goverment Statistics for 1970-1971 and the Farm Business Survey Data for 1972—1973 ...... . . . . . STMTEEIES Fm INCREASING COI'ION PmDIIITIm IN NORIHERN ZAIRE ..... . ....... . ..... Introduction......... .......... Barriers to Improving the Performance of the CottcnSubsectorinNorthernZaire. . . . . . . . Strategies for Developing Northern Zaire ..... Strategies for Expanding Cotton Production inNorthernZaire Revising Pricing Policies and Retoving the Diffusion of Improved Agronomic Practices . . . Overhauling the Cotton Extension Service . . . A Pilot Cotton Development Project in Bas—Uélé. Accelerated Ibsearch am New Varieties . . . . . demicalFertilizers Review of Experiences with Fertilizers inZaire.......... ..... Major Constraints on Using Fertilizers . . . 'IheHighCostofFertilizers . . . . . . LowProductPrices Lack of Fertilizer Response Data . . . . Sociological Factors . ........ . The Agricultural Extension Service . . . Conclusion ................. Pesticides ..... . . . . . ........ Review of Experiences with Pesticides . . . Major Constraints on Using Pesticides . . . Conclusion................. xi 259 265 268 271 271 271 r 274 275 276 279 280 283 285 287 288 292 293 298 299 300 301 301 302 302 305 306 Chapter Page Mechanization......... ........ 307 Review of Experiences with Mechanization . . 307 Major Problems of Mechanization . . . . . . 312 Conclusion ..... . . . ......... 314 Agricultural Credit . . . ........... 314 Experiences with Agricultural Credit . . . . 314 The Administration of Agricultural Credit and Cooperatives . . . . . ......... 320 Ccnclreiom.................323 X SWAIN, POLICY PRESCRIPTIONS FOR INCREASING PRODUCTION ANDFCRIMPROVING'IHEMARKEYPINGPERFORMANCECFTHE COI'ION SUB-SECTOR OF NORTHERN ZAIRE AND FUTURE RESEARCH NEEDS ......... . . . ............ 325 Summary... ..... ............325 Policy Prescriptions for Increasing Production Improving the Marketing Performance of the Cotton Subsector of Northern Zaire . . ......... 340 Revising Pricing Policies and Removing the Imposition System . . ............. 340 Diffusion of Improved Agrmomic Practices andNewInputs ..... ..... 341 Overhauling the Cotton Extension Service . . . 343 Establishing a Pilot Cotton Developrent Project in Bas-Ue’lé . . . . . . . ....... 344 Technical Cotton Research . . . ........ 344 Research on Improved Varieties ....... 344 Research on Better Agronomic Practices andImproved Input Packages . . . . . . . . 345 Improving the Performance of Cotton Marketing . 346 Operational Efficiency . . . . ...... . 346 mducing Transport Costs ....... . 346 ReducingGinningCosts 347 Improving Organization and Administration 347 Diffusion of Information . . . ....... 348 Rational Utilization of Cotton By—Products . 348 CounterpartTraining... ..... ....348 FutureResearchNeedsW. 349 xii Chapter Farm Management Research on Cotton Production Ckmmmmdcations Research for Improving Cotton Extension and Facilitating an Information Feedback to Researchers . . . . .Marketing Research .MacrOPEconomic Research . . . . xiii Page 349 350 350 351 352 Table 1. l 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 3.1 3.2 LIST OF TABIES Republic of Zaire: Production, Exports and Imports of Cotton Fiber (in Tons) ............... Republic of Zaire: Composition of Gross Danestic Pro- duct, 1970-1972 (Value Added in Million Zaires at Current Prices) ................. Republic of Zaire: Value Added in Camercial Agriculture, 1970-1972 (Current Prices in Million Zaires)............. ..... Republic of Zaire: Canposition of Agricultural Production, 1969-1971 (Current Prices) . ....... Reptblic of Zaire: Exports, 1968-1972 (In Million zaireS) o o o o o o ooooooo o o o o o o o o o o 0 Republic of Zaire: Agricultural Exports, 1965-1972 (In Million Zaires) .................. Republic of Zaire: Agricultural Imports, 1968-1973 (In Thousand Metric Tons) . . . ............ Republic of Zaire: Imports of Agricultm'e, 1969-1971 (c.i.f. Values in Million Zaires) . . . . ....... Republic of Zaire: Total Government Expenditures on Agriculture (Current and Investment Budget), 1968-1972 (In Thousand Zaires) ......... . ....... . Republic of Zaire: Goverment Expenditures in Agri- culture, 1968-1973 (In Thousand Zaires) . . ...... Reprlic of Zaire: Anticipated Total Governnent Expenditures in Agriculture (Current and Payment Credit for Investment), 1968-1973 (In Thousand Zaires) . . . Republic of Zaire: Seed Cotton Production, 1915-1974 . Progress Made Through Cotton and Breeding Selection m zair e O O O O O O O O O ..... O ...... O O Page 17 18 19 21 22 22 23 24 25 26 45 56 Table 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3. 10 3.11 3.12 3.13 4.1 4.2 4.3 4.4 4.5 4.6 Location and Characteristics of Paysannats with Cotton as the Main Cash Crop, 1957 ...... . ....... Pesticide Treatments by the Agricultural Service of ONAFITEX, 1971-1972 .................. Average Prices Paid for Seed Cotton in African Countries 1970—1971 . . . . ................... Republic of Zaire: Seed Cotton Price (Advance Payment) Fixed by the Government, 1950-1975 . y ......... Republic of Zaire: Objectives and Realizations of the FED-CAR Project, 1969—1973 .............. Republic of Zaire: Population of the Ubangi, angala, Bas-Uélé and Haut-Uélé Sub-Regions, 1958-1970 ..... Production of the Principal Textile Fibers in Zaire . . Republic of Zaire: Production of Seed Cotton Per Cotton Region, 1959 and 1973 ............. Republic of Zaire: Characteristics of Cotton Pro- duction before Independence, 1954-1958 Averages . . . . Republic of Zaire: Seed Cotton Production in the Northern Cotton Belt Per Region, 1956-1973 ..... Republic of Zaire: Seed Cotton Production in Equator Region, Per Sub-Region, 1956-1972 ........... Republic of Zaire: Ginning Costs and Profits of the Cotton Companies, 1949-1952 . . . . ........ Republic of Zaire: Cotton Companies , Nurber of Ginning Plants and Their location in 1957 ....... Reptblic of Zaire: Situation of the Ginning Plants on September 1, 1970 ................... Republic of Zaire: Cotton Production and Marketing Costs, 1957-1959 and 1969-1971 . . ....... . . . Republic of Zaire: Cotton Marketing and Processing Costs,1969 ....... ...... . Reptblic of Zaire: Breakdown of Cotton Marketing and ProoessingCosts,1969 ........ Page 61 64 67 68 70 78 84 84 85 86 87 99 107 115 117 119 120 Table 4.7 4.8 4.9 4.10 4.11 4.12 4.13 4.14 4. 15 4. l6 4. 17 4.18 4.19 4.20 4. 21 4.22 Republic of Zaire: Cotton Marketing and Processing Costs for Each of the Institutions Involved and for theCottonFamers,l969......... ...... Cost Structure of Cotton Ginning Firms in Zaire in 1971 ..... O OOOOOOOOOOOOOOOOOOO Republic of Zaire: Average Transport Costs in Cotton Marketing, 1969 (In Zaires per Ton of Cotton Fiber and as a Percentage of Total Cotton Marketing and Processing Costs, Included the Purchase Value of Seed Cotton) ..................... Production of Cotton Textiles in Zaire, 1968-1972 . . Equiprent in the Zairean Textile Industry, 1973 . . . Republic of Zaire: Dorestic Sales of Cotton Fiber to Textile Companies, 1970—1974 . . . . . . . . . . . . . Exports of Zairean Cotton Fiber Per Country of Destin- ation, 1970-1973 ................... Exports of the Principal Textile Fibers in Zaire, 1959-1971 ...................... Republic of Zaire: Average Sales Price of Cotton Fiber for Dorestic Sales and for Export, 1964-1973 . . Republic of Zaire: Purchase Price of Cotton Lint From Farmers in Percentage of the Export Price, F.O.B. Matadi, 1950-1973 Average Cotton Fiber Price in New York 1959-1973 . . . Republic of Zaire: Comparison Between July 1974 Cotton Fiber Prices for Export and for Dorestic Use Incote Elasticities in Kinshasa, 1969 ........ Structure of Monthly Expenditures per Family in Function of Income in Kinshasa, 1969 . ........ The Status of the Cotton Seed Crushing Industry in Zaire,1958 Republic of Zaire: 'Ibtal losses of the Cotton Seed Crushing Industry, 1963-1969 . . . . . . . . ..... Page 121 121 123 130 131 131 133 133 135 136 137 138 138 139 144 145 Table 4.23 4.24 4.25 4.26 5.1 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 6.10 Republic of Zaire: losses of the Cotton Seed Crushing Mills, 1968-1969 .................... Republic of Zaire: Tons of Cotton Seed Processed Per Mill, 1963-1969 .................. Production of the Lubumbashi Cotton Seed Crushing Mill, 1939-1972 . . . .................... Availability of Cotton Seeds in Ubangi-Mcngala, Republic of Zaire, 1971-1973 .............. Sampling Plan Used in the Farm Business Survey of 160 Cotton Farmers in Northern Zaire, 1972-1973 ...... Educational Characteristics of Cotton Farmers in Northern Zaire, 1972-1973 ........... . . Northern Zaire: Reasons Mentioned for not vaing to a City, per Sub-Region .................. Average Family Composition on Cotton Farms in Northern Zaire, 1972-1973 . . ........... . ...... Ranking of Principal Food Products Consumed by Cotton Farmers in Percentages of the Responses per Rank, Northern Zaire, 1972-1975 ............... Responses to Question: "Who Selects New Fields to be Brought into Cultivation?", Northern Zaire, 1972-1973 Reasons for not Participating in a Collective Field, Northern Zaire , 19 72- 19 73 ............... Responses to Question: "mo Imposes Cotton Cultiva- tion?", Northern Zaire, 1972-1973 . . . . . . ..... 'Iypes of Labor Used on Cotton Farms in Northern Zaire in Percentages and Units of Family Labor Available, 1972-1973 ............. . . . ....... Labor Input in Man-Weeks for (he Hectare of Cotton, Northern Zaire, 1972-1973 . . . . . . . ........ Period of the Year in Which Farmers were the Busiest, Northern Zaire, 1972-1973 ..... . ......... xvii Page 145 149 168 170 Table Page 6.11 Authorities Requiring Famers to Work on the Roads, Northern Zaire, 1972-1973 ..... . .......... 181 6. 12 Average Number of Tools and Fquigrent Fcnnd on Cotton Farms, Northern Zaire, 1972-1973 . ........... 182 6 . 13 Percentage of Cotton Farmers lacking Tools and Equigrent in Order to Improve Cotton Production , Northern Zaire , 1972-1973 . . . ........ ‘ ......... . . . . 183 6. 14 Persons Who Determine the location of Cotton Fields, Northern Zaire, 1972-1973 . . . . . . .......... 187 6.15 Persons Who Determine the Size of Cotton Fields, Northern Zaire, 1972-1973 . . . . . . . . . ....... 188 6.16 Total Number of Fields in the Survey and Average Number of Fields Per Farm, Northern Zaire, 1972-1973 ...... 188 6. 17 Average Size of Fields, Average Acreage per Fann and Average Total Farm Size, Northern Zaire, 1972-1973 . . . 190 6 . 18 Average Cotton Acreage Per Farm in Forests and in Savannahs, Northern Zaire, 1972-1973 ...... . . . . 190 6.19 Average Cotton Acreage per Farm for 1970-1971 (Census of Agriculture Data) and for 1972-1973 (Farm BLeiness Survey Data), Northern Zaire .............. 191 6. 20 Cotton Plant Density and Spacing Between Cotton Plants, Northern Zaire, 1972-1973 ............... . 192 6.21 Average Total Farm Size and Standard Deviation, Northern Zaire, 1972-1973 ..... . . ............. 194 6 . 22 Average Cotton Acreage Per Farm, Average Cotton Pro- duction per Farm, Average Yields and Average Cotton Revenue per Farm, Northern Zaire, 1972-1973 . . . . . . . 196 6.23 Sources of Cotton Price Information for Farmers, Northern Zaire, 1972-1973 . . . . . . . . . . . ......... 198 6.24 Respective Prices for First Grade Seed Cotton at which Farmers would Change their Cotton Acreage and Prices which Farmers Deon Reasonable, Fair, Northern Zaire, 1972-1973........................ 199 6.25 Sources of Farmers' Revenues, Northern Zaire, 1972-1973 . 204 xviii Table Page 6.26 Responses to Question: "Who Keeps the Proceeds from the Sale of Cotton?", Northern Zaire, 1972-1973 . . . . 206 6. 27 Percentage of Cotton Farmers Who Owned a Bicycle, Radio, Sewing Machine or Watch, Northern Zaire, 1972-19.73 c o o o o o o o o o o o o o o o o o o o o o o 209 7. 1 Reptblic of Zaire: Estimates of Labor Requirements for Cotton Production in Forest and Savannah Areas . . 220 7.2 Republic of Zaire: Estimates of Labor Requirements for Different Activities in Cotton Production ..... 221 7. 3 Estimates of Returns to Labor for Cotton and Other Selected Comodities in Northern Zaire, 1973-1974 . . . 222 7.4 Index of Real Purchasing Power of Seed Cotton in Zaire,1960-1973...................226 8. l Yields and Acreages: Variables Included in the Multiple Linear Regression Equation . . . ....... 233 8.2 Regression Statistics for the Function YLD = fn(ACC, . . .,M)N) Explaining Cotton Yields Unit of Obser- vation: Ubangi and angala, Zaire, Nurber of Obs vations:64......................235 8.3 Regression Statistics for the Function YID = fn(ACC, . . . ,IDC) Explaining Cotton Yields Unit of (laser-vation: Bas-Uélé, Zaire, Number of wservations: 29 . . . . . 237 8.4 Regression Statistics for the Fimction YLD = fn(PRC, . . . , ERN) Explaining Cotton Yields Unit of daservation Haut-Uélé, Zaire, Number of Observations: 19 . . . . . 240 8. 5 Ibgression Statistics for the Function YLD = fn (ACC. . . . ,EI’N) Dcplaining Cotton Yields Unit of Cbservation Bas-Uele and Haut-Uélé, Zaire, Norber of Goservations: 48 O C O O O C O C O O O O O O O O O O O O O O O O O O 242 8.6 mgmssim Statistics for the Function YLD = f(ACC, . . . ,ERN) Explaining Cotton Yields Unit of Observation: Northern Zaire (Ubangi, Mongala, Bas-Uélé and Haut- Uélé) Norber of Cbservations: 112 .......... 244 8.7 Regression Statistics for the Function ACID = fn(ACO, . . . , TER) Explaining Cotton Acreages Unit of Cbservation: Ubangi and Mongala, Zaire: Norber of mn’atiom: 64 O O O O O O O O I O O O O O O O O O O 249 xix Table Page 8.8 Regression Statistics for the Function ACC = fn (CHF, . . . ,SHL) Explaining Cotton Acreages Unit of (loservation: Bas-Uélé, Zaire; Nmber of Cbservations: 29 O O O O O O O O O O O O O O O O O O O O O O O I I O 250 8.9 Regression Statistics for the Function AC1: = fn(IPO, . . . ,YOF) Explaining Cotton Acreages Unit of Observatim: Haut-Uélé, Zaire; Number of Gaservations: l9 ..... 251 8.10 Regression Statistics for the Function ACC = fn (CHF, . . . ,IPC) Explaining Cotton Acreages Unit of Observation: Base-Uélé and Haut-Uélé, Zaire; Nurber ofObservations:48....... ..... 252 8.11 Regression Statistics for the Function ACC = fn(ACO, . . . ,TER) Explaining Cotton Acreages Unit of (laser- vation: Northern Zaire (Ubangi , angala, Bas-Uélé, and Haut—Uélé); Number of Observations: 112 . . . . . 253 8. 12 Republic of Zaire: Comparison of the Coefficients of Multiple Determination (R2) Obtained in the Analysis of Cotton Yields and in the Analysis of CottonAcreagesperFarm.......... ..... 258 8. 13 Characteristics of Cotton Production in Northern Zaire per Zo1e, 1958—1959 and 1970-1971 . ...... 260 8. 14 Characteristics of Cotton Production in Northern Zaire per Zone, 1958-1959 and 1970-1971 . ...... 261 8.15 Production of Seed Cotton in Northern Zaire in 1958- 1959, 1970-1971 and 1972-1973 . . .......... 262 8. 16 Change in Cottcn Production Determinants Between 1958-1959 and 1970-1971 for Equator Region, Zaire (Change in Percentage of the 1958-1959 Figures) . . . 263 8.17 Change in Cotton Production Determinants Between 1958- 1959 and 1970-1971 fOr Haut Zaire Sub-Regions and the Entire Region (Change in Percentage of the 1958-1959 Figures . . . . . . . . . . . . . . . . . . . . . . . 264 8.18 Northern Zaire: Comparison Between Cotton Production Determinants for 1970-1971 (Official Data) and 1972- 1973(FarmBusinessSurveyData 266 Table 8.19 9.1 9.2 9.3 Corpariscn Between Cotton Production Detenninants for 1958-1959 (Official Data) and 1972-1973 (Farm Business Survey Data) for the Ueles and for Northern Zaire . . . Republic of Zaire: Effect of 100 kg. 20-20-0-6 Compos- ite Fertilizer per ha. in Side Dressing, One Week After Planting, (11 Experimental Cotton Fields in [bangi- Mongala, 1973-1974 .................. Prices for Different Types of Fertilizers and Different locations in the Northern Cotton Region, 1971 ..... Republic of Zaire: Fertilizer Nutrient Cost in Gemena, 19 71 ..... . Page 267 291 294 296 LIST OF FIGURES AND MP8 Figure Page 1 Republic of Zaire: Utilization of Cotton and its By-Products, 1973 . . . . . . . . . . . . . . . . . . . 142 Ivhp 1 Administrative Division . . . . . . . . . . . . . . . . 2 2 Population Density . . . . . . . . . . . . . . . . . . 13 3 Natural vegetation . . . . . . . . . . . . . . . . . . 14 xxii LIST OF WEIGHTS AND MEASURES, CURRENCY EQUIWXIEN'I‘S, ABBREVIATIONS AND ADMINISTRATIVE ORGANIZATION (F ZAIRE Weights and Measures 1 hectare (ha.) 2.471054 acres 1 hectare (ha.) 100 ares - 10,000 square meter 1 kiloreter (km.) = 0.621371 mile 1 meter (m.) = 3.28084 feet 1 kilogram (kg.) = 2.204623 pounds 1 gram (9.) = 0.035274 ounce 1 metric ton (tm) = 2,204.6 pounds = 1.10231 short tons = 0.984207 long ton 1 leter (1.) = 0.264178 gallon Temperatures are given in Celsius degrees. Currency Equivalents «Before November 6, 1971: 50 Congolese francs (FC) = 1 US$ --From November 9, 1963 to J1me 23, 1967: 180 Congolese francs (FC) (selling rate) 1 US$ 150 Conglese francs (FC) (buying rate) = 1 USS “After June 23, 1967: l Zaire (Z) = 1,000 Conglese francs (PC) = 0.50 USS 1 Zaire (Z) = 100 Makuta (K) (unit Likuta, plural Makuta) 1 Likuta (K) = 100 Senghi. (S) Abbreviations CNAFITEX (Office National des Fibres Textiles) : National Office for Textile Fibers UNAZA (Universite Nationale du Zaire): National University of Zaire INERA (Institut National d'Etude et de Recherche Agronomique) : National Institute for Agronomic Studies and Research (NRD (Office National de la Recherche et du Developpement): National Office for Research and Development IRES (Institute de Recherches Ecoxoniques et Scciales): Institute for Ecmomic and Social Research FED (Fonds European de Developpemeut) : European DevelOpment Fund USAID: United States Agency for International Development CE'UI‘ (Cotpagnie Francaise pour le Developpement des Fibres Textiles) : French Corpany for the Development of Textile Fibers xxiii IRCI' (Institut de mcherches du Coton dt des Textiles Exotiques) : Research Institute for Cotton and Exotic Textiles CAK (Commission Agricole du Kasai) : Agricultural Commission of Kasai CNATRA (Office National des Transports): National Transport Office COGERCD (Comite de Gerance de la Caisse de Reserve Cotonniere): Management Committee of tie Cotton Reserve Fund CINEmO (Cotptoir des Ventes de Coton) : Cotton Sales Agency SCFIDE (Societe Financiere de Developpement) : Financial Development SCFIDAG (Societe Financiere de Developpement Agricola) : Financial Agricultural Development Company (EDEZA (Caisse Generale d'Epargne du Zaire): General Savings Fund of Zaire Administrative Organization of Zaire "Before 1972 : the country was divided into provinces , each province into districts, each district into territories, each territory into sectors and each sector into villages . -Since 1972: The country has been divided into regions, each region into sub-regions , each sub-region into zones , each zone into local collectivities and each local collectivity into localities. -'Ihe Department of Agriculture, previously called the Ministry of Agriculture, is headed by the State Commissioner for Agriculture, formerly called the Minister of Agriculture, while the civil service is headed by a Director-General who was previously called Secretary-General. Miscellaneom In the Tables: --A long dash (---) indicates zero or less than half the unit shown, or the nonexistence of the item. -Dots (...) indicate that the data are not available. In the Tables and Text: --A slash between years (1972/1973) indicates a single fiscal or crop year. --A short dash (1972-1973) indicates a period of two or more years, though not necessarily claendar years. XXiV ._._ \~‘ .5» .__ . I. \- s .. . o r ‘4 - 1n 1 . \ . s CHAPTER I INTRODUCTION Problem Setting and Importance of the Study The Republic of Zaire, situated in the heart of Tropical Africa, gained independence from Belgium in 1960.1 The first five years of independence were characterized by civil and political disorders and disruption of the economy. The country recovered slowly after 1965 and is presently engaged in an active period of "nation building. " A map of the administrative divisions in Zaire is presented on the next page. Sore 21 million people live in an area one-third the size of the United States. Only 1.2 percent of the total land area is under culti- vation in annual field crOps and tree crops. About half the land is classified as arable. Shifting cultivation and intercropping are the typical methods of farming. About 30 percent of the population is urban and the proportion is rapidly increasing because of rural to urban migration and higher birth rates in the cities. The Zairean economy is well diversified compared with that of other African countries. Agricultural exports made up 16.7 percent of total exports in 1972 compared with 40 percent in 1958. Zaire generates considerable foreign exchange earnings from a variety of minerals, in lThe Democratic Republic of the Congo changed its name to the Republic of Zaire in October 1971. 1 a 9...}. Can: a... — 22 :43... . .. . L - ~ .11 L I I I 1 _H.Jr\\/ - - n— h 23:3...- m_7FF’UK N J .r) ...ll . . . 1 \ in): _ . [I < _ m (< < N A 2’ a 3n: XX 8 .IW‘JW~ fl/J‘h 44°01“ _ (z A l|11...|.|J .... l 1:... .:..u. I) on. R. 9.. o . 1 (I. _ 1.1.1 . VLI/ 1! q n / Jn\ JFII/ 3:2; 11...: :... .5 M~\1. \V w _ ”HIE... uniu Kuw‘y . . , a i .4 3 . ., A J3).;\ ‘ "4.1% 44.5. :..“... <52: in... 1 .: . MA. 4 aw, Lian: .. .. ...1, . — ,1Y1/1111lx in, f 1.. r 1 fl // 3.. I: .\zu\u «(u $1.... . < m < . r m .. / I x v A m m 1 xx «.50.: If, \ < 4 o o z< m t 5.4. N 4 m . 11.111. 4 ... \ x 920 411)\I~ r.» x n. ... 1.. .1. .EQEUL .:."... \(J .- w \, \ x3... . F. . .1 ..I...,Lr\ ... /. s. . .;M N :33... .1. :.. :ahuQ /. 82:; ....1................ / 33262.4. £131.: .111! 3:311 eixcuucs , 1w \/.1 £2 .\ uni-.8. . (N14111n011mm12. / 1. (.mmlaN maxi“ . . ‘ ......Eho .... mS<~ «.2523 _ ..:.... N 4...... in. game L, s. we \7- 332x 7 H \J . Egan.” \1_. ..AU 1.. . / (J .9.» / k r.1\ H1 «:52; O ..x . u 1/1. . I I / F/ \x 2.55.. $4. .1 \. _ i «232...: ecu/.11 \ 0d .\ r. 522.an it. o? A \ WESMA 09 44/ L16 ...... ...4 o a w. .111.‘ 44.9.5 /\. . . ymu 11 l.— . 4824;: x Q 73/ . M {£13. .1341 a» O A .L ./\.?.\w§ ..., 04 2030 l S. i o x 43243: i a; w ‘32.! x 9.02 ./ . 3:2. :3: . a. .83 . t . x L 3.51 7 Av marquee . in; . 5:39.90 ..\( J) H x :S\.\ ... “1121:33 o .0» W0 r)- . \. one}. .lanJ4 114 0.. “M4\1\\\1\\.. 4... u \1N. . 4 x4 ...\ ./11. U . . i=1 . P x»... w Wis; <1N .14.._.7D4.)1J113 A/ o .1111» Mr (I (I. 4.3.58.3..51111 .J .4 r111 (4 :0 £35 1311 , tovcaoa _o_u¢;n£ . 4. 99.01I(. o . 102<¢3 {856: .:..o:.o.slu.l f .\ .9 Pd 1 7..” ... ...—r. 31.4 R 33 .256 o . —. . ) . O :nfiv _n ue.:o_ vl\r\\fl wow. I n .4. . 9..qu .l.\....\{l\ 36.2..” .993“ " ./. x. 1, .. mzo_m_>_o m>:u<32(_. _ . 3 ~ \. Think / I I1 .3 . on m 7 rlk J B :.....3... a. F\ 1.55:4: _ an. R. S. 92 8 n m ~ J 1 I~.«° Mitht‘ a. . Manama \w.22 .1. .1 x. ( 1 T: ezc- coo Z :2: .KIGALI) ~4 v- . _ ‘5, \ ) . , A 7 , )_ "x” __‘ RWANDA) REPUBLIC or z \. 1 \ I Pu‘mo“ ’7” I ,) /] ‘IF—‘U “we ”\j k . L ‘ . CONGO 8"qu “cu, m f/ V 7— NASA: ORIENTAL 1/ .W I) MAI ll‘ilLLE \ x “ /\ \SOM.“ 'L) ‘1 ' BURU N/D' “,5 ‘5’,» \\ BANDANDU Argue" no»... \ 1 f O f’\>\ v)“ ,\ VFNSHAS‘ 7" P \“5 I .4.- r‘ O , I . ”of? WW“. I °w on... \KASAI occmc NTAL’\ ome ‘X _. - _ 7- - .w - __. _. -.." v.4 I o I, 1 “'4. “\ / (o‘- A 90-;0 °_q A .S__ 2_ '_E_ l > a. .1) F») g) ’WWI‘O .o . h.“ x ,/ LULquuohéa" K (dy‘WM L h, . a V . _-- ,. 1 cut-o ‘1 f , .\ *M ,—/ \ ‘YANZANIA ‘3 K “ YIh-Mbv‘uo V’) \ ' ‘3 \LLH / (K’C A N G O L A a (— [ .._. _ :1 I, /~\/ y ’ I . r :v4\/_( J O . . -—~. A ~ J Lu—J f \’ \.‘. '\ x... ’ __, S ‘ Oran-m ”I... ‘u \ 1:: '~ (I U REPUBLIC OF ZAIRE TANGA K A . Yo Allonhc \ [’4- M Itch-w I Ocean w )1.’." ("(51.0 ( \ fowywusu 7‘ “'0‘“ 1 x 11; / \\"\_1._‘11 _.._.._._.S 2 A M BIA \\ : K-"QJ 7h MD. M- n- PL- .0: uh .- :..-r. 1W .... o—vv ’ "'v 0 I00 I” no 00. Iouuuo—vmmm nun-uni ° :- '1- :- Ir t Inu- T '1‘- '1' 1LQ~ '5' |“7" A IB.D 3°". 15 are cultivated. Generally speaking, the overall quality of the soils is average for tr0pical countries. Zaire is almost completely in the Guinean ecological zcme . Gross Domestic Product, Sources, Growth Rates and Agricultural Value Added Perhaps the most important influence on the Zairean economy in recent years was the political crisis in the 19603. All economic grmth indices fell drastically from the 1959 levels and only in 1969 did they return to pre- independence levels. 5 Farm output fell sharply after 1960 as insecurity, unrest and rebellion spread through the country when Zaire gained independence from Belgium. The Gross Domestic Product (GDP) was estimated at 1,014 million Zaires (8) ($2,028 million 0.8.) in 1970 or 475 ($94) on a per capita basis. The economy of Zaire is divided in two basic sectors: the mining industry and agriculture. Cooper, cobalt and diamonds are the most important mu'neral resources, with the copper mining industry accounting for one-fifth of the GDP, practically two-thirds of total export earnings and one-half of government revenues. The GDP grew at a rate of 5.4 percent in 1971, 4.5 percent in 1972 and 6.6 percent in 1973. These rates were in sharp contrast with rates of 9 to 10 percent experienced in 1969 and 1970. 'Ihe agricultural value added at current prices in 19 72 was 15. 3 percent at GDP, or 172.4 million Zaires. The share of commercial agricultural value added in GDP at market prices was 8. 2 percent in 19 72 , 5Except copper production indices which never fell drastically. 16 while the share of subsistence agriculture was 7.1 percent (Table 2.1) . Services make up the largest part of GDP followed by trade, mining and metallurgy. Table 2. 2 gives a breakdown of value added in agriculture, in agricultural exports, darestic nonfood crops and foodstuffs. Structure of Agricultural Production and Marketing It is important to point out the clear distinction between com- mercial and subsistence agriculture in Zaire. The commercial sector is dominated by plantations, many financed by Europeans, which are highly capital intensive and produce palm oil, coffee and rubber for export and sugar and livestock for domestic markets. Since December 1973 only Zairean nationals can own and operate the ccmmerical agri- cultural plantations, ranches and fanms. The subsistence sector, (11 the other hand, is made up primarily of traditional small land-holder farms, with usually less than one hectare under cultivation, cultiva- ting food staples, manioc, rice, plantains, maize, beans and groundnuts in well established rotations. These operations are subsistence oriented but provide supplies of food staples to urban markets . The carposition of agricultural production is illustrated in Table 2. 3. Total agricultural production increased rapidly from 1969-1970, but stagnated or even declined from 1970 on. This stagnation in pro- duction can be partially explained by reduced areas planted due to out-mirgation and serious drops in yields. Along with the rapid increase in population and the large migration to urban areas , such lagging agricultural production has created a serious gap between aggregate demand for farm products, especially food, and aggregate supply. Table 2.1. Republic of Zaire: Composition of Gross Domestic Product, 1970-1972 (Value Added in Million Zaires at Current Prices) Composition of GDP 1970 1971 1972 Percent 1972 Camercialized Agriculture 79.1 82.6 91.0 8.2 Subsistence Agriculture 79.8 80.6 81.4 7.1 Mining and Metallurgy 211.5' 152.0 162.7 14.3 manufacturing 77. 5 89 . 2 102 . 1 8. 9 Construction and Public Works 30.0 39.4 33.6 2.9 Water and Electricity 8.7 9.5 9.6 0.8 Transport and Telecommunications 75.9 86.4 97.3 8.5 Trade 116.0 142.4 168.3 14.8 Servicesz 222. 7 276 . 4 316. 2 27. 6 Banking (imputed)3 -4.8 —5.6 -5.7 —o.5 Import taxes 51.4 59.8 67.6 5.9 Nonccmmercialized Construction 15.0 19.7 16.7 1.5 Gross Domestic Product 962.8 1,032.4 1,140.8 100.0 lBanque du Zaire, December 1973. Rapport Annuel 1972-1973, Kinshasa, p. 68. 2Including government services 3Equal to the difference between interests received and interests paid by financial institutions. 18 Table 2.2. Republic of Zaire: Value Added in Commercial Agriculture, 1970-1972 (Current Prices in Million Zaires)l Composition of Commercial Agriculture 1970 1971 1972 In %, 1972 A. Agricultural Exports 32.2 28.7 30.2 33.2 Oil Palm Products 11.7 9.8 5.0 5.5 Coffee 9.4 9.7 14.7 16.1 Rubber 4.1 3.0 3.1 3.4 Cotton 3.1 2.9 2.8 3.1 Timber 0.8 1.1 1.6 1.8 Tea 0.7 0.9 0.8 0.9 Cocoa 0.8 0.5 0.8 0.9 Other 1.6‘ 0.8 1.4 1.5 B. Darestic Nonfocd Crops 13.3 18.6 22.9 25.2 Oil Palm Products 3.0 3.5 3.0 3.3 Timber 2.6 2.5 2.7 3.0 Cotton 3 9 6.9 9.3 10.2 Other 3.8 5.7 7.9 8.7 C. Foodstuffs 33.6 35.3 37.9 41.6 Crops 22.8 24.2 26.6 29.2 Livestock 5.3 5.4 5.4 5.9 Fishing 5.5 5.7 5.9 6.5 Total 79.1 82.6 9.10 100.0 1‘Bancnle du Zaire, December 1973. Rappprt Annuel 1972-1973, Kinshasa, p. 69. \nsl. 19 . memo» cognizant, 5 mcflmflmum m5 mmmuoxm $33 NEH we. 83383 umfiufl an. ”Emma no: 8 mung mo 628 an... no 8H3”. Hmoflmflfim #3 9n. "362 .8 d .825 .83.;an H852 fiommmm .meH HOD—mg .OHHMN o6 magnum Em. whims. .Q .mmmnmcg .:malonma H0592. PHOQQMm 22.3 “395000 .OHMMN ac ”HEMP.” .352 0:64; 0.2: Emma 82: To: 38.8 ~38. mi. 3% - T? 93 .23 3: 88868.1 82333.5 3w 6.3 m.m mi manna 6:6 8.08153 m6 fifl Tm TE 390 8882 638:8 Gem 8?. m.- TR mood Page. Hi TR DE fimm . mmmfimcoom 92 98 23 H5 macho m.mm :3 Ham 3:: 9mm 6.: coflusooum 68268qu 36th 66.43 Ammaa H38. no 8335 138. no 8333 889 mo 8.33% ommunwonom Saws mmmucmonwm 83> wmmucmoumm 83> 2288on :2 23 $3 333622 «0 Sflamoeso AHmOOHHm quHHSUV .:.mHImmmH .coflubncoum nguaaahmd mo COGUHmOQEDO "0.3an m0 OHHnsmmm .m.m canon. 20 The marketing system consists primarily of small-scale traders operating at primary and wholesale levels with extreme fragmentation at the retail level. The serious deteriation of the transport network, especially since independence, and high storage and handling losses contribute to the very high distribution margins. At the end of 1973 nine agricultural product marketing offices were created along the lines of the cotton marketing office (ONAFITEX) with broad responsi- bilities for the production and marketing of the principal food and export crops. Agricultural Exports and Imports In 1959, agricultural exports accounted for 39 percent of total exports while in 1972, they had dropped to 17 percent. They increased in value rapidly between 1968 and 1970, and then stabilized in 1971 at about 60 million Zaires (Table 2.4), a stabilization attributable to the drop in world prices for the main export crops. Coffee, palm oil, rubber, forest products, cotton, tea and cocoa are the principal export crOps. Over the 1965-1972 period, oil palm exports declined rapidly, mile rubber exports decreased slowly and coffee exports increased as shown in Table 2.5. Note that only the exports of coffee are above the pre-1960 level. In recent years Zaire has been a net importer of several major food products (Table 2.6). The portion of food imports in total imports increased from 14.8 percent in 1970 to 17.6 percent in 1972. Imports of foodgrains are increasing, especially maize and rice. In fact, the Department of Economics in Zaire estimates that maize production has to increase at an annual rate of 6 percent to keep up with increasing 21 demand. Zaire also imports sugar, meat and fish. Fish being the most important animal protein sourCe. Imports of supplies and equipment for agricultural developrent are also increasing, as shown in Table 2.7, but do not reach 3 percent of the value of all imports. Table 2.4. Republic of Zaire: Exports, 1968-1972 (In Million Zairesl) Exports 1968 1969 1970 1971 1972 Agricultural Exports 46.9 41.5 61.5 63.7 57.6 Total Exports 298.5 336.6 400.4 348.2 345.7 Agricultural Exports as a Percentage of Total Exports 15.7 12.3 15.4 18.3 16.7 1République du Zaire, Novembre 1973. Conjoncture Economique, Année . 1972 et Ier Trimestre 1973. Kinshasa: MinstEre de l'Economie Nationale, Nr. 13, p. 14. The Role of Government in Agricultural Development Objectives for Agricultural DeveIOpment and Planning In December 1970 the government identified the following target variables for agricultural development:6 1. increasing agricultural incomes of farmers, 2. self-sufficiency in basic food products and in cotton and a diminishing of the country's dependence on meat imports,7 3. agricultural diversification to~ increase export opportunities, 4. creation and expansion of agricultural processing industries. 6Presentation by the Department of Agriculture to the Consultative Group of the World Bank, November 1973 in Lumpungu Kamanda, March 1974. "Les Problemes Actuels de l'Economie Agricole Zairoise." Kinshasa: I.R.E.S., Séminaire sur 1e Développement Agricole du Zaire. 7If possible, to becare an exporter of these products. Table 2.5. Republic of Z 're: Million Zaires ) 22 Agricultural Exports, 1965-1972 (In Agriculuiral Ecports 1965 1968 1969 1970 1971 1972 Percent 1972 Palm Oil Products 12.9 22.0 13.1 19.6 19.1 13.8 24.0 Rubber 4.5 6.9 7.3 7.2 6.0 5.6 9.7 Coffee 8.6 12.3 12.8 23.5 24.6 28.0 48.6 Cocoa 0.6 1.4 1.4 1.5 1.3 1.1 1.9 Cotton and By-Products -- --- 3.4 5.7 4.2 2.3 4.0 Tea 0.9 ' 0.9 1.0 1.5 1.8 2.3 4.0 ForestProducts 3.0 2.4 2.3 1.7 2.2 2.9 5.0 Other Agricultural Products 1.0 1.0 0.2 0.8 4.5 1.6 2.8 Total Agricultural ~ Exports 31.5 46.9 41.5 61.5 63.7 57.6 ‘ 100.0 J'Republique du Zaire, Novembre 1973. Conjoncture Economique, Année 1972 et Ier Trimestre 1973. Nationale, Nr. 13, pp. 18—19. Table 2.6. Republic of Zaire: Kinshasa : Ministére de 1 'Eccncmie Agricultural Imports, 1968-1973 (In Thousand Metric Tcnsl) Agricultural Imports 1968 1970 1971 1972 1973 Maize 57.7 87.5 107.0 108.5 138.5 Rice 17.7 24.2 25.0 30.0 . . . Sugar 0.9 15.0 20.2 20.0 . . Beef 6.5 12.5 12.0 12.5 . . Fish & Fish Products 20.0 24.8 25.0 25.1 . . Wheat and.Wheat.Flour 49.4 83.5 85.0 77.4 . . . malt . . . 37.3 . . . . . . . . . lReptblique du Zaire, Novembre 1973. Conicncture Economique, Année £912 et Ier Trimestre 1973. Nationale, Nr. 13, pp. 52-54; Nassee, et a1., 1973, 14-35; Banque du Rapport Annuel 1972-1973, Kinshasa, p. 80. Zaire, December 1973. Kinshasa: Ministére de l'Economie 23 Table 2. 7. Republic of Zaire: Imports of Agriculture 1969-1971 (c.i.f. Values in Million Zairesi) Imports for Agriculture 1969 1970 1971 Supplies for Agriculture 1. 7 2.5 4.5 Agricultural Equipment 2.9 3.1 3.1 Total 4.6 5.6 7.6 Total Imports in Zaire 197.6 267.8 309.6 J'Nasse, et a1., p. 14. It is clear that trade-offs are involved in these target variables. However, the question of priorities for target variables remains unanswered. Projections have been made on production and consumption for the major food and export crops and livestock over the 1971-1980 period in the ten year agricultural program [République Démocratique du Congo , Mai, 1971] . The department of Agriculture first prepared a five-year (1971-1975) program for presentatim to the first meeting of the Consul- tative group of the World Bank in April 1971 and later expanded this into a ten-year (1970-1980) agricultural program. The main emphasis of this program is on production objectives. Other important issues and problems, houever, such as price policies, agricultural research and extension, infrastructure, staff training, etc. are not covered. Agriculture in the Government Budget Although agriculture was declared the priority of the priorities in economic and social deveth in 1972 the expenditures for agricul- ture were less than 3 percent of total government expenditures as shown 24 in Table 2.8. However, it should be noted that substantial expenditures on agriculture are incurred by other departments and agencies . For example, part of the apprOpriations for regional expenditures are devoted to agriculture8 and road building and maintenance is the respon- sibility of a government agency, the "Office des Routes." In addition agricultural education is the responsibility of the Department of Education and agricultural research is financed directly from the Presidency. Table 2. 8. Republic of Zaire: Total Government Expenditures on Agriculture (Current and Investment Budget), 1968-1972 (In Thousand Zairesl) Government Expenditures 1968 1969 1970 1971 19722 Agricultural Sector 2,247 5,244 3,650 4,501 7,512 Total Expenditures 195,229 265,132 326,951 347,643 316,379 Percentage of Agriculture in Total Expenditures 1.2 2.0 1.1 1.3 2.37 1Table 2 . 9 and compilations by the author. 2Only first half of 1972 for investment expenditures. For 1974 Agence Zaire Presse, the national press agency, indicated that seven million Zaires were budgeted for agriculture. In 1973 the Office of the Presidency directly controlled 30.2 percent of total government payment authorizations and 40 . 2 percent of total government payments for development projects. Sate of these projects are agricultural development schemes or scheIes which directly benefit agriculture. Thus, the real government expenditures for agricul- ture are higher than those shown in Tables 2.8, 2.9 and 2.10. Nevertheless, ‘ _ 8'Ihe regional Department of Agriculture administration is flmded in part through the regional budget. c Q .I\ .... Cc... . 25 .vHHLHHH .ma ..HZ .mflmcofl—mz wgcoom; mp meumHGHS "mama .thH .95 H9598 umHfimN .mnma gmgha HmH um NEH owccm .Qagcoom wnouocohcoo .mp3 89382 sedan so magnum? pom m d .mp3 71.. pm .mmmmzH . . . . ca.~ ma.~ ~e.m ae.c me.a musuasuaumc How mmmucmoumm mwudufiucmmcnm . . . . mrm.am mc~.me mem.mc Hem.~m eac.m~ mmuauaccwaxm amuse msauocmum . . . . mmam mmo.~ cma.~ mee.m Hem.H musuHsUauec me.~ am.e eH.m Hm.e em.e ea.aH casuasoauec How mmficmoumm eaa.mm mmc.ma amm.ec eem.am eee.ea eea.ma rectum sauce mmuauaccmaxm Nee.~ ecm.m mee.~ aH~.~ cac.m ec~.m muauaaoaums acmeuumacH mow uacmso racemes cm.e Hm.~ mH.e ec.m ce.e e.eH cuspasuauuc How mcmucwonwm moc.cea ame.mc ac~.caa ema.eaa eme.me cao.ae acoaumsauocuac_acuce umuauaccmaxm acc.H mmm.a pea are.c cme.e cea.e was» conceaoaucc 1.33% new 953 now Huocuaé pawn . . . . ec.~ mm.e em.e em.c ~m.o mucuasuaumc How mmmucmouwm mmuwwgcwwwlw . . . . cae.me~ mam.mem mem.cc~ Hmc.mam mac.aca ummcsm Hence “pom . . . . mac.e ace.~ cam.a cmm.H cam cusuasoaumc umuauaccwaxm Rd afim mac me.~ 2.4 e.~ 6953633 85an mammao Hoe temperance ummmwmaoaucc mam.~c~ ecm.emm cec.ra~ mmc.mca eee.mca Hae.~aa umucsm Hares . . as~.e mmc.m ace.a mam.m Hmm.~ Hm~.~ «senescence mead mama Heaa chad coma mesa mueuaccwaxm . «0 when. Aammuflmn 23.905. Cd mnmalmwma .muouaooflumd 5 mouse? agendas/00 5.53 no Deanamm .m.m manna 26 Table 2.10. Republic of Zaire: Anticipated Total Government Expendi- tures in Agriculture (Current and Payment Credit for Investment), 1968-1973 (In Thousand Zairesl) Government 1968 1969 1970 1971 1972 1973 Expenditures Agricultural Sector 5, 491 5, 145 5, 532 3, 716 8 ,963 9 , 319 Total Expenditures 130,950 185,000 214,955 284,899 330,001 346,088 Percentage of Agriculture in the Total Expenditures 4.2 2.8 2.6' 1.3 2.7 2.7 J‘I‘able 2. 9 and compilatiols by the author. government expenditures for agriculture are low, and a rather drastic shift in budget priorities is needed if agriculture is to becore the priority of the priorities. Recently, for instance, an agricultural development bank has been created (SOFIDAG: Société Financiere de Développement Agricole) , an institution which could play a vital role in channelling more funds into agriculture. The current expenditures on agriculture increased from 2.03 per- cent in 1970 to 2.7 percent of the total government budget in 1973 as shown in Table 2.9. However, the Department of Agriculture can not fully utilize its current and investment allocation because of a short- age of qualified staff to execute programs. The investment budget is aimed at the achieverent of long-term development goals and is divided in two cateogries: payment authoriza- tions which reflect multi—year commitments and payment credits which are the annual disburserent of investment funds . Payment authorizations for agriculture have been declining since 1968 in absolute and in rela- tive terms. Nbreover effective investments have been declining in rehtive terms over the same period and in absolute terms since 1969. 27 It is clear that agriculture holds a low priority in government policies. Indeed, agriculture occupies a low status position of all Zairean society. However, the picture is beginning to change somewhat. Current anticipated and effective government expenditures for agriculture have been increasing absolutely and relatively since 1971 and a further increase is expected. Pricing Policies, Wages and Taxation The Department of National Economy is responsible for price and wage policies. In October 1967, with the currency reform, new price and wage regulations took effect. The present agricultural pricing policy is characterized by minimum producer prices for almost all food and export products, maximum whole- sale food prices, fixed producer prices for crops sold to government marketing agencies and fixed marketing and profit margins for the whole- sale and retail levels.9 Thus, three different price regulations are established: 1. Minimum producer prices are fixed for all export crops and for maize, rice, manioc and groundnuts. Regional authorities are permitted to establish minimmmm producer prices for maize, manioc and groundnuts for other products , regional authorities can fix prices at levels not exceeding ten times actual prices in June 1960. 2. Fixed producer prices: for sad cotton and coffee; a uniform price is fixed for all regions. 3. Maximum wholesale prices: for rice, maize, wheat, flour, palm oil, sugar and cattle. 9The autl’or published an article on the effect of these price regulations on domestic production and imports in the maize subsector [Tollens and Kamuanga, 1975] . 28 Minimum producer prices and maximum wholesale prices remained largely unchanged until 1972 in spite of a rise of 44 percent in the retail food price index for Kinshasa from December 1967 to February 1972.10 With the creation of several government commodity marketing offices in 1973-74, however, most prices were adjusted upwards. The price of seed cotton was raised in 1970, 1971 and 1974 and the price of coffee was increased in 1972 and 1974. In 1969, maximum wholesale and retail margins were established for imported products. It is generally believed that the farm gate prices tend to stabilize near the fixed minimum price level set by the government because of the oligopsonistic or monopsonistic power of traders and middlemen. Regula- tions concerning maximum prices and margins have generally not been respected. Indeed, such legislation is difficult to enforce. The goverment fixes minimum wages for workers according to skill and region. Comercial (plantation) agriculture is practically the sole employer of salaried labor in agriculture, and across the board salary increases of 30 percent were decreed in 19 71. In addition, Zaire '5 social security system, probably the oldest and most advanced system in Africa [Werbi, 1966] , has the general effect on employers of doubling wage costs as a result of increases in social security contributions and fringe benefits such as housing, medical care, transportation, schooling. Taxes are administered by the Department of Finance. A net profit tax of 40 percent applies to all agricultural firms and corporations, 10There are a number of different price indexes in Zaire. This information is derived from the "Retail Price Index of Current Consump- tion Goods in Kinshasa Markets" published by I.R.E.S. , Kinshasa. 29 while the personal minimum tax (C.P.M. : contribution personnelle minimum or minimum personal contribution) is fixed at 2 Zaires per year for all able-bodied adult males who do not have a paid job. Employed males pay an inch tax on the basis of their wage. Excise taxes are levied on beer, wines , alcoholic beverages, tobacco and sugar. These taxes generate the largest revenue of all agricultural taxes. Taxes on agricultural exports are second in gen- erating revenue from agricultural products. Several types of these taxes are levied: 1. export duties: are ad valorem taxes levied on the basis of the F.O.B. value at the principal port of export, Matadi. Rates differ depending on the product and, for some products such as palm oil, on the export price. 2. turnover tax (T.C.A.: taxe sur 1e chiffre d'affaires): is similar to a sales tax and also applies to exports. The rate is 6.75 percent of agricultural export proceeds. 3. statistical tax: ad valorem rate of 1 percent and applies to all exports to cover the cost of collecting trade statistics. 4. research tax (taxe de selection): a tax designed to foster agricultural research. This tax is levied on all agricultural products excqat forest products. The rate is 2 percent of the F.O.B. value with the exception of cotton to which a 1 percent rates applies. For most agricultural products, total export taxes are between 10 and 20 percent of the export value, depending on the year (as export levies Change from year to year) . 30 Imported inputs for agriculture are also subject to taxes: custom taxes, fiscal taxes, turnover tax and statistical tax. Up to 1973, the effective rate of fertilizers, insecticides and farm machinery was about 22 percent. Since January 5, 1973, import taxes (custom taxes) for vital foodstuffs (beef, maize, rice, wheat flour) and for agricultural inputs (fertilizers and insecticides) have been abolished while luxury products have been taxed more according 1y . Agricultural Development Projects and External Aid In this section only the major projects related directly to agri- cultural develqment will be discussed. A major agricultural develop- ment schete for cotton and food crops is underway at Gandaj ika in Eastern Kasai. This project, considered as one of the most successful in Zaire, is financed by the European Development Fund (FED) . Other FED projects are a tobacco project in Shaba (formerly Katanga) , an oil palm and cocoa schele in Ubangi and a tea development project in Northern Kivu. USAID is committed to bringing the "Green Revolution" to Zaire. They operate a maize breeding, seed distribution and fertilizer applica- tion project in Shaba. A pilot agricultural credit project in lower Zaire region financed by USAID was terminated in 1972. In addition, about twenty Zairean crop scientists are being trained at CDMYT in A large mechanized maize production project at Kasese in the. Kaniama area is now in its third year. A large cattle ranch is associ- ated with this project. The project is supported by Belgian technical assistance and is aimed at an annual production of 50,000 tons of maize C11 18,000 hectares and a cattle ranch of 13,000 head by 1980. In the 31 North of Ubangi a Belgian team operates a general rural development pro— ject with agricultural.extension, tractor mechanization and a distribu- tion of small farm implements and improved seeds to farmers. Belgian tedlnical assistance personnel are also working in the Faculty of Agronomy at Yangambi and in the Higher Institutes of Agricultural Education at Bengamisa and Mendongo. INERA receives substantial Belgian assistance for its administra- tive and research work, while the Federal Republic of Germany is aiding INERA in developing livestOCk research and extension in the Ituri region. The Wbrld Bank is developing three livestock ranches in Shaba, with a project goal of 79,000 head of cattle. Thelflorld Bank also has a major stake in road maintenance programs which receive additional financial support from FED, USAID and the other bilateral donors. Rice production and extension schemes in several parts of the country are supported.by the Mainland China.Agricu1tural Mission, and the French technical assistance program has a vegetable production pro- ject near Kinshasa and a fisheries project at lake Mai Ndorbe. Moreover, the2Canadian government is sponSoring a major forestry project, including a detailed inventory of part of the central basin equatorial rain forest and the training of Zaireans in forestry at the University, the Depart- ment of Agriculture and in a Higher Institute of Agricultural Education. Finally, the Department of.Agriculture receives technical assis- tance from several agencies: a Belgian team operates in several divi- sions; a French team from BDPA is engaged in economic research and planning; a USAID team is setting up an Agricultural Economics Research Center; and UNDP/EAO experts provide assistance in the statistical division and in other offices. 32 Agricultural Marketing and Transportation Zaire is almost entirely within the drainage system of the Zaire River. The transport network is built around the Zaire River and its affluents with 1,700 miles of navigable waterways. The whole Zaire basin river system totals about 7,200 miles of such waterways. Rail- roads bypass unnavigable portions of the rivers or feed into the river ports. There are 3,125 miles of rail lines, and roads act as feeder lines for both rail and river transport. The transport system is geared to the export port of Matadi via a rail coinecticn with Kinshasa and to the port of Lobito in Angola via a rail connection with Dilolo. Connections between regions, especially in The North-South direction, are poorly developed making interregional exchanges difficult. The government operates the National Transport Office (ONATRA) which assures traffic on the waterways, the exploitation of the ports and the Kinshasa-Matadi, Tshela—Boma and Kisangani-Ubundi railroad. The second public transport institution is the Zaire Office for the Railroads of the Great lakes (C.F.L.) which connects Kindu to Kalemie and Kabalo to Kabongo by railroad, assures transport on the Kivu Lake and operates the ports of Kalundu, Bukavu and Gone. Air Zaire, which recently acquired an impressive fleet of modern jet planes, is also a government corpany. Road repair and maintenance on the major roads is carried out by a goverment agency, the Office des Routes, which receives substantial aid from the World Bank and USAID. Feeder roads are to be maintained by the local communities alongside the roads. The goverment intervenes in agricultural marketing via parastatal organizations. These organizations Operate under the authority of the 33 Department of Agriculture and depend on the government for financing. Prior to November 30, 1973, there were five parastatal organizations concerned with agricultural marketing: 1. the Office of Robusta Coffee (OCR), 2. the National Coffee Council (CNC) , 3. the National Textile Fibers Office (ONAFITEX) , 4. the Office for Agricultural Products of the Kivu (OPAK) , 5. the Office for Agricultural products and Ranching of Kisangani (OPAEKI) . (mam was created in 1971 to replace COGERCO and COVENCO, which dealt with cotton ginning and selling. In July 1972 OCR and CNC were merged in a new institution, the National Coffee Office (ONC) , modeled after (NAFI'I‘EX, while in 1973 the National Livestock Development Office (ONDE) was established primarily to coordinate the managerent of govern- ment ranches. In tlne same year a National Ivory Office (ONI) was insti- tuted with monopoly power for buying and selling of ivory in order to reduce illegal exports. On Noverber 30, 1973 President Mobutu Sese Seko announced a govern- ment take-over of all plantations, ranches, farms and agro—industrial corplexes held by foreigners for distribution to Zairean nationals and to the government. The main purpose of this move was to place the agricultural economy under tlne direct control of the Zairean people. In implementing this order, seven new parastatal marketing offices were created. They are: 1. the National Forest Office (ONB), . 2. the National Cocoa Office (OM?) , 34 3. the National Fisheries Office (ONP) , 4. the National Sugar Office (ONDS) , 5. the National Natural Rubber Office (ONCN) , 6. the National Oleaginous Office (ONO), 7. the National Cereals Office (CNACER) . This list brings the total number of government marketing offices controlled by the Department of Agriculture to ten. 1]" These product marketing offices include export crops as well as crops or foodstuffs for dorestic consumption. They are basically one-channel marketing institutions with a buying and selling monopoly . However, their activities are not restricted to buying, selling and exportation. They also manage directly plantations, ranches, farms and processing plants. Thus, they have an important stake in production, transportation, pro- cessing , storage and export of agricultural products . The government is now directly responsible for several of the most important agricultural production and marketing activities . The newly created structures and institutions are still very young and subject to drastic changes, overhaul or simple abolishment. Several of the new offices were recently abolished or merged with other offices and the situation is not clear as to what offices are now effectively operating. Certainly, a bold step has been taken by the government in the field of agricultural production and marketing, but only time will tell if the newly created vertical structures under direct govern- ment control will be able to foster increased production, incores and employment for the millions of small traditional Zairean farmers. * . ”The National Ivory Office is not included in this list because it is part of the Customs Service. J . 35 Major Problems Facipg the Agricultural Sector The Zaire government now places priority on agricultural develop- ment, although this has not yet been clearly translated into, budget allocations. «Nbreover, large rural to urban migration seems to continue unabated, and most cities will double their population every fifteen years or less if present urbanization trends continue. Still, the demand for foodstuffs through comercial channels is strong, and this demand has been met by increasing imports. This leads to the major problem now facing the agricultural sector in Zaire: providing more food for urban areas. Achieving Self-Sufficiency in Basic Food Products Imports of maize, rice, sugar, meat and fish have been increasing rapidly, yet there is no technical reason why this derand cannot be met from dotestic sources. The country has been self-sufficient in these products in the past, and with a larger share of foreign exchange reserves now being spent on oil and energy-intensive imports, there is an urgent need to decrease food imports. Sugar is grown on large industrial plantations and new production units are now being planned although smallholder production of sugar cane is presently not considered as feasible. Still consumption of sugar is increasing rapidly, at a rate of about 10 percent per year. Cattle are kept on large ranches in the North and South of the country. There is also smallholder livestock production in the Ituri and in the Kivu region. Tsetse fly infestation limits cattle production in many forested areas . Several projects are underway to increase beef production with the aid of the World Bank, the Federal Republic of Germany, 36 USAID and Belgium. The major effort is focused on increasing the number of cattle on government ranches and on providing veterinary services to smallholchrs in the North-East. By 1980, Zaire could be self-sufficient in beef. Poultry accounts for 12 percent of total meat production and 25 percent of meat consumption. Indeed, rapidly increasing consumption has been met by increasing imports (in 1972 Zaire imported 2,746 tons of poultry meat). But poultry production in large production units could be developed rapidly as an important source of meat for the cities , the only major constraints being the availability of concentrated feedstuffs and disease control. For sugar, beef and poultry, then, prospects for meeting donestic demand by 1980 are relatively good because of the comercial, capital intensive nature of these products which call for direct government intervention in production and marketing. Such is not the case with other food products. Fish is the most important animal protein source, yet only about half of the fish catch passes through comercial channels. The major method of augmenting fish production is by increasing fishing on lakes . In fact, pisciculture was well developed before independence, but dis- appeared almost corpletely afterwards. A French technical assistance team is currently surveying fishing potentials in Lake Mai Ndorbe. But although such potentials for increasing fish production are good, little effort is being directed toward realizing them. Maize and rice offer the greatest potential for rural erployment 37 creation. Both cereals are grown solely on traditional farms,12 with maize being grown primarily in the South and rice in the North. Internal detand for both comodities is strong and is increasing. In fact, the Department of National Economy estimates that maize consumption will increase at an annual rate of 6 percent. Rice production, according to FAO estimates, reeds to double between 1970 and 1980 to cover dorestic consumption. Major constraints to tie developnent of maize and rice production needed to meet these demands are to be found in declines of yields, lack of production incentives and marketing problems . It is the lack of fertilizer use corbined with the planting of traditional varieties which explain, in part, the low yields obtained for both maize and rice. Sub- stantial progress could be made by the use of modern inputs. This, lnowever, will require a renovation of the agricultural extension service which has not been retrained since 1960 and which lacks a means of transportation. Expansion of Smallholder Export Crop Production Smallholder export crops with the greatest potential for expanding production are coffee, cotton and cocoa, with coffee ranking now as the first agricultural export crop in Zaire followed by palm oil. Coffee production and exports have increased substantially since 1968, mainly because of an increase in srallholder production. Grown on comerical plantations and on traditional farms, coffee is also the most important 12A notable exception is the large mechanized maize schere in the Kasese-Kaniama area of Eastern Kasai. 38 export crop of the traditional sector. In fact, coffee exports, pri- marily Robusta, surpass slightly the export quota assigned by the International Coffee Organization. Coffee diversification has been called for by the government although coffee is without a doubt the most profitable cash cm of traditional farmers, and although the technical conditions for growing coffee in Zaire are excellent. But with the present uncertainties about the continuation of the International Coffee Agreement, it is doubtful whether Zaire should undertake efforts to diversify away from coffee and prorote other, less profitable cash crops. 1 Cotton is exclusively produced by smallholder peasants. It is grown in rotation with food crops , and its production is widely dispersed in the North, South and East of the country. Zaire cotton fiber is of high quality because of hand picking, and demand prospects, for the dotestic market and for export, are briglnt. Cotton and coffee are the typical cash crops introduced during colonial times in paysarnnats which were created from 1945 onwards on a large scale in an effort to rationalize traditional agriculture. After 1960, however, paysannats collapsed as the discipline required for such a settlerent schere broke down with the upheavals of independence . The goverment now wants to revive some of these paysannats . Cocoa has never been grown by smallholders in Zaire in contrast with West-Africa. Still, the ecological conditions for cocoa production in certain areas of Zaire are favorable and export opportunities are good. However, this crophas never received the attention it merits. Cocoa does require relatively more farm processing than coffee, and this may 39 constrain the introduction of cocoa in traditional farming. But the West African experience teaches us that this problem can be overcote. The planting material presently used in Zaire produces yields inferior to those in West Africa, but high yielding strains are available from INERA. Thus, the introduction of cocoa in smallholder production could provide opportunities for crOp diversification and erployment creation. The Transportation System Rural areas in Zaire are sparsely settled with sore exceptions.l3 In fact, only about 1.2 percent of all land is occupied. Villages are widely dispersed over much of Zaire, and the dispersed pattern of village settletent makes it extrerely costly to connect villages with urban areas . Major attention is now being devoted by the "Office des Routes" to the major roads and to the bridges and ferries, though while the major road connections are being retabilitated, the feeder roads retain a major bottleneck. And it is the feeder roads which are important in providing an outlet for agricultural comodities and in bringing improved inputs and consumer goods into rural areas. Ferries and bridges on these roads are often deficient. Feeder roads deteriorate very rapidly due to erosion by rains. The continual poor condition of the roads results in high transportation costs due to breakdown of trucks , accidents, lack of spare parts, etc. In the northern cotton belt, for example, according to CNAFITEX management, it presently takes three times l3High population densities are found in parts of Lower Zaire, Kwilu, Eastern Kasai, Northern Kivu, Lower Uélé (Isiro) and Ubangi (Genera) . 40 the nurber of trucks than necessary with well maintained roads to haul seed cotton from farmers in a corparable time span. Thus, for some time, poor transportation is likely to remain one of the principal obstacles to agricultural develogrent . Improving the Perfonmance of the Marketing System The urban population in Zaire is increasing rapidly, straining the urban food marketing systems. Kinshasa, with a population of over 1.3 million, typifies the problems of the other cities. Marketing margins are high and an increasing reliance is put on food imports. A prolifer- ation of handlers and middlemen Operating at the wholesale and retail level and high transportation costs result in a wide spread between prices paid by consurers and prices received by producers. In addition, losses in handling and in storage are quite high due to spoilage, and insect and rodent damage. Price signals at the retail level are buffered by the large margins and, thus , inadequately transmitted to the producers . On the other hand, only a limited range of consuter goods is offered to the farmers, and because of high transportation costs and a small number of suppliers, these consumer goods are relatively expensive. Most plantations have their own retail stores for such goods . The government is aware of the inefficiencies in the urban food marketing systers and has intervened with the announcerent of maximum or fixed retail prices, the building of modern market facilities and the opening of government operated retail stores ("the Econcmat du Peuple") . The recent creation of government marketing offices for most of the agricultural products entering commercial channels is designed to coordinate marketing activities by centralization, to increase efficiency 41 of the marketing system and to put control of commercial agriculture in the hands of Zaireans. These offices offer unique opportunities for vertical coordination and economic integration of the country . For pro— ducts which are mainly consumed locally (rice, maize, sugar, meat) , the offices should be able to provide consumers with better quality produce at lower prices . A major constraint on improving agricultural marketing performance is the lack of adequate data and detailed studies on pricing, marketing margins arnd denand. This is related to the shortage of qualified staff at the Department of Agriculture and the general lack of concern with agricultural developrent. However , since agricultural developrent became a top priority, several actions have been taken to strengthen the insti- tutions serving agriculture and to generate data on the agricultural economy. inner The agricultural potential of Zaire appears to be enonmous. The natural endowment generally favors agricultural production , with soils of average fertility for tropical land, plentiful rainfall and extended dry seasons only in the North and South. Seasons alternate between the North and tie South as the country is astride the Equator, and forests cover 45 percent of tie country , containing a trerendous potential output of timber. However, only 1. 2 percent of the land is cultivated and agricul- ture contributes only 18 percent to GDP. Several major foodstuffs--rice and maize-which technically can be grown in the country are imported. 42 Seventy percent of the population are rural. Agriculture has a low value in Zairean society. The predominant share of field work is dore by woten. Also, emphasis on agricultural development has been weak as government revenues and foreign exchange are mainly derived from the export of minerals. Then, too, the Department of Agriculture has a shortage of qualified staff and is not able to utilize fully the budget earmarked for its operations. Moreover, yields for several crops have been decreasing because of degeneration of seeds, disease attacks and poor crop maintenance, and the agricultural extension service is largely ineffective because of lack of training and lack of means of transporta- tion. " Another constraint is that agricultural production is widely dispersed over the country. The transportation system is deficient and constitutes a major impediment to the expansion of agricultural production. The high marketing margins have an adverse effect on agricultural production. High taxes on exports of agricultural products and taxes on imports of improved inputs also have a negative effect on agricultural output, though recently , product marketing off ices have been created to solve this and other major problems in agricultural production and marketing . CHAPTER III CDTION PRODUCTION IN NORTHERN ZAIRE History of Cotton Production in Zaire Introduction of Cotton1 The first European explorers found cotton growing wild in Zaire . Undoubtedly, it was introduced by slave traders from Egypt and Sudan and by Portugese traders from Angola. From 1890 on, missionaries and settlers tried to introduce cotton without success . In 1908 experiments were undertaken in Bas-Zaire (Kitobola) by a government agronomist, J. Claessens who concluded that the area was not suitable for cotton production. In 1913 the Director General of Agriculture in the Ministry of Colonies, Edmond Leplae, charged Mr. Edward Fisher, an American cotton specialist, with the systeratic undertaking of cotton growing experi- ments in Zaire. From 1915 to 1920 he introduced the American variety, Triumph Big Boll, in the Sankuru-Kasai and in Maniéma (Kivu) and obtained satisfactory yields in regions with moderate altitude and a pronounced dry season.2 As a result, tie colonial government decided to extend the production of cotton to other regions. In 1918 cotton growing became compulsory in all provinces except Leopoldville under a 1917 decree 1This section is based on Banneux [l938]and Brixhe [1958]. 2Cotton cultivated in Zaire is of the species Gossypium Hirsutum. 43 44 on corpulsory crops (cultures obligatoires-cultures éducatives) . The long term aim of this law was to stabilize the rural population, improve their nutrition, educate them and increase their incones [Leplae, 19 33] . Cotton was considered a rich product; it was easily transported and stored and it would rapidly generate incone for farmers. It would also supply the well developed Belgian textile industry. 3 When cotton was first introduced, the government bought the crop . In 1918 the colonial government ordered two ginning mills from the USA. On February 10, 1920 COIOMEO, the first important private cotton company. was set up to develop the cotton industry in Zaire. For the East and South where the mining industry was developing rapidly, the qnvernment introduced limits to the expansion of cotton production in order to save labor for the mining activities and to avoid food shortages for mine workers. However, it was discovered that the cotton growing regions became the main food producing areas as cotton was well adapted to a food crop/cotton rotation, especially manioc, maize, groundnuts, rice, beans and bananas. As a result, cotton pro- duction was extended to the mining regions . Growth of Cotton Production The growth of cotton production is presented in Table 3. 1. Between 1915 and 1930 seed cotton production increased rapidly from one ton in 1915 to over 30,000 tons in 1930. Over the next decade production more 3Cotton was introduced in the Central African Republic in 1924 by Félix Eboué. This crop was also corpulsory until 1954. However, it has remained as an imposed crop in the eyes of farmers. [De Dampierre, 1960] . Comercial cotton production was introduced to Nigeria in 1902 by the British Cotton Growing Association [Kriese1, 1968] . 45 Table 3.1. Republic of Zaire: Seed.Cotton Production, 1915-1974 (In Tbns) Year Source Seed Cotton Year Source Seed.Cotton of Data Production of Data Production 1915 1 1 1946 2 117,852 1916 l 12 1947 2 116,353 1917 1 106 1948 2 123,904 1918 1 320 1949 2 143,081 1919 1 450 1950 2 138,389 1920 2 1,527 1951 2 133,402 1921 2 1,770 1952 2 158,347 1922 2 3,105 1953 2 136,411 1923 2 2,610 1954 3 143,348 1924 2 5,130 1955 3 145,726 1925 2 9,167 1956 5 154,481 1926 2 14,928 1957 5 129,829 1927 2 17,639 1958 4 142,507 1928 2 20,207 1959 6 179,660 1929 2 21,754 1960 6 140,077 1930 2 30,600 1961 6 64,630 1931 2 44,799 1962 6 38,873 1932 2 26,775 1963 6 42,820 1933 2 46,400 1964 6 40,210 1934 2 59,160 1965 6 18,700 1935 2 77,781 1966 6 19,960 1936 2 92,105 1967 6 24,910 1937 2 106,496 1968 6 45,210 1938 2 122,468 1969 7 47,400 1939 2 ”112,228 1970 7 53,800 1940 2 132,104 1971 7 59,500 1941 2 138,024 1972 7 52,980 1942 2 117,836 1973 7 '66,915 1943 2 128,796 1974 7 64,840* 1944 2 90,559 1945 2 111,253 Source of Data: [Ministere du.Congo Belge et du Ruandarurundi, 1960]. [Brixhe, 1958] and [INEAC, August 1954]. [van den Abeele and vandenput, 1956]. [IRES, 1968]. [Ministére des Colonies, 1957]. DMinistére de l'Agriculture et du Dévelcppement Rural, 1970]. [ONAFITEX, personal communication]. \lO‘U‘lubWNl-J *estimation 46 than quadrupled and reached 132,104 tons in 1940. However, production fell drastically in 1932 in response to falling world prices. Prices paid to tie farmers in 1932 (0.60 franc/kg) were only half of the 1931 level. The government abandoned cotton production in the northern areas because of the large distances separating these areas from the main transportation routes. Cotton farmers protested in these regions and again received cottonseed the next year. The development of an effec- tive agricultural extension service by the government and by the private cotton corpanies was largely responsible for the rapid increase in production. During World War II production dropped considerably as cotton farmers were directed to tap natural rubber from liana's in forests and to work in "crash" rubber plantations . Cotton production in the lake Mai Ndotbe area was abandoned and the cotton extension effort slowed down. After the war production recovered slowly at first and then increased steadily to‘a peak of 179,660 tons in 1959. During the first years of independence , production tumbled arnd reached a low point of 18,700 tons in 1965, about one-tenth of the 1959 level. With improved security urnder the Mobutu regime , production gradually recovered and has retained at 60,000-65,000 tons since 1971. Ecological Conditions Governing Cotton Production According to Sinclair [1968] , a minimum rainfall of 500 mm. per year is required to provide adequate soil moisture for cotton production. For optimum growth, an average growing season tetperature of just over 21° C. is also required as well as adequate sunshine. Brixhe [1958] 47 states that cotton is only sensitive to extrete temperatures below 5° C. or above 38-40° C. Thus, if we look at terperatures only, most of Zaire would be suitable for cotton production. However, the distribution of rainfall encourages production immediately to the North arnd to the South of the equatorial rain forest. Cotton requires a dry season about five months after planting for the maturing of cotton bolls. Hence, the central basin and the high plains in the East which have a regular rainfall are unsuitable for cotton, while in the northern or southern savannahs or bush savannahs, tlere is a marked dry season for several months which is ideal for the maturation of cotton. Between 400 and 800 mm. of rainfall is recorded in these areas in the five months preceding the dry season, an amount which is needed for the vegetative development of the cotton plants . The northern and southern borders of the equatorial rain forest have a dry season of one to three months which is sufficient altholgh not ideal for the cotton harvest. ‘ In the rnorthern zone cotton matures by the end of the year, while in the southern zone cotton is picked in June-August. Hence, there are two cotton harvests a year in Zaire and a fairly even flow of fiber to the textile mnills. Soils play a crucial role in determining yields. Generally, the soils in Zaire have a very low retention capacity. The clay fraction is predominantly kaolinite which is characterized by a low retention power. Thus, water and nutrient retention in the soils is poor. Mucln of the soil fertility depends on humus content, whicln is rapidly broken down because of the low retention power of the soil, the high terperatures and the 1200 to 1800 mm. of annual rainfall which leaches the soils thoroughly. 48 This explains why soils are very rapidly depleted, on the average after two to three years of cultivation, and need long fallow periods to regain their fertility, normally a period of ten to twenty years. Principal Cotton Production Regions There are two major cotton production regions: the northern and the southern belt. The production seasons alternate between these regions as they are located on each side of the Equator. Van de Walle [1960] distinguished three cotton regions in Zaire according to a productivity criterion and the possibilities for crop intensification: a high yielding region A, a marginal region B and an uneconomical region C. In region A farmers produce a minimum of 42,000 tons of seed cotton per zore with an average yield per hectare exceeding 375 kg.; in region B the minima are 1,000 tons and 300 kg-/ha. and in region C, less than 1,000 tons and less than 300 kg./ha. C(XSEKI) estimated in 1958-1959 that 874,698 families grew cotton: 561,943 in region A, 72,324 in region B and 240,421 in region C. After independence, the number of cotton farmers drOpped and is now roughly estimated by ONAFITEX at 500,000. If we accept that region A farmers continue producing cotton and that region B and C farmers dropped out , which corresponds roughly to the present situation, then we arrive at the figure advanced by ONAFITEX. In 1958-1959 the number of region A cotton farmers in Ubangi , Mongala , Bas—Uélé and Haut-Uélé was equal to 246,285. De Plaen [1957] defined five distinct cotton production regions in Haut-Zaire region based on average first year seed cotton yields. These regions have distinct soil characteristics in terms of soil color, 49 texture, acidity, exchangeable bases and organic matter. In Region I, average yields exceeded 800 kg. per ha. This is the central cotton producing region in Haut-Zaire, including the following urban centers: Buta, Titule, Bambesa, Dingila, Poko, Rungu, Isiro and Wamba. Soils in this region contain over 40 percent of clay, have a high percentage of exchangeable bases and a pH of 6 . 0 . Region II includes the northern savannah's with yields from 400 to 600 kg. per ha. Region III is to the North-West of Region I, mainly a forest area, with yields similar to Region II but with distinct soil characteristics. Region IV is to the South-West of Region I. Yields were quite variable but the number of observations was small. Region V is located East of Region IV and to the South of Region I. Yields exceeded 1000 kg. per ha. but again tie number of observations was limited. The regioral administrative organization of ONAFITEX is as follows: General Headquarters: Kinshasa 1. North-West Regional Branch for Equator region; sub-regions Ubangi and angala; seat: Gemena. 2. North-East Regional Branch for Haut—Zaire region; sub-regions Bas-Uélé, Haut-Uélé and the zone of Mahagi in Ituri. 3. Central Regional Branch; sub-regions Manie’ma and the zone of Shabunda in South-Kivu; seat: Kasongo. The sub-region of Sankuru and the zone of Uvira are also located in this area but are responsible to the general headquarters in Kinshasa. 4 . Southern Regional Branch; the regions of Shaba, Eastern and Western Kasai; seat: Gandajika. 5. Zone Uvira; zone of Uvira, Fizi and Walungu; seat: Uvira. 50 6. Zone Sankuru; sub—region of Sankuru; seat: Kutusongo. Each regional branch reports directly to the general headquarters in Kinshasa. However, the regional branches have substantial latitude in the managetent of their operations . Each regional branch corresponds more or less to a homogeneous cotton growing area. It should also be noted that there are two special areas pro- ducing long staple cotton in the eastern part of Zaire: the Mahagi area and the Ruzizi Valley. Their cotton crop is exported Via Dar es Salaam. In the Mahagi area near lake Nbbutu Sese Seko (formerly lake Albert) , cotton was introduced in 1938 from Uganda. The first variety was S47 characterized by long staples (1 1/8 inch). Present varieties are of the Satu type, also from Uganda. This area produces about 1,000 tons of seed cotton which is ginned at the mill in Mahagi, the only ginnery in Zaire equipped with roll gins instead of saw gins. The Ruzizi valley is located along the Ruzizi River on the Burundi frontier, connecting lake Kivu with lake Tanganyika. The soils in this valley are very fertile, probably the richest soils in Zaire, but the area is insalu- brious. Cotton yields are the highest in the country and the fiber is of very high quality; the staple length is 1 3/32 inches. Human Factors in Cotton Production Cotton, in many ways, is a unique crop. Unlike most of the other export crops, it is an annual crop grown exclusively by smallholder Zairean farmers. Cotton also is one of the most regulated crops, with cotton legislation dating back to 1917. Cotton was the main crop grown in the paysannats. The farmer gets free seed every year from 51 the ginneries. Thus, it is easy to introduce improved seeds. The most unique feature, however, is that cotton production was made com- pulsory by the government in 1918 because it was assured that farmers would find cotton to be profitable and, as a result, they would expand production on their own initiative and the government could then drop the cotpulsory requiretent. However, the colonial government retained the requirerent until it was abolished in 1958. As shall be pointed out later, this research has shown that regional governments still utilize the imposition teclnrnique today . Historically, before tle advent of the colonizers, Zaireans were the victims of inter-tribal wars, slave razzias and frequent migration. Their settletents were always tetporary. These nomadic features are largely responsible for the poor knowledge of agriculture. Their V warrior- 1ike behavior made them predominantly hunters and fishermen, leaving the cultivation of the land to their wives. Today the bulk of the labor for agricultural production, processing and marketing of agricultural products is done by women, while the men clear land, build houses and hunt and fish. Moreover, rural people have marked urban aspirations. These aspirations, both existential and material, are unconditional and urgent [Fox, 1970] . Agriculture, in the sense of "working the land," occupies a very low status. It was never a positive value. 4 According to Fox, kinsmen and tribesmen are the fundamental sources of security, power, prestige, prosperity, fecundity and physical 4The Faculty of Agronomy in the National University of Zaire has great difficulty in recruiting students. Also, the Department of Agriculture has a very low budget and has great difficulty in recruiting and retaining staff. 52 and psychic well-being. High levels of distrust, suspicion and anxiety characterize personal as well as impersonal relations. Beliefs and customs require the individual who does well financially to support relatives who are less fortunate. Fox hypothesizes that the tribal structure, kinship and tie extended family system act as a severe con- straint on producing much above subsistence. The colonial approach attempted to overcome these cultural factors by imposing corpulsory production requirements . The colonizers believed that cbligatory cultivation of certain crops was a necessary part of the "educational" process. In Zaire it was introduced by the decree of December 5, 1933; it required 60 days per year (45 after 1955) of corpulsory cultivation or other public works for each able-bodied adult male (H.A.V.)5 This law was generally applied until 1957 and was still legally authorized although largely abandoned in 1960 [Young, 1965].6 A number of stern measures were used to maintain the obligatory system. For example, the 1947 Senate commission reported that about 10 percent of the adult male population had Spent sore time in prison, in good part because of violation of the agricultural requiretents. Thus, to the rural population, cotton was an extrerely unpopular crop; many 5Each able-bodied adult male was also subject to the personal mini- mum tax (C.P.M.) which tied tax collection to cotton growing. Usually, the C.P.M. tax is collected on the same day and at the same location as the purchase of seed cotton from farmers. 6The imposed surface per H.A.V. varied between 15 and 50 ares with a mean of 30 ares. ' 53 peasants considered it simply a tribute levied by tine European ruler7 [Young, 1965]. Imposed cultivation on food crops started during World War I when troops were sent to eastern Zaire to guard against a possible attack from the German Colonies. Since more than 260,000 soldiers had to be fed the government required each farmer to grow certain food crops. This operation was quite successful and, encouraged by this experience, the government extended required cultivation when the war was over to the rural hinterland adjacent to urban centers and mining camps in order to supply the African pOpulation with staples. When cotton began to be emphasized, it became a required crop in many areas of Zaire [Ieplae, 1933] . This system required a vast administrative organization of the countryside; there was at least one agricultural officer per zone, supported by several African agricultural assistants and monitors.8 Also there were the cotton extension agents supported by the private cotton companies. Thus, in the perception of Zaireans, it was the "government" which made them grow cotton for the benefit of the cotton companies. The sharp fall in cotton production after 1960 can be partially attributed to the unpOpularity of the cr0p and tlne constraint 7Yomng [1965] noted that many of the low cotton yield areas, such as North-Shaba, Maniéta. much of Sankuru and parts of Haut-Zaire coin- cided with the zones of radical nationalist response during the rebel- lion years. 8an the average , there was one European extension agent for every 2,000 to 3,000 cotton farmers [Brixhe, 1958]. In 1959, there were more than 6,000 agricultural extension workers in Zaire, of whom over 1,200 held diploras; they were supervised by 200 Congolese agricultural assistants and 418 Belgian specialists [Jurion and Henry, 1969] . 54 used to enforce its cultivation. The question now is whether fifteen years after independence, obligatory cultivation is still in effect and enforced. The answer is yes, as will be documented in Chapter VI. Cotton Research Cotton farmers in Zaire do not utilize any cotton for domestic purposes. Farmers sell nearly 100 percent Of their harvest. Fraud is nearly impossible as ONAFITEX has a monopoly on the purchase Of seed cotton. Each year, the farmer sells his cotton to the ginnery which then issues the next season's seeds. Thus, new seeds can be introduced easily via the ONAFITEX ginneries in one Or two years. In 1933 the National Institute for Agricultural Studies of the Congo (INEAC) was created as a quasi-governmental independent organiza- tion,9 replacing the former Department Of Plantations.lo In 1959 INEAC Operated seventeen research stations, fourteen experimental centers, five experimental plantations and a veterinary laboratory. INEAC has made a number Of major contributions to Zairean agriculture, especially in increasing productivity on perennial plantations , in launching paysannats and in the diffusion of selected planting material [Drachoussoff, 1965] . The institute's fundamental research on the ecological enviroment and factors influencing plant growth and 9In Belgium a good part of policy is implenented by "para-statal" organizations, a sort of national public service institution, legally granted a monopoly status, distinct from the state itself and with vary- ing degrees of autonomy with relation to the central government. INEAC was such a type of institution. It depended directly from the Ministry of Colonies in Brussels. 10In 1970 INEAC changed its name to INERA, the National Institute for Agricultural Studies and Research, with headquarters at Yangambi , 60 miles west of Kisangani. 55 production is well known.11 Table 3.2 illustrates the progress made by INEAC in the selection and breeding of cotton from its introduction in 1915 up to 1960. INEAC research on cotton was concentrated in the following stations: 1. Northern belt: Bambesa as the principal station for the Equator and Haut—Zaire region with the Boketa station in Ubangi assisting Bambesa. 2. Southern belt: Gandajika in Easternn Kasai for the entire southern zone. 3. Eastern zone: Lubarika, serving the Kivu and Ruzizi valley. Edward Fisher introduced the American variety, Triumph Big Boll from Texas, in Zaire in 1915. This variety was diffused throughout Zaire, except for the Ruzizi valley where Allen long staple was intro- duced. Triumph, the principal cotton variety for over twenty years , was purified by the Department Of Plantations and later by INEAC through mass selection, and only plants well adapted to soil and climate were selected. No other variety proved superior or better suited for Zairean conditions for over twenty years. In 1933 fusarium disease was accidentally introduced in Zaire with American seeds. Triumph was very sensitive to this disease. Its fiber length was also inferior to that in many other varieties and its ginning percentage was less than 33 percent, compared to 36 percent in the USA. (1710100, the largest private cotton cotpany, provided INEAC with several researchers to strengthen cotton research through which 11In June 1960, INEAC employed 420 Europeans of which more than l'alf were university graduates, and 12,000 Zaireans [Drachoussoff, 1965]. 565 .xx manna 5 Endowed Oman mum Smog wo ouofioe ocmemun ocm mow: .xwofi Museum: .93on one coins Seascapes"; moflmeuouomumoo no .595: £85 xoammmuo .muooam moan uo moflcmoa snowman monogamous 303302203 “ .mp3 Show: a :35: media: .o .xx Some. can owe/.mp6}. as a NQH a on) waxes on do H on on a do on oH\mH mem a 3) a on) H mm) H mm) H NQHM 3me mm} H mmhm time 8 saw” 188:: 59.3 node was TR com o.~m osm mam 9mm Ton mam fem o6... oofifiouoo 825d o3 m2 m2 2: NE a: SH o2 «3 m: 2: nounfiwow 5933 pounds some» node 838 83 I- 22 22 $3 $2 32 $3 $2 22 22 88835 do» madman confide dependence: 88 canine 83 m2 3 one was 3 :32 82 8 woo a. command m 235on Soc: stone: on: 888 5 nodes. manor mode 838 55:8 oc8 coupoo Eofiwoz 985m to 88288390 a whims Ce .5300me oregano ocm c9500 smog mom: mmmuoonm .N .m wanna. 57 INEAC bred and selected several important cotton varieties to carbine resistance to fusarium disease, longer staples and higher ginning returns. The BaIrbesa station introduced the Stoneville variety from Mississippi after local adaptation and purification in the northern cotton belt. In the South the Gandajika station introduced GAR hybrids, fran breeding Triumph with U4, imported from Transvaal, South Africa. Other varieties were diffused later, in the North, B49 and Reba BSD and in the South, C2 and NC8. At the Lubarika station, varieties adapted to the Ruzizi valley were developed and~ diffused: 14,125 and 14,125 purified and 1,021. All these varieties were resistant to wilt. --In the North: Reba B50 replaces B49 and Stoneville. It was developed in 1960 by the French "Institut de Recherches du Coton et des Textiles Exotiques (I.R.C.T.)" on the Banbari station in the Central African Republic.12 Presently, Reba B50 is grown on a large scale in northern Zaire, Central African Republic and Paraguay [Centurion et a1. , 1972 and Roux, l974] . This variety was introduced in the Uélés in 1967 by Mr. Canivet, a COIONCO agent. Reba B50 is genetically resistant to Bacteriosis (bacterial blight) and tolerates Fusaricsis, diseases typically prevailing in hunid areas. The ginning yield is also excellent; 37 percent in Central Africa, and staple length is 1 1/16 inches; other fiber characteristics are good. However, the main challenge facing INERA is to keep Reba BSO frcm degeneration; purebred seeds are not available in Zaire nor in the Central African Republic. A reduction in staple length from 1 1/16 inches 121?an a cross between Stoneville B 1439 and Allen 50 T. 58 to 1 1/32 inches was reported in 1974 indicating a loss of varietal purity which can be preserved by careful mass selection in isolated fields. —In the South: DEB, replaced C2 in 1959. NC8 was develOped at the Gandajika station and is characterized by longer but somewhat less resistant fibers. --In the Ruzizi valley of the East: 1,021 which replaced 14,125 in 1960. This variety was developed at the Lubarika station. Juricn and Henry [1969, p. 183] indicated the following maximum yields of seed cotton observed in intensive cultivation:13 Yield in Kg. per Ha. Forest central basin region 1, 200 Derived savannah without cattle 1,200 Eastern mountain region 1,800 Eastern mountain region with total plant protection 2 , 700 In 1973 the major responsibility for cotton breeding and selection was cmcentrated in Gandajika. 'Ihe Ban'besa station limited itself to nultiplication of Reba B50 seed, auto-fecundation of Reba B50 and BJASl9 and determination of crop and ginning yields. In Boketa the influence of climate on selected cotton characteristics was studied and Reba 350 was multiplied on a large scale (5 hectares). Six other purebred varieties were also multiplied on a small scale (0.5 hectares). INERA has cane under the direct auspices of the Presidency since 1970. Previously, it had been controlled by the "Office National de la Recherche et du Dévelcppement" (ONRD) and, before that, by the Ministry 13'Ihe yields given are the highest obtained over several seasons on a fair-sized plot cultivated by advanced methods. 59 of Agriculture. Presently , the Department of Agriculture is linked with INERA in the direction committee (Comité Directeur) in which the state carmissimer for agriculture and the director general of the Department of Agriculture have a major voice. However, it is clear that INERA has weak links with the agricultural extension service, a weakness which hampers the effectiveness of the research activities of INERA. Paysannats and Cotton Production After World War II a serious effort was undertaken by the colonial administration to rationalize and intensify agriculture. The dispersion and semi-nanadism of the pOpulation constituted a fundamental obstacle to agricultural developrent. 14 The beginning of a rural to urban exodus was a major reason for the establishment of paysannats in an effort to settle the farmer on the soil [Young, 1965]. Paysannats are farm settlanent schemes which are organized around a new or regrouped village. An agricultural plan is prepared and land use , rotation and cropping patterns are specified and supervised by the agricultural service. 15 Fields are allocated and rotated in a systematic manner. Mechanization was feasible because all land plots were grouped in one block. 14'lhis semi-nanadism is typical of a system of shifting cultiva- tion. There are two kinds of shifting cultivation: those who rotate land while staying in permanent villages and those who move and establish new villages after a period of years. The first kind is most canton in Zaire. The Azande tribe in the Uélés of northern Zaire is known for mving from village to village [de Schlippe, 1957] . 15On the average, there was one agent per 320 farmers in a paysannat [Young, 1965] . 6O Paysannats are located along roads for easy transport of produce and easy access by the agricultural service. 'Ihe discipline required in paysannats included observation of rotations and fallow periods, use of better planting material, correct spacing of crops and timely planting. Often mechanization was tried and measures were taken to prevent soil erosion. Cotton was the principal cash crop introduced in paysannats on an obligatory basis, and it became part of the food crop rotation. Table 3. 3 illustrates the location and characteristics of paysannats with cotton as the main cash crop in 1957; of the 200 paysannats, over (me-third were in Haut-Zaire region. Paysannats were the cornerstone of the 1945-1960 agricultural development efforts of the colonial administration. 'Ihey succeeded in many ways: yields in paysannats were well above yields outside paysannats , mechanization and new crops were introduced, erosion and soil fertility were controlled and generally, levels of living were improved. However, the ten year plan goal of 385,000 farmers by 1960 was not met as only 200,000 had been established by the beginning of 1959 [Young, 1965]. A fundamental problem of paysannats was that the agricultural plan was not organized around the economic and social objectives of farmers. The rigidities of the system had always been unattractive to the farmers. In fact, most paysannats disintegrated after indepen- (18106 when the European supervisory agricultural personnel left and when the government extension services deteriorated. Several government efforts are now Lmderway to revive paysannats, the best known and most 61 Table 3. 3. Location and Characteristics of Paysannats with Cotton as the Main Cash Crop, 1957l Region Number of Farmers Total Cotton Acreage in Hectares Haut-Zaire 65,000 482,000 Equateur 24,000 166,000 KivurManiéma 28,000 258,000 Kasai 54,000 486,000 Shaba 29,000 261,000 Tbtal 200,000 1,653,000 J‘Brixhe [1958] . successful of which is a project at Gandajika in Eastern Kasai. This project is receiving substantial aid from the FED. The Gandajika paysannat was one of the showpiece settlements of the colonial administration [Young, 1965] . Agricultural Extension In the colonial period a large agricultural extension network prmbted cotton production. 'Ihe extension service consisted of govern- ment agents from the Ministry of Agriculture and the territorial service and agents from the private cotton companies. 16 There was an average of one Belgian agricultural officer for every 2,000 to 3,000 cotton farmers and one agricultural monitor (extension agent) for every 250 farmers. 16After independence the private cotton companies continued at a reduced scale to provide technical assistance to the producers until W in 1970 ceased to refund expenses for this purpose. 62 In 1959 the nutter of cotton farmers was estimated at 800,000 and about 300 agricultural officers concentrated on cotton production. They were aided by an estimated 3,200 Zairean extension agents. Much of the extension work was concentrated in paysannats . After independence the extension service , stripped of its European supervisors , rapidly disintegrated. Since independence extension agents have not been retrained. And although most do possess a bicycle, the supervisory staff lacks vehicles and travel allowances making it difficult to visit agents in the field. Many agents were not paid for months in the early 19605, though this problem has now been solved with sore exceptiors . 17 An institutional constraint compounds these problems. Cotton farmers are visited by three different types of extension agents, (1) the Department of Agriculture (moniteurs agricoles) , (2) the "Collectivités Locales" (moniteurs agricoles adjoints)18 and (3) ONAFITm (moneteurs coton) . O These three typos of extension agents are poorly coordinated. The Agricultural Service of ONAFITEX (he of the objectives specified in the decree establishing ONAFITEX is to provide technical assistance to cotton farmers and to stimulate the formation of cooperative societies, capable of ginning and marketing their products (Article 2, Objective 1 and 4, Decree 17Sa1aries of all government employees are now handled by the data processing center at the Department of Finance, using an IBM 370 computer. 18"Collectivités locales" were formerly named "Secteurs" and are the smallest administrative units in the comtry. Usually they include several villages . 63 No. 71/077 of August 12, 1971) . To this effect, the state commissioner of agriculture created an agricultural service in OWITEX on September 13, 1971 and a Zairois miversity trained agronomist was named director of this service. After an initial assessment of the agricultural pro- blems confronting the cotton subsector, the service started a large scale pesticide treatment project to combat cotton diseases and insect attacks. As Brixhe [1958] stated, no other important crop is so much attacked by insects and diseases as cotton. Table 3. 4 shows the efforts deployed by the agricultural service to fight cotton diseases. Yet the reeds largely surpass the present capacity for pesticide treatments. For Ubangi only (north-west region), for example, it is estimated that 2,000 tons of pesticides and 1,000 sprayers are necessary for three treatments per crop and an average dose of 20 kg. per treatment. The agricultural service of ONAFITEX also works closely with the agricultural extension service. However , this service depends largely from the Department of Agriculture and the local collectivities. Thus, ONAFITEX has no direct control over these agents, a fact which seriously handicaps the effectiveness of extension work by ONAFITEX. Recently, (NAFIM hired additional extension personnel, especially supervisors , to improve its extension program. It is obvious that the agricultural service of ONAFITEX has an important role to play in the development of cotton production in Zaire . Goverment Policies Affecting Cotton Production One Channel Marketing Office Before the creation of ONAFITEX on August 12, 1971, several private cottm marketing companies were buying and ginning cotton for 64 Table 3. 4. Pesticide Treatments by the Agricultural Service of ONAFITEX, 1971-1972 (NAFIT'EX Regions Shipment of D.D.T. Shipment of Pesticide Cotton Dust (In Tons) Sprayers (Nurber) North-West 700 520 North-East 700 480 South 605 60 Central 1,018 60 Uvira 100 62 Total 3,123* 1,182 *mly 1,500 tcns were actually shipped to the different regions at the end of 1972. lAgricultural Service of ONAFITEX. C(IERCO m a custom work basis (see Chapter IV). COTONCO was by far the largest company, covering about 90 percent of cotton production. Thae companies , under the leadership of COIONCO, did not cotpete with each other as each controlled a well delineated cotton growing area. They paid the farmers the official minimum cotton advance price. With the creation of ONAFITH, a government controlled institution was set up with sole responsibility for buying, ginning and sale of cotton. ONAFIM is a para-statal institution, a one—channel govern- ment marketing office with monopsony and monopoly powers . ONAFITEX was able to substantially raise seed cotton prices paid to the farmers. Uniform Fixed Producer Prices The decree of 1921 specified that minimum prices had to be paid to 19 farmers. The price was set annually on the basis of the prices obtained ‘— 19Artic1e 10 of the Cotton Decree of August 1, 1921. 65 for cotton fiber in the previous year. The farmers also received salt in proportion to their production, and cash gifts were given to village chiefs who had actively participated in the cotton extension campaign. Cotton prices fluctuated widely in the 19308 and in 1938 a price stabilization fund (Fonds de Réserve Cotonniére) was established. The seed cotton price paid to the farmer was derived from a potential cotton price Whidl was determined automatically from a table with two entries: the average cotton fiber price realized in the preceding year and the total cotton fiber production of that year. The difference between the "potential" price and the price received by farmers was in large part deposited in the price stabilization fund (Fonds de Reserve) . The cotton decree of June 18, 1947 stipulated that cotton farmers retained ownership of their cotton until the final sale of cotton fiber. (IDGERCO (a Cotton Management Cormittee) acted on behalf of the cotton farmers until farmer cooperatives were formed to take over its functions. On delivery of his seed cotton, the farmer received an advance or down- payment which was eventually supplemented after the global results of fiber sales for the year were known, several months later. The advance payment was based on the expected sales price of cotton fiber and marketing costs. Only in 1950, 1951 and 1952 did farmers receive a supplement. From 1953 on supplements were discontinued. Nevertheless, the stabilization fund increased its reserves considerably (reserves were 1,308,000,000 francs or 26,160,000 dollars on June 30, 1955) and spmt large amounts of "cotton money" for road construction and maintenance, mechanization, creation of paysannats, cooperatives, purchase of small hand tools, etc. 66 G. Peeters [1958] criticized C(ISERCO for the large disparity between prices paid to farmers and prices received for cotton fiber in the world market and the utilization of the stabilization fund reserves for road construction and public works . Although OOGERCO stabilized producer prices, the prices paid were extremely low. For example, the Zairean price was the lowest of all the countries for which price data could be obtained in 1970 and 1971 (Table 3.5) . The 1947 regulations were effective until the creation of ONAFITEX on August 12, 1971 with slight modifications and additions in 1951, 1955, 1962 and 1965. Presently the management committee of ONAFITEX proposes the advance payment and the supplement to the state commissioner for agri— oilture (Minister of Agriculture). The evolution of the advance payment since 1950 is given in Table 3.6. Seed cotton is paid according to quality. There are two grades-- first and second-with about 90 percent of seed cotton classified as first quality. There has always been a uniform seed cotton price in Zaire, regardless of transport cost differentials between regions.20 The principle of a flat seed cotton price favors cotton production in even the retotest or least accessible regions. Thus, cotton farmers in areas of low cost of production are actually subsidizing those growling cotton in high cost areas. This principle, "equal pay for equal work," is not followed, however, in the fixing of regional minimum wagss.21 20Only the seed cotton prices in the Kivu (Ruzizi Valley) and in Mahagi were always above the nationally set prices because of the superior qualities of cotton fiber (lcmg staple) produced in these areas. 21Fuel prices are uniform throughout Zaire. 67 Table 3. 5. Average Prices Paid for Seed Cotton in African Countries, 1970-19711 Comtry 1970 1971 Makuta/kg. U.S. Cents/kg. Makuta/kg. U.S. Cents/kg. Zaire2 4.31 8.62 4.31 8.62 Central African Republic 5. 40 10 . 80 6 . 00 12 . 00 Cameroun 5.76 11.52 6.40 12.80 Dahomey 5.04 10.08 5.60 11.20 Tr‘had 4.68 9.36 5.20 10.40 Ivory Coast 6.30 12.60 7.00 14.00 Uganda 7.72-8.40 15.44-16.80 . . . . . . Ethiopia 12.60 25.20 . . . . Nigeria 8.75 17.50 . . . . . . lONAFITEX, Kinshasa and personal communications with students from African countries. 2Assuming 90 percent is first grade cotton and 10 percent is second grade. Cotton Production Development Projects In this section current developtent projects directly affecting cotton production in Zaire will'be reviewed briefly. Presently, three cotton projects are underway: l. The FED-CAK project in eastern Kasai 2. The CFDT‘ mission to relaunch cotton production in the Equator region 3. The FIWA—BILI project of the Belgian Aid Mission in northern tbangi. 68 Table 3. 6. Republic of Zaire: Seed Cotton Price (Advance Payment) Fixed.by the Government,1 1950-1975 Years First Quality Cotton Second Quality Cotton Francs/kg. or U.S. $/kg. Francs/kg. or U.S. $/k9. Makuta/kg . 2 Makuta/kg . 2 1950-1951 5.50 .110 4.00 .080 1952 6.00 .120 (4.00 .080 1953-1958 5.50 .110 4.00 .080 1959-1962 6.0 .120 4.00 .080 1963 8.00 .123 6.00 .092 1964-1966 18.00 .120 12.00 .080 1967 24.00 .151 16.00 .106 1968-19693 3.60 .072 2.40 .048 1970—19714 4.50 .090 3.00 .060 1972-19745 6.50 .130 5.00 .100 1975-...6 7.00 .140 5.00 .100 1 ONAFITEX, Kinshasa. ZFrancs per kg. till 1967, Makuta per kg. from 1968 on. 3 4From January 1, 1970 on. Ar'rété 422/AE/0024 of October 7, 1967. 5When ONAFITEX was created on September 11, 1971. 6 Oommmication with Mr. Vandewalle , ONAFITEX in July 1974. 69 The FEDbCAK Project in Eastern Kasai This project started in 1968 in Gandajika and Kalenda-Kashile in eastern Kasai and is funded in a major part by the European Development Fund (FED: Fonds European de Développement) . It is an integrated agricultural developrent scheme on a former paysannat , which is con- centrating on cotton and food crops . The paysannats of Ganda j ika were created in 1954 and were among the most successful. The Gandajika population is dynamic and receptive to innovations . Before 19 60 this area was known as the breadbasket (maize production) for the mining regiore of Kasai and Shaba. The project is jointly managed by C.F.D.T. (Conpagnie Francaise pour le DévelOppement des Fibres Textiles), a French agency for technical cooperation specializing in the development of cotton production, and the Agricultural Commission of the Kasai (CAK: Commission Agricole du Kasai). The budget for the period 1968-73 was 1,875,000 8 and the five year extension.will cost 3,179,000 8, of which, 1,580,000 S is from FED, 466,200 3 from the government, 284,000 8 from OWIT‘EX and 848,800 8 from CAK. The objectives of the project are the introduction of drying and insect coitrol equipment for maize seed, the continuation of the agri- cultural program which was started in 1968 and the training of Zairois project management capable of taking over the project in 1978. FED also proposes the creation of a Regional Agricultural Development Office in Kasai. As demonstrated in Table 3.7, the 1969-1973 project objectives have been largely surpassed, except for groundnuts. The gross value of cotton and maize production increased from 255,600 8 in 1968 to 1,472,500 8 in 1972. Table 3.7. Republic of Zaire: 70 Objectives and Realizations of the FED-CAK Project, 1969-19731, Objectives and Realizations 1969 1970 1971 1972 1973 No. Farmers in Project Objective 4,000 8,000 12,000 14,000 14,000 Realization 3,700 12,036 17,458 18,411 18,495 Acreage.Maize (Ha.) Objective 4,000 8,000 12,000 14,000 14,000 Realization --- 15,000 19,550 22,270 --- Maize Production (Tons) Objective 2,600 6,000 10,000 12,700 13,700 Realization -- 12,000 ’16,300 20,000 19,519 Acreage Cotton (Ha.) Objective 1,000 2,000 3,000 4,500 5,500 Realization 2,120 6,100 10,241 10,135 -- Cotton Production (Tons) Objective 700 1,400 2,300 3,950 5,250 Realization 1,098 2,901 7,117 7,500 9,000 Acreage Groundnuts (Ha.) Qajective 1 , 000 3 , 000 5 , 000 S , 500 5, 000 Realization -- 1,400 3,400 2,800 -- Groundnut Production (Tons) Objective 6,000 2,200 4,100 5,000 5,100 Realization -- 900 2,000 1,900 -- 1 C.F.D.T. [1972] and CAR-Commission Agricole du Kasai Oriental, "Relance Agricole des Paysannats de Gandajika et Kalenda-Kashile." Rapports Trimestriels, unpublished documents. Gandajika: 71 The FED-CAK project is characterized by a balanced, integrated "paCkage" of technical innovations: improved seeds, fertilizers, pesti- cides, agricultural credit, tractor mechanization and improved small hand tools.‘ These innovations are communicated to the farmers by trained agricultural extension workers. A.type of farmer cooperative, called "groupement de producteurs" was created for the sale of agricultural products, particularly manioc and maize, on the markets of Gandajika and Mbuji-Mayi. In these markets, manioc prices are 4 to 6 times the prices in local markets in the producing areas [vasos, 1973]. TherAgricultural Commission of Kasai provides transport services for the "groupement" at a charge of 8K. pertonandperkm. foraminimumof 6 tons. The 1973 gross incore per farm in the project was 154 B, and the average net famuly income was 148 B. On the basis of 300 working days per year, this amounts to 49.3 K. per day, which is almost double the minimum wage rate of 26 K. per day for unskilled labor in the Gandajika area. The C.F.D.T. Mission to Relaunch Cotton Production.in.the Equator Region In 1972 a technical assistance agreement was signed between the RepUblic of Zaire and the French government that called fbr a C.F.D.T. mission.t0‘work with.ONAFITEX in order to relaunch cotton production in the Equator region. In 1972-1973 the C.F.D.T. teamlincluded a teamtleader, university trained agronomist and three agricultural technicians. In 1974 two more technicians joined the project. The project objectives , broadly 72 stated, were to develOp cotton production in Ubangi and Mongala, mainly through improved agricultural extension work. This project is a feasibility experiment for a much larger project covering most of the northern cotton belt in which the World Bank would take a major responsibility. The World Bank has already corpleted several pre-investment studies in the Ubangi and the Uélés and has held back action pending a reform of the cotton marketing institutional set-up. A 19 70 World Bank pre-investment study proposed to develop cotton production over 38,000 hectares in Ubangi and the Uélés by improving cropping practices, using fertilizers and pesticides, undertaking applied cotton research and good maintenance of feeder roads . The project cost was estimated at 7.2 million dollars over six years, half of the cost being in foreign exchange. The project returns were esti- mated-at 27,170 tons of seed cotton per year, corresponding to a net increase in value of agricultural output at farm gate prices of 843,000 8 and foreign exchange earnings at full development of 4. 5 million dollars per year. The rate of return to the economy was estimated at 25 percent. Undoubtedly, the C.F.D.T. mission is pointing out the essential elements for increasing cotton production, the main obstacles being institutional. This project merits more attention on the part of ONAFITEX, the Department of Agriculture, INERA and the National University of Zaire. If the C.F.D.T. experiment proves conclusive, World Bank financing will probably be forthcoming. 73 The FIWA-BILI Project of the Belgian Aid Mission in Northern Ubangi Since June 1965, just after the rebellion, a team of eight to ten persons was given broad responsibilities for rebuilding the Ubangi and the Moigala SLb-regions. The team started with purely administrative tasks such as overseeing salary payments of all government employees- including back pay, payment of pensions, reorganization of the terri- torial police force, Opening of schools , etc. Other important achieve- ments were repair and maintenance of roads and ferries, construction of houses, rehabilitation of a coffee paysannat in the Kungu territory. Nbreover, over 100,000 coffee plants were distributed. Beginning in November 1967 the team, responding to a demand from the government , abandoned its administrative functions and began engaging in rural develogrent activities. They grouped the rural population in larger entities and proroted food crop and small animal production. Mechanization trials were set up first in FIWA and later in BILI. At the same time, a ranch was created in FIWA. Further, the agricultural statims of B010 and Likimi were reactivated, and improved seeds of groundnuts, maize, rice, sesame and sorghum were multiplied and dis- tributed to the farmers. Small farm tools were sold at cost to farmers. Farmers pay 8 3 per hectare for tractor plowing, half at plowing time and half at harvest. The operating expenses have been estimated at 6 5 per hectare in 1972, excluding parts and repairs, and the total costs have been estimated between 15 and 20 5 per hectare. With tractor plowing, average cotton field sizes increased from about 0. 5 hectares to 1.0 hectares. The team also treated fields with pesticides (DDT cotton dust) with the spraying or dusting implements and the pesticides being 74 darated by ONAFITEX. Cotton yields in Fiwa with tractor plowing and pesticides treatment are between 800 and 1,000 kg per hectare. The main problem with this project is its lack of focus and orientation. The objectives of the project are broad and vague. Systetatic agricultural extension is not an integral part of the project. No data have been collected on the economics of tractor mechanization or pesticide treatments . Counterpart training is neglected; only one university trained agronomist is working with the project but he is mainly concerned with the building of a cattle ranch. Thus, it is hard to evaluate the impact of this project. Cotton Production in Northern Zaire The area studied includes the following sub—regions of the Equator and Haut-Zaire region: 1. Equator region a. sub-region Ubangi; administrative center: Gemena b. sub-region Mongala; administrative center: Lisala 2. Haut-Zaire region a. sub-region Bas-Uélé; administrative center: Buta b. sub-region Haut-Uélé; administrative center: Tsiro. Climatic Conditions The climate in the northern cotton belt is tropical and humid with the main rainy season occurring from mid-July to mid-November and a short rainy season between March and half of June. Average annual rainfall is between 1,700 and 1,900 mm. with an average minimum tempera- ture of 18° C. and an average maximum terperature of 30° C. During 75 the short rainy season food crops are cultivated. Cotton is then planted after the harvest of food crops and is harvested during the main dry season from December to February. In the northern savannah the short dry season is less pronomced and cotton is grown as the first crop in the rotation. During the rainy season temperatures are relatively low and variations between day and night are about 10° C. In the dry season, however, temperatures are higher and fluctuate more, 14° to 16° C., which favors the maturing of cotton bolls. Thus, the climate is generally favorable for cotton production. Soils The northern cotton belt is to the North of the large equatorial rain forest and includes rain forest areas, savannahs and mixtures of both. INEAC has publist detailed soil, soil utilization and vegeta- tion maps for the Ubangi and Uélés [Jongen, et a1. , 1960; Jongen, 1968; and Frankart, 1960] . Generally, soils under forest coverage are more fertile than savannah soils since forest soils contain more organic material, have more horns, a better structure and better moisture retention. Nevertheless , depending on the location, cotton yields in some savannah soils may equal or surpass yields in forest soils. The soils of northern Zaire are classified as ferralsols on not-- differentiated rock. They are well suited for cotton production, but are fragile and rapidly lose their fertility if exposed to the eletents. Some of the best soils of Zaire are fond in this region. Certain parts of the Uélés (for instance, the Bambesa zone) are renowned for their good soils. However, poor soils are also comon and, thus, there is a considerable variation in soil quality from one area to another. 76 It should be noted that cotton is very demanding on the soil and is lmom to exhaust trOpical soils rapidly. Nbst of the cotton farmers are well aware of this. ‘ Population and Demography22 Vansina [1965] estimates that there were about 250 ethnic groups in the Congo around 1900. The tribes of northern Zaire, where shifting cultivation is comron, are predominantly of the Sudanese type, while the Ubangi is inhabited mainly by three different ethnic groups: the Ngbandi in the Mobaye area and to the East, the Ngbaka around Gemena and the Mbandja to the North and East.23 In addition, several smaller enclaves of different tribes are found such as Buraka, Gbanziri, Ngbaga, anjotbo and Furu. In the Uélés, the principal ethnic groups are the Zande in the northern savannahs and the Mangbetu and their neighbors , the Mamva and Mangutu, in forest areas , people who are known for their devotion to land cultivation. Zande agriculture has been studied in detail by de Schlippe [1956] . The Zande often move their horestead in search of fertile land and live in isolation or in small villages. The main cr0ps of the Zande are millet, sorghum, maize, finger millet (eleusine) , sesame, manioc, taro, yams, sweet potatoes, groundnuts, voandzea, beans, peas, gourds and ricinus. The Mangbetu, on the other hand, cultivate mainly oil palm, bananas and manioc. 22This section is based mainly on Vansina's book: Introduction 5 l'Ethnographie du Congo, [Vansina, 1965]. 23In the literature different ways of spelling are found: Ngbandi (Ngwandi); Ngbaka (Ngwaka, Bwaka); Mbandja (Mbanza, Banza). We follow Vansina [1965]. 77 Table 3. 8 depicts the demographic evolution of the four sub—regions and gives the results of the 1958 and 1970 censuses of population for each of the zones in the sub-regions. Several zones in the Haut— Zaire region have low to negative rates of population growth because of disease, out-migration, infertility and a dominance of older people. The prevailing social structure in several tribes is polygynie with a large age difference between spouses.24 Finally, we should not forget that the rebellion devasted this area. Infrastructure The northern region is drained by many navigable rivers such as the Ubangi, Mongala and Itimbiri that flow into the Zaire river, and unnavigable rivers such as the Uélé, Uéré and Ituri and a multitude of smaller ones. These waterways are linked with many bridges and ferries and are major bottlenecks for rapid and reliable road transport. The Ubangi and Mongala sub-regions are served by a fairly dense road network feeding into the main embarkation ports of ONATRA, the rational transportation office, on the Lbangi , Mongala and Lua River. On the Ubangi river ports of particular importance to the transport of cotton bales are Libenge (port of Batanga) and the upstream ports of 25 201190, Pandu, Dula , Mobaye-Mbongo and Yakorra . Ports on the angala 24In the Zande and Mangbetu tribes girls becote eligible for marri- age at puberty or even before while men rarely marry before the age of thirty. 25The upstreamports are open 4 to 5 months a year, when the water level is high, and are served by a French shipping corpany Operating from the Central African Republic. 78 Table 3.8. Reptblic of Zaire: Population of the Ubangi, ala, Bas-Uélé and.Haut-Ué1é sub-Regions, 1958 and 1970 sub—Regions and Zones 1958 1970 Annual Rate of Increase Predicted Effective Natural Total sub-region.Ubangi Gemena 187,003 264,584 3.3 3.1 Kungu 102,155 146,387 1.9 3.2 Libenge 86,072 108,965 1.6 2.0 Bosobolo 69,450 89,659 2.1 2.2 Budjala 108,216 167,385 2.6 3.9 Total 552,896 776,980 2.4 3.0 sub-regionhMongala Iisala 91,581 137,164 3.1 3.6 Bumba 168,432 235,618 2.6 2.9 Banzyville 97,936 118,772 2.1 1.7 Businga 97,508 132,356 2.8 2.7 Bongandanga 69,490 115,903 0.9 4.6 Total 524,947 739,813 2.4 3.0 sob—region Bas-Uélé Buta 52,097 63,568 0.1 1.8 Bondo 109,686 174,277 -0.5 4.1 Aketi 85,912 ’82,984 0.7 -0.3 Ango 59,021 65,962 -0.7 1.0 Bambesa 78,549 85,404 -0.3 0.7 Poko 109,032 116,573 0.2 0.6 Total 494,297 588,768 -0.1 1.5 subwregion.HautrUélé Rungu 126,671 180,702 0.3 3.1 Ndangara 67,964 60,345 -l.3 -1.0 Dungu 121,548 155,326 -1.1 2.2 Whmba 137,455 180,068 1.1 2.4 watsa 69,428 93,433 0.1 2.6 Faradje 88,238 125,745 1.4 3.1 Total 611,304 795,619 0.1 2.3 1de Sainthoulin, L., 1971. "Les Statistiques Démographiques en Kinshasa: Congo.Afrique, 377-385. RépUblique Démocratique du Congo." 79 river are Akula and Businga, and the Mogalo port is on the Lua river. 2 Several of these ports are operational during only the rainy season when the level of the rivers is high. In the Uélés the road network is less dense than in Ubangi and migala, and cotton production is extended over a large area making collection and transport more expensive. A single track railroad net- work connects the main ginning plants with the port of Aketi on the Itimbiri river and Bumba on the Zaire river. 'me Aketi-Bumba portion of the railroad, corpleted only in 1973, bypasses river transport on the Itimbiri which was hindered by low water levels in the dry season. The Uélé raierad connects Mungbere, East of Isiro, with Bumba via Isiro, Buta and Aketi with extensions to Titule and Bondo. Despite service. potential, however, the railway company, Vici—Zaire, which meets the area's needs, has been handicapped by poor management and organization, resulting in irregular service and undue delays . A new railroad connecting Banalia with Kisangani is being planned. Although the main purpose will be to transport iron ore from Banalia, it may also serve to transport cotton to the new textile mill in Most feeder roads are in a poor condition, a fact which is reflected in the high transportation rates charged by transport companies and the high cost of collecting and transporting seed cotton. 27 26On the Lua there is only shipping from July to December with barges with a maximum capacity of 400 tons. 27In April 1972 the Ministry of Agriculture increased transport rates per ton and per km. from 12.5 K. to 14 K. for hauling seed cotton, 7.5 K. to 11 K. for cotton fiber and 9 K. to 11 K. for cotton seed. These rates are considerably lower if return freight is available. The ‘80 land Tenure, Crop Rotation and Cultural Practices In Zaire, as in many parts of Africa, land has traditionally belonged to a lineage, a large family in which many members are dead, sore alive and many more to be born [FAO, 1962]. In 1973, however, the Zairean government decided that all land belonged to the state. Farmers now have a tenfruct right: they own all the crops and trees they planted. A farmer gains rights to a plot of land by putting it under cultivation, and he maintains his rights until he abancbns the field [Harms, 1974]. Since population density is low throughout most of Zaire (9. 2 persms per square km.) , new plots usually are easily acquired. The present land tenure system does not ccnstitute an obstacle to agricultural development [Lurpungu, 1973] . A land market does not seem to be functioning in Zaire. The crop rotations recomended by ONAFITEX are: --In forest areas: lst year: March-April-—planting maize July 1—15—-p1anting cotton 2nd year: March-Apr il--p1anting maize , groundnuts or rice July 1-15--p1anting manioc and bananas 3rd and 4th year: manioc and bananas 5th year and following: fallow. -In savannah areas: lst year: June l-lS—planting cotton 2nd year: March-April—planting maize or groundnuts handling of seed cottm at collection points costs 16 K. per ton; loading and unloading of trucks costs 20 K. per ton; and unloading of cotton seed costs 11 K. per ton [République du Zaire, Avril 1972] . 81 July l-lS—planting manioc and bananas (with more manioc and less bananas than in forest soils) 3rd and 4th year—manioc and bananas 5th year and following--fa11ow. In forest areas a pre—culture of food crops is necessary. This helps to clear the land and eliminates the excess nitrogen which stimulates vege- tative developrent of cotton and hinders the developrent of cotton bolls. R. De Ccene [1956] of INEAC recomended the "Bafwasende" rotation for forest areas and Pennisetum savannahs. In this rotation manioc and baranas are the first crops, and rice can be grown between the rows of these crcps. The recommended sequence of crops is as follows: lst year: bananas with rice and manioc interplanted 2nd year: bananas plus manioc, begirming of harvest 3rd year: bananas plus manioc, end of harvest 4th year: March to June—maize, gourds or other local food crops July to December--cotton 5th year: March to June—groundnuts July to December—cotton 6th to 20th year: fallow. This rotation was tested out in the Ubangi and the Uélés and proved satisfactory. H. DuBois [1957] confirmed the superiority of this rota- tion for the Bambesa region. The recommended calendar for cotton is as follows: -In forest areas : land clearing: November to March Land preparation: March 1-15 Planting of maize: March lS—April 15 Harvest of maize: June lS-July 15 82 Planting of cotton: July 1-15 Thinning: July 20-August 5 Ridging: July 25-August 10 Weeding: August 20-October 5 Harvest: December-January land clearing: January to May land preparation: May 15—31 Planting of cotton: June 1-15 Weeding: June 20-July 5 Thinning: June 20-July 5 Ridging: June 25-July 10 Weeding: July 25-September 10 Harvest: November-December—January (ptimal spacing is 80 on. between and 30 on. within the rows with 6 to 8 seeds per pocket.28 About 40 kg. of cottonseed is recomended per hectare. The first weeding should be done three weeks after planting, followed by two more weedings, a month apart. Thinning to two plants is done about three weeks after planting. Harvest starts when the cotton bolls spring open. First quality cotton is kept separate from second quality. The harvest is dried in the sun on trays made with sticks and is stored in large hand woven baskets in a shed. Destruction of the cotton stalks , and hence the tabitat of insects and fungi, after cotton picking has been corpulsory in Zaire since 1921. For a detailed description of the cultural prac- tices in cotton production, we refer to Brixhe [1958] or Van den Abeele and Vandenput [1956] . 28This should provide 41,625 pockets per hectare; in savannahs, often 60 cm. x 25 om. is recomended. 83 Role of Cotton Production in Northern Zaire Cotton is by far the most important textile fiber produced in Zaire (see Table 3.9). Other fibers produced in Zaire are Urena Iobata (African jute), agave-sisal, kapok, jute, raphia, manniophyton, punga, abrora, a. o. The northern cotton belt has gained in importance since 1960. In 1959 the Ubangi and the Uélé cotton regions produced 42 percent of the total cotton crop and in 1973 they produced 50.6 percent (Table 3.10) . In particular, Ubangi has become more important since Ubangi now contri- butes l9. 0 percent of the total production as contrasted with only 9.6 percent in 1959. The largest relative drOp in cotton production since independence occurred in Shaba, while the Sankuru sub-region in Kasai tas increased its share from 1.8 percent to 4.5 percent. Finally, in 1964-1965, when the Congo rebellion was at its peak, the Equator region (Ubangi and Mongala) produced 42 percent of the national cotton crop. The craracteristics of pre- independence cotton production in the North and South are given in Table 3.11. The evolution of cotton production in the Equator and Haute-Zaire regions of the northern zone since 1956 is presented in Table 3.12. This table shows that cotton production fell in the Uélés relatively more than that in Ubangi-Mongala. Thus , the potential for increasing cotton production in the Uélés is much higher tlan that in Ubangi- Mmgala. Table 3.13 reveals that production in the Mongala sub-region has decreased relatively more since independence ttan that in Ubangi. In fact, the decrease in the Equator region since 1959 is almost cotpletely 84 Table 3. 9. Production of the Principal Textile Fibers in Zaire (in Tons) Type of Fiber 1959 1966 1970 1971 Cotton fibers 59,280 6,590 17,250 19,930 Urena fibers 9,035 3,388 1,857 3,113 Punga fibers 2,828 2,622 1,993 961 Sisal fibers 263 216 . . . . . . lMinistére de l'Agriculture, 1968. "L'Agricultur Congolaise en Tableaux Statistiques. " Kinslrasa: Direction Etudes et Politique Agricole; Ministére de l'Agriculture et du Développerent Rural, Avril 1970. Rapport Préliminaire 1959-1969. Kinshasa: Direction Etudes et Politique Agricole; République du Zaire, Novembre 1973. Conjoncture Economique. Année 1972 et Ier Trimestre 1973. Kinshasa: Ministére de l'Economie Natiorale, No. 13, p. 18. Table 3.10. Republic of Zaire: Production of SeedlCotton Per Cotton Region, 1959 and 1973 (In Tons) Cotton Regions 19 59 Percent 1973 Percent Ibangi 17 , 177 , 9 . 6 12 , 709 19 . 0 Uélés 58 , 018 32 . 4 21 , 178 31. 6 . f r Kivu 15,102 8. 4 11,635 17.4 Maniéra 17,415 9.7 L L ( Nord-Shaba 19,943 11.1 V 6,319 9.4 Sud-Shaba 12,409 6.9 k L Kasai 35, 992 20 . 1 12 , 090 18. l Sankuru 3,038 1.8 2,984 4.5 Total 179 , 114 100 . 0 66 , 915 100 . 0 1ONAFITEX, Comercial Department, Kinshasa. 85 Table 3. 11. Republic of Zaire: Characteristics of Cotton Production before Independence, 19 54-1958 Averages]- Cotton Production Acreage Yields Production Percentage of Regions Hectare Kg. /Ha. Tons Total Production North 165,291 419 69,235 47.00 South and Central 182,717 430 78,069 53.00 Total 347 , 462 424 147 , 304 100 . 00 1"Van de Walle, B. , 1960. Essai d'Une Planification de l'Economie Agricole Congolaise. Bruxelles: INEAC, SErie Technique No. 61, p. 65. attributable to a reduction in Mongala where production is now only about one-fourthof the 1959 level. slim Cotton was introduced in Zaire in 1918 as a required crop and it has remained so ever since. During the colonial period a dense agricul— tural extension network promoted cotton production, particularly in paysannats where cotton was the principal cash crop. Presently, however, the extension service is largely ineffective. The agricultural calendar is poorly observed, seeding is retarded, fallow periods are too short and many cotton fields are poorly maintained. Since independence exten- sion agents have not been retrained; they lack travel facilities and are insuffiently motivated. To these problems is added an institutional constraint since the extension agents come under the auspices of three different institutions. The agricultural service of ONAFITEX now has its own extension agents who started large scale pesticide treatments on cotton fields in 1972. 86 Table 3.12. Republic of Zaire: Seed Cottm Production in the Northern Cotton Belt Per Region, 1956-1973 (In Tons)1 Year Equator Region Haut—Zaire NAHV2 Total Region Company North 1956 18,929 55,477 --— 74,406 1957 15,416 38,934 -—- 54,350 1958 15,583 & 39,869 & NAHV 4,066 59,473 1959 18,003 57,752 -—- 75,755 1960 14,198 50,357 --- 64,555 1961 9 , 443 30 , 472 --- 39 , 915 1962 7,604 & 14,012 & NAHV 1,566 23,182 1963 9,447 & 17,976 & NAHV 1,772 29,195 1964 7,261 & 21,923 & NAHV 2,260 31,444 1965 7,336 & 1,179 & NAHV 85 8,600 1966 6,397 & 2,342 & NAHV 679 9,418 1967 5,807 & 2,444 & NAHV 261 8,512 1968 5,284 & 11,527 & NAHV 220 17,631 1969 8,714 13,268 —-- 21,982 1970 11,155 9,023 --- 20,178 1971 12,948 9,778 - --- 22,726 1972 6,346 11,127 -- 17,473 1973 .12,709 21,178 —-— 33,337 1Derived from seed cotton collection figures of the following corpanies: SACCIJM), C(NACO, BEIGIKA, NAHV and VINCHENI‘, ONAFITEX, Kinshasa. ZNAHV operated 5 ginneries: EKG/IA in Equator region and AMADI, LEBO, TAPILI and ZOBIA in Haut-Zaire. YAKOMA' 5 reception never exceeded 1, 000 tons of seed cotton. It was not always possible to split up NAHV' 5 seed cotton reception between Equator and Haut-Zaire. Where this was not possible, NAHV's share appears separately. 87 Table 3. 13. Republic of Zaire: Seed Cotton Production in Equator Region, Per Sub—Region, 1956-1972 (In Tons) 1 Year Ubangi Mongala Equator Sub-Region Sub-Region Region 1956 11,731 7,198 18 ,929 1959 10, 644 7, 359 18,003 1963 7,076 2,371 9,447 1967 4,383 1,424 5,807 1969 7,231 1,483 8,714 1970 9, 377 1,778 11,155 1971 10,590 2,358 12,948 1972 5,4452 900 6,345 1[République du Zaire, Avril 1972, p. 12]. 2The sharp drop for Ubangi is attributable to a lack of rain in October 1971 and major insect damage, especially from he10peltis, acaridosis, dysdercus and the pink bollworm. Since colonial times there has always been a uniform seed cotton price paid to farmers in Zaire. In 1970 the price of seed cottm paid to producers was the lowest of all African countries for which price data could be obtained. Presently , three cotton production projects are underway in Zaire: (1) the FED—CAK project in eastern Kasai which is an integrated "paysannat" development schere; (2) the CFUI' project in Ubangi-Mongala which has as its broad objective the development of oottm productim mainly through improved agricultural extension services to small farmers. Presently, the project is engaged in varying research aCtivities which are basically the domain of INERA and ( 3) the FIWA-BILI 88 project of the Belgian Aid Mission in northern Ubangi which is con- centrating on tractor mechanization and pesticide applications in cotton fields. Ecological conditions in the North are well adapted to cotton production. Soils in the Uélés are generally better than those in (bangi-Mongala. Land tenure is no constraint on cotton production. The cultural practices used by farmers are primitive and very labor intensive. The only source of power is human labor. Only simple hand tools are used and no fertilizers or pesticides are applied by the farmers. Feeder roads are in a poor condition in northern Zaire. The Uélés railroad has been handicapped by poor management and organization. The population in Ubangi and Mongala sub-regions is increasing at a rapid pace but several zones in the Haut-Zaire region have low to negative rates of populatim growth . The northern cotton belt has gained in importance for cotton pro- duction since 1960. The Ubangi cotton region now contributes 19. 0 percent of the total cotton production in Zaire as contrasted with 9.6 percent in 1959. The decrease in cotton production in the Equator region since 19 59 is almost completely attributable to a reduction in angala sub—region where production is now only about one-fourth of the 1959 level. Cotton production fell relatively more in the Uélés than in Ubangi-Moigala. The potential for increasing cotton production in the Uélés is much larger tl'an that in Ubangi-Mongala. CHAPTER IV CUI'ION MARKETING AND PIDCESSING The purpose of this chapter is to describe and analyze cotton marketing and processing in Zaire in order to diagnose the major barriers constraining the present system and to formulate policy rec- ommendations for improved performance. A historical review of the cotton industry will provide background for a better understanding of tlne present cotton marketing structure in northern Zaire . The cotton seed crushing industry and the textile industry in Zaire will be briefly discussed as they are an integral part of the total sub-sector. History of the Cotton Industry in Zaire Development of the Cotton Industry The first private cotton cotpany , the Cotpagnie Cotonniére Congolaise (CO'IOBCO) , was founded in Zaire on February 10, 1920, and the first two ginneries were ordered by the government in the U.S.A. in 1917 and installed in Kibombo (Manie’ma) and Lusambo (Sankuru) in 1921. In 1925 the first cotton textile mill was established in Kinshasa. From 1920 to 19 30 several private cotton corpanies were created in Zaire, and by the end of 1937 there were 119 ginning mills installed in the country. The first legal text referring to cotton production is the decree of August 7, 1918, specifying measures to protect cotton against diseases, 89 90 to conserve fiber quality and to fix a minimum price for the purchase of seed cotton. The decree of May 28, 1920 guided the organization of official markets for cotton and introduced cotton licensing. The license was contingent upon guarantees for adequate ginning and fiber quality. The decree of August 1, 1921 introduced reserve cotton areas for each mill equipped with mechanical cotton gins , and divided the cotton producing areas into geographical buying monOpolies. Its purpose was to secure adequate supplies for each mill and to limit the maximum distance in each reserved area to 40 km. This limitation helped to reduce the dis- tances whidn cotton farmers had to carry their seed cotton to the mill. On December 10, 1930 the delimitation of reserved buying areas became subject to the authority of the provincial governors . In 1929 the Congolese Cotton Committee (Comité Cotonnier Con- golais) was formed, bringing together representatives from all the cotton corpanies. This committee coordinated the efforts of the cotton com- panies, standardized work methods and acted as a lobby. It had a strong voice in the design and impletentation of cotton policies until the creation of ONAFITEX in 1971. The ordinance of October 31, 19 35 introduced open areas for the newly developed cotton regions in which any one could buy cotton, sub- ject to minimum price legislation and licensing. The purpose of this ordinance was to test an alternative to geographical buying monopolies. .The ordinance of April 19, 1938 terminated this practice, however, because of fierce cotpetition for cotton produced near railroads and abandonment of retote cotton areas. The quality of cotton fiber declined because of the mixing of cotton seeds of different quality. 91 The Fonds de Remploi: 1924-1935 In 1924 the governor general of Zaire was authorized to collect a cotton tax which would constitute a "Fonds de Remploi. " or reinvest- ment fund, to be utilized for agricultural education and extension, cotton seed improvement and general agricultural developrent of cotton regions. In 1930 the tax was lifted because of adverse economic con- ditions, but it was reinstated in 1935 and in 1936 a price stabilization fund, "Fonds de Réserve Cotonniére" (Cotton Reserve Fund), was created to replace the "Fonds de Remploi. " The Cotton Reserve Fund: 1936-1942 The Reserve Fund was established by the government to stabilize seed cotton prices, improve road maintenance and distribute farm tools. The seed cotton price was determined from a "potential" cotton price, and the Reserve Fund was credited with the difference between the "potential" cotton price and the actual price which farmers received. A two-way (baréme) table was used to determine the "potential" cotton price which could be paid to the farmer. The seed cotton price paid to farmers was computed by deducting the following from the potential price: cost of small farm tools and road improvetent, contributions to local village chests, new taxes which increased tlne cost price of cotton ginners and deposits in the stabilization fund. Banneux [1938] , although favoring the "potential" formula, argued for a regular revision of the data used to calculate the "baréme" table in order to keep abreast of the changing economic and social situation. And, in fact, in 1941 the table could no longer be followed because of the special economic conditions created by World War II. A new formula 92 for tie calculation Of the potential cotton price was adopted by the government and the private cotton corpanies in 1943. COGERCO: 1943 In 1943 the Cotton Reserve Fund was brought under the manath of a committee entitled the "Comité de Gerance de la Caisse de Reserve Cotonniere" (COGERCO) . This committee was corposed of four civil ser- vants and two representatives from the cotton conpanies . Ordinance Nr. l82/Agric. of June 12, 1943 established a Cotton Reserve Fund. The cotton corpanies bought seed cotton and financed their own Operations. The potential formula was used to determine . the remunera- tion Of cotton farmers, "potential" referring to the price whidn theoretically could be paid to farmers after deduction of the following elements from the realized price of cotton fiber: export duty and per unit cost of the cotton corpanies including assembly, ginning, trans- port, storage, sales costs and profits. The potential formula was as follows: P = (SF - ED) - (CG X 1.15) - (P X 0.15) Where : P = Potential SF Total receipts from the sale Of cotton fiber ED = Export duties CG = Costs incurred by ginning mills after each campaign, includ- ing 6 percent interest on invested capital and depreciation charges Thus, profits for the cotton companies were equal to 0.15 (CG + P) or 93 15 percent of ginning costs , including transportation, and 15 percent of the potential. If the price paid to farmers was lower than the potential price, the difference was deposited in the Reserve Fund, if the price paid to farmers was higher than the potential price,then funds from the Reserve Fund were used to maintain the price to farmers. This formula was used, with various modifications in 1950, 1955 and 1962, until the creation Of ONAFITEX in 1971. From 1940 to 1948 the Reserve Furnd spent 3.28 million dollars on road improvement in cotton areas. All cotton areas received 0.20 fr. per kg. seed cotton annually for maintenance and improvement of roads, an amount which grew to an annual sum Of 25 million francs. The Cotton Decree Of June 18, 1947 After World War II the geographical monopoly rights granted under the decree of August 1, 1921 expired, setting the stage for an overhaul of all cotton legislation. The 1947 cotton decree constituted the legal framework for post World War II cotton policies until the creation Of ONAFITEX in 1971. The central feature of the 1947 decree permitted the cotton farmer to remain the owner of his crop until the final sale to domestic or foreign textile cotpanies.l As a result, the role of private cotton cotpanies was reduced to that of a contractor providing services to farmers until farmer OOOperatives were able to take over their funnctions. The creation and transfer of ginning plants was made subject to government approval. J'Nbst producers in Africa sell their cotton (with transfer Of ownership) either to a cooperative society as in Sudan and Tanzania, to the goverment or a goverrnment marketing board (Egypt, Nigeria), to a marketing cotpany such as CFDT in the former French colonies or to private companies (Ethiopia). 94 The remuneration of cotton companies was fixed by the governor general of Zaire by taking the fol lowing into account: the sales value of cotton fiber F.O.B. Matadi,the amount of seed cotton produced and the costs incurred by cotton conpanies. The "potential" formula was used to calculate the advance payment and the supplement to cotton farmers. The Cotton Reserve Fund was given legal status and an independent budget. The Fund was managed by a management committee (COGERCO: Comité de Gerance de la Caisse de Reserve Cotonniere) composed of six civil servants and four representatives from the cotton cotpanies . Each year (IJGEICO proposed and the governor general fixed the amount of advance payment to be paid to farmers upon delivery Of their seed cotton, and decided upon the utilization of tte remaining Fund balance after each cotton campaign. The functions of CDGERCO were as follows: -Financing, through private cotton companies, the purchase of seed cotton from farmers at a fixed guaranteed price. --Financing ginning and transport costs: refunding to cotton companies all expenses (on a cost plus basis) incurred in their activities. -—Supervision Of sales of cotton fiber. --Developrent Of cotton production through the supply of improved seeds, insecticides, farm equiprent, etc. The 1947 decree became effective on Noverber 15, 1949 in the Equator, Haut-Zaire and Kivu provinces and on Decerber l, 1951 in Kasai and Shaba. In 1950 the reserves of OOGERCI) amounted to 25 million dollars, and, over the 1951-1953 period a total of 12.84 million dollars 95 was returned to farmers as cash bonuses and distributed according to their production.2 Although from 1953 on cotton farmers never received a cash supplement beyond their fixed advance price, OOGERCO continued to spend part of its surpluses for cotton development purposes, mainly maintenance and repair Of roads and the hiring Of extension workers to stimulate cotton production. The total disbursetent of COGERCO between 1950 and 1955 was 32 million dollars. Because Of depressed world market prices for cotton , COGEROO disbursed the following to support the seed cotton prices: 1.4 million dollars in 1957, 4.3 million dollars in 1958 and 5.1 million dollars in 1959. From 1950 to 1960, COGEICO reserves exceeded 20 million dollars but these reserves declined rapidly from 1960 to 1969 and only about one million dollars was left by 1969. During World War II the Congolese Cotton Committee formed a pool for the sale Of cotton fiber. The pool was quite successful, and on January 3, 1946 a central sales organization, COVENCO (Corptoir de Vente des Cotons du Congo) was established with its headquarters in Belgium. COVENCO had tlre legal status of a cooperative with the cotton cotpanies as cooperative merbers and a representative from the governor general as watchdog. COVENCO received a sales commission of 2.75 to 3.00 per- cent on He F.O.B. Matadi price Of cotton fiber. Then, with the conven- tion of June 2, 1962, CDVENCO became a mixed company, legally a non- profit making cooperative, 50 percent owned by COGERCI) and 50 percent 2Tire large surpluses were accumulated as a result of high world market prices for cotton fiber, especially in 1950, 1951 and 1952, the modest prices paid to the farmers. 96 by the cotton companies, with monopoly rights for the sale of Zairean cotton. In 1949 224 European agents from the cottm companies were engaged, at least half of the year, in cotton extension. 3 The cost for these agents'was included in the ginning cost and was paid indirectly by the farmers . The Convention of May 30, 1950 The 1950 convention permitted farmers to remain owners of their crop until its final sale was consummated. The cotton companies became "service" companies and were remunerated on a custom work basis. CIJGERCO financed the purchase of seed cotton. The old potential formula was reinstated, although the profit rate was reduced from 15 to 13 percent. Since this formula was used until the end of 1962, a closer analysis is warranted. The potential is the sum of money which can be paid to farmers for their prodmtion of seed cotton, after all expenses are paid for processing and marketing. It consists of an advance payment at the time of seed cotton delivery and a supplement, arrived at after all cotton bales are sold. This surplement could be used to pay farmers a bonus proportional to their cotton production or could be spent by COGERCO on economic and social developtent projects in cotton regions. If the advance payment exceeded the potmtial, COGERCO reserves were used to cover the difference between the advance and potential payment. The following potential formula was adopted: P=SF-ED-CG-O.l3(P+CG) (1) 3111a allocation of agents by region was as follows: Kivu-l9, Shaba-29, Kasai-50, Equator-38, and Haut-Zaire—88. 97 “here: P = Potential SF Total receipts from the sale of cotton fiber ED = Export duties CG = Costs incurred by ginning mills as accounted for after each campaign. This cost includes depreciation charges but ex- cludes the advance payment made to cotton farmers at the time of delivery of their seed cotton. Thus, profit of the ginning mills was fixed at 13 percent of the potential plus ginning costs. Formula (1) can be transformed as follows: (P+CG)=SF-ED-0.l3 (P+CG) l.l3(P+CG)=SF-EDorP+CG=-%—g-g-(SF-ED) Thus, .. _ _ -l.3__.l.99 _ P—SF ED CG 100 113 (SF ED) or P=SF-ED-CG-l'§-—(SF-ED)0r ll3 P=SF-ED-CG-0.115 (SF-ED) (2) We find that profits of the ginning mills were equal to 11.5 percent of the net sales value of cottoi fiber with the net sales value equal to gross sales value (SF) minus export duties (ED). Profits are indepen- dent of ginning costs. When ginning costs increase, the potential de- creases accordingly. Yet regardless of ginning costs, ginners realize their profits which depended only on the net sales value of cotton fiber.4 The ginners benefitted directly from an increase in the world market price 4It is assumed that for a given seed cotton production, there is no link between per unit ginning costs and the net per unit value of fiber exports. 98 of cotton fiber, and even when the world market price fell, their pro- fits were still guaranteed at a fixed percentage of this price. As a result, ginners had no incentive to reduce ginning costs. However, ginners did have strong incentives to increase the net value of fiber sales by: --higher seed cotton production --improving the quality of cotton fiber to obtain higher per unit prices --increasing ginning percentages “reducing export duties --increasing the sales price of cotton fiber. This system permitted the cotton companies to overexpand and to over- commit resources in ginning plants. It also encouraged them to push for imposed cotton production in areas where, because of low yields, the cotton price did not offer a reasonable incentive to farmers. This process also explains why the companies financed extension work and agricultural research. Since COGERCO reimbursed all costs incurred by the ginning companies, every additional kg. of cottoi produced increased the profits of the cotton companies, regardless of the cost of producing this additional kg. of cotton. The profits of the cotton corpanies have beei excessive as Table 4. 1 illustrates.5 If the remuneration of ginners had been set at a fixed amount per bale of fiber ginned, (pen to competitive bidding and annual review, 5The cotton companies were not required to present a detailed breakdownof theircoststothegovenmentortoCOGEKEO. COGERCOor the government could not properly audit the accounting of the cotton companies. The governor geieral could oily refuse to accept certain costs presented by the cotton corpanies to (IJGERJO for reimbursement. 99 efficiency would have been fostered. 6 Alternatively the profits of ginning companies could have been computed as a percentage of the differeice betweei the net sales value of cotton fiber and the ginning costs . Table 4.1. Republic of Zaire: Ginning Costs and Profits of the Cotton Companies, 1949-1952 (In Francs per Kg. of Lint)l Year Ginning Costs (CG) Profit Profit as Percentage of Ginning Costs (CG) 1949 8. 4819 3. 8633 45. 55 1950 9.0920 4.1623 45.78 1951 11. 4698 5. 1311 44. 74 1952 11. 9158 4. 5762 38. 40 lDepy, G. , 1954. "Etape de la Convention avec les Sociétes Cotonnieres." Kinshasa, Unpublished document. The Cotton Seed Coivention of April 19, 1951 Prior to the convention of April 19, 1951, the crushing mills, on the basis of their balance sheets, were reimbursed for depreciation and received 6 percent interest on invested capital. The balance was equally spit between the Cottm Reserve Fund or the chiefdom which had produced the cotton seeds and the crushing mills. But it was the 6111 French Equatorial Africa to the West and North of Zaire, the remuneration of the French cottm corpanies was determined by the decree of Decetber 12, 1948 in which they shared 3 percent of the export price F.O.B. of cotton fiber and 15 percent of the difference between the export price F.O.B. and the cost paid by the corpanies, taxes included, with a maximum of 3 francs per kg. of cotton fiber exported. The cost paid by the cotton cmpanies was also subject to the governor's approval. [Sladden, 1949] For Nigeria, Kriesel [1968] reported that the British Cottm Growing Association ginned, baled and classified cotton at a fee fixed annually on a per unit basis. lOO convention of April 19, 1951 betweei COGERCO, the government and the Congolese Cotton Committee (representing the cotton and cotton seed crushing industry) which determined how the value of cottoi seeds and their derived products would be distributed between cotton farmers , owners of the seeds, and the cotton seed crushing industry. The oil mills pledged to process all cotton seeds except in areas where assetbly and transport costs were prohibitive. The mills also agreed to finance the processing and to market the cotton seed products. After each production season the profits and losses were equally split between the companies and COGERCO. In case of a net loss, COGERCO paid half of it. Table 4.22 shows that the crushing industry frequeitly closed its accounts in the red. The Convention of January 31, 1955 This convention extended the use of the potential formula which was inaugurated in the early 19405 and formally adqated by the conven- tion of May 30, 1950. However, the formula was only applied when the sales price F.O.B. Natadi (after deduction of export duties) was between 35 and 40 francs per kg. A separate calculation was made for the northern and southern cotton belt. The ginning costs used in the calculation of the potential were those of the COTUNCO corpany except for transport costs which were secured from public transport companies (ONAT‘RA or the railways). COGERCO advanced the funds necessary for the purchase of cotton by the cotton finms. However, the firms financed the operations of trans- port, ginning, baling, storage, insurance, etc. 101 The COVENCO-Convention of June 2, 1962 This cowention between the Zairean government, COGERCO and the cotton companies transformed COVENCO, the central sales organization of the cotton companies, into a mixed company with the legal status of a cooperative; it was owned 50 percent by COGERCO and 50 percent by the OOttCIl companies. Thus, COGERCO, representing the cotton farmers, gained 50 percent ownership in OOVENCO and reimbursed the cotton com- panies for the transfer of titles. COVENCO transferred its corporate headquarters from Brussels to Kinshasa and retained monopoly rights for the domestic and export sale of cotton fiber. The proceeds from the sale of cotton fiber were distributed by COVENCO as follows: 1. Reimbursement of COVENCO operating costs. 2. Maragement fee for COVENCO 3. Reimbursetent of costs and remunerations of cotton companies for reception, transport and ginning. 4. The rerainder (if any) was transferred to the COGERCO stabilization fund . The Convention of June 18, 1962 This agreement between COGERCO, the Congolese government and the cotton ginning mills updated the earlier agreements of 1955 and 1950. The 1962 cowention restricted the role of the cottoi companies to the cottoi ginning phase (Phase 2), and until farmers’ cooperatives were established, to the purchase and transport of seed cotton from farmers to the ginning mills (Phase 1) . A new system of budget control for the cotton companies was also introduced by which the cotton companies had 102 to present a budget plan before each campaign instead of presenting all costs to (DGEKD after the campaign. COGERCI) loaned the cotton corpanies money for the purchase of seed cottm. Furthermore, the cotton companies were refunded all expenses incurred in the transport and ginning of cotton up to its delivery to transporters where COVEMIO took over respmsibility. In addition, the cottoi corpanies were paid for their services by adopting a new formula for corputing the potential: P=SF-ED-CG-C.CWE1\CO-C.CG3EICO “here: P = Potential SF = Total receipts from the sale of cotton fiber ED = Export duties CG = Costs of ginning mills: this included four elements (see below) C.COVENCO = Costs incurred by COVENOO for transport, insurance , managerent, etc. (See Convention of June 2, 1962) C.COGERCD = Costs incurred by COGERCO for managemeit and assis- tance to cotton farmers The costs of ginning mills included the following elements: 1. all costs for transportation, ginning and baling accounted for by the cotton ginning mills after each campaign. 2 . remineration for the depreciation (amortization) .of equipment and buildings equal to 2 percent of average sales price F.O.B. Matadi, export duties included, of cotton lint. 3. remuneration of capital employed according to a percentage calculated on the average sales price F.O.B. Matadi, export 103 duties included, of cotton lint. This percentage is between 2 percent and 4 percent and is calculated as follows: total seed cottm production in tons 110,000 %rate= X3 4. remuneration of managerent for the financing of the cotton ginning phase according to a percentage of the total costs accounted for by the cotton ginning mills (prix de revient industriel) as in (1). This percentage should be between 5 and 7 percent of total costs and is calculated as follows: % rate ___ total seed cotton production in tons X 6 110 , 000 Based on the realized production levels, only the rates of 2 percent and 5 percent were used. Thus, the potential formula is: P = 0.96(SF - ED) - 1.05 CG - C.COVENCO - C.C(XSERCO and profits, of the ginning mills are equal to: 0.04(SF - ED) + 0.05 06 In the 19 50 convention depreciation was determined by the ginning mmpany and was included in ginning costs . In this formula depreciation charges are fixed at 2 percent of the average sales price F.O.B. Matadi, export duties included, per kg. of cotton fiber. The profits of the companies increased directly with ginning costs, regardless of the volume of cotton ginned. Thus, the ginning mills had an incentive to drive up ginning costs and to increase cotton production which, in turn, 104 increased their relative share of cotton earnings. The cotton farmers , or the other hand, benefitted from lower ginning costs. In making the percentage rate of remuneration of capital a func- tion of the total seed cotton production, any production increase between 73,333 and 146,666 tons led to a higher relative and absolute return on capital. Similarly, any production increase between 91,666 and 128,333 tons led to a higher relative and absolute return to managerent and financing. Depreciation and capital remuneration were a function of the sales price F.O.B. Matadi, export duties included. There was no sound basis for making depreciation and capital remuneratioi a function of the net export sales price. In fact, the cotpanies applied their profit rate of 4 percent to the sales price of cotton fiber sold to dorestic mills.7 As usual, the cotton farmer was the receiver in last resort; he received what was left over after paying for all marketing functions, on a cost plus basis. The bargaining position of farmers was very weak vis-a-vis the cotton ginning corpanies who enjoyed a near—monopoly situation. It is then not surprising that corpared with several African countries, the Zairean cotton farmer received the lowest price paid per kg. seed cottoi, although the quality of Zairean cotton (strict middling l l/16 inches) is considered to be excellent (see Table 3.5) . On January 22, 1963 the Ministry of Finance provided COGEKIO with an annual guarantee of 500 million francs of private bank loans for the purchase of seed cotton. On Noverber 15, 1963 the Ministry of 7Domestic mills paid about 25 percent to 40 percent more than the net export price. Because of large cotton fiber imports in 1965, 1966 and 1967 the government enacted a policy whereby domestic textile mills SLbsidized domestic cotton production. 105 National Economy decided to offer a 50 percent subsidy on the sale price of cotton lint sold to dorestic mills. This had the effect of increasing the sale price of cotton lint to domestic mills by 50 per- cent, and raised the price paid to farmers (see Table 3.6) as well as encouraging cotton production in marginal areas . However, cotton production continued to fall until 1966.. The Decree—Law (Décrét - loi) of August 13, 1965 This executive order focused on Zairisation of COGERCO managerent and it called for increasing participation of cotton farmers in the managerent of COGERCI). The new COGERCO maragetent committee was corposed as follows : l. 6 civil servants--secreta.ry generals—of Ministries interested in cotton production 2. a minimum of 6 and a maximum of 11 representatives of the cotton farmers , nominated by the provincial governors 3. 3 representatives from the cotton corpanies 4. a president, proposed by COGERCO and nominated by the President of the Republic. This decree reduced the influence of the private firms in designing cotton policies . The Private Cotton Corpanies COTOMIO (Cotpagnie Cotonniére Cogolaise) , a Belgian cmpany, controlled eight cotton companies in Zaire and three vegetable oil (including cottm seed oil) corpanies. These companies controlled about 90 percent of cotton marketing in Zaire and operated 74 cotton ginneries 106 with an estimated ginning capacity of 355, 300 tons of seed cotton. COTONCO and seven other private companies in Zaire operated a total of 115 ginneries with an annual capacity of about 400,000 tons of seed - cotton. Presently, however, only about 60 ginning plants are in opera- 8 Thenunber tion with a total annual capacity of about 300,000 tons. of cotton companies and the number of ginning plants and their location by the end of 1957 are shown in Table 94.2. Seed cotton production in 1973 amounted only to 66,915 tons; thus, less than one-fifth of theoretical ginning capacity is presently used. Since independence , tre cotton corpanies have made very few new investments. The Creation of ONAFITEX on August 12, 1971 Since mGERCO's losses were projected to be extrerely high in 1970 the Ministry of Agriculture decided in 1970 to drastically reduce the budget for the cotton corpanies in order to reduce ginning costs . All firms were forced to accept the lowest rates for particular cost items, unless a firm could substantiate a higher rate . However, the cotton companies refused to accept budget restrictions, and on June 28, 1970 the Minister of Agriculture renomced tre convention between comma and the cotton companies. The government established ONAFITEX (Office National des Fibres Textiles or National Office for Textile Fibers) under ordinance law 8m 1958 a total of 353 technical assistance agents were paid by the cotton conpanies and reimbursed by CDGERCO. In 1970 there were only 54 left, but they did not perform any extension work because COCERCO did not pay them for this service. 107 Table 4. 2. Reprlic of Zaire: Cotton Corpanies, Number of Ginning Plants and Their Location in 19571 Cotton Companies Sub-Regions No. of Plants Northern Cotton Belt Belgika Ué lé 8 Societé Cotonniere du Borokandi (Socobom) ' Uélé 5 Nieuwe Afrikaanse Handels— Vennootschap (NAHV) Uélé-Ubangi 6 Vinchent Ituri (Mahagi) l COIOI‘CO Haut-Uélé 6 Bas-Ué lé ll Ubangi ll Subtotal 48 Southern Cotton Belt Societé Cotonniére Coloniale (Colocoton) Lorami-Sankuru 3 Campagnie Comerciale Belgo- Africaine (Corbelga) Lotami-Sankuru 3 Societe’ Cotonniére du Tanganika (Cotanga) Tanganika 7 Campagnie de la Ruzizi Sud-Kivu l Societe’ Cotonniére de la Luisa Kais 1 Cotpagnie du Lubilash Katanga l Societe Congolaise Bunge (Buncolo) Haut-Lorami and Haut-Katanga 5 Societe Agricole, Comerciale et Industrielle du Kasai (Sacominka) Kasai l COIODCO Tommi-Kasai ll Sankuru 5 Kivu 5 Maniéma 7 Dilolo 2 Sub-Total 52 Total 100 lBrixhe, 1958.p. 17. 108 No. 71/077 and ordinance No. 71/224 of August 12, 1971. ONAFITEX replaced (IDGERCD and COVENCO and terminated all previous cotton legis- lation and conventiore. ONAFITEX cotbines all the different bodies concerned with cotton production, processirng and marketing under a single organization. It was created as a public institution subject to the authority of the Ministry of Agriculture with monopoly rights for the marketing of seed cotton and its by-products and any other textile fiber. The main purpose of the ONAFITEX reform was to reduce tie profit margin of cotton companies and eliminating them if an agreerent coild not be reached. In that event they would be replaced by (NAFITEX mntil cotton farmers coJld assure responsibility for cotton marketing. The ordinance-law erpowered the Minister of Agriculture to authorize private conpanies, producer cooperatives or individuals to receive, gin and bale cotton on a custom work basis. Under the ordirance— law the Minister of Agriculture fixes the advance payment to farmers at the time of collection of seed cotton while the balance is paid after the final sale of all cotton fiber. ‘Ihe balance is determined by ONAFITEX. The Ministry of Agrioilture sets the sales price of cotton fiber to local textile mills after corn- sultation with the Ministry of Natioral Economy. The new legislation charged ONAFITEX with the responsibility for Phase 1 of cotton marketing: collection of seed cotton at local assenbly points and transport to the ginning plants . The cotton cotpanies retained responsible for Phase 2: ginning, baling and trans- port of bales from ginneries to the pick-up points by public transport 109 agencies. (NAFITEX then took over and arranged for the sale of cotton fiber to local textile firms or for export. The Zairean government negotiated a contract with the cotton cotpanies for their Phase 2 operations. However, both parties could rot agree on the amount of the custom work fee. The government first proposed 37 Zaires per ton of cotton fiber ginned and the ginneries finally agreed to 55 Zaires per ton. However, after only a month of work, the ginneries stOpped all activities and detanded a higher fee. The government then intervened by taking over their facilities, supplies and manpower, both Zairean and expatriate.9 (Law 72-006 and ordinance No. 72-235 of May 8, 1972 specified the terms of the requisition.) Ginners received 4 8 per ton of cotton fiber for the use of tleir facilities, while treir supplies were reimbursed at cost. Manpower was paid the same salary ‘ as before the take-over, but about 90 percent of tie expatriate personnel were laid off, resulting in considerable savings.10 The principal reasons which led the government to a take-over (with compensation) of the facilities of the private cotton cotpanies can be summarized as follows: 1. endnaustion of (DGERCO reserves and imminence of bankruptcy of the cotton industry 2. government efforts to put the economy under Zairean control 9Offices and houses of personnel of the cotton conpanies were also requisitiored, but cotton seed crushing mills were excluded arnd remained under control of the cotton companies. 10Savings in the order of 650,000 8 per year were reported. 8. 10. 11. 110 stubbornness of the cotton corpanies in yielding to government pressure for clanges inflation of ginning costs slow Africanization of ginning managerent lack of new investments in ginning mills since 1960 persistent losses of the cotton seed crushing industry (see section on "Cotton Seed Crushing") , 50 percent of which were borne by COGERCO difficulty experienced by the government and COGEIRCO in con- trolling ginning costs because of the lack of knowledge of real processing and transport costs conflicting incentives and interests placing cotton farmers and the government in opposition to tl'e cotton conpanies a levelling off of cotton production private cotton companies generally ignored cotton legislation. Structural Organization of ONAFITEX The managerent committee of ONAFITEX is corposed of a president, tie director general, deputy director general, a representative from the mninistries of agriculture, national economy, finance, foreign trade, transport, public works and interior, and two representatives from the cotton farmers. The managerent committee is appointed by the President of tie Republic upon rnomination by the Minister of Agriculture. Directors and regional directors of ONAFITEX are appointed by the Minister of Agriculture. Several special funds are set up by ONAFITEX. They are: 111 a renewal fund: for tie replacerent of buildings, installa- tions, machinery and materials an insurance fund: to cover costs resulting from fire, losses, damage I a training fund: for the retraining of personnel a reserve fund a provisions fund: to finance the purchase of equipment needed for the development of tie activities of ONAFITEX a price stabilization fund: to guarantee stable prices to cotton farmers . The reserve and provisions fund are financed by appropriations from the annual net profits of ONAFITEX. ONAFITEX draws its incore from the difference between the sales value of cotton fiber and by-products and the amount paid to cotton farmers plus all Operating costs. Role of ONAFITEX T'te ordinance- law creating ONAFITEX requires it to prorote the economic arnd social develOpment of tie regions interested in producing textile fibers. The roles assiged to ONAFITEX are to: l. 2. provide technical assistance to Zairean cotton farmers assure collection, transport, ginning and sale of cotton and its by-products, belonging to Zairean farmers assure financing of cotton marketing organize cotton farmers into cooperative societies in order to gradually turn the responsibility for processing and commerci- ali zation over to them 112 5. acquire financial participation in colpanies which have a similar, allied or conplerentary role to that of ONAFITEX 6. conclude conventions with cotton companies in order to take over the processing of cotton or of its by-products 7. create a price stabilization fund for the purchase of seed cotton from farmers 8. control fiber quality and its conditioning for export 9. certify quality standards and origin for export 10. carry out any industrial, comercial, financial, real estate and other operations related to its role. Cotton Marketing and Processing in Northern Zaire Grading and Collection of Cotton Seed cotton is graded by tie farmer in two categories: first and second class. There is a significant price differential between tiese grades (see Table 3.6). About 90 percent of the Zairean seed cotton production is first quality. Second class is any seed cotton that is not white light creamy or contains impurities. It is recormended that farmers sort cotton while picking into two classes. However, many famers sort only at hone after the cotton is sun dried. Following the sun drying of seed cotton, it is then packed in large hand woven baskets whidn are stored in a specially made wooden shed. The baskets are then carried to the local collection points at the time of sale. They are emptied in large standardized bags or weighed in the original baskets. At this time, the buying agent inspects the quality of the produce. After weighing with a roran balance, the farmer receives his money and a receipt. Seed cotton is then transported by truck to the ginning mill. 113 ONAFITEX first buys up first grade seed cotton. The second buying period usually occurs one to three months later and both first and swond quality cotton are purchased. Cotton seeds for the next season are distributed to tie farmers at the same time. In 1973, considerable time was lost in most collection centers because of a shortage of standardized bags. In the collection centers in Lbangi cotton is weighed in its original hand woven baskets. Qn tie average, only one truckload of 5 tons is received per day and per reception team. The C.F.D.T. mission in Ubangi reported that in Crnad, Dahoney and Upper Volta a collection team conld receive up to 120 tons per day, using a canvas spread ont on a large scale on which He famers dumped their cotton crop, up to 80 kg. of seed cotton per canvas. Several canvasses can be put on the scale at one time, thus weighing tl'e farmer's total cotton crop. In Zaire each basket or bag is weighed separately and a sales receipt is issued after each weighing. Sore farmers have several baskets and thus receive as many receipts. In 1973 33 collection teams Operated in Ubangi. In the northern cotton belt about 60 percent of the cotton is strict middling to good middling, average fiber 1 l/16 inches, white light creamy; 30 percent is strict middling, 1 1/32 inches white off- white to white light creamy; and 10 percent is middling to full middling, 1 1/32 inches, creamy dull to off—white grey. Cotton is baled in the ginning plants by a hydraulic press which cotpacts tl'e bales with pressures up to 320 kg. per cmz. This assures a high weight to volume ratio and adequate protection from fire hazard. Each bale weighs 100 kg. and is packaged in a jute or synthetic fiber 114 cloth with steel or nylon bands. Cotton bales are classified in the ginning mills in three qualities: superior, E and BA. Each category can be further divided into subcategories such as superior , average , inferior, etc. The comercial department of ONAFITEX (previously COVEBCO) makes a final grading of the cotton bales according to universal grade standards and on the basis of a sample taken from each bale.11 Nmber, Size and location of Ginning Plants Nbst of the presently used gins were installed between 1945 and 1960, except those equipped with 60 saw gins which date back to before World War II. All gins are from Contirental Gin Company, Prattville, Alabama (The Pratt Gin), with saws and usually pneumatic feeding.12 The gin presses are all "Velghe" with one or two bins, with mechanic or steam Operated rams. In 1970 only 57 of the 115 gins were in operation as shown in Table 4.3. In 1974 67 ginning plants were in operation in Zaire, 11 in the North-West, 21 in the North-East, 33 in the South and 2 in the East and Uvira. The number of ginneries in Zaire is relatively high and tie scale of Operation is low because of high transport costs for seed cotton due to the poor condition of feeder roads . And the higher the transport costs, tie more gins are needed and the smaller the size of gins for Optimal economic efficiency. uFiber quality is determined by the grade (color, impurities and preparation), staple length and character (resistance and stretching as measured by a Pressley dyramoreter) and fiber diameter as measured by the porosity and expressed in the micronary. 12Only the gin in Mahagi is of the roller type. "115 Table 4.3. Republic of Zaire: Situation of the Ginning Plants on.September 1, 19701 Corpany Ginning Plants Ginning Plants in Operation Closed Down (DRINK!) 37 40 BEIGIKA l 9 NAHV 6 0 WA 5 2 RIDGE 3 2 CCMBEIEA 2 1 COM 2 2 Cie du LUBILASH l 0 SOC. Cot. de la LUISA 0 1 3mm 0 1 (being reopened) Total 57 57 + l ‘ 1 COGERCO, Kinshasa. Economies of size are important in cotton ginning [Looey and Wilmot, 1971] . Ginneries with one 80-saw gin erploy a crew of 10 workers while 14 are employed in two 80-saw gins arnd 17 in three 80-saw gins. However, labor is not an important cost elerent in ginning in Zaire as workers earn only between 30 and 50 Makuta per day (0.60 cents to 1.00 dollar per day). On the average, a crew of 14 workers produces 28 bales of cotton lint (2,800 kg.) per shift of 8 hours with two 80-saw gins. Most of the ginneries presently in Operation are in fairly good working condition. Equiprent-wise, most ginning mills currently have delinters to separate linters from seeds, although these delinters are not being used because of tire low export prices for linters. 116 In 1973 ONAFITEX received a special investment budget from the presidency and ordered 30 new, modern, high capacity gins from the U.S.A. Several of these are already being installed in Zaire. They are HARWICK—E'ITER gins, with 120 saws and a capacity of 7 bales (700 kg.) of seed cotton per hour, about double the capacity Of a 90-saw Contin- ental gin. These new gins cost about $30,000 each. There are indications however, that they are not well adapted to tropical African conditions as they are elctronically operated and require the drying of seed cotton. The purchase of new gins seens unwarranted because ginning has not been a bottleneck in cotton marketing. Althougln the high cost of ginning depleted COGERCI) reserves ,this was mainly due to high transport costs and expatriate salaries. The ginneries still belong to the private cotton conpanies altlnough they have been requisitioned by the government for a remuneration of 4 8 per ton of cotton lint.13 Thus, the policy of ONAFITEX is in line with the "economic independence" policies proclaimed by the government. Cotton Marketing Costs Table 4. 4 presents cotton production and marketing costs for 1957, 1958, 1959 and 1969, 1970, 1971. Total costs per kg. of cotton fiber are comparable between the two time periods listed. However, export prices F.O.B. Matadi were corsiderably lower in 1969-1971 and, as a result, COCERCO incurred losses. In 1973 and 1974 the export prices were more than (buble the 1969-1971 level. 13For a production of 20,000 tons of cotton lint per year, this amounts to 80,000 8 for the cotton companies. This is only a small price to pay in conparison with the cost of acquiring new ginning equiprent. llf7 .ucmouoo H.mm no menu oceccfio_n tom anaemco osoomm ucmouoo 0N one huflamco umuem ucwouwo om ocaeowmd oeeesedm .>m>bsm uoeomm Hebeeesoesmse .muesooom oczmsoo teameeneo CH czocm mm meuew ma mo uco 0H mo mumoo ocflmmooono Hmcuoc mo oomum>m nounoem3 a mum mumoo N wmmcom .oamm emcee ocm oneomuo .mocmnsmce .mmamn couuoo mo oomuoum ocm unoowcmuu mmmfluosoo m moose “mumucmo coHuuoHHoo cu mmfluoccflo_:ouw women copuoo mo buoomcmuu com mucwoo mhm>eaoc oozm>ou Q» mwamo mo uuoomcmuu ocm ocaamn .ocflccflo mooDHOCH N mmmno “mucmHo 0553 8 uncomcmuu tom 39me .8300 town mo ongouco 05 mwmfluozbo a ommco .953 £0.58 on... tam oommooo consumn cofiucw>coo Nome was cu xomn mount m ocm N .H ommco Ce mumoo ocauoxume mo cOHmH>eo tree .ooemcoo be emuueensm mundane teeuoo Heme ease teemeeumm m N .HH .o .ucmeuumomo muomnuoo amusuadofluoe ..U.o .:oooCflnmmz :.muamw wo .NhoH medh ..¢.D.HI.Q.m.m.H so waCQ .v OHDMB .N OEDHO> CO memm a vv.mN mm.VN om.mN Ho.mN mm.om vv.vm flcmumz .m.o.m moeuo .v mn.~m mm.am oo.mm Ho.mN Hm.oN mo.am eomumz um umoo Hence .m Hm.ma mm.ma mo.HN oo.ma Ho.ma MN.5H announsm om.H om.H 5H.N III In: in: oozm>OOln m ommno mn.a mn.a mo.a in: II: in: mono mm.oH vv.HH em.oa Inn nun uni mumoo meufim coppoonn mN wmmco III II: In: OH.N Ho.N mN.m meuem coauOOIn vm.m no.v vm.m om.HH oo.ma mm.ma ooomoooli we mmmno mumoo meeumxumz .N hm.HH hm.aa hm.HH om.va om.va mv.va mmumeumm 0» muCME>mo .H .ox\x .mx\x .ox\x .ox\mocmum .ox\mocmum .ox\mocmum NHBmH onma meme mmma mmma nmma Home Humanoooa new omoanhmoa .mumOU oceumxumz one COHuOSoOum :OquU alumnae couuoo mo .mx “whens mo Ueanomo "v.v manna 118 Marketing costs appear to be high. Cotton firms had no incentive to process cotton cheaply nor to erplcy their equipment at capacity as all costs were reimbursed and their fees were in proportion to their average total production cost. The percentage distribution of costs incurred in cotton marketing for 1969 is shown in Table 4.5. Total cotton marketing and processing costs were composed of 34.63 percent fixed costs and 65.37 percent variable oasts . Fixed costs are irndependent of the volune of production. By increasing processing volume per unit fixed costs decrease. Thus, one way of reducing marketing and processing costs is by increasing cotton production or vice-versa; one of the main reasons why cotton marketing costs in Zaire are high is because tl'e processing volume is low relative to ginning capacity, staff, personnel and other fixed elements. A further breakdown of cotton marketing and processing costs for 1969 is given in Tables 4.6 and 4.7. It appears from Table 4.7 that the cost of ginning firms represented 37. 46 percent of total marketing and processing costs, more tlan the share attributed to cotton farmers, 33.54 percent. Salaries for personnel were high which is illustrated more clearly in Table 4.8 made up for 1971. Salaries totaled 37 percent of marketing and processing costs Of cotton ginning firms in Phase 1 and 2. Expatriate salaries represented 18 percent of total costs, a cost eletent which ONAFITEX reduced considerably. As appears from Table 4.6 financing d’arges were also a substantial cost in cotton marketing because COGERCO had to borrow all of its funds from private banks for the financing of each cotton campaign. 119 Table 4.5. Republic of Zaire: Cotton Marketing and Processing Costs, 19691 Sources of Marketing Costs Costs Zaires Per Ton Percentage of Total Marketing Costs CDGERCO Costs: 11.9 percent fixed costs and 88.1 percent variable costs for managetent, financing, reception and transport of seed cotton from local assembly points to the ginning mills, called Phase 1 costs. Costs Cotton Companies : Costs cotton corpanies consists of 70 percent fixed costs and 30 percent variable costs. Expatriate personnel account for a large part of the fixed costs. Costs for ginning, baling, storage, trans- port Of cotton bales to public trans- porters and packing and transport of cotton seed to the local assembly points plus remuneration of capital, depreciation, financing charges and a managerent fee constitute tie costs of the cotton conpanies, called Phase 2 costs. COIENCO Costs: 9.33 percent fixed costs and 90.67 percent variable costs for managerent, insurance, storage, sales cost and transport of cotton bales to their final destination. This con- stitutes Phase III costs. Cotton Farmers: 100 percent variable costs . Advance payment for tie seed cotton crop plus payments in kind (tools , insecticides, etc.) . Total 76.83 119 . 39 15. 50 106.94 24.14 37.46 4.86 33.54 318. 66 100 . 00 lNelis and Wieers [1969]. 120 Table 4.6. Republic of Zaire: Breakdown of Cotton Marketing and Processing Costs, 19691 fi/Ton B/Ton B/Ton Cotton Fiber Cotton Fiber Cotton Fiber 1. Phase 1: Reception and Transport of Seed Cotton from local Assembly Points to the Ginneries. 1. 1 Fixed Costs Salaries of Personnel 13.37 Travel Expenses 3. 82 Other Fixed Costs 2.88 20 .07 1.2 Variable Costs Purchase of Seed Cotton 104.40 Reception of Seed Cotton 5.20 Transport of Seed Cotton 38.23 147.83 1. 3 Financing Costs 11. 67 COGERCO Costs: 179.57 2. Phase 2: Ginning, Baling and Transport of Cotton Bales from Ginnery to Public Transporters; Bag- ging and Transport of Cotton Seeds from Ginnery to local Asserbly Points 2. 1 Fixed Costs Salaries of Personnel 40.26 Travel Expenses 5.73 Purchase and transport of Supplies 9.69 Bagging and Transport of Cotton Seed 5.18 Other Fixed Costs 10. 32 71.18 2.2 Variable Costs Ginning 5.00 Baling 7.00 Transport of bales 18.30 Other Costs 0.11 30.41 2.3 Depreciation (2 percent of export price F.O.B. Matadi): 6.36 2.4 Capital remuneration (2 percent of export price F.O.B. Matadi): 6.36 2.5 Management Fee (5 percent of CG): 5.08 Cost Cotton Corpanies (CG): 119. 39 3. Phase 3: COVEINKI) Costs: Transport Of Cotton Bales to Textile Factories or Port of Export; Storage; Insurance 3.1 Variable Costs 13.34 13.84 4. Management Costs: 4.1 CCXERCO 1.66 4.2 (INENCD 1.66 3.32 5. Cotton Entrension Costs: 2.54 2.54 Total Costs per Ton of Cotton Fiber 318.66% lNeils and Wieers, [1969]. 121 Table 4. 7. Republic of Zaire: Cotton Marketing and Processing Costs for Each of the Institutions Involved and for the Cotton Fanrers, 1969l Source of Costs In Z/T'On of Cotton Fiber Percentage of Total Cotton Marketing Costs C(XZERCO Costs 76.83 24.14 Cost Cotton Cotpanies 119.39 37.46 (DVEINCO Costs 15.50 4.86 Cost Cotton Farmers 106.94 33.54 Total Costs 318.66 100.00 1Derived from Table 4.6. Table 4.8. Cost Structure of Cotton Ginning Firms in Zaire in 1971 (In Thousand Zaires)1 Type of Cost Cotton Cotton Gin- Total Percentage Assenbly ning, Baling Costs of Total (Phase 1) and Trans- Costs port (Phase 2) Fixed Costs of which 473.3: 1,610.0 2,083.3 60 Salaries for Expatriates 227. 9 424. 3 652 . 2 18 Salaries for locals 109.9 557.0 666.9 19 Variable Costs 764.0 592.6 1,356.6 40 of whidn Transport 649.4 289.9 948.3 27 Tbtal Costs ,237.3 2,202.6 3,439.9 100 Note: Fixed costs cotprise: wages, salaries, travel costs, equipment purcl'asee, utilities, office supplies, transport of cotton seed; vari- able costs corprise: transport of seed cotton and cotton fiber handling and ginning costs, insurance, etc. lI.B.R.D.-I.D.A., June 1972. "Agricultural Sector Survey, Republic of Zaire." Washington, D.C., Agricultural Projects Department, Vol. II, Annex 6, p. 13. 122 Salaries and transport costs together made up about 64 percent of total cotton marketing and processing costs in 1971 for Phase 1 and 2 (Table 4. 8). This is usually high. The cotton corpanies were also engaged in comercial activities , plantation crops , ranching , etc. in aidition to their cotton Operations . Undoubtedly , a large proportion of personnel, especially expatriates, were assigned to noncotton departments and the transport facilities in the dead season were utilized for the comercial activities, but C(IiERCO paid for all these operations. In Zaire in 1969, according to Table 4.6 and assuming that the final value of the cotton crop was equal to total cotton marketing and processing costs, 318.66 5 per ton of cotton lint, the farmer's share was equal to 104.40 8 plus 2.54 B for cotton extension or 33.54 percent.14 Transport costs represented 79.65 8 per ton of cotton fiber in 1969 or about 25 percent of total marketing and processing costs, including the purchase of seed cotton. A breakdown of transport cost is given in Table 4.9. Transport costs as a percentage of total cotton marketing costs were much higher for the Uélés (31 percent) for Ubangi and Mongala (21 percent); transport distances are much longer in the 14In fact, the average sales price of cOtton lint, dorestic and for export in 1969, was eqnal to 308 8 per ton (Table 4.15) or less than total cotton marketing costs. In Nigeria farmers realized 65 percent of the final value of their cotton crop, F.O.B. Iagos or local textile mills. The total costs for processing and marketing raw seed cotton and cotton seed were on the order of ore—third of total farmer receipts [Kriesel, 1968] . 123 Uéles became cotton production is spread out over a much longer area. 15 Table 4.9 Republic of Zaire: Average Transport Costs in Cotton Marketing, 1969 (In Zaires per Ton of Cotton Fiber and as a Percentage Of Total Cotton Marketing and Processing Costs, Including the Purchase Value of Seed Cotton) 1 Source of Costs Z/‘I‘on of Percentage of Total Cotton Fiber Cotton Marketing Costs Phase 1:38.23 + 2.66 (financing) = 40.89 12.8 Phase 2: 18.30 + 4.24 (cottonseeds) + 2.38 (Supplies) = 24.92 7.8 Phase 3: 13.84 4.4 Total 79.65 25.0 lDrived from Table 4.6. T'te regions with the highest transport costs per ton of cotton fiber actually produce the most cotton. This is particularly true for the Uélés whidn produce about one-third of the total seed cotton. Trans- port costs are very high because of the poor roads which account for M“u~~w-l_fi_-_ "n1/ high transport rates and the lorng distances involved in moving seed \ "”M “ * - --. ...—.....— cotton and cotton fiber. Trucks Often travel long distances to pick up small loads, and most of their runs are one—way pay-loads. Vehicle 15cm, 1970. "Prix de Revient Industriel 1968/1969." Kinshasa: unpublished docurent. 124 life is usually short; amortization is orer three years . And most trucks have only a five ton capacity. The transport rates for seed cottonrange from8tol4K. pertonandperlqm., ontheaveragelZK. in 1969. Rates for cotton fiber are usually lower, on the average 9 K. per ton and per km. in 1969. The distance for seed cotton transport to the ginneries is on the average 106 km., implying a round trip journey of 212 km. -[Nelis and Wieers, 1969]. At 12 K. per ton and per km., this amounts to 1,272 K. per ton or 1.3 K. per kg. of seed cotton or about 3.8 K. per kg. of cotton fiber. The trucks also distribute cotton seeds for the . 16 next cenpaign. In 1972 the World Bank indicated that road and rail transport rates inZairewereveryhigh, muchhigherthanthose inotherAfrican comtries [I.B.R.D.-I.D.A., 1972]. By’abandoning marginal production , ......— ..—-. M..- w—— areas, overall transport costs could be reduced considerably. Another M possibility Eflflllm. producer prices in marginal or isolated areas to compensate for high transport costs from these areas. -..—1....“ —.- ...“. Cotton fiber transported by (NAT‘RA or VICIZAIRE is charged at the class B, luxury goods rate. In fact, the calculation of transport fares on public or semi—public transport facilities is corplicated. The tariff sturcture has remained largely unchanged for over 30 years . The government controls transport fares and each corpany has exclnsive transport rights in its territory [Huybrechts, 1970] . 16Wieers [1969] estimated that 11,000 tons of cotton seeds are distributed each year to the cotton farmers. Tteoretically, this would be sufficient for a seed cotton production of 110,000 tons, far above present production levels . 125 Transport fares change according to the value of the product , type of product, distance, dorestic price relative to world price, destination of the product (<1de use or export), inflation and other factors. ONATRA calculates the fare as follows: F=M(B+T) (8+1) +H Where: F = total transport fare M = moretary coefficient B = base rate T = transit cost S = sliding scale H = handling cost. The base rate B is the basic cost for transporting a good over a given distance. This rate is expressed in Congolese francs and has retained unchanged since independence. The base rate B also differs according to the classification of the product. For agricultural comoditiss, products are classified from A to E according to their unit value. Class A comprises the goods with lowest value per kg. such as maize, manioc, wood. Class E contains those with highest value per km. such as cotton and coffee, the rate here being about twice that of Class A. Further- more, the base rate declines on a ton-kiloreter basis as distance increases. In general, the 600th km. costs only half as munch as the first one. Thus, rerote or isolatedregions benefit from lower ton-km. rates to facilitate export of their products. For agricultural comodities consumed in the country, the base 126 rate is calculated for the entire distance from origin to final destin- ation. If omnodities are further processed at another location, as is the case for cotton fiber used in dorestic textile factories, the base rate is taken from the origin to a point of transfer to the processing firm, and then recalculated from this point to the processing factory. Thisbreaksuptl'edistanceintwopartsandthus increases theton-. kiloreter transport rate. Transit costs T refer to charges for reloading goods when the mode of transportation changes , for instance from rail to river barge. Handl- ing costs H are charges for loading and unloading the goods at origin and destination. Tie sliding scale S provides a method of adjusting transport costs to monthly changes in world price levels of agricultural comodi- ties. If the world price of a comrodity rises above its July 1, 1960 level, then the transport fare for domestically traded goods rises as the world price of the comrodity rises. This isolates the farmers sore— what from increases in world market prices for products that are eported. The monetary coefficient M compensates for inflation since 1960. B, T and S are expressed in 1960 francs. Successive devaluations have reduced the franc to one-tenth of its value and, in 1967, the monetary unit was changed to Zaire and Makuta with one Makuta (called Likuta) equal to 10 franes. A monetary coefficient of 10 is applied for agri- cultural comrodities to be exported, 6 for agricultural comrodities used for industrial processes and 4 for agricultural conmodities to be consured locally. Thus, the rate structure favors dorestically produced 127 arnd consuned products and is biased against local processing and exports. In fact, products that are exported are charged a much higher trans- portation rate than for dorestic use . For example , cotton bales for export are charged more tlan twice the rate of bales used in dorestic textile factories. Thus, exports are penalized relative to locally consuted goods [Wieers, 1969] . The fare bears little relation to the actual cost of transport. Cotton, classified in Class E, pays well above real transport costs and corpensates for losses incurred on the transport of Class A and B comodities . This is particularly important for rail transport as rail fares are on the average three times higher than river transport fares [Van de Velde, 1974]. The Uélés produce about ore-third of the Zairean cotton crop. And because of its retote location, transport and handling costs are among the highest in Zaire. Cotton produced in the Uélés is brought to the local assenbly points in large baskets, carried on the head. From time it is transported in bulk with trucks to the ginning mills. Cotton bales are then transported by truck from the ginning mills to the Mungbere- Bumba railroad. In Bumba tie bales are trans-shipped from railcars to river barges for transport to Kinshasa. If the bales are for export, they are again put on railcars in Kinshasa for transport to the port of Matadi. This description of transportation chain illustrates wlny trans- port and handling costs are high for cotton produced in the Uélés . The Opening of a large textile factory in Kisangani will considerably improve the economic efficiency of cotton marketing in the Uéle’s. ONAFITEX does not publish its balance sheets nor cost accomts 128 and, thus, marketing costs from 1971 on are not known. However, the director general of ONAFITEX reported that there have been major 17 savings in cotton marketing since the establishment of ONAFITEX: -a consicbrable reduction in labor costs, mainly through lay-Offs of expatriate personnel who formerly worked for the cotton companies. Ninety percent of the expatriate staff and techni- 18 This also included the elimination of cians were fired. of representation costs of the cotton conpanies in Kinshasa and in Europe of over 110,000 8 in 1970. «reduction of transport costs from 14 K. to 8-12 K. per ton and km. -financing costs were reduced by repayment of outstanding loans for the buying up of seed cotton, cotton ginning and transport after nine months instead of twelve. -reduction of baling costs from 13 K. per bale to 6 K. The establishment of ONAFITEX, however, also created sore new problers. For example, the buying period of seed cotton in 1972 and in 1973 was extended over almost the whole year instead of the usual three to five months. The main problers were (a) the inability of the ONAFITEX office in Kinshasa to adequately supply funds to its regional branches, (b) seed cotton production projections were underestimated, 19 (c) poor transport facilities and (d) difficulties with the transfer Of facilities from the cotton firms to ONAFIT'EX. l.IRenrarks by the director general of ONAFITEX at a 19 73 cotton conference at the National University of Zaire, Kinshasa campus. 18Savings in the order of 650,000 8 per year were reported. 1911331: of the vehicles used by Vici-Zaire in the Uélés for trans- Port of seed cotton were acquired before 1960. / \ / -- . O 129 Since chueucclalection of seed cotton was spread out over the year, trucks arnd vehicles were in Operation fora full year. Sore trucks /’w_e£e making frequent runs of more than 100 km. for a payload of one- unw— -"'-Iv~n.-nq “’4‘- ‘ 20 third or one-fourth of. load capacity in one-way. Thus, transport and ww*aw—Mnu ...... -. — ' salary costs were excessive. The Evaluation of the Textile Industry The first textile company in Zaire (UTEXCO) was created in 1925, and in 1928 the first textile mill went into operation. Three other conpanies Opened a textile mill during or shortly after World War II. Presently, four cotpanies spin and weave cotton in Zaire and a fifth mill (SOTEXKI) is starting production. Most of the textile mills also dye arnd print a large part of their production, and several have clothing manufacturing shops attacked to the mills. Since 1950 two corpanies also spin and weave urea and punga fibers, mainly for bags used for the export of coffee, groundnuts, palm kernels, and for carpets, strings, tarpaulins, etc. Blankets are produced by two corpanies. Pres ertly, the Zairean textile industry is well developed and consumes abont 16,000 tons of cotton fiber or three-fourth of the domestic cotton production. 21 In 1966 Lacroix [1967] estimated that 80 percent of the Zairean textile imports could still be replaced by omestic production and thatcotton fiber consumption would then increase to about 30,000 tons per year, more than the present production. Since 20(he regional director made it a point of honor to buy up every quintal of seed cotton produced in his region, whatever the cost of l'auling it. 21In Nigeria the first modern textile mill began operation in 1957. During 1968 it was estimated that the dorestic use of cotton was equal to the dorestic output of lint [Kriesel, 1968] . 130 1972 a British company (CPA) produces "wax" textiles, 16 million square meters in 1973, which are in great derand in Zaire. In 1972 2.3 million square meters of textiles were produced from imported synthetic fibers. The major textile mills in Zaire are located in Kinshasa, with other mills in Lubumbashi, Kalemie and Kananga. Tables 4.10, 4.11 and 4. 12 provide statistics on the Zairean textile industry. Of particular interest to the northern cotton belt is the opening of a large mill in Kisangani in 1974. The mill will initially cornsnme about 3,500 tons of cotton fiber per year from the Uélés and will pro- duce about 18.5 million square meters of textiles annually. The mill has 456 loons, 24,000 spindles and will enploy 1,100 workers. Table 4.10. Production Of Cotton Textiles in Zaire, 1968-1972 (In Thousand Square Meters) 1 Year Natural Colored and Bleached Source of Material Textiles for for Printing Direct Printing Total local Imports Total Sale Production 1968 26,887 28,865 55,572 28,865 7,632 36,497 .1970 35, 078 34,049 69, 127 34, 049 9 ,968 44 ,017 1971 38,733 36,789 75,522 36,789 15,265 52,054 1972 35,369 37,546 72 ,915 37,546 21,437 58,983 J‘République du Zaire, Novembre 1973. Conjoncture Economimlew Année 1972 et Ier Trimestre 1973. Nationale, Nr. 13, p. 245. Kinshasa : MinistEre de l'Economie 131 Table 4.11. Equipment in the Zairean' Textile Industry, 19731 Oorpany Number of Spirndles Nurber of Icons Utexco 70,084 1,382 Filtisaf 16 , 660 360 Amato 10,000 196 Solbena 10,000 198 Sotexki 24,000 456 1Republique du Zaire, 99. cit., 246-250. Table 4. 12. Republic of Zaire: Dotestic Sales of Cotton Fiber to Textile Conpani es , 1970-1974 Conpany 1970 1971 1973 19742 Utexco 8,400 8,400 8,350 9,000 Filtisaf 1,700 1,700 3,000 3,000 Amato 1,200 1,200 1,200 1,200 Solbena 1,800 1,800 1,800 2,000 Zaitex 200 200 500 600 Sotexki -- -- --— 1,200 TOtal 13,300 13,300 14,850 17,000 lOINIAFIT‘EX, Kinshasa, Comercial Department. 2Estimated, based on current contracts Note: Data for 1972 per corpany were not available; total dorestic sales were 11,649 tons. 132 Sale of Cotton Lint About three-fourths Of Zairean cotton production is used in domestic mills. The rest is exported, mainly to the comm market countries of Europe (Table 4.13). Cotton fibers are the principal textile fibers exported in Zaire (Table 4.14) . Since 1962 Zairean cotton lint has been sold to dorestic spinners at a higher price than for export. This premium price was designed to cover COGERCI)'s deficits. In 1965, 1966, 1967 and 1968 Zaire had to import cotton lint from the U.S.A. at a price of 369 8 per ton to sat- isfy oxtestic needs. Zaire started exporting cotton again in 1968. However, from 1968 until ONAFITEX took over the Operations of COGEKI) and COVEICO in 1972, the average price received from the sales of cotton fiber, domestic and export, was below the average total cost of production, estimated at 320 8 per ton in 1970 (Table 4.4). To counter these losses, several measures were taken. On Decenber 14, 1.968 exports of cotton fiber were exonerated from export duties (3 percent) and from the turnover tax of 7.50 percent.22 On Novenber 15, 1969 the Ministry of Natioral Economy decided to increase the price of cotton lint sold to dorestic mills by 25 percent in order to safeguard CDGERCO reserves and to support local cotton production. Thus , local textile mills were penalized in order to subsidize exports. In fact, the Ministry Of National Economy fixed the sale price of cotton lint to domestic textile factories on the basis of the anticipated export 22Tte government still collects sore minor taxes on exports: -statistical tax: 1 percent --se1ection tax: 1.60 K. per 100 kg. --cotton tax: 0.60 K. per 100 kg. Table 4. 13. Exports of Zairean Cotton Fiber Per Country of Destination, 1970-1973 (In Tons)l Destination 1970 1971 1972 1973 Belgium 1,843 3,511 2,440 3,015 France 1,082 -- 61 377 W. Germany 2,075 253 366 356 Netherlands 1 , 302 677 1 , 024 754 United Kingdom 824 881 191 371 Italy 80 26 122 771 Japan 1,004 658 261 633 Bong-Kong 645 50 -- 546 Total Exports 8,855 6,056 4,465 6,823 lONAFITEX, Kinshasa, Comercial Department. Table 4.14. Exports of the Principal Textile Fibers in Zaire, 1959-1971 (In nbns)1 Fibers 1959 1965 1970 1971 Cotton Fibers 52,790 89 8,806 5,429 ”’9’“ Fibers 2'9“ 3,220 3,153 461 Punga Fibers 1,193 Sisal Fibers 100 166 . . . . . . 1nainistere de l'Agriculture, 1968, 2. _c_i__t.,, Ministére de l'Agriculture et du Développenent Rural, Avril 1970, Q. cit.,; République du Zaire, Novelbre 1973, g. cit. 134 price F.O.B. Matadi after deducting all duties and taxes. The 25 per- cent surcharge was added to this so-called "export parity price", and the theoretical transport costs from Matadi to the local textile indus- tries were deducted to arrive at the sale price to the local textile factories. The average prices of cotton fiber paid by donestic textile factories are given in Table 4. 15. These prices remained remarkably stable over the 1964-1973 decade. The average export prices F.O.B. Matadi and the overall average sales price of cotton fiber for dorestic use and for exports are also shown in Table 4.15. The purchase price of cotton lint from the farmers as a percentage of the export price F.O.B. Matadi over the 1950-1973 period is presented in Table 4.16. This ratio fluctuated widely over the years . When the world market price dropped, the farmers ' share in the export price increased and vice-versa which illustrates the role played by the price stabilization fund. For 1974 the percentage again declined because of rising export prices for cotton lint. In 1973 and 1974 world market prices for cotton lint reached unprecedented heights (see Tables 4.17 and 4.18) . However, dorestic prices remained relatively low (see Table 4. 18) . Thus, the policies from 1968 to 1972 whereby the Zairean consumer supported tlre local textile industry arnd local cotton production have been reversed. local fiber prices are now kept low to protect the Zairean textile industry and the purdasing power of Zairean consuters. Since January 1, 1974 sale prices for donestic textile mills are fixed ex-cotton ginnery and thus, transport costs from ginning mulls to local textile factories are paid by the textile corpany. The 135 Table 4. 15. Republic of Zaire: Average Sales Price of Cotton Fiber for Dorestic Sales and for Exports,11964- 1973 (Price in K. per kg. of Cotton Lint) Year Domestic Sales Export Price, weighted.Average Price in K. F.O..B.Matadi2 Sales Price in Per kg. in K. per Kg. K. per kg. 1964 34.96 . . . . . 1965 34.22 . . . . . . 1966 34.76 . . . 34.76 1967 22.54 . . . 22.54 1968 32.60 . . 32.60 1969 34.81 25.50 30.80 1970 34.81 24.97 29.76 1971 34.14 29.44 32.62 1972 34.15 33.48 33.97 1973 34.99 36.70 35.52 1ONAFITEX, Commercial Department, Kinshasa. 2 This price includes export duties; in 1957, 1958 and 1959, the price was respectively 34.44, 30.93 and 26.61 K. assum- ing 1 Belgian franc = 1 Likuta. Charges fOr handling, transport and storage are about 6 K. per kg. of cotton fiber. Previously these charges were supported.by ONAFITEX. Looking at the domestic market for locally made textiles, one notes that the prospects for expansion are bright. Houyoux [1973] undertook a cross-sectional survey of 1,471 Kinshasa families in 1969 and found that on the average clothing represented 7.3 percent of total expenses and had an incore elasticity of 2.33 (see Table 4.19) . low incote groups spent only 1.7 percent of Hair meager incone on clothing and 79.9 percent on food, while high incone groups spent 9.6 percent of their incore on clothing and 56.8 percent on food (see Table 4.20). 136 Table 4.16. Republic of Zaire: Purchase Price of Cotton Lint fran Farmers in Percentage of the Export Price, F.O;B. Matadi, 1950-1973 Year Average Purchase Average Purdhase Average Export Purdhase Price Price of Seed Price of Cotton Price, F.O.B. in Percentage Cotton2 in K. Fiber in K. per Matadi, of of Export per kg. (1) kg.3 (II)=(I)x3 Cotton Lint Price (IV) 1:33:11) 1“ 1" Per IV;(II) x 100 ° (III) 1950 5.35 16.05 40.28 39.85 1951 5.35 16.05 52.14 30.78 1952 5.80 17.40 47.34 36.76 1953 5.35 16.05 37.22 43.12 1954 5.36 16.09 39.07 41.08 1955 5.35 16.05 39.70 40.43 1956 5.35 16.05 34.09 47.08 1957 5.35 16.05 34.44 46.60 1958 5.35 16.05 30.93 51.89 1959 5.80 17.40 26.61 65.39 1960 5.80 17.40 28.56 60.92 1961 5.80 17.40 30.62 56.83 1962 4.44 13.32 32.84 40.74 1963 6.00 18.00 39.40 45.69 1964 4.83 14.50 33.24 43.62 1965 4.83 14.50 35.22 41.17 1966 4.83 14.50 35.77 40.54 1967 3.22 9.67 24.90 38.84 1968 3.48 10.44 36.27 28.78 1969 3.48 10.44 25.50 33.68 1970 4.35 13.05 24.97 52.26 1971 4.35 13.05 29.44 44.33 1972 5.85 17.55 33.48 52.42 1973 5.85 17.55 36.70 47.82 1 CDGEROO, unpublished document dated 9/8/1969 and own calculations. 2 Assuning 90 percent of seed cotton is first quality and 10 percent is second.quality. 3 Assuning 3 kg. of seed cotton yields 1 kg. of cotton lint. Note: conversion rates until 1960- 1 franc - 1 K; fron11960 to 1967 conversion at the official exchange rates assuming 1 USAS = 50 Belgian francs = 50 K. 137 Table 4. 17. Average Cotton Fiber Price in New York 1959-1973 (In U.S. cents/1b. and K./kg.)l Year Fiber Price in Fiber Price in U.S. cts/lb. K/kg. 1959 34.57* 38.11 1963 38.83* 42.80 1964 32.44* 35.76 1965 30.37* 33.48 1966 24.98** 27.54 1967 23.28** 25.66 1968 27.65** 30.48 1969 25.41** 28.01 1970 25.69** 28.32 1971 29.67** 32.71 1972 35.12** 38.72 1973: lst Semester 44.37** 48.91 2nd Semester 76.41** 84.23 e du Zaire, Rapport.Annue1, 1967; 1969/1970; 1972/1973 and Bulletin Trimestriel, 4 éme Tri- mestre 19 73 , Kinshasa. * C.I.F. New! York ** Cbntract Nr. 2 138 Table 4.18. Republic of Zaire: Comparison Between July 1974 Cotton Fiber Prices for Export and for Danestic Use (In K. per kg.) Type of Cotton3 Average Staple Export Price, Danestic Length no.3. Matadi Price2 in in K. per kg. K. per kg. Reba B50 (M to full GM 1 1/16" 81 45 Reb 1er SM to full SM 1 1/16" 80 44 810 1er Special SM to Shy 94 1 1/32" 79 44 Rebel M to full M 1 1/32" 76 42 l ONAFITEX, Kinshasa, Ccnmercial Department. 211113 price includes 6 K/kg. for transport, handling and storage from the ginnery to the textile factory. 3 Cotton fibers from the northern cotton belt. Table 4.19. Inccm'e Elasticities in Kinshasa, 1969 l Expenses for In Percentage Inccme Elasticity of Total Expenses 1. Food 67.4 0.78 2. Housing 14.9 1.21 3. Miscellaneous 10 . 4 1. 53 4. Clothing 7.3 2.33 1I-louyoux, J. , 1973. Bugets Ménagers, Nutrition et Mode de Vie a Kinshasa. Kinshasa: Presses Universitaires an Zaire, p. 111. 139 oH-HH om sow-HIM 0mm sch. ‘5.»ng ooa N¢.Hm v.oa mm.m m.> om.~ m.¢a mm.v v.ho ma.am m.m Hanna ooH wo.mm ~.ma mo.ma m.m Hm.m ¢.o~ om.om m.mm m~.mm ¢.m m cm A ooa Hm.vv H.AH mm.¢ N.oa Nm.¢ m.mH me.m «.mm hm.m~ m.w m mm.mm I mm ooa m~.mm m.m an.m m.> vH.m o.ma mm.m v.Hh mm.om o.o m mm.vm I mm ooa mo.- >.m hm.a m.m am.o «.ma om.m m.vh m>.ma m.m m mm.vm I om ooa mw.ha o.n om.H h.m mv.o «.ma mm.~ H.>> mm.mH m.m m mm.ma I ma ooa mm.oa o.m mm.o h.H mH.o ¢.NH mm.a a.mu mn.m m.v m ma v m N m N m N w N m N mmmcwmxm mooocmaamomflz. mcflcuoao mcflmcoq coom you maflemm\mc0muwm meHmN ca Hope? you mmmcmmxm you mmmcwmxm you mmmcwmxm mmmcwmxm mo umnepz mEOOCH mmma .mmmnmcax CH meoooH mo coauocom cw mafismm mom monsuflocwmxm manned: mo muouoouum .oN.v magma H 140 It is also interesting to note that females spent nore on clothing than males in all income groups except the lowest. Consequently, when personal incones increase over time, expendi- tures cn clothing and, thus, the effective demand for textiles will increase more than twice as fast as income. Although this study was undertaken mly in Kinshasa, the results should be substantially the same for rural areas. Before the petroleum crisis in 1973 and 1974, it was expected that the world demand for cotton would expand mainly through the growth in population. Man-made fibers were lTDSt likely to benefit from increased per capita textile ccnsumption as a result of rising living standards. Cotton's share in world fiber use in 1968 was estimated at 57 percent and was predicted to drop to 47-48 percent in 1980 because of carpeti- tion fron man-made fibers [Magleby and Missiaen, 1971] . Because of the rapid increase in petroleun prices the price of synthetic fibers has increased significantly and, as a result cotton lint is now carpetitive with textile fibers. However, this also led to an upward pressure on world cotton prices. Since May 19 72 world market prices for cotton lint have been increased fron 36 U.S. cents per 1b. C.I.F. Liverpool to 87 U.S. cents in February 1973 for strict middling 1 1/16 inches and 84.5 U.S. cents on Decerrber 21, 1973.23 The dorestic textile factories in Zaire are using larger and larger quantities of cotton and, if cotton production stagnates, no fiber will be available for export in the near future. Since Zaire 23Cotton Outlook Supplement, C.I.F. Liverpool quotation for principal growths, Liverpool Cotton Services Ltd. , Liverpool. 141 is a very small cotton exporter in relation to total world cotton trade, it is reasonable to assure that no difficulties will be encountered in disposing of increasing cotton production. Moreover , Zaire is a price taker and a quantity adjuster in the world market for cotton. Cotton By-Products 'Ihe ginning process divides seed cotton into about one-third of the weight in cotton lint and two-thirds cotton seed. The 1973 utiliza— tion of seed cotton is presented in Figure 1. In 1959 the sales value of cotton lint in Zaire represented 90. 7 percent of total cotton receipts; cotton seed oil made up 4.1 percent and cotton meal and cake, 5.2 percent. Distribution of Cotton Seed to Farmers Of the seed production, 35 to 40 percent is returned to the farmers for planting in the next seasons by way of seed distribution to them during the purchase of second quality cotton. The seeds are not sorted according to weight and are not treated with an insecticide- fungicide. 24 For an average yield of 300 kg. seed cotton per ha., 200 kg. are seeds. 0: the average, 40 to 50 kg. seeds are needed per hectare or 20 to 25 percent of total seed production. Zairean farmers receive an excess of seeds and they use this surplus as organic fertilizer or burn the seeds to produce snoke which keeps away insects and baboons fran the fields. 24'Jhe C.F.D.T. mission in Ubangi-angala recommends sorting plus treatment with an appropriate seed disinfectant; this could increase yields by an estimated 10 percent. 2 4 l @5603 H25 9.3 3.3 5. neon memd finned capo—5900 0.358 ANN. ..:. 6an :..n 0:05 muhomxm .mp3 .muoonouoLE m3 com. c038 mo coflugflflu: wmé mode wodm Wmfia @5530.“ Hoodoo pom 93m Slocumcmom Ii ... Hand "3.3: illumpmom $82 1! 9.2.6 com Home comm couuoo do one... 838 I Ill) 88538 038351 80685 03 we managed .l 353 5. moo» d: 3.59.0 so» ~33: n 38 donning :95 .956 "cofiuchummol‘ was ooo.ma H mmvmvwm T5 wO wov Cu mm EHUCMHQ HON 68ml 88 coma 82 was oeflnfio ”mung mo 038mg .H Paco: mac» 086 “3%.: Awméwv sou email “3:: 33x8. Uflummabo Awmémv 93» mmm.m “mqume I. 98 MSJN End .638 .232. QN anew: m) I||I|| 5:: 89682 .835. so 83383 can £8.85 n3 .8221 "838 mcon mac . mm :858 Bow ||.|I|I 143 Cotton Seed Crushing Between 1936 and 1938 three cotton seed crushing mills were created in Zaire and currently most of then are located along a navigable river or railroad. Most mills have large storage facilities. It was only after World War II that the crushing mills prospered because of increased chnand for cotton seed oil and neal. For exanple, between 1950 and 1960 all of the seeds not used for planting were processed for oil. In 1958 ten cotton seed crushing mills were active in Zaire, 6 of them under COI'CNCO control with 71.4 percent of total capacity (see Table 4.21) . In 1959 Zaire exported 629,0003 cotton seed oil and 780.0005 cotton seed meal. All cotton seed oil extraction in Zaire is done with nechanical expellers after the seeds are cooked. The linters are separated fron the seeds before expelling the oil, and are used for the manufacturing of blankets, bandages, etc. Remote ginning mills use the cotton seeds as fuel, sell the seeds to mining companies as fuel, or simply burn the seeds. Cotton seed oil, after deodorizing and neutralization, is used dorestically, eSpecially for the prodmtion of margarine. Cotton seed meal and cake are used locally in cattle feeding and are exported mainly to Zanbia and South Africa. Cotton seed flour might be of special interest in supplying needed protein to the populatim. Generally, the Zairean diet is adequate in calories but lacks protein. Cottm seed flour could be mixed with starchy foods such as manioc or bananas to upgrade the diet. The pro— cessing of cottcn seed Heal into flour for hIman consumption, however, requires the renoval of gossypol, a toxic phenolic pigment. Such removal is said to be econonically feasible [Hirsch, 1971] . 144 Table 4.21. The Status of the Cottm Seed Crushing Industry in Zaire, 19581 Carpany Localization Daily Seed Yearly Seed Capacity in Capacity in Tons Tons3 Huilco (C)2 Lubmbashi (Shaba) 60 15,000 Huilco (c) Samba (Mamiéna) 45 11,250 Huiltinda (C) Isiro (Haut-Ue’lé) 60 15,000 Huiltinda (c) Tinda (Bas-Uélé) 60 15,000 Cotonco Businga (Mongala) 30 . 7,500 Bunge Tende (Kasai) 10 2,500 Corbelga Kabinda—Kanukmgu 20 5,000 (Larami) . Colocotcn Lodja (Sankuru) 8 2,000 Colocoton Katanda (Lomami) 12 3,000 Vinchent Mahagi (Ituri) ' 10 2,500 1Brixhe, 92. _c_i_._1_:_., p. 130. 2The (C) indicates that these carpanies are controlled by Cotonco. 3Based pm 10 nonths of operation per year and 25 days per month. There are three cotton seedcrushing mills in the North: the Businga mill for Ubangi and angala, located almg tl'e Mongala river, the Tinda (Aketi) mill for Bas-Uélé and the Isiro (Gossamu) mill for Haut-Uélé, both located along the Mmgbere-Burba railroad. 'Ihe Isiro mill was closed shortly after independence because of a lack of seeds. The Businga mill was shut down in 1968. Since 1969 only two major oil nu'lls have been Operative in Zaire, Tinda for the entire North and Lubutbashi for the South and Central. Since independaice the financial results of the cotton seed cashing industry have been negative (see Table 4.22 and 4.23). And, 145 Table 4.22. Republic of Zaire: Tbtal Losses of the Cotton Seed Crushing Industry, 1963- 1969 (In Francs or in Zaires) Year Loss in Francs or in Zaires 1963 24,844,607 Fr. 1964 14,686,675 Fr. 1965 10,861,155 Fr. 1966 24,713,866 Fr. 1967 72,060.95 8 1968 224,200.58 5 1969 75,596.74 8 1 COGERCO, Kinshasa. Table 4.23. Republic of Zaire: Losses of the Cotton Seed Crushing Mills, 1968-1969 (In Zaires) Locatim of the Mill 1968 1969 Tinda 101,843 38,872 Businga 49,153 12,502 Lubumbashi 60,343 833 (profit) Samba (idle) 12,862 25,056 Total 224,201 75,596.74 1 COGERCO, Kinshasa. 146 since the cotton decree of 1947, cotton fiber and seeds, after ginning, have remained the property of the cotton farmers until the mment of final sale. The convention of April 19, 1951 imposed obligatory cotton seed processing for the ginning companies, except where transport costs to the oil mills were prohibitive. It also specified a 50-50 distribution of profits and losses between COGERCO and the cotton ccmpanies. From gross sales, production costs, sales costs and amortization charges were deducted and the balance was shared 50-50 by the cotton firms and 009mm. It is difficult to understand why private firms would maintain an operation which is unprofitable for more than a decade . However, the nature of the losses explains this seemingly irrational behavior. Total cotton seed processing per mill for the 1963-1969 period is presented in Table 4.24. Since 1964 several oil mills were left idle or cperated far below capacity because of insufficient cotton seeds . However , COIONCO and the other cotton companies continued to deduct depreciation charges for these mills. The Lubumbashi mill is a modern mill which produces first quality oottm seed oil for hunan consumption (see Table 4.25) .25 In 1973 the mill paid 14 8 per ton for cotton seeds. The mill processes most of the seeds from Shaba, Mamiéma, Kivu and Kasai. On January 29, 1972 ONAFITEX fixed the price for cotton seeds at 1.50 K. per kg. for seeds within a radius of 300 km. fron the mill and 0.75 K. for seeds outside this radius. 25For cotton seeds with 20 percent oil, extraction rates of 16 percent and above are realized by modern mills. Cotton seeds in Zaire contain an average of 18 percent oil. 147 Table 4.24. Republic of Zaire: Tons of Cotton Seed Processed Per Mill, 1963-1969 (In.Tons)1 Year Lubutbashi Businga Tinda Kabinda—Kamukungu 1963 4,2582 3,380 9,164 792 1964 2,9862 1,935 3,432 124 1965 3,8972 -- -- -— 1966 3,4742 1,493 -.. -.. 1967 3,3072 1,221 4,503 --— 1968 1,8383 1,6284 3,156 -- 1969 8,965 -- 6,647 —-- 1COGEch), Kinshasa. 2 Fran July lst to June 30th of the next year. 3From July lst to December Blst of the same year. 41“an September lst to December Blst of the same year. Table 4.25. Production of the Lubumbashi Cottcn Seed Crushing Mill, 1939-1972 (In Tons) Year Cotton Seeds Cotton Oil Meal and Linters Processed Production Cake 1939i 6,665 677 2,729 310 19451 8,379 940 3,365 287 19501 13,296 1,446 5,649 758 1955 1 17,241 2,036 6,596 723 1959-1960l 21,998 2,462 3,117 1,135 1964-1965 6,577 627 2,578 428 1968; 8,052 601 3,044 327 19692 15,612 2,007 6,142 842 19702 12,011 1.295 4,741 633 19712 13,427 1,983 5,170 656 1972 9,821 1,324 3,548 402 llhmpungu: Kamanda, Novedore 1970, p. 78]. 2[RépIIJblique du Zaire, Novembre 1973, p. 61]. 148 ONAFITEX did not requisition the cotton seed crushing mills and does not own any, but it did sell the seeds directly to the mills or exported them. On November 30, 1973, President Mobutu announced that all agricultural plantations, ranches and angro-industrial complexes had to be ovmed by Zairean nationals. Thus, the cotton seed crushing mills were taken over by the government and distributed to Zairean nationals who have ten years to reimburse the original owners if the acquisition proves profitable. The construction of a high capacity oil extraction mill in Gerena for (ININGEM (Cotbmat Agro—Industriel de Gerena) is significant for tte Ubangi-Mongala area. This mill will extract oil by pressure and them with solvents. It represents an investment of 13.2 million dollars, 65 percent from a Zairean industrialist (Moleka) , 25 percent from the Zairean government and 15 percent from Krupp—Helm from West Germany. melve hundred workers will be employed by the mill. In Table 4.26 the availability of cotton seeds over the 1970—1972 period is given. Seed cotton production in Ubangi-Mongala is expected to increase slowly, but not at a rate which will satisfy COVIINGEM's demand for cotton seeds. It is possible to bring in cotton seeds from the Uélés, but the transport cost will probably be prohibitive as CDMIBEEM is located about 115 km. from the nearest port (Akula) . With this new mill, it is evident that the Businga mill will remain inactive. 'Ihe Businga, Tinda and Isiro crushing mills are old mills, with outdated equipment and a low oil extraction rates. 149 Table 4. 26. Availability of Cotton Seeds in Ubangi-Mongala, Republic of Zaire, 1971-1973 (In Tons)1 Year Seed Cotton Cotton Seed Seed Available Production Production Requirements Cotton Seeds 1971 12,950 7,770 2,100 5,670 19 72 6 , 350 3 , 810 2 , 300 l, 510 1973 12,000 7,200 - 2,700 4,500 1 C.F.D.T. Mission, Gemena. Eports of Cotton Seed In 1973 Colocoton processed 1,500 tons in lodja and Katanda, Huilco processed 8,000 tons in Lubumbashi, and the rest of the cottcn seeds were destroyed or exported: 3,000 tons to Japan and 3,000 tons to Greece. All of the seeds not needed for planting were burned in the Uélés as the railroad could not assure transport of the seeds to the mill in Tinda. The railroad company Viol-Zaire had only two locaro- tives in operation which were used for passenger trains and transport of high value produce such as cotton bales, coffee and supplies. Since the creation of ONAFITEX, a serious effort has been under- taken to find export markets for cotton seeds . The main handicap for exporting cotton seeds is the high transport cost from the ginneries to the export port of Matadi and the difficulty of finding maritime transport. A 19 70 World Bank Study reported that cotton seed exports to Japan could be quite profitable if the price were in the neighborhood of 98 dollars per long ton C.A.F. Japan. With the opening of a he» ginning mill in Gemena and the improve- ment of transport conditions in Zaire and, in particular, rail transport 150 in the Uélés, Zaire should be able to process all of its seeds. The demand for vegetable oil within Zaire and for export is strong, and since palm oil production in Zaire is decreasing, cotton seed oil could replace more and more palm oil for the production of margarine, vegetable oil and soap. m Historically a number of conventions with foreign controlled cotton ginning firms regulated cotton processing and marketing on a custom work basis . These conventions provided little incentive for ' operational efficiency and awarded generous profit margins to the ginners. By 1970 the reserves of COGERCO, the price stabilization fund and regulatory agency, were exhausted. In 19 71 the government created a national cotton marketing office (ONAFITEX) in order to rationalize and improve the efficiency of the cotton marketing structure. In 1970, half of the 115 ginning plants in Zaire were closed. Although sufficient ginning capacity is available, 30 new ginneries ’ are now being installed by the government in Zaire. There is no economic justification for the purchase of these new mills. Cotton marketing costs are very high. Transport costs represented ‘ one-fourth of total marketing costs, including the purchase of seed cotton in 1969 . Transport costs were mmlch higher for tte Uélés than for Ubangi-Mangala. Transport costs are very high because of the poor state of the roads, the long distances involved in hauling seed cotton and the luxury goods rate charged by ONATRA and VICI-ZAIRE for the tranSport of cotton bales. 151 ONAFITEX reduced marketing costs by laying off expatriate personnel and through the reduction of transport rates, financing costs and baling costs. However, the collection of seed cotton in 1973 and 1974 started late and was extended over much of the year. This reduced the farmer's willingness to plant cotton, it occupied collection agents, many of whom are cotton extension workers, for a long time on collection, and it increased considerably transport costs. The Zairean textile industry is well deve10ped. A recently opened textile mill in Kisangani will considerably improve the marketing efficiency of cotton from the Uélés. The average sales prices of cotton fiber to domestic textile factories remained remarkably stable over the 1963-1973 decade. Presently, local fiber prices are kept low to protect the Zairean textile industry and Zairean consurers while export prices are high. The prospects for expansion of the datestic market are bright. Export opportunities are plentiful as Zairean cotton has an excel lent reputation in export markets . The cotton seed crushing industry is well developed but suffers from outmoded equipment and high transport costs. Since 1960 most mills have been operating in the red. In 1973 most cotton seeds in the North were either burned or dumped; a high capacity oil extraction mill is being opened in Gemena in Ubangi and will process all the seeds from Ubangi—Mongala and probably some from the Uélés. CHAP'IERV THE ORGANIZATICI‘I OF A FARM BUSINESS SURVEY OF (DTIVON FARMERS IN NORI‘HERT ZAIRE Purpose of the Survey The original plan of this research was to conduct a study of a small number of selected cotton growing villages in Northern Zaire. The cost route method would be used to collect precise information on resource use and returns, especially with respect to labor, farm incates and the relative profitability of different crops. Linear programming would be used to calculate the optimal enterprise combination on tradi- tional farms and then carpare the results with the present resource allocation to determine needed adjustments in the farm businesses. Al- though the Department of Agriculture thought such an approach was use- ful, they preferred that the study pursue a policy—oriented study of the entire northern cotton zone.1 The Department believed that such a study would yield valuable insight on how to increase cotton production in the North as well as providing information on the future cmparative advantage of cotton production in specific locations in nortlern Zaire . Description of Survey Area The Congo rebellion raged in the northern cotton zone from 1960- 1967, particularly in the Uélés. The Presidency recently focused 1A study of all cotton growing regions in Zaire was not possible because of the size of the country, the wide dispersion of cotton grow- ing areas and budget and time constraints. 152 153 attention on the North and announced that a serious effort would be made to industrialize and prorote agricultural development in this area.2 The North-East and North-West are well known for their agricultural potential. The Department of Agriculture is now devoting more attention to this area and is in the process of identifying and implementing agricultural development projects. Little research has been devoted to the northern cotton belt although several pre-investment studies for a cotton project in this area have been done or are underway by FAQ, World Bank and C.F.D.T. The field survey area includes the Equator and Haut- Zaire regions with the Ubangi and Mongala sub-regions in Equator and the Bas-Uélé and Haut-Uélé sub-regions in Haut-Zaire. The Kibali—Ituri and Tshopo sub-regions in Haut-Zaire were not included in the survey as cotton production is marginal in these areas. 3 The survey area is bordered on the west and north by the Ubangi river, separating Zaire from the Popular Republic of the Congo and the Central African Republic, and on the north and east by the border with Sudan and Uganda. The southern border corresponds to a straight line from Budjala in Ubangi to Wamba in Haut-Ue’le’. The distance east to west of this area is 800 miles while the distance from north to south is 220 miles. 2The recent move of tle Faculty of Agronomy from the capital city to Yangambi in the northern area is part of this effort by the Presidency to revitalize agriculture in the North. 3In 19 70, according to the Annual Agricultural Report of the Haut- Zaire region, out of an estimated 41,051 ha. of cotton grown in Haut-Zaire, only 1,603 ha. were in Tshopo and 1,003 ha. in Kibali-Ituri [République Dérocratique du Congo, 1970a]. In 1958, 105,470 ha. of cotton were reported for the Haut-Zaire region with 9,344 ha. in Tshopo and 3,291 ha. in Kibali-Ituri. 154 T'He cotton belt is at the northern limit of the equatorial rain forest and includes equatorial rain forest and savanrehs with the transi- tion from rain forest to savannah: forest galleries along rivers, savannah woodland, tree savannah and shrub savannah. Presently, there are no security problems but long distances, poor roads, dangerous bridges, unreliable ferries, poor communications and a lack of car repair facilities and board and lodging seriously handi- cap field research in this area.4 Survey Method Several methods exist for collecting primary data on traditional farms in Africa [Collinson, 1972; Yang, 1965; Upton, 1973; Spencer, 1972]. The purposes of the survey, the available budget and the time constraint usually determine the choice of survey method. This researcher used the farm business survey and visited 160 cotton farmers twice, one visit during or after planting and one visit after the harvest. The 160 farmers were selected by a three stage sampling design. Since tra- ditional farmers keep no written records, heavy reliance was placed on the farmer's recall. Thus, observational errors in this type of survey can be quite high [Spencer, 1972]. Determination of Sample Size The main criteria used in the determination of sample size are budget and time. considerations and the representativeness of the sample . A representative sample makes it possible to apply the results of the 4The major roads between cities are in fairly good shape and ade- quate board and lodging can be found in the larger cities. 155 analysis to the population from which the sample was taken. However, although statistical theory helps us to determine the degree of repre- sentativeness of a sample, it is hard to apply this theory to a farm managerent survey in tre African context. The field research budget was only large enough to hire and equip four field enumerators for the entire northern cotton zone. It was estimated that enumerators would spend about half their time on travel and the other half on farmer interviewing, measuring fields and weighing cotton. Each enumerator was able to interview only 40 farmers during planting and harvesting periods. Hence,the sample size contained 160 farmers of an estimated 250,000 cotton farmers in the survey area which gives a sampling percentage of 0.064. Selection of a Sampling Frame The research budget and time constraints did not permit drawing a primary sampling frame of cotton farmers in the northern cotton belt. "5 A list of farmers who were under "crop imposition and a list of famers who had to pay a personal minimum tax6 could have been utilized. However, in 1970 the Department of Agriculture with the aid of FAO organized an Agricultural Census in Zaire. All farmers who reported a cotton field in the 1970-1971 Census of Agriculture in the Lbangi, 5All HAV's (home adulte valide) or able-bodied adult males who are unetployed in a rural area are required to grow particular crops as prescribed by tle regional authorities . Failure to comply may result in a prison term or payment of fines. 6The personal minimmm tax for each HAV, called C.P.M. (contribution personnelle minimum) or minimum personal contribution is fixed at 2 Zaires per year, equivalent to 4 dollars. The tax list is usually prepared on the basis of the HAV list. The government agronomist in each zone, aided by the extension agents and the village chief, draw up the list of all HAVs. 156 Mangala, Bas-Uélé and Haut-Uélé sub-regions were listed by computer printrout. This list served as a frame for the farmIbusiness survey. Farmers with more than one field were listed more than once.7 In total, 875 farmers were listed, 262 in Ubangi, 153 in Mongala, 282 in Bas-Uélé ' and 178 in Haut-Uélé. _ A cross-check revealed that the 1970 Agricultural Census list contained the same errors as the crop imposition list (list of all HAVs) from which it was probably derived.8 Some farmers were not included in the list, others appeared more than once but under different names, and still others could not be traced in the village. Since time was pressing to start.the survey, we used the Agricultural Census frame and accepted its inaccuracy.9 This frame offered the opportunity to gather information on the same cotton farms at two points in time, 1970-1971 and 1972-1973. Since the seed cotton price increased by 50 percent foam 1971 to 1973 while food grain and staples' prices remained constant, the unique Opportunity to study the price responsiveness of cotton farmers was presented. A drawback of this frame was the possibility of ending up with less than 160 farmers as some may have migrated to other areas, died or simply stopped growing cotton. In fact, the survey yielded only 129 usable questionnaires (see Chapter VI). The frame is also biased 7Iess than 5 percent of cotton farmers had more than one cotton field. 8The frame for the 19 70 Agricultural Census was also prepared under the responsibility of the government agronomist in each zone. 9Using an inaccurate frame introduces a bias in the survey. It was judged that the bias was acceptable taking into consideration the oast of drawing up a frame for this research. 157 against young cotton fanrers or recent cotton growers as only farmers with a cotton field in 1970 were included in the frame. SamplinLProcedure The 1970 Census of Agriculture included a complete survey of all agricultural holdings in tie comercial sector and a sample survey of (the traditional farms. The purpose of this survey was to gather reliable statistics on farming systems, land tenure and type of holding, land utilization, input use, crop acreages and yields, production, storage, sales and erployment. The sample survey of the traditional sector 10 There are an esti- included 16,800 holdings clustered in 446 areas. mated 2,941,294 traditional farms in Zaire [République du Zaire, 1973b]. Each farm in the Agricultural Census was visited three times at selected intervals over the March 1, 1970 through March 1, 1971 period. The questionnaire used in the Census was guided by the 1970 FAO World Census of Agriculture, Regional Programme for Africa. The farmers included in the sample survey were selected by a two- stage random sampling procedure. The country was stratified in eight regions and each region into two, three or four sub-regions following tle administrative organization. In each region a coIplete list of ”groupeten "wasdrawn up, excludingurbanareas andminingareaswith little or no agricultural production. 11 These "grmpements" are the 10Information supplied by Mr. Jansonius, FAO, expert for the Census of Agriculture . lJ'A "groupement" is the smallest administrative unit in Zaire con- taining ten or more villages. In the Agricultural Census it is the primary sampling unit. 158 primary sampling units. In each region from 56 to 62 "grOLperents" were selected by weighted random sampling with probabilities propor- tional to the estimated number of holders per sub-region and per "grouperent. " The number of agricultural holders in each "grouperent" was estimated on the basis of the total number of men and unmarried womm listed in the latest population census. The number of "groupe- ments" was determined by allotting two "groupelents" for each of the enurerators in tie region. On the average, there were 28 enumerators in each region. The "groupement" was split up in two or more primary units when tre number of holders exceeded 800. For each selected primary unit the enumerator compiled a list of all agricultural holdings. He then drew 35 holdings in each "groupemen " by systematic random sampling . Sampling Plan The cotton farmers for the field survey were selected by three- stage random sampling: two—stage random sampling in the Census of Agriculture which yielded the frame and one-stage random sapling from the frae. The farmers were selected by weighted random sapling with weights proportional to the 1970 seed cotton production in each 51b- region (Table 5.1) . For each sub-region the farmers were selected by random sampling.12 The results of this sapling procedure are shown in Table 5.1. One hundred sixty cotton farmers were selected, dispersed in 109 different villages and clustered in 33 primary sampling units or "groupements . " 12Four farmers were selected mice because they appeared more than once in the 1970 agricultural census frame as they had more than one cotton field. However, the next farmer on the list was selected as a replacerent for each of the four. 159 Table 5.1. Sampling Plan Used in the Fanm Business Survey of 160 Cotton Farmers in Northern Zaire, 1972-1973 Sub-Region Seed Cotton Nmber of Number Of Nutber of Production, Cotton Prmary‘ Villages in Tonsl Farmers Sampling Selected Selected Units Selected tbangi 9, 377 72 ' ' 9 46 mngala l, 640 13 5 12 Bas-Uélé 5,942 45 11 3o Haut—Uélé 4 , 024 30 8 21 Total 20,983 160 33 109 lRépublique Detocratique du Congo, 1970a. Rapport Annuel, Service Provincial de l'Agriculture, Province du Haut-Zaire. Kinshasa: Ministere de l'Agriculture; République DéIocratique du Congo, 1970b. R_apport Annuel , Service Provincial de 1 'Agriculture, Province de Tmteur. Kinshasa: Ministére de l'Agriculture. Selection, Training and Supervision Of Field Enumerators Field enumerators were not solicited from the Department Of Agri- culture because they were engaged in a major survey on coffee. Instead, this researcher looked for enuterators with an agricultural background and with sore formal training in agriculture. The Higher Technical Agricultural School of Mondongo (E.T.S.A.-Mondongo, 20 km. west of Lisala) and the Higher Institute Of Agricultural Education of Bengamisa (I.S.E.A.-Bengamisa, 60 km. north of Kisangani) provided us with the addresses of their recent graduates and drop-outs. TwO andongo graduates were hired in Gemena. Both were from the region in which they were supposed to work. Then, three drop-outs from Bengamisa were asked to participate in enumerator training in Kisangani. '1‘!!!) were hired after 160 they passed a theoretical exam and a field test. All four field enumerators had corpleted secondary school and had some agricultural training. The enurerators were posted in Ubangi, in Ubangi plus wngala, in Bas-Uélé and in Haut-Uélé. They were all fluent in French and Lingala. The enumerators for Ubangi and Mongala were also fluent in Ngwaka and the enumerators for the Ueles in Kiswahili. Each enumerator was assigned specific farmers, on the average forty, the number depending mainly on the distances of travel required, his type of transport vehicle (bicycle or motorcycle) and his surveying skills. The field enumerators were paid a fixed monthly salary Of 30 Zaires and a variable bonus Of Lp to 10 Zaires per month, depending won the quality of work during the month. Two field enumerators were trained at Bwamanda, a Catholic mission 100 km. south-west of Gemena, and the other two at the O.N.R.D. Office in Kisangani.l3 The Office training which took about ten days emphasized the purpose and importance of the survey, the significance of each item in the questionnaire, metrods of field measurement, tech- niques of approaching farmers and administrative procedures. These lectures were supplemented by on-the-job training in the field which included interviewing techniques, field size measurement, crop density counts , disease identification and bicycle or motorcycle repair and maintenance. Each enumerator was given a survey manual, in which the 13At these training sessions assistance was provided in Bwamanda by kaako Kma E. Madjo, student in the Department of Agricultural Ecolomics at the National University Of Zaire in Kinshasa, and in Kisangani by Bazola Muangu, researcher at the National University of Zaire in Kisangani. 161 ontents of the course were summarized. Throughout his field work, he was instructed to record any problems which might occur in his survey manual. Preliminary interviewing in the Bwamanda area served as pre- testing of the survey questionnaire and field survey methods. Some items in the questionnaire were found irrelevant and redundant and were deleted. Other questions were too (sensitive and were dropped. Since the initial questionnaire required more than two hours of inter- viewing,the schedule was administered in two sections on different days. Eadh enumerator was provided the essential materials and equipment needed for field research. Besides the questionnaires and a survey manual, he was given rubber boots, raincoat, bicycle or motorcycle with a repair kit and tools, a briefcase, travel bag and Office supplies (writing materials, a writing pad, protractor, compass, envelopes and paper, postage stamps). He was also given drugs for malaria, diarrhea and a few other diseases. For the measuring of fields, he received a measuring chain or tape and a measuring wheel, a surveyor's corpass and a 3m. by 3m. crop density square cord; he was given a spring balance to weigh cotton. The author supervised the enumerators personally and visited them on unannounced days on an average Of once a month. During the visit this researcher paid their salary, supplied the necessary survey materials, and attempted to solve other problems. Questionnaires were checked for errors and omissions or discrepancies were corrected on the spot by the enumerator. An effort.was made to accompany each enumerator during some Of the interviewing and measuring of fields. 162 rIhis helped in getting to know how data was collected to make contact with farmers and local extension personnel; it was also good for morale Of the field enumerators and farmers. locating Farmers When arriving in a new survey village, the enumerator always con- tacted the local agricultural extension agents or agricultural Officers who then introduced the enumerator to the farmers. This helped in locating the farmers and insured farmer cooperation. Often the extension agent assisted the enumerator with his survey work. Questionnaire Design, Tabulation and Handling Of the Data The questionnaire was prepared in French and the questions were worded and listed exactly in the way that they should be asked. The enmerators then fraed the questions in the vernacular language familiar to the respondents (Lingala, Ngwaka or Kiswahili) . The questions were cross-checked to detect the accuracy of the information supplied by the farmers. Sore items in the schedule turned out to be irrelevant such as questions on farm mechanization, fertilizer and pesticide use because none Of these was used by the farmers. Questions on prices received for seed cotton were Often misunderstood by the farmers as the concept of "price" was unfamiliar to many of them. For example, several understood "price" as the total amount of money received for their cotton crop. Questions on the farmer's possessions and on borrowing and lending behavior were found to be quite delicate and had to be handled with caution at the end of interviewing. Specific questions on personal possessions had to be deleted as they were too . ah: ‘1. Le II! 163 sensitive. Finally, Open-ended questions were included at the end of the questionnaire to determire beliefs, values and thought patterns affecting agriculture and cotton production. The data in the questionnaires were summarized and hand tabulated on separate worksheets . Answers to quantative questions were tabulated to determine frequency distributions, averages, percentages and dis- persions. Qualitative answers were grouped in classes, each class being a collection of similar answers. The classes were also tabulated and a frequency distribution was made. A large amount Of time was required for this activity because of the cotprehensive nature of the 14 questionnaire. The tabular method and regression analysis were then employed to study relationships between various farm managetent factors. seen A farm business survey was conducted in northern Zaire over the 1972—1973 cotton production year. The farmers for the field survey were selected by three-stage random sampling. As frame, we used a list of all cotton farmers included in the 1970 Agricultural Census in rorthern Zaire. One hundred sixty cotton farmers were selected, dis- persed in 109 different villages and clustered in 33 primary sampling units. Four field enumerators visited each farm twice a year, once before harvest and once after harvest. l4Tollens [September 1974b] describes a method to eliminate most Of the tabulation and corputation work by using pre—coded survey forms which allow keypunching to be done directly from the form. CHAPTER‘VI ANALYSIS OF THE FARM BUSINESS SURVEY DATA Characteristics of the Survey Sample Nonresponse and Drop-Out The original sample consisted of 160 respondents chosen from the frame of the 1970-1971 Agricultural Census in Zaire. However, only 129 respondents produced usable questionnaires . Nonresponse and drop-out occurred for the following reasons: 9 had gone to work for a corpany or a private person since 19 71; 8 had moved to another rural area; 5 were sick and could not be interviewed; 3 had died since 1971; 2 had left for a city; 2 were too Old and could not be interviewed; l was in prison; and 1 could not be located. Forest and Savannah Areas A farmer was classified as being in a forest area when his cotton fields were located in the forest. Often farmers have their fields in the forest although they live on a strip of savannah. Since much of the northern cotton belt is on the fringes of the dense rain forest, farmers Often have tl'e Option of locating their fields in forests or in savannahs. 164 165 Of the 129 respondents, 102 or 79.1 percent located their cotton fields in the forest and 27 or 20.9 percent located them in savannahs, distributed as follows between the four sub-regions included in the survey : _Ub_a_ng_i_ anqa la Bas-Uélé Haut-Ué lé Total North Forest areas 56 7 30 9 102 Savannahs 11 3 1 12 27 Our sample included a relatively large number of cotton fields in forest areas and relatively few in savannahs. Hence, it was difficult to derive statistically valid inferences for savannah areas. It was thought that random sampling from the 1971 Agricultural Census frae would have yielded a random distribution of cotton fields in forests and in savannahs, but this did not occur. In retrospect the sampling should have been stratified according to ecological area, forest. or savannah,because of the large differences in environmental conditions and farming systems between these areas. Farmers in Paysannats Seventeen respondents (13.2 percent), all from Ubangi, were located in a paysannat. A location in a paysannat should be considered a criteria for stratification because these farmers had been exposed to improved production technologies and intensive agricultural extension efforts in the past. Farmers Possessing a Cotton Farm Book Eighty out Of 129 farmers (62.0 percent) possessed a cotton farm book, a small book which each cotton farmer used to receive from COGEKZO 166 or ONAFITEX in which the amount of cotton seed received and the sales of seed cotton per grade were registered by the cotton extension or cotton collection agent. Payments in kind (salt, small farm tools, etc.) were also recorded in this book, which proved to be a valuable source Of information for this study. Farmers in the Congo Rebellion The Congo rebellion swept over the eastern part of the northern cotton belt from 1961 to 1965. As a result of the hostilities, many farmers fled their villages and hid in the bush. Sixty-one out of 129 farmers interviewed in the survey experienced rebellion activities in their villages and 49 abandoned cotton production during these difficult years . Characteristics Of Cotton Farmersl General Characteristics The Ngbaka and Zande make up over 50 percent of the farmers. In total, the farmers belong to 13 different ethnic groups. About 70 per- cent Of the farmers indicated that they were Catholic. The remainder were Protestant, Kimbangist or had no religion. 2 The average age of the respondents was 44 years. For Ubangi and Mongala, it was 41 years; for J'As there were only ten farmers from Mongala included in the sur- vey, usually no separate results will be; presented for this sub-region. They will be added to those of Ubangi. ZKimbanguism is a prophetic and messianic movement which was started in 1921 in Lower Congo province by Simon Kimbangu and whidn developed into a religion recognized by the state. 167 Bas—Uélé, 47 years and for Haut-Uélé, 48 years.3 The survey questions were addressed to the household head who is usually the head of the family. The average age seems high, but one should bear in mind that the head of the family is usually the Oldest person in the family. On the average, 91.5 percent Of the farmers derived their incore exclusively from agriculture, including hunting and fishing. The figures per sub—region are 92.2 percent for Ubangi and angala, 93.5 percent for Bas—Uélé and 85.7 percent for Haut-Uélé. The most important secondary activities were maintaining a small shOp for the sale Of consurer goods and being a part-time counselor or judge in the village administration. Nbst farmers interviewed in the survey were illiterate. On the average, they had received 1.4 years Of formal schooling; 57.0 percent Of the farmers never went to school. Those who went to school spent on the average 3.1 years in school. Detailed data are presented in Table 6.1. Of the farmers in the survey, 19.4 percent had resided in a city in the past. Residence in a city was defined as having established one's residence in a city for at least one month. A city was defined as each urban center considered a "city" by the farmer. Of the 129 farmers, 110 lived in the villages in which they were born. Those farmers who had not yet resided in a city were asked why they had not moved to a city. Their responses are given in Table 6.2. Mast farmers preferred to stay with their families in the villages and 3There is the question about the extent to which farmers know their age. In Zaire each adult person has a natioral I.D. card. When the date of birth was not indicated on the citizen's card, the enumerator could usually derive his approximate age by recalling historical events to the farmer. ' 168 mentioned that they had no families in a city. Only one of the 129 farmers in this sample wanted to move to a city. Table 6.1. Educational Characteristics Of Cotton Farmers in.Ncrthern Zaire, 1972-1973 SubPRegion Average No. Percentage Average No. Years in Years in Who went SChool for Those‘Who School to School went to School Ubangi and Mongala 1.6 50.6 3.1 Bas-Uélé 1.0 29.0 3.4 HautrUélé 1.1 38.1 3.0 Average for the NOrth 1.4 43.0 3.1 Table 6.2. NOrthern Zaire: Reasons.Mentioned for not Moving to a City, per Sub-Region (in Percent) Reasons Mentioned Ubangi and Bas-Uélé Hautr Average .Mongala , Uélé for the NOrth I could not find work in the city. 14.9 3.8 0.0 9.0 life is too expensive in the city. 17.0 23.1 6.3 16.9 I prefer to stay with.my family in the village. 51.1 57.7 56.2 53.9 I prefer to stay in agriculture 17.0 15.4 37.5 20.2 Tbtal Percent 100.0 100.0 100.0 100.0 169 Employment History Farmers were asked whether they had been employed outside their farms since 1960 and had been paid in money or in kind, above food and lodging, and for at least one month; 41.4 percent reported that they had been employed in sore way in the past (51.3 percent in Ubangi and Mongala, 16.1 percent in Bas-Uélé and 42.9 percent in Haut-Uélé) . Thus, many farmers have had sore erployment experience in the past. Family Cotposition On the average, each family was composed of 5.05 persons, and each farmer had 1.16 wives and 3.04 children. The age distribution per sub-region is given in Table 6.3. Average faily size in Ubangi-Mangala was much larger than that in the Uélés with a higher degree Of polygamy and a larger number Of children per family. The demographic pyramid in the Uélés is quite different from Ubangi-Mangala; the [Jélés pyramid has relatively more peOple in the higher age brackets and, apparently, a much lower fertility rate. This confirms the findings of the 1970 demographic survey in Zaire which indicated a precarious demographic situation in the Uélés (see Chapter III) . The survey found on the average one person living with the family who did not belong to the family by birth (ascendents, descendants) or marriage. However, these persons usually have ties with the family in that they are, fiancés, orphans, cousins, etc. In a broad sense, then, they could be counted as part Of the faily. 170 Table 6 . 3. Average Faily Colposi tion on Cotton Fame in Northern Zaire, 1972—1973 Age Ubangi and Bas-Uélé Haut-Uélé Total North Mongala Under 15 years 2.71 1.19 0.62 2.01 16-64 years 2.99 2.39 2.38 2.74 Over 65 years 0.45 0.03 0.14 0.30 Total 6.15 3.61 3.14 5.05 Nurber Of wives 1.25 1.00 1.10 1.16 Number of children 3.66 2.65 1.33 3.04 Other Persons Living with the Family Under 15 years 0.36 0.26 0.00 0.27 16-64 years 0.70 0.68 0.52 0.67 Over 65 years 0.05 0.00 0.14 0.05 Total 1.11 0.94 0.66 0.99 Food Consmption Habits Farmers were asked to enumerate and rank , according to their importance in the diet, the principal food products which they consume. In Table 6.4 the percentage of farmers who ranked a food product in lst, 2nd, 3rd, etc. class is given. The 4th, 5th and 6th class are of lesser importance and are less accurate as not all farmers could rank beyond the 3rd class. Maize flour, manioc (roots and leaves) and bananas are the principal food staples with maize and manioc the most important food product in the North-West and manioc and bananas the principal food products in the North-East. Rice, groundnuts, yas, sweet potatotes, vegetables. meat and fish are Of lesser importance. (’1 'r1 171 land Tenure System in the North land Occupation Rights All sample farmers Operated their farms for their own account. Ore farmer considered himself an "independent farmer. "4 In addition, all farmers except four held land under the customary land tenure system.5 The nature of the indigenous land tenure systems varies widely among the different groups of people. According to Harms [June 1974], the cor- porate group that holds the land is usually a small group—a lineage which may be a village or part of a village or, in some cases, a political d'Iiefdom. Generally, the man responsible for land is relatively low in the political hierarchy. Furthermore, the land held by the group has fixed boundaries which are either clearly marked by geographical features or not marked at all. There are several kinds of rights in land, such as hunting rights, gathering rights and cultivation rights. All the members Of the corporate group generally hold hunting and gathering rights in the uncultivated areas Of the group's land, whereas plots under cultivation belong to the individual cultivators or their immediate families. The individual cultivator gains rights to a plot 4"Independent farmers", called "Planteurs indépendants," are farmers who are not subject to crop imposition. Usually, these farmers are the most progressive in their region and are fairly well integrated in the ecoromic system as they produce primarily for sale. They almost always Operate a coffee plantation. They also pay more than the minimum head tax (C.P.M.) . Soretimes they possess an identification card delivered by the local government agronomist with their status of "independent farmer" being marked on the card. 5In Ubangi-Mongala one farmer, a village head, held land in private ownership. Another farmer stated that he held land in concession from a village chief. 'IWO other farmers rented land from somebody else against payment Of a small fee. 172 Table 6. 4. Ranking of Principal Food Products Consumed by Cotton Farmers in Percentages of the Responses per Rank, Northern Zaire, 1972-1975. Food Products lst 2nd 3rd 4th 5th 6th Rank Rank Rank Rank Rank Rank Maize flour 41.2 13.4 2.5 -- 3.0 -- Manioc leaves 22.1 21.2 9.2 19.6 4.5 8.3 Manioc flour 21.4 29.8 26.2 14.9 10.4 4.2 Bananas 12.2 20.5 37.9 12.6 9.0 -- Grolndnuts 1.5 2.4 --- 5.7 7.5 8.3 Vegetables 1. 6 —- 0 . 8 -- 14 . 9 8. 3 Meat --- 0.8 --— 4.6 14.9 29.1 Fish --- 2.4 3.4 9.2 14.9 18.8 Palm wine --- 0.8 0.8 -— 1.5 4.2 Rice --— 8.7 10.9 12.6 3.0 2.1 Sweet potatoes ---- --- 2.5 3.4 13.4 4.2 Yas —- -—- 3.4 3.4 1.5 10.4 Other --— --— 2.4 4.8 1.5 2.1 Of land by putting it under cultivation and maintains his rights until he abandons the field. Farmers had mixed feelings about cultivating land Of other groups even when they were authorized to do so. In Ubangi-Mala about 80 percent indicated that they would cultivate land belonging to other clans as compared with less than 20 percent in the Uélés. This is probably related to papulation densities which are much higher in lbangi-Mzngala than in the Ueles. Renting Of land and Land Utilization Charges None of the farmers included in the survey leased their land, although two farmers were renting land. Farmers holding cultivaticn rights 173 to land do not have to pay anything in cash or kind to those with authority over land. However, most farmers Occasionally do have to work for the head Of the village or head of the clan which may or may not be related to securing cultivation rights for land. Selection of New Fields Farmers were asked who selected new fields to be brought into cultivation. their own choice . Only 20.4 percent of the farmers picked a new field by However, this question is somewhat vague and lacks precision, for there are three distinct elerents involved in the choice of a new field: choice Of crop rotation. location, size Of the field and, associated with these, Data in Table 6.5 points up the importance Of the agricultural monitors and extension agelts in the choice Of a new field. Table 6.5. Into Cultivation? ", Northern Zaire, 1972-1973 Responses to Question: "Who Selects New Fields to be Brought Who Selects New Fields Ubangi and Bas-Uélé Haut-Uélé Tbtal North Mongala Agricultural Monitorsl 59.0 64.7 76.0 63.5 Farmers themselves 29.5 14.7 0.0 20.4 Village or Clan Authorities 3.8 20.6 0.0 11.7 Territorial Agronomist 7.7 0.0 24.0 4.4 Tbtal Percent 100.0 100.0 100.0 100.0 JAgricultural monitors (moniteurs agricoles) are agricultural extension agents . 174 Communal Crops and Fields Communal crops or fields are defined as crops grown on a field for which all labor inputs are communal and all field proceeds are distributed among the merbers of the group. In Lbangi and Mongala 22.1 percent of the farmers had commlnal crops or fields. They were all small coffee plantations. About one-fourth (23.8 percent) Of the farmers wolld agree to join a communal field. Those who refused gave as tie principal reason the difficulties in distributing the proceeds from the field (Table 6.6) . Thus, farmers were not in favor Of commmal crops or fields. Table 6.6. Reasons for not Participating in a Collective Field, Northern Zaire, 1972-1973 I Reasons Ubangi and Bas-Uéle Haut-Uélé Total angala North Difficulties in distributing the proceeds 62.5 81.0 46.2 64.6 Others won't work hard 31.3 19.0 46.2 30.5 Cannot trust the others 6.2 0.0 7.6 4.9 Total Percent 100.0 100.0 ’ 100.0 100.0 Crop Imposition In 1972-1973 we found that on the average, 86.0 percent Of the sample farmers in the North were under Obligation to grow cotton: 83.1 percent in LIbangi—Ibngala, 83.9 percent in Bas-Uélé and 100.0 percent in Haut-Uele . The average acreage imposed was 47 . 0 ares in Ubangi-Mongala , 47.8 ares in Bas-Uélé and 49.7 ares in Haut-Uélé. However, only 52.7 percent Of the farmers knew the acreage imposed. Regional decrees 175 signed by the regional commissioner provide the legal frae-work for required cultivation.6 Execution is up to the local authorities, mainly the territorial agronorist and agricultural monitors (see Table 6.7) . Usually, the territorial agronomist comrands the agricultural monitors and the authorities of the local collectivity. Those generally exempted from cotton growing were people with a paid job , independent farmers, single persons, widows and people who were ill or Old. Ignoring the cotton growing requirements leads to stiff fines and/or jail terms enforced by the local police force. All the farmers interviewed were well aware of the consequences of not growing oatton. The fines ranged from 0.50% to 5.008,with a modal value Of 3.508. These fines may be combined with jail terms ranging from one week to 60 days, with a mode Of 25 days. It is up to the local law enforcerent units to deter- mine the exact punishment tO be applied. Enforcetent is usually carried out shortly after tle usual planting time Of cotton. The size of the imposition of the cotton field is usually depen- dent On the family situation Of the farmer. Single farmers need only to plant a small field, while married farmers with more than one wife are required to grow a large field, up to ore hectare in Haut-Uélé. According to Zaire Mggazine of Novetber 4, 1974 [Nr. 326], the Agricultural Commission Of Haut-Zaire region recomended a reinforcetent of "educative crops " in order to stimulate cotton production. On the _ average, 30.2 percent Of the cotton farmers were also required to grow other crOps such as coffee, rice, maize, manioc and groundnuts. 6Formerly tle governor Of a province. 176 Table 6. 7. Responses to Question: "Who Imposes Cotton Cultivation?" , Nortlern Zaire, 1972-1973 Who Imposes Ubangi and Bas-Uélé Haut-Uélé Total Mongala North Territorial agronomist 72.6 3.8 76.2 36.7 Agricultural monitors 24.2 96.2 0.0 56.9 Authorities Of the local collectivity 3.2 0.0 23.8 6.4 Total Percent 100.0 100.0 100.0 100.0 Labor Utilization Types and Units of Labor Used on Cotton Farms The majority Of labor used on cotton farms in northern Zaire was family labor. On the average, 3.89 units of labor were available per farm (Table 6.8) . All outside labor hired on cotton farms was day-labor, and day-laborers were only engaged on farms in the forest areas, except for one farm. Day-laborers were only hired for the task of felling trees and clearing the forest. For the North, cotton farmers hired on the average 0 . 43 day- laborers who worked, on the average, 3. 24 days on the farm on which they were engaged. The average wage paid was 15. 9K per day (15.6 K in Ubangi-Mongala and 20.0K in Bas-Uélé) . They always received food and drink during their work. labor in mutual exchange or communal labor is fairly comron in Zaire. On the average 6.05 persons participated in the mutual exchange of labor for an average duration Of 3. 1 days per person. 177 Table 6.8. Types of Labor Used on Cotton Farms in Northern Zaire in Percentages and Units of Faily labor Available, 1972-1973 Types of Labor Ubangi and Bas-Uélé Haut-Uélé Total Mongala North Olly family labor 79.1 90.0 100.0 85.5 Predominantly faily labor 17.9 10.0 0.0 12.8 Predominantly hired labor 3.0 0.0 0.0 1.7 Units Of Faily Labor Availablel 4.57 3.00 2.76 3.89 1?. labor unit is a unit-measure of labor capacity in a family and is defired as equal to 1.0 for adults between the ages Of 16 and 64 years and 0.5 for those under or equal to 15 years or above or equal to 65 years. Man-Weeks of Labor Input in Cotton Production In Table 6.9 the labor inputs for cotton production in man-weeks per hectare are presented. A man-week is defined as five days Of labor performed by an adult male or female. Saturdays are usually reserved for self—help activities on roads, schools, etc. , Sundays are for rest, although most village markets Operate on Sundays . The data are rough estimates as they were based on only two visits per farm in the survey, me before and one after cotton harvest. It is impossible to collect accurate labor figures for a crop with only two visits per farm.7 The data in Table 6.9 are to be considered upper limits. When a farmer declares that he and his family worked for three weeks weeding a cotton field, it is understood that he also carried out his regular 7Collinson, however, argues that limited visit surveys are adequate for the collection Of labor data for planning purposes when the farming system is relatively simple. More on this point is found in Collinson [1972], Collinson [Decerber 1974] and Spencer [1972]. 178 Table 6.9. Labor Input in Man-Weeks for (he Hectare Of Cotton, Northern Zaire, 1972-1973 Sub-Regions land Planting Maintenance Harvest and Total Preparation Seeds Sorting Ibangi and 5.0 4.5 18.7 34.5 62.7 Lbngala Bas—Uélé 20.0 12.3 . 22.6 49.4 104.5 Haut-Uélé 23. 9 18. 7 36 . 3 69 . 7 148. 6 Tbtal North 10.9 8.1 21.6 42.0 82.6 subsistence activities such as tapping palm wine, hunting, fishing, harvesting food for home consumption, maintaining his food crop fields, repairing his house, visiting friends and neighbors, etc. The smaller the cotton field, the larger the effect Of overestimation is likely to be as famers are less occupied on small fields corpared to large ones, other things being equal. It does appear from Table 6.9 that the smaller the cotton fields, the higher the labor inputs per hectare. Since the labor figures were converted to a per hectare basis, this may help to explain the seemingly excessive labor inputs for farms in Haut-Uélé where field sizes are the smallest, an average 0.25 hectares compared to 0.40 hectares in Ubangi-Mongala. 8 We will not utilize the data from Table 6.9 for the calculation Of returns to labor nor for the relative profitability of different crops . 8Labor data were only derived for 19 farms in Haut—Uélé coIpared to 67 farms in Ubangi-Mongala. 179 Arduousness of labor and Labor Bottlenecks on Cotton Farms Famers were asked when they were the busiest not only because Of cotton but also as a result Of work for other crops grown on the farm. The purpose of this question was to collect information on labor bottle- reeks and peak seasonal labor demands. The results are presented in Table 6.10. Table 6.10. Period of the Year in Which Farmers were the Busiest, Northern Zaire, 1972—1973 Months Ubangi and Bas-Uélé Haut-Uélé Total Mongala North December, January, February 80.6 30.0 36.8 60.3 July, August, Septetber 19.4 70.0 63.2 39.7 Total Percent 100.0 100.0 100.0 100.0 The question was not well understood by the respondents. Several farmers equated "most" labor with "arduous" labor or answered that they were busy all year round. Unquestionably, although felling trees and clearing land in the savannah areas are arduous tasks which are per- formed by men, it is not necessarily the busiest time. Since most respondents were men , the answers were probably biased towards tasks performed by men. Felling trees and clearing land are done from mid-Novaber to the end of February during the dry season, coinciding with thecotton and coffee harvesting season. Typically, harvesting and sorting Of cotton is done by women and children who are occasionally aided by men. The majority Of the farmers in Ubangi and angala indicated that this was 180 the busiest season. In the Uélé's July, August and Septerber were believed to be the months with most work, followed closely by Decerber, January and February. In June and July cotton seeds are planted during the short dry season. Crop maintenance , especially thinning and weeding during the first two months after planting, is critical and is believed to explain a large part Of the variation in yields. When thinning and weeding start too late, the fields rapidly becore overgrown with weeds, and yields drop considerably. Moreover, weeding in cotton fields coincides with weeding of food crops (maize, manioc, rice, groundnuts, etc.) and coffee. Although farmers consider harvesting and sorting Of cotton and clearing of new fields in December, January and February as their most difficult tasks, tle most critical activity affecting cotton yields is undoubtedly thinning and weeding Of cotton fields in July, August and September . Of the farmers questioned, 58.8 percent indicated that they could grow more cotton with the labor available on their farm. Farmers gave the fol lowing reasons why they did not produce more cotton: low cotton prices, old age and poor health were the most frequent answers. Road Maintenance Farmers are required to help repair the roads which pass through their villages. (h the average, 83.7 percent Of the farmers had to per- form this service for an average duration of 6 . 7 weeks per year. The authorities who order farmers to work on the roads are predominantly from the local collectivity or from tte village or clan (Table 6.11) . Work on roads Often conflicts with activities in cotton fields ; 88.1 181 percent of the farmers responded that road work interfered with their cotton Operations. Work Performed by Farmers for the Head of the Village or Clan Some cotton farmers also had to work for the chief Of the village or tl'e head of the clan; 41.4 percent Of the farmers had performed duties for these authorities during the past year for an average 10. 5 weeks. These duties consisted of working on the fields of the chief, repairing his house, etc. Again, over three-fourths of the farmers declared that these duties interfered with their cotton schedule. Table 6. ll. Authorities Requiring Farmers to Work on the Roads, NOrthern Zaire, 1972-1973 Authorities Ubangi and Bas-Ué lé Haut-Ué lé Total Mongala North local Collectivity Authorities 60.3 46.7 41.7 53.3 Village or Clan Authorities 26.5 53.3 50.0 37.7 Territorial Agronomist 10. 3 0.0 0.0 5.7 Territorial Comissioner 2.9 0.0 8.3 3.3 TOtal Percent 100.0 100.0 100.0 100.0 Off-Farm Paid Labor and Legal Minimum Wages Only six farmers out Of the 129 respondents were employed outside their farms for part of the year during the survey. These farmers received an average Of 17.2K per day plus food and drink. They held a variety Of jobs: mason, judge, night guard, coffee sorter and day-laborer for felling trees . 182 Table 6 . 12. Average Number of Tools and Equiprent Found on Cotton Farms, Northern Zaire, 1972—1973 Tbols and Ubangi and Bas- Haut- Total Percentage Average No. Equipment Mongala Uélé Uélé North Of Farmers For Farmers Having Tool Having Tool Machete 2.86 1.94 2.38 2.56 100.0 2.56 Hoe 2.08 2.03 2.57 2.15 100.0 2.15 Axe or Hatchet 1.10 1.58 1.62 1.31 81.4 1.60 Spade 1.01 0.71 0.19 0.81 48.1 1.68 Coupel Coupe 1.29 0.36 0.48 0.93 55.0 1.69 Basket 1.68 3.36 3.86 2.43 62.8 3.88 1 A long knife bent at the end for cutting grass. Farm Tbols and Equipment NO machinery operated with a motor was utilized on cotton farms in- cluded in tl'e survey.9 As in most Of Zaire, animal traction does not exist in the North. Nbst farmers possessed only a small number of tools and equipment. Tte average number of tools and equiprent found on cotton farms is given in Table 6.12. The machete, hoe and axe were present on nealy all farms. Weeding is Often done by hand or with tie machete instead of the hoe. Most Of the equipment is fairly old and part Of it had been received from the Government, COGEICO of ONAFITEX. Farmers paid on the average 18.9K per year for repairing tools and equipment. Of the farmers responding, 45.7 percent knew a blacksmith in their village. Farmers had difficulties in buying new tools and equiprent not produced by local black- smiths. Several farmers complained about the high prices of new equipment 9Except Platz Sprayers powered by a two stroke engire used by ONAFITEX for spraying or dusting pesticides on some cotton fields. 183 and the absence Of improved tools in local stops. Obviously, ONAFITEX could play a useful role in this respect. Farmers were asked what tools and equipment they lacked most in order to improve their cotton production. Their responses are sumarized in Table 6.13. Table 6. 13. Percentage of Cotton Farmers Lacking Tools and Fquiprent in Order to Improve Cotton Production, Northern Zaire, 1972-1973 Tools and Ubangi and Bas-Uélé Haut-Uélé Total Equipment angala North Machete 80.5 80.0 71.4 79.1 Hoe 68.8 64.5 81.0 69.8 Axe 58.4 77.4 47.6 61.2 Spade 72.2 29.0 23.8 54.3 Coupe Coupe 32.5 6.5 33.3 26.4 Wl'etstore 23.4 9.7 0.0 16.3 Rake 16.9 9.7 0.0 12.4 Hammer 10.4 12.9 0.0 9.3 long-Handled Hoe 11.7 0.0 0.0 7.0 Pickaxe 6.5 3.2 0.0 4.7 Fertilizer and Pesticide Use No farmer in my sample had ever applied chetical fertilizers On his cotton farm. Nbreover, tley were generally unaware of chemical fertilizers. Chemical fertilizers are available in Zaire cnly in the major toms and are used in coffee and Oil palm plantations. Farmers are more familiar with pesticides and, in fact, confuse pesticides with chemical fertilizers. "drug" to make the cotton plants healthier. When the enumerators They call them both "Nkisi" or 184 explained the use and effect of chemical fertilizers and pesticides to farmers, most expressed an interest in securing these products. Eighteen percent Of the farmers recalled that COTONCI) had applied pesticides on their cotton fields before independence. A total of 10.5 percent of tle farmers in Ubangi—angala reported that their cotton field(s) had been treated with pesticides by ONAFITEX, without charge, during the 1972-1973 campaign. Many farmers complained that only the fields along the roads were treated. Several agricultural monitors also rerarked that He treatment came too late in the season to be effective. Of the famers questioned, 19.4 percent recalled that someore had explained to them low to use and how they could benefit from pesticides. In surmary, cotton farmers in the North know very little about clemical fertilizers and pesticides . Cotton Seed Delivery of Cotton Seed Agents of ONAFITEX and agricultural monitors are responsible for the distribution Of cotton seeds to farmers. The seed is usually stored in a sack, suspended from the ceiling Of the cotton storage shed. On the aver- age, farmers received cotton seed in 1972 during the first week Of May. Sore farmers received seeds in March while others had to wait Imtil June or even July. A total Of 12.9 percent of the farmers complained that they had received the seeds too late, on the average 4.8 weeks too late. Only 3.9 percent of the farmers knew the seed variety they had received. 185 Planting Of Cotton Seed No precise data could be gathered on the date of planting since farmers had difficulty in recalling the day or week of planting. Some fields were planted by the end of June, most were planted in July and the rest in early August. Gererally, however, fields were planted one to three weeks too late for Obtaining maximum yields. According to tie C.F.D.T. Mission in Ubangi-angala, every day after Optimal planting time reduces seed cotton yields by 5 kg. per ha. In addition, late planted cotton is more vulnerable to insect attacks, particularly dysdercus, at the end Of the growth cycle during the dry season. Farmers usually had a surplus Of seeds left after planting.10 Farmers put in, on the average, 5.4 seeds per pocket. Crop Rotations and Duration of Fallow Nbst farmers followed the recommended crop rotations in their region. In forest areas cotton is grown as the second crop in the rotation after a food crop, usually maize but sometimes groundnuts, while in savannahs cotton is planted as the first crop in the rotation. The basic crop rotation patterns found in the North are: Forest Areas Main Crepe Growing Season Variations Maize March-June or groundnuts Cotton J uly—December Groundnuts March-June or maize or manioc Manioc (and bananas) July-. . . sometimes rice be- tweel groundnuts and manioc In areas where rice is an important crop, the rotation is Often as follows: loSome farmers burned the surplus seeds on their fields to drive away insects and monkeys. 186 Main CrOps Growing Season Rice JIme-Decelber Groundnuts or maize March-J1me Cotton July-December banioc (and bananas) February-. . . Fallow Savannah Areas Main Crops Growing Season Variations Cotton June-Decetber GrOLmldnuts March-June or maize Manioc (and bananas) July-. . Fallow Cotton is almost never grown in a pure stand but always mixed with other crops, such as manioc, vegetables, Oil palm, papaya, pireapples, gourds, sesame, maize, groundnuts, sugar cane, bananas, tabac, etc. Still 8.4 percent of the farmers have practiced tre same rotation year after year since the days of the colonial administration or since they began cultivating their own fields. The average fallow time was 6.5 years which is far below the recomrended time Of 10 to 20 years. The relatively short fallow time is a factor which helps to explain the declire in yields since 1960, parti- cularly since cotton cultivation rapidly depletes the soil Of its nutrients. Farmers in forest areas return to the same field much more freqrently than they do in savannahs. The Cotton Field The location Of Cotton Fields Cotton fields were located on the average 1.9 km. from the tome of the farmer. The main reason for such retoteness of many cotton fields seers twofold: a search for the most fertile soils and an attempt to dis- courage agricultural monitors and authorities from visiting the field. 187 Twenty-eight percent Of the cotton fields in Ubangi-Nongala were grouped in one large bloc in the village, 0.0 percent in Bas-Uélé and 15.0 percent in Haut-Uélé. Such grouping greatly facilitates treatment with pesticides and visiting by agricultural monitors. It was predaminantly agricultural monitors and village or clan authorities who decided where the cotton field sl'ould be located. In fact, only about one-third Of the farmers determined the location Of tleir own cotton fields (Table 6.14) . Table 6.14. Persons Who Determine the location of Cotton Fields, Northern Zaire, 1972-1973 A Ubangi and Bas-Uélé Haut-Uélé Total Who Determines Mongala Village or Clan Authorities 44.4 12.9 16.7 32.4 Farmers Themselves 30.9 41.9 20.8 31.6 Agricultural Monitors 16.1 45.2 54.2 29.4 Tefiitorial Agronomist 8.6 0.0 8.3 6.6 Total Percentage 100.0 100.0 100.0 100.0 Number and Size Of Cotton Fields Farmers could Obtain as IIILlCh land as desired for cultivation, and re farmer reported any difficulty in getting as much land as he wanted. The size of cotton fields was predominantly determired by agricultural monitors and, to a lesser degree, by the territorial agronomist and village or clan authorities. Farmers themselves determined the size of their own cotton fields only in one case out of every ten as illustrated in Table 6 . 15. 188 Tle total number of cotton and other fields included in the survey and the average number of fields per farm are presented in Table 6.6. by the enurerators . Table 6. 15. Zaire, 19 72-1973 Only 7.9 percent of the cotton fields were recognized as pure stands In Bas—Uélé most farmers had only one field, the Persons Who Determine the Size of Cotton Fields, Northern Who Determines Ubangi and Bas-Uélé Haut-Uélé Total Mongala Agricultural monitors 72.2 87.1 85.7 77.9 Farmers Themselves 11.4 12.9 0.0 9.9 Mritorial Agronomist 13.9 0.0 4.8 9.2 Village or Clan Authorities 2.5 0.0 9.5 3.0 Total Percentage 100.0 100.0 100.0 100.0 I Total Number Of Fields in the Survey and Average Number Table 6. 16. of Fields Per Farm, Northern Zaire, 1972-1973 Sub-Regions Number Number Number Total Average Average Cotton Cotton Other Number Number of Nmber Fields Fields Fields of Fields Per of Cotton in Pure with Fields Farm Fields Stand Inter- Per Farm cropping Lbangi and 4 71 97 172 2.5 1.1 mngala Bas—Uélé l 31 2 34 1. l 1. 0 Haut-Uélé 5 15 26 46 2 . 3 1. 0 Tbtal North 10 117 125 252 2 . l 1. l 189 cotton field. Thus, in order to provide their family with food, they had to mix food crops with cotton, although this practice is discouraged by cotton extension workers. In other sUb-regicns farmers usually had at least one field with.food crOps in addition to the cotton field. Average field size, average cotton acreage per farm.and the acre- age in other crops are given in Table 6.17. The largest fields and the largest number of fields per farm were in Ubangi and angala. There were, on the average, more than 50 percent more labor units available on these farms than in the Uélés (see Table 6.8). The average cotton acreage per farm was small, 35.7 ares less than one acre (40.5 ares). There was a large variation in cotton field sizes between farms, as evidenced by the large standard deviations. The average cotton acreage per farmeas larger in savannahs than in forest areas (Table 6.18). The difference was more pronounced in Ubangi and Mongala than in the Uélés and was prObably related to dife ficulties Of land clearing in forests. A comparison between the 1970-1971 cotton acreage data per farm from the Census Of Agriculture in Zaire and from the 1972-1973 farm business survey for the same farms is given in Table 6.19. The data are quite similar, except that cotton acreages per farmlwere much larger in Baa-Uélé in 1970-1971 than in 1972-1973. Cotton acreages per farm were also larger in forest areas than in savannahs, tte opposite Of the 1972-1973 data. Termite Hills Ttere were termite hills in 33.6 percent of the cotton fields, on tle average 2.2 hills per field. Tie termite hills, mostly fossil 190 Table 6.17. Average Size Of Fields, Average Acreage per Farm and Average Total Farm Size, Northern Zaire, 1972-1973 (in Ares)1 Sub-Regions Average Average Average Average Average Size Size Cotton Acreage Total Farm Cotton Otler Acreage Other Size Fields Fields Per Farm Crops Per Farm Lbangi and 35.1 (20) 52.8(108) 40.0(20) 73.6(144) 113.6(160) angala Bas-Uélé 32.1 (21) 15.8( 8) 33.2(21) 1.1( 4) 34.3( 27) Haut-Uélé 25.1 (19) 23.7( 17) 25.1(19) 28.6( 20) 53.7( 30) Total North 32.8 (21) 46.5( 96) 35.7(21) 47.3(111) 83.0(126) 1The figures between brackets are the standard deviations . Table 6 . 18 . Average Cotton Acreage Per Farm in Forests and in Savannahs, Northern (in Ares) Zaire, 1972-1973 Sub-Regions Forest Zones Savannah Zoes Lbangi and Mongala Bas-Uélé and Haut-Ué lé Total North 38.6 47.2 29.5 31.7 35.0 38.4 191 Table 6 . 19. Average Cotton Acreage per Farm for 1970-1971 (Census of Agri- culture Data) and for 1972-1973 (Farm Business Survey Data), Northern Zaire (in.Ares)l SUb-Regions 1970-1971 1972-1973 Ubangi-Mongala 41.3 (18) 40.0 (20) Bas-Uélé 49.4 (31) 33.2 (21) Haut-Uélé 22.7 (17) 25.1 (19) Tbtal North 40.1 (24) 35.7 (21) Forest Areas 40.8 (23) 35.0 (25) Savannahs 37.8 (27) 38.4 (25) 1Data for the 1970-1971 Census of Agricul- ture in Zaire is derived from a preliminary computer print-out Of tle results of the Census; the figures between brackets are the standard deviations . and soletimes up to 4 or 5 meters high, are not cultivated as most Of the soil in the hills is sterile. The hills were also serious obstacles in measuring field sizes, and will hamper mechanization. Plant Densities Cotton plant densities on cotton fields in northern Zaire were fairly uniform as illustrated in Table 6.20. Sore farmers used a knotted rope to determine the spacing between plants and between rows. The average distance between rows was less than the recommen®d ore (80 cm.) in tle Uélés. The average distance between plants in the row in all sub-regiors was much greater than the recomtended distance (30 cm.) particularly in the Uélés. 11 11A spacing Of 80 cm. x 30 on. gives theoretically 41.625 pockets per ha. With two plants per pocket, this is 83,250 plants per ha. 192 Tie C.F.D.T. Mission in Ubangi-lbngala recommends a minimum density Of 40,000 pockets per hectare and thinning to two plants per pocket. Higher densities did not increase yields. Below this density, yields in kg. per ha. decreased. Thus the plant densities Observed in cotton fields in the northern belt were well below tle recomended density. Table 6.20. Cotton Plant Density and Spacing Between Cotton Plants, Northern Zaire, 1972-1973 Sub—Regions Number of Distances in om. Distances in cm. Plants r Between Lires Between Plants Hectare in the Line lbangi and angala 32, 778 79 . 6 40 . 1 Bas-Uélé 32, 889 71. 5 55. 8 Haut-Uélé 38,556 60,0 57.7 Total North 33,556 77.9 42.6 lMeasured in a square of 9 m2 (chosen at random in the field) and converted to a field size of 10,000 m2 (1 hectare). Field Maintenance and Harvesting Farmers thinned cotton plants on the average 2 .7 weeks after planting. No cotton fields were ridged. On the average, fields were weeded 2.5 times. Farmers generally pick their cotton crOp too late. As a result, the ripe cotton bolls are exposed to rain and insect attacks and many cotton bolls fall on the soil and are lost. Most Of tlre farmers, though not all, burn cotton stalks and debris after harvest as prescribed by law. 193 Cotton Diseases Although GTAFITEX has been treating some cotton fields with a combined insecticide—fungicide beginning in 1972 many fields are left untreated, and the timing and the rate Of application are far from ideal. Tle dispersion Of cotton fields is a major bottleneck in pesticide treatment. A systeratic study of the evolution of cotton pests in various regions has not been undertaken although ONAFITEX recognizes the impor- tance of such a study. Ore important lesson leaned during the inten- sification of cotton production before 1960 and, more recently, in Eastern Kasai (FED Project) and in FIWA-BILI is that cotton pests tend to become more important when production and yields rise in an area. Number and Size of Otter Fields From Table 6 . 16 we can derive the average number Of other fields per farm as 1.0 fields. The average size of other fields is presented in Table 6.17. Coffee is the most popular crop besides cottcm with an average 0.46 fields per farm, while the most important food crops are rice and manioc. The average field size for the North was 46 ares, larger than the average size Of cotton fields which was 33 ares. Farm Sizes There are several ways of defining farm size in African agriculture. (he way is to defire farm size as the sum Of all acreage under cultivation at a given point in time. Table 6.21 provides us with data on fanm sizes Of cotton farms in nortlern Zaire according to this definition. Farm sizes vary widely between sub-regions and between farms, and the standard 194 deviations for farm size are high, particularly in Ubangi-angala (160 ares) where the standard deviation is larger than tl'e average farm size. One farm had 1,020 ares Of coffee and 1,000 ares of Oil palm and a total fanm size Of 1,276 ares. If this farm is left out of the sample, however, average farm size for Ubangi-angala drops to 85.4 ares with a standard deviation of 58.9 and for the North, to 66.17 with a standard deviation Of 53.0. Table 6.21. Average Total Farm Size and Standard Deviation, Northern Zaire, 1972-1973 (in Ares)1 Sub-Regions Average Total Standard Farm Size Deviation tbangi and Mongala 113.6 160.0 Has-Uélé 34.3 26.5 Haut-Uélé S3 . 7 30 . 3 Bas- and Haut—Uélé 41.0 29.2 Total North 83.0 126.0 lMeasured after planting of cotton and before harvest. Since farmers always have fields in fallow we can include them in anotl'er definition Of farm size. We can define farm size as the acreage Of land under cultivation plus the acreage Of land in fallow. If we defire r, the coefficient of rotation, as the ratio Of acreage under cultivation and acreage under cultivation plus acreage Imder fallow, then r = 0.20 means that only 20 percent Of the land Of a farm at a given point in time is cultivated. In the section on "Crop Ritations and Duration Of Fallow" we listed average fallow times per 195 sub- region for crOp rotations with cotton. 'Ite average cultivation time of a crop rotatiol with cotton was about three years. The rota- tion coefficients per sub-region are, tlen, as follows: Ubangi-Mmgala: 3.0 / 3.0 + 5.2 = 0.37 Bas-Uélé: 3.0 / 3.0 + 9.0 = 0.25 Haut-Uélé: 3.0 /,A 3.0 + 8.0 = 0.27 Total North: 3.0 / 3.0 + 6.5 = 0.32 If we assume that these rotation coefficients apply equally to all fields on cotton farms, then farm size is as follows: Ubangi-Mangala: 113.6 x 1 / 0.37 = 307.0 ares Bas-Uélé 34.3 x 1 / 0.25 = 137.2 ares Haut-Uélé 53.7 x l / 0.27 = 198.9 ares Total North: 83.0 x l / 0.32 = 259.4 ares With this definition, the differences in farm sizes between sub-regions are less pronounced because small farms in terms of acreage under cultivation (for instance in Bas-Uélé) have longer fallow times than the larger ones. The larger the farms, the shorter fallow times tend to become. Cotton Yields Table 6.22 presents the average physical yield per farm in kg. per ha. and in kg. per farm together with average cotton acreage per farm. The average yield in kg. per ha. is coIputed by dividing the survey production by the survey acreages per sub-region . T‘te average yield per farm in kg. per ha. differs widely between sub-regions. Yields in Bas-Uéle’ are more than double those in Ubangi- Mongala. Yields also vary considerably between farms as shown by the 196 .839?an some you concoct hogan on tongues coflosoono >o>nom . magnetron Emocmum our momentous c0203 monsofl draw H mom Hmm ammo mom 8m: some 33 5mm 5.82 H38. mmm mom Game ems Amm V am 35 1mm WHWquomm mama mm 8va 5o :.va 0mm 33 m.mm mansion can as some amm 83 m2 ~88 cos 8889‘ not Honda .M :..” 6mm Hod 0555mm ado Hod .ox .om woo .om couuoo 9." page :.." 6mm mom EHmm woo soapstone 5mm moo communes monomers connotes new; dosage. couuoo dosage c0300 connotes mCanothm memalmhma deans fiofiuoz .:fimm woo mooQfim couuoo mooHo>< one macaw woman?» .Enmm Moo cogooum couuoo downward .cfimm moo moomho< coupon cookware 197 high standard deviations. And tlere is an inverse relation between cotton acreage per farm and yield : the higher the cotton acreages, the lower the yields in kg. per ha. The higlest yields were Obtained in the Buta and Babesa zone in Bas-Uélé with an average yield per farm for these two zones Of 908 kg. per ha. (standard deviation 467 kg.; 16 observations). The highest farm yield was in Kmongino, locality of ange, Bambesa zone with 1,704 kg. per ha. (201 kg. on a field of 11.7 ares). Of tle fields in the survey 2. 7 percent had a zero yield as ttey were completely over- grown by Imperata Cylindrica. The average cotton revenue per farm in K is also presented in Table 6. 22. Revenues in Bas-Uélé were more than double those in lbangi-angala and Haut-Uélé, mainly because yields in kg. per ha. were much higher in Eds-Uélé. On the average, farmers earred with cotton production a gross revenue of 8. 79 8 Cotton Prices Knowledge of Cotton Prices Of the farmers interviewed, 17.6 percent believed that the seed cotton price was a fixed price which did not change from year to year, while 30.2 percent Of the farmers could not recall tle price they had received during the previous year. Several farmers did not understand the concept of price; instead they considered the total amount Of morey which tley had received from their cotton sale as tle "price. " The agricultural monitors and the territorial agronomists were the principal sources of information about cotton prices (Table 6.23) . Nbst farmers leaned about the cotton price on the day Of the sale or only a few days be fore . ‘ 198 Table 6. 23. Sources Of Cotton Price Information for Fanrers, Northern Zaire, 1972—1973 Sources Of Information Ubangi and Bas-Uélé Haut-Uélé Total angala North Agricultural Monitors 48.1 60.6 50.0 51.6 Territorial Agronomist 36. 3 --- 20. 0 24.6 Nobody 10.4 33.3 -- 14.6 Village or Clan Authorities 1.3 6.1 30.0 6.9 ONAFITEX Agents 2.6 -- --- 1.5 Neighbors 1.3 -- --- 0.8 Total Percentage 100.0 100.0 100.0 100.0 Importance and Effect of Cotton Prices The price of cotton is an important factor in determining cotton acreages. Indeed, 67.7 percent of the farmers affirmed that the price over the last five years hai influenced tie size of their cotto'I holdings. Tte farmers were asked, respectively, at what price Of first grade cotton they would increase their cotton acreage, at what price they would dorble their cotton acreage, at what price they would decrease their cotton acreage , at what price trey would reduce their cotton acreage to half and at what price they would stop growing cotton. These questicns were posed assuring that there was no Obligatory cultivation. The enumerators were carefully trained to explain these questions to the farmers. Nevertleless, sore responses had to be discarded because sore of the farmers could rot answer the questions properly. The results Of this inquiry are presented in Table 6.24. Tiese data should be interpreted with caution. However, we do believe that 199 mo «:3 NA in m.m c5 52 5.82 H88. on S: c.m Tm ma TS ms outshone mo S: a; as ca 3: 5.3 newsroom panacea me can cg o.~ mm TS one can amend seesaw Hon 8 mo mane communes communes commune commune on road moans wane tombstone oouuoc couooc eonuoc cocooc Hoodoo . unannounom counoc doom 858m onpouooc menace unwound ncoacwmhsm cor woo .M 3 mmoaumv mhmaImsmH .mHHcN fimfiuoz .Hnmm .manmcommmm Econ mum—Eco fiance magnum can commence couuoo Hung cosmfi 3:03 magnum cog um couuoo noon dunno Hannah MOM mound 9,30%QO .sm .m manna 200 the data are indicative Of a gereral picture which can be portrayed as follows: farmers conceive present cotton prices as low as indicated by the high prices needed before they will increase and double their cotton acreages. Also, prices would have to decline by a large amount before they would decrease their cotton acreage . Prices could decrease by 25 percent and cotton acreage would not change very much; but if the price dropped 50 percent (from 6.00K.On the average to 3.00 K. per kg.) , the acreage would fall by the sae magnitude. Thus, at present cotton prices, farmers seem insensitive to small price changes. / Farmers reported that they considered 16.3 K. per kg. a fair or reasonable price for cotton which is more than double tle current price. The Uélés are more sensitive and responsive to changes in prices than Ibangi-Mongala. Farmers in the Uélés would increase cotton acreages sooner wten prices are rising, reduce acreages earlier when prices are falling and stop growing cotton at a higher price than the farmers in Ubangi-Nongala. They consider 10.7 K. per kg. a reaso'lable or fair price, at less than half the level which farmers in Ubangi-Nongala deem reasonable. The Risk of Growing Cotton For many farmers cotton is the only cash crop which procures a fair aount Of cash and which enables them to pay their head tax and other taxes. About half tl'e farmers interviewed (48.8 percent) believed tlat cotton had eaned tl'em a lot of money during the past years. But 96.1 percent of the farmers also considered cotton a risky crop, i.e. a crop which may yield a lot Of money in one year but not in another year. 201 Sale Of Seed Cotton Time and Conditions of Sale Most farmers (60.9 percent) sell their cotton at two periods of time: a first market for first grade cotton and a second market for the remaining first grade and all Of second grade. Those who sell all their cotton in one market either have no second grade cotton to sell or simply missed or skipped one of tie markets. The majority Of the farmers, 94.8 percent,sold all the cotton they had harvested. The reasons for not selling all cotton were several: absence of the farmer during a market, refusal to buy because of poor quality, late harvest or reroteness of a buying center from the hore- stead Of the farmer. Only 1.6 percent of the farmers did not store their cotton crop in baskets or on a drying rack in a cotton shed but instead dumped it on the earth floor. All farrrers but ore transported the cotton them- selves to the buying center. The buying center was on the average 1.88 km. from the house of the farmer. NO farmer knew in advance the weight of seed cotto'I he was bringing to tie buying center, and 2.0 percent Of the farmers saw their grade classification changed by ONAFITEX buying agents from first grade to second grade. In 1972 first grade cotton was sold on He average around mid-March, with extreres running from mid-January to October. Indeed, farmers never know exactly when their seed cotton crOp will be bought. On the average, 85.9 percent of the cotton harvest was classified as first grade and 14.1 percent as second grade. 202 Willingness to Join a Sales Cooperative Only 21.4 percent Of the farmers expressed an interest in joining a sales cooperative for cotton which would collect and weigh seed cottol, transport it to the ginning mill, distribute seeds and pay the farmers.12 Tte main advantages farmers expected from a cooperative were timely buy- ing-up Of seed cotton and tte services of a truck, while the principal reasons for rejecting a cooperative were examples of cOOperatives that had failed in the past and mistrust of cOOperative managerent. The Aggicultural Extension Service On the average 87.5 percent Of tie farmers knew an agricultural monitor (or agricultural extension agent) , either from the Department Of Agriculture, ONAFITEX or the local Collectivity. For those who knew me, 78. 6 percent believed that they had helped them in producing cotton. Help offered by agricultural monitors included recomrendations for finding a suitable location for cotton fields, determining the size of the field, crOp rotation, planting time, spacing, nurber of seeds to putineachpocket, timingandmethodof thinning andweeding, timeof harvest, sorting method and, in general, for every aspect Of the cotton growing process. They also inspected cotton fields and encouraged farmers to produce more. Sore agricultural monitors treated cottol fields with pesticides. Still, the single most important advice was the date of planting. 12This type Of farmer cOOperative, assuming the Phase 1 operations of cotton marketing, was specifically proposed in the cotton decree law of August 13, 1965 and is referred to in the ordinance law Of August 12, 1971 creating ONAFITEX. In this latter ordinance, farmer cooperatives are also prOposed for cotton ginning and, thus, for Phase 1 and Phase 2 Operations. ' 203 The territorial agronomist, a powerful administrator in each zoe, is less well known by the farmers. On the average, 67.2 percent of tie farmers knew him. Of those who knew him, 74.7 percent indicated that he had helped them in cotton production. His help consisted mainly of control of agricultural monitors and enforcerent Of cottol growing requirerents, particularly time Of planting and weeding. He was also closely associated with the fines imposed for negligence of cotton growing requirerents; thus, farmers feared him. Of the farmers questioned, 16.4 percent also knew other persons who had l'elped them in cotton production. These were mainly foreren in a village who worked under instructions from the agricultural monitors. Sore farmers also mentioned village or clan authorities who recomrended the same practices as agricultrmal monitors or territorial agronorists. Finally, 83.6 percent of the farmers believed that the help Of agricul- tural monitors and the territorial agronomist and capitas was needed to produce more cotton. The rerainder Of the farmers, 16.4 percent, believed that their help was redundant. Livestock and Poultry Farmers in northern Zaire generally do not keep large animals. Some village chiefs keep a small herd of cattle, but this is rather exceptional. The area is infested with tsé-tsé flies except in He northern savannahs where there are large cattle ranches. No farmer in the survey had cattle. Farmers in Ubangi-Mongala kept more animals than those in He Uélés. A possible explanation is that most farmers in the Uélés have not been able to build 1p their animal stock because Of the massive 204 slaughtering Of dorestic animals during the Congo rebellion: Farmers in savannahs also kept more animals than farmers in forest areas. Goats were the most popular large animals on farms with an average of 0.91 goats per farm. Hogs, guirea pigs, rabbits and hares were found only in tbangi—Mongala. Poultry was very popular in all regions with an average of 9.43 units per farm. Ducks were popular in Ubangi and pigeons in Bas-Uélé. Farmer's Revenues, Elpenditures and Savings Sources of Farm Revenue Food crops and nonfood crops are about equally important in pro- viding reventes for farmers. In Ubangi-Mongala food crops are the primary source, while in Bas—Uélé nonfood crOps, particularly cotton, are far more important than food crops. Table 6.25 provides a ranking of the major sources of sales revenue from food and other crOps. Table 6.25. Sources Of Farmers' Revenues, Northern Zaire, 1972—1973 Source Of Ubangi and Bas-Uéle’ Haut-Ué lé Total -Revenue mngala North Food Crops 59.0 22.6 47.6 48.5 Other Crops 41.0 77.4 52.4 51.5 Forty-two percent of the farmers ranked cotton as their first and 23 percent as their second most important source of sales revenue. For 18 percent of tre farmers, coffee was the main source Of revenue. Cotton is by far the primary source in Bas-Uélé where 84 percent Of the farmers ranked cotton in first place, while in (bangi—angala coffee 205 was judged more important than cotton (29 percent versus 23 percent). Of the farms in Ubangi-Molgala, 22.1 percent are in a coffee paysannat and this may explain why coffee ranks first. Taxes Paid by Farmers All able-bodied adult males have to pay a personal minimum tax. On the average, 81.4 percent of the cotton farmers paid the head tax. The following are exempt from the tax: Old farmers, sick or handicapped farmers , divorced wives , war veterans and tte village or clan authorities . This tax is collected once a year and the collection usually coincides with the market for seed cotton. The tax levied was 2.00% or 2.80%, depending on the location. Farmers with two wives paid double the tax rate. Tl'e average tax collected was 2.478 per farmer. Most farrrers also pay otl'er taxes in addition to the head tax. These taxes are either a merbership fee in the political party,usually 20 K. per person, a road tax (taxe de voirie), 20 K. to 50 K., a bicycle tax, from30K. tolOOK, orataxforthesaleofpalmwire, from 100 K. to 600 K. Ore Of these additional taxes was paid by 64.3 percent Of the farmers, and tie average amount paid per farmer was 40.7 K. per year. The tax was usually collected once a year. Sore Of the farmers, 8.5 percent, also paid taxes in kind. Usually, trey were gifts in kind to authorities from the village or from the local collectivity. Exchange of Nbrey and Products with Family Merbers in Cities Farmers Often exchange gifts with members of their failies or clan who are residing in cities. In fact, 10.1 percent of the farmers 206 received gifts from members of their families living in a city. These gifts were usually cash, on the average 9.85% in 1972. In return, farmers give food products to their family mothers in cities; Of those in the survey, 15.5 percent gave focd.prcducts to family members in the city. The usual gifts were poultry, slain wild animals (especially monkeys), bananas, termites, goats, fish, gromdnuts, paddy, palm Oil, beans, etc. The average total value Of these gifts in 1972, at local prices, amounted to 2.38% Savings, Lending, Borrowing and Expenditures The proceeds from the sale Of cotton are either kept by the farmer himself or tte farmer and his wife split tie cash (Table 6.26) . Table 6. 26. Responses to Question: "Who Keeps the Proceeds from the Sale of Cotton?" , Northern Zaire, 1972-1973 (In Percent) SUD-Regions Farmer l Wife | Splitting Percent lbangi and Mongala 63.3 11.7 24.7 Bas-Uélé l3. 3 16.7 70.0 Haut-Uélé 11. 8 0. 0 88 . 2 Total North 44.3 11.3 44.4 The men spent tie money mainly on clothing, taxes and drink, the women on clothing and food. All farmers but one kept their money at here; one farmer in Haut-Uélé deposited his cash in a savings account in a savings and loan institution (CADEZA) in the nearest city. In many areas Of Zaire and particularly in the cities, a mutual 207 savings plan, called "likelemba," is very popular. It is a sort Of savings club with rotating disburserent. A number Of individuals pool tleir money at regular intervals and, in turn, each one Of the parti- cipants receives all the cash; 21.9 percent of tie farmers participated in such a scheme. T're average number Of participants was 8.15. Tle average contribution was 1 . 42% j . Farmers were asked about their lending and borrowing practices . On the average, 32. 6 percent Of the farmers had loaned morey to other persons in the past. No interests were paid, in fact, He concept Of interest was never encountered in these transactions . Tie average amount of money advanced was 2.14%. With respect to borrowing, 14.0 percent of the farmers had borrowed morey in the past. The average sum borrowed was 1. 24% and repayment usually occurred after tle sale of agricultural products. When asked what trey would do if trey ear-red an additional 10%, or 30%, or 1008, farmers frequently mentioned, among other things, tte hiring Of day-laborers for their cotton or coffee field and, for tie 1008 question, the Opening Of a trading business. Visits to local Markets Tre market visited most frequently by farmers was, on the average, 9.6 km. from his house; 80.6 percent Of them regularly returned from the market with unsold products. Twenty-six percent Of the farmers went to the city on a regular basis to buy and sell products. he city was on tlre average 35.5 km. from the village Of the farmer. Nbst farmers (67.2 percent) believed that the shops and markets in their area were sufficiently stocked with products. 208 Dowry Traditionally, the bridegroom has to pay a dowry to the parents of the bride. Part of it is usually paid at the time of marrying and tie remainder is paid after children are born. Although the amount and cmposition of the dowry vary widely from region to region and from ore ethnic group to another, the domyusually consists of a cash gift and gifts in kind. Farmers in the survey were asked what they would cbmand as dowry for one of their daughters. Answers seemed to depend on the wealth and starxiing of the bridegroam's family. Also, if the bride had been to scl'ool, the dowrywas usually more than double tle rormal mount. 'ne following average figures were obtained for the cash share of the cbwry: for Lbangi-angala, 48.88, for Bas-Uélé, 14.85, and for [hut-Uélé, 13.38. The average for He North was 34.8%. Thus, the dowry represents a considerable outlay of money relative to the farmer's incate. Manyycungmenare unablempay thedcwryorcanmlymarry in their late 203 when they have amassed enough cash. Farmer's Possessions Allfarmers livedinhousesmadecfmmdandwoodstickswitha thatchroof. Norehadatinroof. Masthouseshadawoodenentry chor which could be locked. 'Ihe degree of material wealth could be measured in the possession of tte following goods: bicycle, radio, sewingmachineandwatch. Sale famersalsohadshotgtms, altl'ough to accurate information could be collected on this item became they hid the possession from authorities. The percentage of cotton farmers whoamedmeofthesegoodsisgivenin'rable 6.27. ' 209 Table 6.27. Percentage of Cotton Farmers Who Owned a Bicycle, Radio, Sewing Machine or Watch, Northern Zaire, 1972-1973 Sub-Regime Bicycle Radio Sewing Machine Watch Ubangi and Mongala 14.3 5.2 7.8 1.3 Has-Uélé 35. 5 12.9 9.7 -- Haut-Uélé 19 . 0 4 . 8 ‘ --- --- 'Ibtal North 20.2 7.0 7.0 0.8 On tre average, one out of every five farmers owned a bicycle . These bicycles are always used to transport produce to markets. They increase the range of sales opportunities open to farmers and bring them closer to urban centers, and are usually a precondition for opening a small shqa for petty trading which may eventually evolve into a larger shOp. Seven percent of the farmers also owned a small transistor radio. 'Ihis is probably very important in shaping the beliefs and values of farmers. It is also a potential source of information for better methods of agricultural production, although, presently, the radio is not much used for agricultural extension. Farmer ' s Attitudes Regarding Cotton Production Based on (pen-Ended Questions At the end of tie survey farmers were asked a series of open-ended questions in order to learn about their attitudes, beliefs and values regarding agriculture and, in particular, the production of cotton. Do People Like to Grow Cotton? A large majority viewed cotton as an important source of money, while some farmers also thought of cotton as a source of food. Cotton 210 is part of the food crop rotation and morey from the sale of cottm enables them to buy dried fish, salt, and other food products. Across all sub-regions the majority of the farmers indicated that they like to grow cottm. ally 16.3 percent of the farmers complained that the cotton price was too low. Why 0011' t You Produce Dbre Cotton? The answers varied among sub—regions . In Ubangi—mngala low seed cotton prices were given as the main reasons for not producing more. The other main reasons given were poor soils, bad seeds, illresses and old age ard a variety "of other constraints. In the Uélés illnesses and old age were viarled as the main reasons; farmers complained about their poor health. In Bas-Uélé 34.6 percent of the farmers believed that trey work enough already and can not produce more . Why Do Same Farmers Abandon Cotton Production? This question overlaps with "Do People Like to Grow Cotton?" Of the farmers in the survey, 67.8 percent said low cotton prices were the main reasm for abandoning cotton production. A similar reason was their preference for coffee which yielded higher returns to labor. Such respmses confirm the main reason why peOple like to grow cotton: for the money it provides. The second most important motivation for abandoning cotton, in tte eyes of tl'ose who grow cotton, was lack of discipline and motivation of farmers. Farmers in the Uélés saw this as the primary reason. In Has-Uélé 28 percent of the farmers also mentiored late sale of cotton as a reasm for abandoning production. 211 What Should Be Dore to Encourage More Cotton Production? This question follows logically the question of "Why Don't You Produce Mare Cotton?" Again, farmers in Ibangi-Mongala were the most price conscious; about 60 percent requested an increase in price. All farmers in tle North asked for farm tools, machinery and gifts from the goverment. During the (1710120 days, farmers regularly received free farm tools or at half price in addition to gifts of salt and otler products in proportion to their cotton production. ONAFITEX abolished this paternal practice although, daviously, farmers appreciated these gifts. mly me out of every ten farmers mentioned pesticide treatment as a stimulus to producing more cotton. In Bas-Uélé one out of every four farmers asked for measures to protect his fields from wild animal damage, mainly elephants and cynocephalus (monkeys). What Kind of Help Do You Like to Receive Fran the Government, ONAFITEX or Others to Produce Nbre and Better Cotton? 'Ihis question is complementary to the previous one. Again, farm tools, a bicycle and gifts received priority. Increasing the cotton price ranked low (3.2 percent of the farmers) but, tren, morey or credit were asked for by one-third of the farmers. Pesticide treatment was requested by only 7.1 percent of the farmers. No farmers mentiored better extension services, probably because they didn't like the paternal amroach of extension agents. 212 ém One—hundred and twenty-nine cotton farmers were interviewed twice- once before harvest and once after harvest--during the 1972—1973 cotton production year. The average age of the farmers was 44 years. The average years of formal schooling was 1.4. Each family was composed of 5.0 persons. Land tenure was no constraint on cotton production. On the average 86.0 percent of the farmers were under obligation to grow cotton. Stiff fines and/or jail terms were given to farmers who ignored cotton growing requirements. Only one—third of the farmers decided where to locate their cotton fields and only one-tenth determined their cotton field sizes. Agricultural monitors and to a lesser degree territorial agronomists and village or clan authorities were primarily responsible for determining the location and size of cotton fields. The average cotton acreage per farm was small: 35.7 ares, less than one acre. There was a large variation in cotton field sizes between farms. Average farm size was 83.0 ares (113.6 ares in Ubangi-bbngala, 34.3 ares in Bas-Uélé and 53.7 ares in Haut—Uélé) . The average seed cotton yield per farm in kg. per ha. was 399 kg. (257 kg. in Ubangi-Mongala, 687 kg. in Bas-Uélé and 435 kg. in Haut-Uélé) . Yields differed widely between farms and between sub-regions. On the average, farmers earned a gross revenue of 8.79% frcm cotton production. Cotton was ranked as first source of sales revenue by 42 percent of the farmers; 23 percent also ranked cotton as the second most important source. For 18 percent of the farmers, coffee was the first source. Farmers paid an average head tax of 2.478 per year. Most farmers also 213 had to pay other taxes . The average cash portion of the dowry was 34. 8%. Most farmers had to help repair the roads for an average duration of 6. 7 weeks per year and about half of the farmers also had to work for the chief of the village or head of the clan for an average of 10.5 weeks per year. This was a constraint on cotton production. The farmers possessed only a small number of machetes, hoes and axes. Chemical fertilizer was unknown to farmers. Sate farmers (12.9 percent) received cotton seeds too late. Fields were planted one to three weeks after the optimal time of planting. The recommended crop rotations were generally observed although cotton was almost never grown in a pure stand. Fallow time was on the average 6.5 years, far below the recommended period of 10 to 20 years. Cotton fields were located an average of 1.9 km. from the compound of the farmer. Cotton plant densities were fairly uniform but well below the recommended density. Cotton maintenance was generally poor and farmers harvested too late. Of the farmers interviewed, 87.5 percent knew an agricultural extension agent. The territorial agronomist was less known by the farmers. Most farmers learned about the cotton price on the day of sale or only a few days before. They were generally unaware of the day of collection. The majority of the farmers (67.7 percent) affirmed that the producer price over the last five years had influenced their decision to plant more or less cotton. Farmers conceived present cotton prices as low. However, fanmers reported that prices would have to fall considerably before they would reduce their cotton acreage . At present 214 cotton prices farmers seemed insensitive to small price changes. Farmers in the Ueles were the most sensitive to price changes. Almost all famers considered cotton a risky crop. Only (me-fifth of the farmers expressed an interest in joining a sales cooperative for cotton. Most farmers like to grow cotton for the cash incare it provides. However, they do complain about the low seed cotton price and the small amount of money they earn with cotton. They expect the government to raise cotton prices and provide farm tools, machinery and gifts as incentives to expand production. CHAPTERVII ANANAIXSISG‘THEREIURBB'IOIABORFCRCOI‘I‘NAND GIT-[ERSEUEXTIEDCQTDDITIESINBDKBIERNZAIRE Introduction The objective of this chapter is to carpare the returns to the factors of production on traditional farms in northern Zaire. In particular, the returns to labor will be analyzed because labor and land are tie dominant inputs in traditional agriculture in northern Zaire. A framework will be presented for comparing tl'e returns to labor for different crops. The real price of seed cotton since 1960 will be derived in order to carpare tre returns to labor in cotton production . overtime. Methodology The factors of production that are traditionally taken into consideration by farm management and production economics researchers are land, labor and capital. Managarent is viewed as the art of com-- bining these different factors. We assure that "average" management is incorporated in the labor input. No separate returns to management will be considered. Returns to land In this survey of 129 farmers none of the famers experienced difficulty in acquiring more land. As a result, we will consider land 215 216 a nonlimiting factor; farmers can obtain as much of it as they desire. Thus, no returns will be imputed to land in our analysis. Land is not a homogeneous input and, in reality, returns do accrue to fertile land above those to "average" land. It is reasonable to assure that farmers or those who select plots of land will choose the most fertile soils at the beginning of the crop rotation. These are really soils which have been under a lorg forest or savannah fallow. The longer land remains under fallow, the more labor is needed to clear it and the better is soil fertility. Returns to Capital Under preeent production technologies , very little capital is used in traditional agriculture in northern Zaire. This is true for food cmp Production as well as for perennial and export crops grown in traditioral agriculture. Capital items such as hoes, hatchets, machetes, axes,baskets, etc. are extremely limited. Sore of the equipment was supplied by C(IERCO as a bonus for cotton production or has been made In; the farmer himself such as baskets, storage sheds, drying radcs, etc. In the entire northern cotton belt we did not encounter any cotton farmers who applied demical fertilizers or pesticides. ONAFITEX did treat cottcn fields with pesticides, but the farmers did not bear the costs for these operations directly. For these reasons, we will assure that no returns accrue to the capital input. For perennial crops in modern plantations, capital is very important resource and cannot be neglected. Capital is used in tre form of soil improvererrts, infrastructure works, mechanization, pesticides, fertilizers, improved seeds, etc. and, in gereral, in forms which substitute for labor. 217 Crops in modern plantations will not be analyzed in this chapter as we are focusing on crops grown with traditional production practices. Returns to Labor labor is the mly factor of production which we will consider in this aralysis. It is the most important productive resource in tradi- tional agriculture in northern Zaire at this time and it constitutes tte only "cost" of production in our analysis. We will carpute the returns to this factor for cotton and other competitive crOps. labor Assurptiore It is assured that the expected returns to labor in different crops are the most important factors in tie decisim-draldng process of traditioral farmers. This is in conformity with the economic rationality of traditional farmers which has been docurented by several authors [Verhaegen, June 1967; March 1968 and Jones, May 1960]. I Labor inputs are measured in terms of duration and not in labor characteristics of arduousness and urgency of the task being performed. The latter are unccmron and difficult to measure. Thus, "normal" labor inputs on a given plot of land over time are considered. The unit of measurerent of labor is man days on a field of ore hectare (ha.). Men and waren are given the same weights in estimating standard man days, namely 1.0. The labor of children is not considered in deriving standard labor inputs. For annual crops grown in a rotation, the labor requirements for land clearing and land preparation are divided evenly among the four crops which are usually grown in a rotation: cotton, maize or paddy, 218 groundnuts or paddy and manioc. Thus, each crop is allotted one-fourth of total labor requirements for land clearing and land preparation on a new field. Detenmination of Standard labor Inppts for Different Crgps The farm business survey on cotton production in northern Zaire did not generate accurate labor data for different crops grown on cotton farms. This is one of the main, if not the main drawback of farm business surveys or limited visit surveys. We will utilize standard labor rates derived by INEAC and published in 1958 as "Normes de Main d'Oeuvre. . ." [INEAC, 1958]. No other precise labor input data are available for Zaire.l These labor norms were derived by INEAC over the years based on observations of hand labor methods in its agricul- tural experiment stations, and are considered "normal" labor inputs for a crop and averages for a production system. Actual rates of work within and between farms over time will vary according to local ecological conditions, agricultural methods, nonfarm activities, health of tie farmer, family composition, etc. The advantage of utilizing labor data based on experiment station conditions is that they were derived under controlled and uniform conditions. But there is no doubt that the actual labor inputs on farms are different from those derived under experiment station condi- tions. Although the actual values of data generated under experiment station conditions may be somewhat suspect, the relative picture revealed er. Paul Perrault, Stanford Food Research Institute, derived reliable labor data for banana production on traditional farms in tle Kisangani area over the 1973—1974 period. 219 is unquestionably an accurate one. As a corollary, the returns to leaor as calculated in this exercise prcbably do not reflect actual returns, but do indicate the relative returns to labor for different crops. (lily single crops or pure stands will be considered in this analysis.2 Since virtually no labor data is available for the dominant crop mixtures in Zaire we are forced to restrict our analysis to single crcps . Estimates of Iabor Requirements for Cotton Production The labor requirements for cotton production in forest and savannah areas are given in Table 7.1, according to different sources of data. This table clearly shows that cotton fields in forest areas require more labor than in the savannah zores. The data from the "Nonres de Main-d'Oeuvre" on which our analysis is based seem to be on the high side. Table 7. 2 presents labor requirerents for different activities in cotton production. There is considerable variation among authors. The requirerents for land clearing according to tie "Nonres de Main-— d‘Oeuvre" seem high. These are only ore-fourth of the total requirerents. Harvesting and sorting data appear low for forest areas . (bviorsly more research is needed on actual labor allocation under field ccnditiore . Determination of Yields There are several problems involved in the determination of yields of crops grown under African conditions [Spencer, September 1972; Norman, 1973]. The fanm business survey on cotton provided us with accurate 2The theoretical prdalem of allocating labor inputs to crops within a crop mixture has not adequately been resolved [Norman, 1973]. Most authors regard the dominant crop mixtures as joint products in their analysis. 220 Table 7.1. Republic of Zaire: Estimates of Iabor Requirerents for Cotton Production in Forest and Savannah Areas (In Man-Days per Hectare) Source of Data1 Forest Savannah Areas Areas Normes de Main-d'Oeuvre 225 198 Depy G. 209 174 Lurpungu Kamanda . . . 189 Van den Abeele & Vandenput 187 178 OOGERCO . . . 130 Be’guin H. 140 . . . VandeVelde G. 224(2) (1)Assuming six hours of work per man-day. (2)Type of area not specified; calculated for a yield of 400 kg. seed cotton per ha. lNormes de Main-d'Oeuvre [INEAC, 1958]; Depy G. [1954]; Lumpungu Kamanda [Noverbre 1970]; Van den Abeele & Vandenput [1956]; CIDGEIRCO [1970]; Béguin [1958]. data on cotton yields but not yields of other crcps. Therefore, we used the yield data from the "Normes de Main—d'Oeuvre," wherever possible and from tie renowned book on crops in Zaire by Van den Abeele and Vandenput [1956] for the remaining cases. The source of yield data is indicated in Table 7.3. Perennial Crops In order to compare the profitability of perennial crops with annual crops , several working hypotheses have to be adopted. We prOpose the following: “Total production of a farm with perennial crops is the product 2221 .E.m OOHHH cu .E.c omuw soap mxcouqm: .53: .358 :03: 0858 KEma: usacmncg a 3684 can 53 :23 89563 manage @593 ima: .u .58 :33 .052: 6E8.nufi£ on motto} 8:960 5 confine: add—.0 wfi 59E E .mc non couuoo ooom .ox oov wo pace» c ace omucaooamo “ooHMHcoom co: coho mo come Ame .coflumuou och cfl moouo mace Ham coozpoo coaumocadm Anson wo omccooo mucoemuaooou Hooch mo ouHoOwlmco waco ucomouomu coco momma Ame .mccccm>cm ca .ox com com woman ummHOm ca .ox OOQH mo moaoflx HOW occ eaco moocuoe Honda occ: cuflB moofluomuo Honouaco oooo ace omcmflaocumm coco Honda AHV . .. ... ... mm mm HVH on mmmn< nmccc>mm .. ... ... .. .. we me am mva mound ammuom A3 coemmfiz had.no o.vmm .. o.m m.ma o.mm o.mm In: m.vm o.o~a Amy .0 moam> mo cm> wna OH OH OH mm mm ma mm ma mmou< noccm>om nma we ma OH ov mm OH Hm mm goaacm Ecummflccom ewe «H ma OH ov mm OH mm om momma umouom AS dd nsmcmcqg tag a dammed coo cm> Aoanv vea .. om ca mm om om mowed coccm>mm and mom om 2 mm om mm . manna ummnou .lurflmlo o.mma o.nw o.vw o.mm o.vm o.mH mmmuc coccm>mm m.mmm . o.vm o.ov o.mH o.ea m.maa mound ammuom ou>coo.oncamz mo mMEuoz ocflcucm mamm can one oocccoucamz coflumucomum ocaunmao Hooch uncomcoue ocfluuom ocflMOOHQD umo>ucm occ monomoz ocauccao ocmq coma coco mo mouoom H Amucuoom moo meteredz.cwv cofluccoono couuou CH mmflufl>auc¢ ucoHOMMHQ HOw monotouflooom uoncq mo moumbflumm ”ouaoN mo ongomm .NK canoe 2:22 .H mega en m «can ocunn>fln en encamuno ma xnm manqm ...ogwotlfimz up mmEHOZ: moucofiocfl sz vac isms“; US.» .3084. coo 5?. Eco wo 00.38 ”#58ch A>v “cauuoo occ macho coco Hocccc new common cofluccoOnu moo u0e oaoam cocuc>c occ macho Hoflcconoc ace mucox ma wo penumm m uo>o pace» oocuw>m acjcc<~ .couuoo occ macho coop Hmoccc new common coHuUJUOua oco new and macho Hmaccouom ace mumox om mo oofluoo m ace muomenacowu “HondaH oo.a oa.~m QN.H~ oq.a Hm.v on.m~ om.m NH.NH ao.oa mH.Ha om.mo mm.¢m mxmoucme nma x ow mocw>ou mmouo maa.m oma.~ omm.m coo.m ooo.m oom.oa ooo.e oov.~ oov.~ ooo.om oo~.mam mvo.mv m an mscm>mn mmono o.H m.m m.m o.m o.~ o.v o.v o.o o.o o.o~ o.og mm.o .mx nwa x no manna mkH.m omm.m oom.g ooo.a ooo.H oon.~ coo.H ace oov oom.v omm.v~ om~.nma .ma non .mx an coauoanonm lzcmea.m Azcomm.~ l>coom.a l>cooo.g l>cooo.a lzcooa.m l>cooo.H l>vooe r>coov l>coom l>co~m.g A>Vomm.m ~.gc nun .mx an ngmnw m.Hma ~.hq~ ~.nvm m.mgv m.mav H.vmv H.ama o.mmH m.m- H.mvm.a N.~mm.m a.vo~.H H.gz nma memoucgz nmcco>om ammuom Amnmnsec “annual Annemac an an mwnuoo conga: mung: gang: mngcecsono muscncsono none none conuoo _ connoo moooo mumsnom Egan goo Axcolcmz Mom cucxmzv vnmalmnma .ouHoN cnocuuoz CH mmHuHoQEtoo oopomamm Macao oco :Quuou new Honda on mcucumm m0 mmoueaumm .m.h canoe 223 of annual average yield and the productive life time of the plantation, fixed at 16 years for all perennial crops. This is to conform with the "Normes de Main-d'Oeuvre" which have been calculated for a total planta- tion life of 20 years. --The per Lmit product price remains constant over the productive life (16 years) of the plantation. --The annual average yield remains constant over the productive life (16 years) of the plantation. Price Data The product prices used in this analysis are the prices fixed or cormonly practiced in 1973. The form of the product is what is usually sold by the farmer to traders, processors, etc. such as oil palm nuts, dried unshelled coffee beans, dried cocoa beans, paddy, dried unshelled groundnuts, dry maize kernels and dried retted manioc chips . Method of Analysis The retm‘ns to labor of perennial crops are corpared with those of annual crops by calculating the gross remrureration of one man-day of work in each crop enterprise. Thus, the basis for comparison is the imputed rermmeration of a man who works one day in each crop. In principle, this is fliewage which could be paid to an unskilled laborer. The total lifetime earnings of a crop or the annual revenues which a farmer receives from a particular cr0p or sequence of crops are not analyzed in this chapter. Such an analysis requires knowledge about the nurber of crcps grown, the crop mixtures and sequences of crops per 224 year, crop rotations, duration of fallow, etc. which are beyond the SOOpe of the present analysis. Results The detailed data and calculations of the returns to labor for different crops can be found in a paper prepared by the author for a conference on agricultural development problems in Zaire [Tollens, l974a] . The results are summarized in Table 7.3. Analysis of the Results The difference in gross revenLe per man-day between perennial crops and annual food crops and cotton is very pronounced. The returns to labor for perennial crops are on the average more than double those of annual crops and cotton. The perennial crops—cocoa and coffee--yield the highest returns per man-day. The highest returns to labor are realized for cocoa.3 The returns to labor on cotton are about the same as the returns to food crops. The returns to cotton and food crcps are below the minimum wage rate for unskilled labor in rural areas which varies between 18 and 35 K. per day, according to tie regions.4 Cotton produc- tion in savannahs yields higher returns to labor than in forest areas, under the same yield assumptions. 3In Zaire cocoa is almost completely produced on large commercial plantations. It was never proroted as a smallholder cash crop as in Ghana or Nigeria. It would be interesting to study why in Zaire cocoa (and coffee) were predominantly produced on cormercial plantations and were not introduced on a large scale in traditional agriculture. 4"Salaire Minimum Interprofessionnel pour Manoeuvres Ordinaires, Zone III" in (RDICEC, 1973. Code du Travail-Mesures d'Application. Kinshasa: editions CADICEC. 225 With good yields, maize and rice production provide returns to labor above the minimum wage rate. However, there is often no regular market for food crops in remote rural areas. In northern Zaire farmers often cannot sell maize, manioc nor groundnuts because of the lack of buyers. At the time of this farm business survey, a regular guaranteed market outlet existed only for cotton, rice and coffee. Recently, new product marketing offices have been created which may provide regular sales opportunities for a variety of agricultural products in the remote areas. This analysis considered land and capital as nonlimiting factors under the present technologies of production. No returns are assured to accrue to these factors. Thus, the returns to labor as presented in Table 7.3 are to be considered as upper limits. The Real Purdmasing Power of Seed Cotton Since 1960 Since 1960 the real purchasing power of cotton has declined drastically. In fact, the real purchasing power of seed cotton as measured against the price index of 61 common consumer goods in Kinshasa declined from 149.3 in June 1960 to 100.0 in December 1964. and 72.4 in December 1973 (Table 7.4) . And,as we can safely assert that on the average, prices of consumer goods in the interior rose relatively more than in Kinshasa, the "terms of trade" for cotton farmers deteriorated even more than shown in Table 7.4. Since December 1964 the index of real purchasing power of seed cotton has oscillated widely between 100.0 and 58.2. Rapid inflation has eroded tie purchasing power of seed cotton. At the end of 1973 tie index was at the level of that of December 1966, far below the index for June 1960 which had then reached 149.3. 226 Table 7.4. Index of Real Purchasing Power of Seed cotton in Zaire, 1960-1973 Year Seed.Cotton Producer Price Index of 61 Index of Real Price 1 Price Index Consumer Goods in Purdhasing Grade 1 Kinshasa2 Power of Seed Cotton June 1960 6.00 33.3 22.3 149.3 December 1960 6.00 33.3 25.0 133.2 " 1961 6.00 33.3 33.8 98.5 " 1962 6.00 33.3 60.9 54.7 " 1963 8.00 44.4 87.4 50.8 " 1964 18.00 100.0 100.0 100.0 " 1965 18.00 100.0 108.7 92.0 " 1966 18.00 100.0 137.1 72.9 " 1967 24.00 133.3 229.0 58.2 " 1968 3.60 200.0 284.0 70.4 " 1969 3.60 200.0 310.8 64.4 " 1970 4.50 250.0 325.3 76.9 " 1971 4.50 250.0 338.1 73.9 " 1972 6.50 361.1 369.8 97.6 " 1973 6.50 361.1 460.5 78.4 1Price in francs per kg. until 1967 and K/kg. from 1968 on. 2 "Index des Prix de Detail de Biens de Coneommation.Courante aux Marches de Kinshasa," Kinshasa: National university of Zaire, 3 I.R.E.S., published in "Cahiers Economiques et Sociaux". Obtained.by dividing the second column by the third column. 227 This analysis shows that, other things being equal, the real value of the cotton crop and the real returns to labor for cotton pro- duction are now much lower than during the colonial administration. £92m The returns to labor in cotton production are corparable with those for food crops if a market outlet exists for these products. However, these returns are below the legal minimum wage of unskilled workers in rural areas. Perennial crops (export crops) earn much higher returns to labor than cotton with cocoa the most profitable crop followed by coffee and oil palm. These returns are well above the legal minimum wage. Precise resource utilization and yield data are reeded for a more thorough analysis of the financial profitability of cotton and other crops. In December 1973 the real purchasing power of seed cotton was about half of the June 1960 level. Thus, other things being equal, the real returns to labor for cotton production are now much below those of 1960. This researcher strongly believes that erosion of tre farmers' real purchasing power of seed cotton explains a large part of the lack of interest in cotton production, low yields and the present low levels of production. Presently, cotton farmers can derive much higher labor earnings per day from other crops than cotton, particularly coffee, which has expanded greatly among smallholders since 1965. Maize and paddy can also be more profitable if there is an assured market outlet at the official minimum purchase price. CHAPTER VIII ANALYSIS OF THE MAIN DETERMINANTS OF (D‘I'KN PKWDLCTION IN I‘DRI‘HERN ZAIRE In Chapter VII an attempt was made to determine the returns to labor for cotton production and for other crops using experiment station labor data. Our analysis revealed that the returns to labor in cotton production, at 1973 prices, were well below those for perennial crops and m a par with those for annual food crops. The returns to labor are a function of the price of seed cotton and the physical labor pro- ductivity. We found in Chapter VII that the real price of seed cotton paid to farmers has declined drastically since 1960. This implies, other things being equal, that the real returns to labor for cotton production are now much lower than in 1960. The purpose of this chapter is to analyze the main determinants of cotton production in northern Zaire and atterpt to explain the major decline in production since 1960. Production is t‘re product of yields per unit of acreage and the nurber of acreages planted. In Chapter VI, we identified several factors depressing yields in many areas . They were: short duration of fallow, field plots on marginal soils, poor land preparaticn, low cotton plant densities, late planting of cotton seeds over an extended period of time, poor crOp maintenance, particularly thinning and weeding which are performed too late, insect and disease attacks, planting crop 228 229 mixtures instead of a pure stand of cotton, late harvest of cotton and generally, inadequate incentives to produce cotton. With respect to acreages, we found in Chapter VI that the location of cotton fields and the enforcerent of cotton growing requirements and field size by agricultural monitors, territorial agronomists and village or clan authorities were important factors in determining cotton acreages in rorthern Zaire. Yields will be analyzed in this chapter using a multiple linear regression technique. An attempt will be made to determine the price elasticity of supply (supply in tents of acreages planted), using remunal prices. . The variation in cotton acreages between farms and between sub-regions and regions will be analyzed using a multiple linear regression equation model. Finally, we will compare the nurber of cotton farmers, acreage per farm and yield in kg. per ha. and kg. per farm for 1958-1959 with 1970-1971 and 1972-1973. 222.122 Yields are defined as the physical production either in kg. per ha. or in kg. per farm. In this section, however, we will analyze yields in terms of kg. of seed cotton per hectare; our objective is to explain the differences in yields between fame and between regions and sub-regions. Several factors determine yields and trey can be categorized as / bio-technical factors (soil, climate, diseases, date of planting, weeding, etc.) , economic factors (seed cotton price, returns to labor, etc.), social factors (grouping of fields, communal labor, labor for the chief of tte village, etc.) and administrative factors (crop 230 requirerents, determination of location and size of fields by agricultural monitors, etc. ) . Method of Analysis We have chosen a multiple linear regression equation of the following fonm for the explanation of the variation in yields between farms: 8 +,...,+Bx. +8 Y1 = 81 + 82 x12 +'° ' "+ B m+1xjm+1 n m i x. + m m where the variables are defined as follows for farm i: Yi = production of seed cotton for farm i in kg. per ha. Xi2" . . ,xjm = variables which we believe explain part of the variation in yields between farms. Data have been collected for these variables. X. 1mH-l’ . . . ,xin = variables which we believe explain part of the variation in yields between farms, but no data were collected on these variables . Examples are soil fertility, rainfall, hours of sunshine and insect damage. Since we have no data for these variables, we assure that they remain constant between farms. They will be lumped together in the constant term of tie regression equation. Si = disturbance term for farm i. The regression paramters 81, 82,. . . , 8m were estimated by the method of ordinary least squares (OIS) . The usual problems with a model of this type include high degrees of multi—collinearity of the independent variables. But even with a high degree of multi-collinearity, we still get unbiased efficient estimators for tre regression parameters. 1 Yet lest simple correlations between the variables are quite small. Simple correlations larger than I0.40| are between YID and BUL, 0.49; ACC and ACE, 0.48; LAO and 1AM, 0.50; DEN and LAO, 0.41; UBI and YOF, 0.42; SHL and AGE, -O.55; IPC and ROD, 0.48; UBI and CHF, -0.59; BUL and DDN, -0.42; PRC and ERN, -0.52; BUL and UBI, -O.59. 231 another problem may arise from heteroskedasticity of tie disturbance terms. If this is tie case, then the 01.8 estimators are still unbiased but not efficient and the tests of significance are incorrect since the variances of the estimators are biased. However, testing for l’eteroskedasticity would lead us too far; thus, we will conveniently assure that the model is hcxroskedastic.2 The solution of the multiple linear regression equation by 015 has the following form for farm i: I+BX Y.=§+§2xi2+,... min 1 1 A stepwise add and a stepwise delete OIS corputer routine was used to calculate the multiple regression statistics. This routine calculates the significance probability of the F statistic for the least squares coefficients of a variable to determine whether or not it should be deleted from or added to the equation. The significance level was \/ arbitrarily set at 0.095 for all runs. The following regression statistics are presented for each function: (1) the regressim coef- ficient; (2) the standard error of the regression coefficient; (3) the t value of each regression coefficient; (4) significance level of each regression. coeffitient: ”this is the significance probability level of the F statistic for the DIS coefficient of a variable. It indicates tre confidence level of each regression coefficient. (5) Beta weight: this is a standardized regression coefficient that measures a variable in terms of its standard error. The higher the Beta weight, the more 2Weighted least squares for the heteroskedastic case make tre regressim equation satisfy the conditions of homoskedasticity and thus yield the best lirear unbiased efficient estimators with correct tests of significance. 232 the variable contributes toward explaining the variability of the dependent variable.“ (6) R2 delete: this indicates how much omission of the variable would reduce R2, other things being equal. (7) partial correlation: this coefficient measrmes the degree of correla- tion between a particular indepeident variable and the dependent variable when all other independent variables involved are kept con- stant. The partial correlatim coefficient differs from the simple correlation coefficient. 3 (8) simple correlation: this measures the degree of correlation between two variables when only two variables are involved. (9) R2, the coefficient of multiple determination: this indicates the percentage of the variance in Y, the dependent variable, \/ explained by or "associated with" the variance in the X's, the inde- pendent variables. And, (10) F value: this is the value of an F test in the analysis of variance model equivalent with the multiple regres- sim model indicating the ratio of interaction (or treatment) mean square for the overall regression with error mean square. In Table 8.1 we report all variables included in the multiple linear regression equation. The last three variables, UBI, NDL and BUL. are dummy variables for the Ubangi , Mongala and Bas-Uélé sub—regions , respectively. 4 These dummy variables are a proxy of all variables 3The partial correlation coefficient can be written as a function of the simple correlatim coefficients between the independent variables and the dependent variable and the simple correlation. coefficients between the independent variables [Spiegel, 1961] . 4We did not include a dummy variable for the Haut-Uélé sub-regim to avoid matrix singularity. 233 Table 8.1. Yields and Acreages: Variables Included in the Multiple Linear Regression Equation Variable Description of tie Variable YID1 Seed cotton yield in kg. per ha. ACC Cotton acreage per farm in ares ACC Acreage of other crops in ares 1AM Iabor input in man-weeks per ha. for cotton maintenance LAO Iabor input in man-weeks per ha. for other cotton activities than maintenance 1AA Iabor units available on the farm YOF Years the cotton field remained in fallow KIVE Distance cotton field—house of the farmer in km. REPl Reasonable, equitable seed cotton price in K per kg. 2 DEN Cotton plant density: nurber of cotton plants per 9 m AGE Age of the cotton farmer SHL Years the cotton farmer stayed in primary school FSA Cotton field located in a forest (=1) or in a savannah (=0) TPL Planting time: June-July = 1, August = 0 IPC Cotton growing imposition yes = 1, no = 0 IPO Imposition of other crops: yes = 1, no = 0 ICC Determination of location of cotton field: by the farmer = l, by otters = 0 SIZ Determination of size of cotton field: by the farmer = 1, by others = 0 DAY Hiring of day-laborers: yes = 1, no = 0 MIT Mutual work on cotton fields: yes = 1, no = 0 RDD daligatory work on roads: yes = 1, no = 0 CHF Work for chiefs or otter authorities: yes = 1, no = 0 MN Knowledge of cotton monitors or territorial agronomi st : yes = 1, no = O ERN Cottonearnedthefarmeralotofmoney: yes=l,no=0 POS Possession of a bicycle, radio, sewing machine or watch: 1 yes = 1, no = 0 PES Cotton fields treated with pesticides: yes = 1, no = 0 PRC Seed cotton price influenced cotton acreage: yes = 1, no = 0 T'ER Termite heaps on cotton field: yes = 1, no = 0 GRP Cottm fields in the village are groLped: yes = 1, no = 0 UBI Dimmy variable: Lbangi = 1, other sub-regions = 0 VOL Dummy variable: Mongala = 1, other srb-regions = 0 BUL Dummy variable: Bas-Uélé = l, otter sub—regions = 0 1These variables were not included in the multiple regression equation model eiplaining cotton acreages per farm. 234 which determine cotton yields and which are typical for a srb—regio’r but for which we lack data (the ximrI-l” . "xin variables). Results of the Statistical Analysis Tte results will be presented and interpreted for each srb-region in the study, for tbangi-Mongala and the Uélés and for the entire rorthern cotton belt which is corposed Of four sub—regions. However, to results will be presented for the Mongala sub—region as only eight observations were available for this area. Thus, Lbangi—l'rngala will be viewed as one sub-region. Results for Ubangi-mnqala Table 8. 2 reports the regression statistics for tre function explaining the variation in yields between farms in Lbangi-Mngala. All regression coefficients are significant at the 0.05 level of significance except for LAA (labor units available on the farm) and for TPL (time of planting) which are significant at the 0.10 level and the constant term Whid‘l is only significant at the 0.15 level. We observe that the variables LAO (labor input per ha. for otter cotton activities than maintenance), MN (knowledge of cotton monitors), ACC (cotton acreage per farm in ares) and KM: (distance cotton field- house in km.) contribute most toward explaining the variability of yields as their Beta weights are, respectively, 0.41, -0.33, 0.32 and 0.30. Omission of these variables reduces R2 by, respectively, 0.10, 0.08, 0.05 and 0.05. As indicated by the regression coefficients for MN, knowledge of a cotton monitor reduces cotton yields by 546.41 kg. per ha. , otter 235 338 monotone moooov Hod means a mo. mm mm .1 oo . .. mm . mm . a mooo . ov me z? oo .mfi d. .oamu as am. am. am. R. «Ho. oo.~ mime Rod 26 mo... mm... oo. om... moo. #4.. sods $621 one «N. oo. mm. om. moo. mad 2:3 mm.mm 92 mm. Hm. om. mm. Hmo. om.~ oo.oo oo.mma Hm: so... am. oo. 3. >8. 84 mod 2.3 5: am. mo. om. to. moooov ooé so om.~ 93 Ho... Hmf om. mmf mmo. mm.~.. 3.3 8.87 to mm. R. om. om. m3. mo.~ Rom 863 can oof mm. mm. mm. ooo. mo.m go Sum 8a In om. oo. oo. $1... 84 omoom 8.8m octagon :oflflonnoo 833880 mBmHmo oceans unmaoflmooo mo unmaoauwmoo £95m Hugo mm 33 gangs... 33> a muons Eamon scammonmmm magnate, so uneonatethat no .4352th fiancee 98 S58 unontaaomd to has mouth» 838 goon lg: . :835. u as 83er Eu not Bioflfim eoeommkomm .~.o manta. 236 things being equal. The sign for the coefficient of TPL seers wrong, but this variable is one of the least significant. The large coefficient of the dummy variable for Ubangi sub-region (UBI) , 153.87, illustrates that yields in Ubangi are distinctively higher than those in Mongala. When farmers have to work for village chiefs or other authorities (CHF) , cotton yields are reduced by 193. 87 kg. , probably because farmers are unable to maintain or harvest their cotton field properly. Cotton imposition (IPO) , on tie otter hand, increases yields by 116.03 kg. and, thus, constitutes a positive factor. The larger cotton acreage per farm (ACE) , the higher the yields are, although the effect is weak, 2. 74 kg. per are. Each additional man-week per ha. for cotton acti- vities other than maintenance (IAO) increases yields by 2.90 kg., and the further the cotton field is from the home of tl'e farmer (Km), the higher yields becore, an increase of 35.53 kg. per km. For tie overall regression equation, the coefficient of multiple determination (R2) is 0.62, not particularly high, and 33 percent of the variance in yields remains unexplained. Results for Bas-Uélé In Bas-Ue’le’ yields were on the average the highest (687 kg. per ha.) and tley varied considerably from farm to farm (standard deviation of 453 kg.) (see Table 8.3) . In applying our equation, we obtained an excellent fit as R2 equals 0.96 (Table 8.3) . The partial correlation coefficients of all the variables with yields are high as they all exceed |o.50|. All regression coefficients are significant at the 0.05 level except for ROD (cbligatory work on roads) which is still signi- ficant at the 0.10 level. The Beta weights larger than |0.4o| are DEN 237 .338 mnasomuma mooo.ov Sow mod—m5 .m oo. mm 3. on. .3. mm. Hoo. Nmé duos mmfimm UQH om . om. mm. mm . hmo . mm .m 363 oh .mam ooH hm. oo.: oo. 3.: moo. mo.mt oméma 363.. $6 oo. 3.: mo. 3.: vmo. mm.~| mmdo Home: mm? 21: om... mm. Hm... omo. mm.m| ovoam mvéoml and me. no. em. om. moo. oH.m mo.oo mooom mom 3.: oo.: oo. oo.: oo. no.7 omé mmom I mum om. oo. .:.. om. moooov om.m moo whom zoo oo. om. mo. om. boo. Hod who: om.mm~ com mm. on. mm. mm. moo. moé omoa Honk. 96o mmf mnf mm. mm... moo. «Tm: vméma some: NHm mo. om. oo. mm. mmo. owlm ooém .:..mm .24 mm. no. mo. mo. moooov 36 no.3 «Home 75; oo.: on. am. No. Hoo. mmé ooé dim 25 mm. R. .8. mm. Hoo. ode mméo :63. 0mm ma. mm. oo. .3. mooo.ov mo.m omé go 84 run oo.: oo. oo. oHo . oo.? .3 .oom mm .oool ucmumcoo coflmaounoo coficamfinoo mouoawo 3:363 ucmfloflmmoo mo ucofloemmooo seesaw emanate mm 58 noeooaoeéam seems a nwownm Botswana soennmhmmm manpower, Hem "20.3% Mo .3952 .oncu .wHwolmcm "oceans/Home mo use: mosses sowooo ages 60H. . . .693st u S.» sofloé 9t on noflmfifim 538.com .m.o manna 238 (0.56) whidn indicates that cotton plant density contributes nest toward explaining the variation in yields; (mission of this variable reduces R2 by 0.25. Each additional cotton plant per nine square meters accounts for an increase in yields of 33.72 kg., a fact which illustrates the importance of plant densities. Work for village chiefs or other authorities (CHF) has a strong negative effect on yields; the regression coefficient is -476.19. The Beta weight and the partial correlation coefficient are high, -0.51 and -0.75, respectively, although R2 delete (0.92) is only slightly below R2. Age (AGE) of the farmer also has a negative influence on yields (20.53 kg. per year); the Beta weight is equal to -0.46, the partial correlation is high, -0.80, but R2 delete (0.90) is again close to R2. Knowledge of a cotton Ironitor (MON) strongly affects yields as the regression coefficient is 423.14, positive, the Beta weight is large, 0.42, and so is the partial correlation coefficient, 0.87; R2 delete is also 0.11 below R2. Labor for cotton maintenance (LAM) has a 2 delete is high Beta weight (0.42) and partial correlation (0.78) , R 0.91 and the coefficient is equal to 8.41, clearly not enough to dictate an increase in labor for cotton maintenance. However, we should keep in mind that labor inputs were not accurately measured. Cotton acreage per farm (ACC) has a positive and high partial correlation with yields (0.82), the Beta weight is large (0.41) and R2 delete is equal to 0.89. Each additional are, ceteris paribus, increases cotton yields by 6.91 kg. An analysis of the remaining regression coefficients is also quite revealing. The farmers who reported that seed cotton prices influenced their cotton average over time (PRC) have a yield that is on the average .239 340. 14 kg. higher than for those who answered negatively. Determination of oottcn‘field size by the farner (SIZ) is associated with a lowering in yields of 499.43 kg., while remoteness of fields (KNC) has a positive effect on yields, 75.81 kg. per 100., as does obligatory work on roads (ROD) , 293. 58, although one would normally expect a negative relation- ship here. This latter variable was only significant at the 0.10 level. As expected, possession of a bicycle, radio, sewing machine or watch (P08) is strongly correlated with high yields; the partial correlation is 0.67, the Beta weight is 0.30 and the coefficient is equal to 266.89. Hiring day-laborers (DAY) has a detrimental effect on yields; the coefficient is equal to -504.49. And termite heaps (TER) on a cotton field are also associated with lower yields, the coefficient being -164.15. Imposition of other crops (IPO) and determination of cotton field loca- tion by the farmer (LDC) have strong positive effects on yields, 313.76 kg. and 337.29 kg., respectively. 'lhe number of labor units available on the farm (LAA) also has a positive influence on yields, the coeffici- ent being 55.71 kg. Results for Haut-Uélé (Ply l9 yservataons wefe available for this smb-region and the final regression equation includes 15 variables. R2 is extremely high, 0.99, all partial correlation coefficients are also very high and all regression coefficients seem exaggerated. The results need to be analyzed with caution as not enough observations are available for a powerful statistical analysis (Table 8.4) . All regression coefficients are significant at the 0.05 level except for REP (reasonable, equitable seed cotton price) which is 240 .0038 03.30804 mood Son 88> .... oo..-- . mm mm . mo . mm . 8 4 So . do. . 3 me .me @Hmoc 2mm oo... oo. oo. 2.4 So. 8d 8.3 $.32 08 mm. oo.- ma. 3:? «oo. no.2- 3.8a $.32- 8H Ho. 3. cm. so. So. 36 8.8 moo? owe me. oo.- ca. co.~- moo. am.oa- ao.ae _ea.m~e . arm 3. mm... mo. 8. .. omo. $6 .. moo No.2. u as me. am. om. om.~ Hoo. Ho.ma ma.aoa Hm.aaoa mom oo.: mo.u mm. mm. u moo. am.~ n oo.ea ma.me . cum No. 8.- oo. me. u moo. mg. u 8.8 3.3a .. 02 am... oo.: mo. 8.? 8o. 8.2... 3.3 ”mesom .. so» Nor so. oo. oo. So. So flow moose so 3. oo.- cm. 3. . woo. was .. mmo ow; .. o5 om. oo. so. 84 . Go. 2.2 34 $.30. _. 25 oo.- 3... oo. am. u «mo. ~o.m u 24 one .. .. 8a an. «of co. San... Ro. eoé u omom $68.... one I- mm. mm. oo. aoo. moo $48.8 8.3% 05088 833008 8030880 0800c 3803 088008 mo ufiaoaemxo 083 83000 mm 30m 850$?ch 0385 a 0.65 cementum 8800800 0335/ am." "90308808 mo uwQBZmHHoN .wawaluomm "coflgmmnd no 35 030? :808 messaged 2mm. . . 1883 u 3» 50.950 05 new moflmflfiw 88003 aim. 0309 241 significant at the 0.10 level. The l'tDSt inportant variables are years of fallow (YOF) with an unexpected negative coefficient of -263.65, 2 Beta weight of -4.83 and R delete of 0.62. The years the cotton fanner stayed in primary school (SHL) shows a strong negative effect on yields, coefficient of -425.77, Beta weight of -2.84 and R2 delete of 0.76. However, possessicn of a bicycle, radio, sewing machine or watch (P06) is strongly associated with high yields as the coefficient is 1617.51, Beta weight is 2.56 and R2 delete is 0.50. Inposition of other crops (IPO) has a very pronounced negative effect on yields, the coefficient is -1734.24, Beta weight is -2.54 and R2 delete is 0.75, while location of the cotton field by the farmer (10C) is strongly associated with high yields as the coefficient is 1098.74, Beta weight is 1.74 and R2 delete is 0.69. If the farmer responds that cotton earns him a lot of money (ERN) , yields are likely to be high as the coefficient is 1051.18, Beta weight is 1.66 and R2 delete is 0.55. The effect of the remaining variables can be read frcm Table 8.4. Results for Bas—Uélé and Haut-Uélé This section conbines the results for Bas-Uélé and for Haut-Uélé. A dummy variable for Bas-Ue’lé (BUL) was included as a proxy for factors typical of Bas-Uélé which influence cotton yields. The results are reported in Table 8.5. The coefficient of multiple determination (R2) is 0.68, much lower than for each one of the sub-regions separately. In fact, these two sub-regions appear as one distinct cotton production areas. Yields in Bas-Uélé were on the average much higher than those in Rant-Uélé (Table 6.22) . The regression analysis confirms this as the dummy 242 .828 08808 H mooo.ov Hog. 98> m oo. mm oo. oo. 8. om. So. om.m mmoo omcmm 2% oo. oo. oo. ms. moo. oo.~ 86 3.3 zoo om. mm. om. mo. mooo.ov moé Roma oo.omm 8m oo. oo. oo. oo. moooov ooé snow Hood as oo . S . om . mm . moo . om .m oo . o: mm .oom 29 oo. oo. oo. om. omo. So 8.8 3.8. 45 oo. oo. 8. 8.. So. mm.~ $6 84 05 NH. oo. so. so. moo. oim No.8 mm.mom o8 mo. em. oo. oo. goo. mo; omsma $on 00H 2. so. so. so. moo. 88 mo; Sé cos I: oo.. oo. oo. moooov moo- om.mom $.83- 0.80:8 833088 833088 0808c 8:303 £88388 no £08388 03% H8800 mm 30m 0080886 88> e. 888. 8888 830088 03088 Hmv umcoflugmmno mo HmnoBz .mHHoN .wawoluomm poo WHWDImmm "sconce/.888 m0 is: 0303 888 cfififlcxm 2mm. . . .8958H u a 888m 08 8m ncflnflflm 8800800 .mo 038 243 variable for Bas-Uélé (BUL) emerged as the most important factor in explaining the variation in yields with a coefficient of 538.00, Beta weight of 0.65 and R2 delete of 0.54. The other important variables were cotton plant density (DEN) , already discussed for Bas-Uélé, dis- tance cotton field-house of the farmer (KIVC) which has a positive effect on yields, 115.31 kg. per km., with Beta weight of 0.40 and R2 delete of 0.54 and knowledge of a cotton mnitor (MDN) which is also strongly associated with high yields as the coefficient is 379.25, Beta weight is 0.35 and R2 delete is 0.59. Other factors which have a strong positive effect on yields are ERN (cotton earned the farmer a lot of money) with a coefficient of 226.38, IPO (imposition of other crops) with a coefficient of 213.99 and III: (location of the cotton field by the farmer) with a coefficient of 203.32. Insults for the Northern Cotton Belt, Including (bangip angala, Bas-Uélé and Haut-Ueflé Again, the durrmy variable for Bas-Uélé (BUL) is the nest impor- tant factor in explaining the variation in yields (Table 8.6) . The coefficient is 452.40 and Beta weight is 0.57; dropping this variable reduces R2 by 0.23. Other variables with a Beta weight over 0.20 and which, if deleted, reduce R2 by at least 0.04 all have a positive effect on yields. They are ERN (if cotton earned the fanner a lot of nmey) with a coefficient of 190.37, KMZ (distance cotton field-house of farmer) with a coefficient of 56.92, DEN (cotton plant density) with a coefficient of 10.01, IPO (imposition of other crops) with a coeffi- cient of 166.15 and LAO (labor performed for other cotton activities than maintenance) with a coefficient of 0.89. Other variables with a 244 .888 005.880 H mooo.ov 8.2 88> m om. mm on . mm . ~m . om . mooo .ov me .m 8 .8 on .oS 28 R. mm. mm. mm. 8o. mm.m mod 8.3 28 S... 2. mm. 2. moo. meé 8.3 3.3: 29 NH. om. Hm. om. mooo.ov oo.». oo.: No.8 82 as . om . mm . om . mooo .ov om 4. oo .8 oo .mmo .50 oo.: om. om. om. moo. mo.~ omom 8.8 a om. mm. em. 8. ooo. dam moo coo 08 08 mm. om. oo. «oo. 38 oo.mm mimma 8H 8.- oo.: mm. 3... moo. mo.~.. oméo £288 goo mo... mm. mm. mm. omo. m~.~ mo; Rum cos nl oo ... mo . oo . mooo . ov on .T mm . omH on .oom- 800008 8808.88 880888 0888 00883 0888008 0880 88000 or 0000 00808 88 88> e. 088 808000 8e 000800 800805 and u08in..050080 no .4882 88:80: 0:0 88.000 .0888 . 8:03 0080 80582 80808008 0o 08: 030.9 888 888mg Ag. . . 186.0 n new 8888 08 8.0H 00808000 880088 .m.o 038. 245 high coefficient but 100 Beta weight are NDL (dummy variable for angala sub-region) with a coefficient of -191.75 and Beta weight of -0.14 and LAA (labor units available on the farm) with a coefficient of 29.41 and Beta weight of 0. 20. The overall coefficient of multiple determination for the regres- sion equation is 0.58 leaving 42 percent of the variation in yields unexplained. 'Ihus, several important factors influencing yields in the northern cotton belt were left out of this study and could not be analyzed. 591339518 In this section, we will first attempt to determine the price elasticity of supply (supply in terms of acreages planted), using nominal prices. We will then analyze the variation in cotton acreages between farms and between sub-regions and regions through the use of a multiple linear regression equation model and with as independent variables those listed in Table 8.1. The Price Elasticity of Supply (Acreage Planted) Prices in 1970 were 4.50K per kg. for first grade, 3.00K for second grade; in 1972 the prices were 6.5K for first grade and 5.00K for second grade. Since farmers in the 1972 survey were selected frcm a frame of cotton farmers Surveyed in the 1970 Census of Agriculture in which fields were measured, it was possible to compare the cotton acreages for these farmers and to calculate the price elasticity of supply in terms of cotton acreages which we assune would be positive. The Arc elasticity is calculated according to the following 246 formula: _ S72 S70 P72 + P7o Arc E _ P - P s + S 72 70 72 70 Vhere S = cotton acreage per farm in square meters P = cotton price in K per kg. , assuming 90 percent first grade and 10 percent second grade. We assumed for convenience "all other things equal" between 1970 and 1972 , in particular the prices of other agricultural products such as coffee and rice. We obtained the following results: Sub—region Arc E Ubangi—Mongala -0 . 086 Bas-Uélé -1.027 HautrUélé 0.272 Total North . -O.309 These results do not confirm our hygothesm of a p051tive price” 3835‘ / ticity of 51,129le SGYQIFQL factors may explain these results. The W ”3.”. "fl‘! -q_-r‘r‘ It 1970 acreage data may be unreliable. The 1970 sample census used a“ \/ WED farmermgf “NE":dch enumerator had a heavy work schedule, and it is quite possible that measurement errors were high.5 Yet another explanation for the poor results may be the use of nominal prices and not real prices in our elasticity calculations. It may be that the real seed cotton price did not increase. To calcuate the real 1972 seed cotton price, we deflate the nominal price with the IRES price index of 61 common consurer goods in Kinshasa [IRES, 1973]. This 5The 1972 acreage data were determined with great accuracy by the author and his enumerators. Several field checks were made, all fields were drawn to scale and the closing gap was less than 10 percent of the perimeter. 247 index had an October-December 1964 base of 100. The index was 239.0 in June 1970, and 343.3 in June 1972. Converting this index to the base of June 1970 = 100, we obtain a June 1972 index of 143.6 and a real June 1972 average seed cotton price of 4.42, about the same as the June 1970 price of 4.35. Thus, the real cotton price was practically the same in 1972 as in 1970. 1" Nth—N‘s t ‘ ‘11-»... anal-I. Another factor which mayexplajnthe bizarre price elasticity of ‘/ W ~~ - n... ..,_... "...—.... __n __9 supply figures is that "other things were not equal." In fact, several W _NH WU—In-v l":- ___J“ w... an k—rnr-rvfirn‘wfiw'm“* things changed. The coffee price paid to producers did change and ONAFITEX took over the marketing operations frcm the private cottcn companies which introduced an element of uncertainty for cotton farmers. In some areas cotton was bought late and the distribution of cotton seeds became less regular. This had an adverse effect on cotton planting. In conclusion, the available data and tie quality of the data do not allow us to calculate accurate supply elasticities for cotton in terms of acreages planted. Results of the Multiple Regression Statistical Analysis The purpose of this section is to explain the variation in cotton acreages between farms and between sub-regions and regions through tie use of the multiple linear regression equation model which was developed for the analysis of yields. The underlying hypothesis is that sure of the variables which explain yield variation between farms also explain part of the acreage variation between farms. We specified the same independent variables as for the yield analysis except for those which did not make sense in explaining acreages planted (see Table 8.1) . 248 0.1r approach is exploratory and unsophisticated. Time series data are needed for a proper study of the factors explaining acreages planted in a particular year, in our case 1972-1973. We do not possess such data and, thus, we will limit ourselves to an analysis of tie available cross-section data. The results are reported for sub-region , for region and for tie entire North in Tables 8.7, 8.8, 8.9, 8.10 and 8.11. For all units of observation, the coefficients of multiple determination (R2) obtaired for the analysis of acreages are below those for the analysis of yields (Table 8.12). Results for Ubangi-Mongala Fifty-seven percent of tie variance in cotton acreages per farm can be explained by five variables which are all significant at the 0.05 level. The most important one is ACO (acreage of other crops) \/ with a Beta weight of 0.42 and a regression coefficient of 0.07 which indicate that, other things being equal, cotton acreage on the farm increases by 0.07 ares when acreage of other crops raises by one are. Moreover , cotton acreage and acreage of other crops are strongly cor- related, the simple correlation being 0.60 and the partial 0.50. Onissicn of this variable decreases R2 by 0.15. Other important factors are LAO (labor inputs per ha. for other cotton activities than maintenance) with a negative coefficient of 0. 24 and a Beta weight of -0.28 and R2 delete of 0.49; or (work for chiefs or other authorities) which has a positive effect on cotton acreages as the coefficient is 19.89, the Beta weight is 0.22 and R2 delete is 0.53; TER (presence of termite heaps on the cotton field) which has a 249 e3 mflfifim .8 828 mmgflm H mooo.ov mm.ma msam> m hm. mm hm. hm. em. ma. ovo. oH.m hq.v hm.m mmfi mm. mm. mm. mm. mmo. mm.m em.m mm.mH mmu mm.| om.u me. mm.n moo. ea.mu mo.o v~.o I qu om.| om.l vm. mH.I ovo. «O.NI HH.o mm.o I zed oo. om. me. we. mooo.ov ov.v No.0 no.0 00¢ III we. mo. oo. mooo.ov mv.m hm.m mm.>q penumcoo coaumawuuoo coHumamHuoo mouwamo mucmflwz ucmfloflmmmoo mo ucmfloflwmooo madeflm ameuumm mm muom mOGMOHMHcmHm msam>_e mHOHHm pHmocmum coflmmmummm mannaum> Hem umcoflum>uomno mo umneoz “mHHmN .mammcoz cam Mommas "soapm>HmmnO mo peas mommmuo¢.:ouuoo mCflchmem Amme.. . ..Ouflvcm u 00¢ coauocsm mcu How moflumflumum coflmmmummm .h.m wanna 250 338 gm.” 88.8 8.2 83> m we. mm mm. mm. ee. em. Re. 84 $4 $5 .Em em. E. 3. mm. 38.8 and :.o mes 8» mm. 8. em. 3. :5. Sum 35 Rem 8H em... em... 8. emf 3e. 8.? . 8.0 25 .. 9: we... 8.- mm. om..- N8. ems- 3:3 32- 8a 3.. 3.. me. am. m8. m~.~ 86 3:3 as ...: em. . me. 8. «mm. mm. 8.3 3.2 £588 833880 Beumamuuoc 89.38 $883 9.88388 no 988388, 295m deflate mm Sen 883383 83> a 305 33m coemmmummm 8333, SN .meoflmfimmoc NO 8852 “whee £8-93 "eoflmgmeo no ads 8832 :Buoo efififlea Sum. . . ...Eooam u Be c0383 «5 now 833.3% acemmmumum .m .m manta mumamo mmegdmum 251 a moo. H5.m msam>.m mm. mm oo.: mo. we. mm. maou mm.m oH.H ov.m mow mm.n mm.| mm. mv.l mmo. ov.ml em.h oa.mal mom mm. mo. me. mo. moo. mm.m em.h mo.om 0mm Hm.n mn.u mm. oo.: Hoo. oH.v| oa.o ov.o I. :34 mo. mm. om. mo. owo. mm.m we.ma mo.mm omH Ila mH.I no. oo. emw. He.on mo.ma om.mHI unwumcoo coflumamunoo coflumHmuHoo mmumamo munmflmz ucwfioflmmmoo mo pcmfloflwwmou madeam Hmeuuem muwm moetoemecmem msam> e muonum cementum scammmummm manmeum> “mooflum>ummno mo “mnesz “wuemm .wawolpomm ”coaum>ummno mo uwcb mommwu04.:ouuou mcflcflmamxm Amow.. . ..omchm n uum.coauocsm map How moflumfiumum GOflmmmnomm .m.m magma 838 mfliflma mooo.ov mad 83> .m .mm. mm ma . om . 5v . hm . mmo. mod d. .3 3” ...N UmH mu 2. hm. we. Hm. omo. omé hmé mm.m Arm 2 mm. mm. em. me. mooo.ov mm.m mm.o >N.N mo» mof mmf EV. omf mao. mm.mI No.3” m~.mml mg mmf doll 3. mmf ovo. vodI moo oo.o I 93 STI Nmf ow. Hmf mmo. MH.NI Mao >~.o I 25 oa. am. 3. pm. So. mo.m 31> 3.3 BU III mm . De . oo . moo . .3 .H a» . ma Hm . mm #53 coflmamfiuou dogmawfinoo mmumamo 3:0me ucmwoflmwoo mo ucwflowmwwa 383 Agenda mm 88 833383 83> a muoflm gym :3 883$ 333$ How umcgumgmno no “8952 “8.2% .WHWDI¢SME can WHWDImmm "coflguwma mo p.35 mommmug coupoo magmaxm CmH. . . . ..mmuvcm n 8m mega 0:» How moflmflflm Ewmmmhmmm .oHd magma III-II l l l .'l 253 mumamc mmegmmpm H meee.ov He.ma msam> m em. Nm em. em. me. me. eme. ~H.N me.m Ha.m may me. am. we. mm. mmo. eH.m mm.m mm.~a Hes me. we. we. ea. one. me.H em.m em.ma .eue NH. em. me. am. mooo.ov mm.m mm.o eH.~ now He. mm. ee. em. oeo. me.m m~.m ee.mH emu em.- e~.- we. e~.- eeo. em.~- mo.e mo.o I oeq ee.- em.- ee. e~.- eoo. mm.~- me.o mm.e I seq me. me. am. em. meee.ev me.e No.0 me.o one In: an. we. oo. mme. eH.m Hm.e em.ma memumgpu eoeumammuoo eoeueflmuuoo mmumame magma memeoeummoo we nemeoeeemoo 39% age mm 3 833383 83> a $on Begum 833 3839s HNHH "meoeem>ummeo mo umeesz “imamsuuemm new mawaImmm .mammeoz_.ememeso muHmN eumeuuoz "eoeum>ummeo to pee: mmmmmuum eouuoo memeHmHexm Amme.. . ..ooaocm u oom coauucsm may you monumflumum eoemmmpmmm .Ha.m manna 254 strong positive relation to cotton acreage, the coefficient being 9.37, the Beta weight 0.19 and R3 delete 0.54 and, finally, LAM.(labor for cotton maintenance) which has a negative effect on cotton acreage with the coefficient equal to -0.23, Beta weight -0.18 and R2 delete 0.54. Results for Bas-Ué lé We obtained a good fit: R2 is equal to 0.74 and all variables are significant at the 0.05 level except SHL (number of years in school) which is significant at the 0.10 level of significance and the constant term which is only significant at the 0.35 level. By far the most important factor is the number of years the cotton field remained in fallow (YOF) . The Beta weight here is 0.59, and deleting this variable reduces R2 by 0.33. The regression coefficient is equal to 3.73. The next two important variables are ROD (obligatory work on roads) and IPC (cotton growing imposition) with Beta weights of -0.50 and 0.49 and R2 deletes of 0.59 and 0.57, respectively. Road work has a negative effect on cotton acreages—-the coefficient is 43.39, while cotton imposition shows a positive influence with a coefficient of 34. 35. Other important factors are LAO (labor for cotton activities other than maintenance) with a coefficient of -0.18, CHF (work for chief and other authorities) with a positive effect, coefficient of 15.87, and the number of years in primary school (SHL) with a coefficient of 3.58. Results for Haut—Uélé Only 19 observations were available for Haut-Ue‘le’ and thus, the results should be interpreted with caution. The coefficient of multiple determination (R2) is high (0.69), and all variables are significant at the 0.05 level except for the constant term which is 255 only significant at the 0.48 level. The partial correlation coeffici- ents of the variables are also all high,above I0.50I. As was the case for Base-Uélé, the factor whiCh explains most of the variation in cotton acreages is the number of years the cotton field remained in fallow (YOF) . The Beta weight is 0.92 and omission of this variable reduces R2 by 0.21. The regression coefficient is equal to 3.40, about the same as that for Bas-Uélé (3.73). The variable which reduces R2 the most when deleted is LAM (labor input for cotton 2 maintenance) with R delete equal to 0.28. This factor has also the highest absolute partial correlation coefficient with cotton acreages, |-0.75|; the Beta weight is equal to 0.68 and the coefficient to -0.40. Other important factors are PRC (the seed cotton price influenced cotton 2 acreage) with a Beta weight of 0.68, R. delete of 0.42 and a coefficient of 26.68; IPO (imposition of other crops) with a Beta weight of 0.62, R2 delete of 0.56 and a coefficient of 29.08 and POS (possession of a bicycle, radio, sewing maChine or watch) with a Beta.weight of -0.42, R2 delete of 0.55 and a coefficient of —l8.l8. Results for Bas-Uélé and Haut—Uélé The regression equation explains 52 percent of the variation in cotton acreages. All the variables included in the regression are significant at the 0.05 level except for the constant terms, SHL (number of years in primary school) and IPC (cotton imposition), which are significant at the 0.10 level of confidence. The partial correlation coefficients of the variables with cotton acreages are relatively low except for YOF (years of fallow) [0.52] and ROD (obligatory work on roads) {-0.38}. 256 As can be expected from our discussion of the results for Bas-Uélé and Haut-Uélé, YOF is the variable which explains most of the variation in cotton acreages for the Ue’le’s. The longer fields remain in fallow, the larger are cotton acreages per farm, although the effect is weak, 2.27 ares per year in fallow, ceteris paribus. The Beta weight is equal to 0.43 and deleting this variable reduces R2 by 0.18. ROD (obliga- tory work on roads) is tie second most important variable with a Beta weight of -0.36, R2 delete of 0.44 and a coefficient of —35.23. Clearly, then, work on roads reduces the acreage planted in cotton. The next two most important variables are labor input in cotton pro— duction: labor for cotton maintenance (1AM) with a Beta weight of -0. 31 and R2 delete of 0.46 and labor for other cotton activities (LAO) with a Beta weight of —0.29 and R2 delete of 0.47. The more labor input per hectare of cotton, the smaller the cotton acreages are. The greatest effect is for labor in cotton maintenance with a coefficient of -0.27; labor for other cotton activities has a coefficient of only -0.06. Variables of lesser importance are CHF (work for chiefs or other authorities) which has a positive effect on cotton acreages (coefficient: 15.26, Beta weight: 0.27 and R2 delete: 0.47); IPC (cotton imposition) which has a strong positive influence on cotton acreages (coefficient: 21.16, Beta weight: 0.27, R2 delete: 0.47) and SHL (years the cotton farmer stayed in primary school) which has a small positive effect on cotton acreages (coefficient: 2.82, Beta weight: 0.21, R2 delete: 0.48). 257 Results for the Northern Cotton Belt Including Ubangi , Mongala, Bas-UEIeT and Haut—Uélé Our linear multiple regression model explains just half of the variance in cotton acreages in terms of the variance of eight other variables (R2 = 0.50). .All variables included in the equation are significant at the 0.05 level of significance, except the dummy variable for Mongala (MOL) which is still significant at the 0.10 level. The partial correlation coefficients are generally low, except for the two most important variables, ACO (acreage of other crops: 0.43) and YOF (years in fallow: 0.36) . The major differences between sub-regions are reflected in the inclusion of two dummy variables for sub-regions in the equation: 2 delete of 0.48 and UBI (for Ubangi) with a Beta weight of 0.25, R coefficient of 12.58 and MOL (for Mongala) with corresponding values of 0.16, 0.48 and 15.57. ACO and YOF are the two most important variables. As we have already pointed out, ACO (Beta weight: 0.36, R2 delete: 0.39 and coefficient: 0.08) is the most important variable in Ubangi-Maigala and YOF (Beta weight: 0.31, R2 delete: 0.42 and coefficient: 2.14) in the Uélés. Other variables with a positive effect on cotton acreages are (31F (work for chiefs or other authorities) with a Beta weight of 0.27, R2 delete of 0.47 and coefficient of 13.77 and.TER (presence of termite heaps on cotton fields) with a Beta weight of 0.15, R2 delete of 0.48 and coefficient of 8.11. Variables with a negative influence on cotton acreages are the labor inputs per hectare in cotton produc- tion: LAM (labor for cotton maintenance) with a Beta weight of -0.24, 258 R2 delete of 0.46 and coefficient of -0.25 and LAO (labor for other cotton production activities) with a Beta weight of -0. 26, R2 delete of 0.46 and coefficient of -0.08. Thus, the labor input for cotton maintenance has a stronger negative effect on cotton acreages (coeffi- cient: -0.25) than the labor input for the other cotton activities (coefficient: -0.08) . Table 8.12. Republic of Zaire: Comparison of the Coef- ficients of Multiple Determination (R2) Obtained in the Analysis of Cotton Yields and in the Analysis of Cotton Acreages per Farm1 Sub-regions Analysis of Analysis of Cotton Yields Cotton Acreages Ubangi and Mongala 0.62 0.57 Bas-Uélé 0.96 0. 74 Haut-Uélé _ 0 . 99 0 . 69 Bas—Uélé and Haut-Uélé 0.68 0.52 Ubangi, Mongala, Bas-Uélé and Haut-Uélé 0.58 0.50 1The same significance level of the F-statistic was speci- fied for the study of acreages as for the study of yields, namely 0.095. Cotton Production Production is determined by total acreages planted and average yield per unit of acreage. , 'Ihese elerents have been analyzed in the previous sections. In this section we will concentrate on a comparison of the determinants of cotton production (number 0f cotton farmers, acreage per farm and yield in kg. per hectare and in kg. per farm) for 1958-1959 when production was at its peak, for 1970-71 using government statistics and for 1972-1973 using the farm business survey data. 259 Comparison of Cotton Production Determinants Between 1958-1959 and 1970-19716 Production in 1973 was 45.1 percent of the 1959 level; in 1971 it was 31.8 percent. The purpose of this analysis is to identify why cotton production declined over the 1958/1959-1970/1971 period. The characteristics per zone in the northern belt for 1958—1959 and 1970-1971 are found in Tables 8.13 and 8.14 using ”official government statistics. Table 8.15 presents the production of seed cotton in northern Zaire in 1958-1959, 1970-1971 and 1972-1973. The change in the determinants of cotton production between 1958-1959 and 1970-1971 in percent of the 1958-1959 figures for the Equator region are given in Table 8.16. For Ubangi production was at the 1959 level, but the increase in the number of cotton farmers and in total acreage was completely offset by a large decrease (26.6 percent) in yields per farm in kg. per hectare. For Mongala the large drop in production (68.0 percent) was due to three factors: a decrease in the number of cotton farmers, a large decrease in the acreage per farm and an even larger decrease in yields per farm in kg. per hectare. For the Equator region production was 28.1 percent below the 1959 level because of this large drop in pro- duction in Mongala sub—region. A decrease in yields per farm in kg. per hectare accounted for most of the Equator regions drop in production. Table 8. 17 provides the same information for the Haut Zaire region. The Ituri and Tshopo sub-regions are also included although they were 6We could not use 1972 data as 1971-1972 was an abnormal cotton production year. Lack of rains in October 1971 and major insect attacks caused a sharp drop in cotton production in Ubangi. The 1973 data was not yet available for analysis when this study was catpleted. 260 Table 8.13. Characteristics of Cotton Production in Northern Zaire per Zone, 1958-1959 and 1970-19711 Regions, Number of Cottm Farmers Acreage in Hectares Production of Seed SquRegions t m in Tons and Zones 1958-1959 1970-1971 1958-1959 1970-1971 1958-1959 1970-1971 K tor Regio 112,047 124,685 43,294 48,133 18,003 12,948 -Ubanqi Gamma 25,130 34,987 8,917 14,910 4,526 3,913 Libenge 13,580 11,018 4,650 4,462 1,148 983 Bosobolo 16,228 21,467 7,445 9,813 2,309 3,391 Budjala 9,057 7,240 3,170 3,326 1,474 940 Kungu 7,478 12,702 2,560 3,761 1,187 1,363 Total 71,473 87,414 26,742 36,272 10,644 10,590 -Mongala MUD ‘ C (Ruhr/ville) 18,166 14,392 7,658 4,927 3,213 819 Busiwm 14,549 15,541 5,379 5,099 2,045 1,178 [4111:1211 7,254 7,338 3,264 1,835 1,989 361 1.15.113 605 0 251 0 113 0 Total: 40,574 37,271 16,552 11,861 7 359 2,358 limit/Kiln) 120311311: £7,191 149,906 £93,123 40,276 15,754 10,489 41.5-(101a [\XL‘ ' 13,202 8,846 6,198 2,533 4,016 550 A1110 18,129 11,879 6,618 3,023 4,064 1,153 Damn “I 13,081 6,960 11,187 3,879 8,643 1,849 nut.) 7,990 4,799 4,319 1,586 2,911 I‘okc 23,206 11,038 10,903 3,046 5,647 974 Bondo 22,805 15,980 9,758 4,322 5,457 1,945 Total: 98,413 59,502 48,983 18,392 30,738 6,867 -l {ant-UL" lo. Ulll'YfLJ 37,924 27,064 19,777 5,671 7,956 1,143 Isiro 21,451 11,884 7,823 3,266 2,924 738 Hamlin 25,920 18,595 10,165 4,521 4,615 1,083 Nltlngzir.) 13,687 6,776 4,469 3,860 1,455 106 l-‘arad ]L‘ -—— 3,426 —— 451 --— 45 W. tsa --- 546 —-- 134 --- 21 Total: fi,fi 68,291 42,234 17,903 16,950 3,134 -l turi Mahagi 10,445 16,977 4,067 2,467 3,456 130 Total: 10,4fi 16,977 4,067 2,467 3,456 130 4129110110 Bylfwasunde 9,686 -—- 3,878 —- 2,275 - Banalia 9,865 5,136 3,961 1,514 2,335 358 Total: 19,551 5,136 7,839 1,514 4,610 358 'Ibtal North 339,438 274,591 146,417 88,409 73,757 23,437 chubl 1(1'UL du Zaire, Avril 1972. Projet do Raltlnce Cotonniere dans la Province de 1' Flimgteur. anshas Ministére do 1' Agriculture, annexe 12 for the 1958—1959 figures and Re'publique du Zaire, Rapport Anni Ie,1 Service Provincial de 1‘ Agriculture, Province de 1' Equateur, 1971 and Province du Haut—Zaire, 1972 for the 1970—1971 figures. Note: It is doubtful whether the 1958—1959 and 1970-1971 statistics are accurate. The nurber of mttcn farmers is probably fairly accurate as an enumeration of all cotton farmers is quuarly can led out by agricultural officers and extension agents. We believe that the production figures are accurate as they are derived from the seed cotton purchase figures of the cotton companie . However, we seriously question the accuracy of acreage figures as most territorial. agronomists and cotton extension agents responsible for the collection of this data do not measure the fields accurately. Thus, acreage figures should be considered as rough gm. stimates. consequence, yields derived from the production and acreage data are also a guesttimate, but yields in kg. per farm are probably fairly accura 261 Table 8.14. Characteristics of Cotton Production in Northern Zaire per Zone, 1958-1959 and 1970-197.11 Regions, Yield (kg./ha.) Yield (kg. per farm) Acreage per Farm in Ares Sub—Regions and ZOFKS 1958-1959 1970-1971 1958-1959 1970-1971 1958-1959 1970-1971 Equator Region: 416 270 161 104 39 39 —Ubangi Gamma 508 262 180 112 35 43 Libenge 246 220 85 89 34 40 Bosobolo 310 345 142 158 46 46 Budjala 465 283 163 130 35 46 Kungu 464 362 159 107 34 30 Toml 398 292 149 121 37 41 —lVbnqala Nbbaye (Banzyvillc) 419 166 177 57 42 34 Businga 379 231 141 76 37 33 umba 609 196 274 49 45 25 Lisala 445 0 187 O 41 0 Total: e2 12% fl 2 11 3.2 Maui-.2299 mien: w E 2.42 .72 e a —Bas—Ucle Aketi 648 617 304 62 47 29 Anne 614 381 224 97 37 25 Bambesa 773 476 661 266 86 56 Buta 674 250 364 83 54 33 Poke 618 320 243 88 47 28 Bondo 559 450 239 122 43 27 mm Q 212 22 1L5 :9 .31 -lIaut-Uele ungu 405 206 210 42 52 21 15m) 373 226 136 62 36 27 Wamba 453 239 178 58 39 24 Niangara 325 27 106 16 33 57 Faradje -—— 100 --- l3 0 l3 Watsa —— 157 —— 38 0 25 Total: are E E e e a -Ituri Mahagi 850 53 331 8 39 15 Tom: 822 .52 32 _8 a e —Ishopo Bafwnsende 586 —-- 235 —- 4O 0 Banalia 589 236 237 70 40 29 Tom M 212 2a _79 Q 19. Total Nerd) 504 265 217 85 43 32 1 Derived frcm Table 8.13. 262 Table 8.15. Production of Seed Cotton in Northern Zaire in 1958-1959, 1970-1971 and 1972-19731 Regions and Sub-Regions 1958—1959 1970-1971 1972-1973 ‘Tons Equator Region 18,003 12,948 12,709 Ubangi 10,644 10,590 9,573 Mongala 7,359 2,358 3,136 Haut Zaire Region 55,754 10,489 19,0752 Bas-Uélé 30,738 6,867 10,440 Haut-Uélé 16,950 3,134 7,621 Ituri 3,456 130 1,456 Tshopo 4,610 358 560 lrrom Table 8.13; for the 1972—1973 data, for Equator Region: République du Zaire, 1974, Ragport Annuel 1973, Région de l'Equateur, Division Régionale de l'Agriculture; for the 1971-1972 data for Haut Zaire Region: République du Zaire, 1973 (a) Rapport Annuel 1972, Region du Haut Zaire, Kinshasa: Division de l'Agriculture. 2The data for Haut Zaire are for 1971-1972 as the 1972-1973 data were not available per sUb~region. The 1972-1973 seed cotton production for Haut Zaire was estimated at 21,178 tons. Table 8. 16. Change in Cotton Production Determinants 263 Between 1958-1959 and 1970-1971 for Equator Region, Zaire (Change in Percentage of the 1958-1959 Figures) Cotton Production Ubangi Mongala Equator Determinants Sub-Region Sub-Region Region No. of Cotton Farmers +22.3 . - 8.1 +ll.3 Total Acreage +35.6 -28.3 +11.2 Acreage per Farm +10.8 -22.0 no change Production - 0.5 -68.0 -28.1 Yields in kg./ha. -26.6 -55.3 -35.1 Yield per Farm in kg. -18.8 -65.2 -35.4 1 Derived from Tables 8.13 and 8.14. not part of the farm business survey. importance as compared with the Uélés. Both sub-regions are of lesser In Bas—Uélé the number of cotton farmers, the average acreage per farm and the yield per farm in kg. per ha. decreased about 40 percent, and account for the large drop in cotton production, 77. 7 percent. For Haut-Uélé the same picture is repeated, but the reduction in yields per farm in kg. per ha. is much larger, 73.1 percent. Production there dropped 81.5 percent. For Ituri and Tshopo production declined more than 90 percent because of the combined effects of a decrease in acreage per farm, yield per farm in kg. per ha. and nurber of farmers (except for Ituri) . In summary, for the Haut Zaire region the number of cotton farmers, acreages per farm and yields per farm each declined more than 30 percent and the production was less than 80 percent of the 1959 level. The single most important factor responsible for this decline was the decrease in yields per farm in kg. per ha. 264 .vaé cad mad mmHQeB 500% «00.2.58; 0.00- 6.00- 0.00- 0.60- 0.60- .00 00.0060 000 60000 0.00- 0.0m- 0.00- 0.6m- 0.06- .60\.00 00 60000 ~.00- ~.mm- 0.00- 0.00- 0.00- 0060000000 0.06- m.0m- 0.00- 0.06- 0.06- .0060 000 0060006 0.00- n.00- m.0m- 0.50- 0.00- 0060006 H6009 0.66- 0.65- m.mo+ 0.06- 0.0m- 0000060 000000 00 .02 009.590 coflmmmlhom cameo-now coflmmmlnom dogma-ham 3000:0538 00060 0060 000000 00000 0600-0060 .0600-060 0000000000 000000 10000000 0000-0000 000 00 0060000000 00 0006000 000000 000000 000 606 0000000-000 00060 0060 000 0000-0s00 606 0000-0600 0003000 000600000000 0000006000 000000 00 000600 .m0.0 00000 265 Comparison Between the Official Government Statistics for 1970-1971 and the Farm.Business Survey Data for 1972-1973 The cotton production parameters derived in the farm business survey could not be compared with the official 1972-1973 data as the 1973 agricultural report of Equator region did not contain the desired data and the 1973 report for Haut Zaire region.was not.yet available in July 1974 when the author left Zaire. The 1971-1972 data were "abnormal" because of the October draught in Ubangi and massive insect damage to cotton. Thus, we will compare the survey data with the official 1970-1971 data (Tables 8.13 and 8.14). The cotton production detenninants for Ubangi-Mongala are almost identical for 1970-1971 and 1972-1973 (Table 8.18). For the Uélés the acreage data is crnmerable, but there are marked differences in yields, in terms of kg. per ha. and kg. per farm (Table 8.18). Yields in 1972-1973 were more than double the yields in 1970-1971. The number of cotton farmers per sub—region could not be estimated in the 1972-1973 farm.business survey. If we accept the 1972—1973 farm business survey data for the Uélés, then the implications for our analysis of the reasons why cotton production in northern Zaire in the 19705 was low are: -—yields in kg. per ha. for the Uélés in 1973 were at the 1959 level (somewhat higher for Bas-Uélé, somewhat lower for HautrUélé) (Table 8.19). This contrasts with the official 1971 data.whidh showed a sharp decline since 1959. --acreage per farn1(Table 8.19) dropped in Bas-Uélé (down 34 percent) and in Haut—Uélé (down 42 percent) at about the same rate as for the official 1970-1971 data. 266 6.7m can mad mmanme can 8238 Emma 3; mm em 8 omm m3 m2 2: .3 E Emu 8m Emu; 3m mom sum was mms mum Sm EN .2 “an .3 fl 3m? mm mm mm mm mm gm 8 mm 8% 5 Emma 8m mung mama Head mums Hams mama Head mama shag -msma -osma -mema -osma -Nsma -osma -Nsma -osma mHHmN fimfiuoz c.3952 mpgmfimo Hmuoe .wflmauusmm _wamsummm uflmcmn: cowuosooum copuoc an 3.10me How Himumo mm>usm mmmcfimsm_eummv msmaumsma can Amumo Hmfiowmmov mucmcdfiwpmo coauosconm £0.58 £8.33 somHHmmubo "mnwma cumapuoz .mH.m magma Table 8.19. 267 Carparism Between Cotton Production Determinants for 1958-1959 (Official Data)and 1972-1973 (Farm Business SunreyData)fortheUélésa1flforNorthernZaire Cotton Production Has-Uélé Haut—Uélé Total Determinants Northern Zaire 1958- 1972- 1958- 1972- 1958- 1972- 1959 1973 1959 1973 1959 1973 Acreage per farm in ares 50 33 43 25 43 36 Yield in kg. per ha. 628 753 401 376 504 391 Yield per fann in kg. 312 250 171 94 217 140 lFran Table 8.14 and Table 6.22. --yields per farm in kg. (Table 8.19) dropped in Bas-Uélé (down 20 percent) and in Haut-Uélé (down 45 percent) but much less than in the official 1970-1971 data. Thus, according to the 1972-1973 farm business survey data, the low level of cotton production in the Uélés relative to 1959 can be explained not in terms of yields in kg. per ha., but because of a decrease in acreage per farm and a large decrease in the nunber of cotton farmers. Cotton production has increased rapidly since 1971 in Haut Zaire region. Production is now around 20,000 tons of seed cotton (see Tables 3.11 and 8.15) but, it is not known which factors account for this increase although, probably, the carbined effects of an increase in yields, particularly since the introduction of a near seed variety (Reba B 50), and an increase in the nurber of cotton farmers since the raise in seed cotton prices in 19 72 brought about the increase in production. 268 annex A multiple linear regression equation model was used to explain the variation in yields between farms. For the northern cotton belt a coefficient of multiple determination (R2) of 0.58 was obtained and the dummy variable for Bas-Uélé turned out to be the most important factor. Other important variables with a positive effect were cotton revenues, the distance from the house to the field, cotton plant density, imposition of other crops than cotton, labor performed on other cotton activities than maintenance and labor units available on the farm. In the analysis of factors explaining acreages, an attempt was made to determine the price elasticity of supply in terms of acreages planted. The results do not confirm our hypothesis of a positive nominal price effect. The following factors probably influenced the unexpected outcome of our supply response analysis: the real cotton price remained almost constant over the 1970-1972 period, the 1970 acreage data could have been inaccurate and other things may not have remained equal over the time period considered. The variations in cotton acreages between farms were analyzed with a linear multiple regression technique and one-half of the variance in cotton acreages was explained in terms of the variance of eight other variables for the North (R2 = 0.50) . 'IWo dumy variables were included in the equation, one for Ubangi and one for Mongala, reflecting the major differences between sub—regions. The most important variables were acreage in other crops, particularly in Ubangi-Mongala, and years of fallow, particularly in the Uélés. Both variables had a positive ‘effect on cotton acreages. Another variable with a positive effect was 269 work for chiefs and other authorities . Variables with a negative influence were the labor inputs per hectare in cotton production with labor for cotton maintenance having a stronger negative effect than labor for other cotton activities . A comparison of the cotton production determinants between 1958- 1959 and 1970-1971 revealed that production in 1971 in Ubangi was at the 1959 level but that the increase in the nurber of cotton farmers and in total acreage were offset by a large decrease in yields per farm in kg. per ha. For Mongala, production dropped 68 percent as a result of a decrease in the number of cotton farmers, a large decrease in acreage per farm and an even larger decrease in yields per farm in kg. per hectare. In Bas-Uélé the number of cotton farmers, the average acreage per farm and the yield per farm in kg. per hectare decreased all in the same degree, about 40 percent,and accounted for the large drop in production, 77.7 percent. In Haut-Uélé the same picture was repeated but the decline in yields was much larger, 73. l percent,while production fell by 81.5 percent. A comparison between the cotton production determinants derived in the 1972-1973 farm business survey and the 1970-1971 official data yielded almost identical data for Ubangi-Mongala but marked differences for the Uélés, except for the acreage data. Yields in kg. per hectare for the Uélés in 1973 were at the 1959 level, sale-what higher for Has-Uélé and somewhat lower for Haut-Uélé. This contrasts with the official 1971 data which showed a sharp decline in yields since 1959. As a result, this author questions the 1970-1971 yield data for the Ue’le’s. According to the 1972—1973 farm business survey data,the low 270 level of cotton production relative to 1959 can be explained only in terms of a decrease in acreage per farm and a large decrease in the number of cotton farmers in the Uélés. CHAPTERIX \/ STRATEGIES FOR IMIREASING common PKJDUCI'ION INNORI'HERNZAIRE Introduction The purpose of this study is to understand the functioning of the cotton subsector in northern Zaire, to identify barriers to improved perfonmance, including production and marketing performance ,and to formu- late policy prescriptions for improving the performance of the subsector . In the previous chapters , we have described the cotton subsector in northern Zaire and we have identified a number of important barriers to improving the performance of the subsector. These barriers are diagnostic analytical statements about the performance of the subsector. They are discussed in the next section. We will then analyze alter- native strategies for expanding cotton production in northern Zaire within the overall political strategies for developing northern Zaire. In the final chapter, we will advance systematic policy prescriptions for increasing cotton production and for improving the performance of the cotton subsector in northern Zaire. Barriers to Improving the Performance of the Cotton Subsector in Northern Zaire Under present pricing policies -Cotton is not profitable for most farmers in the North relative to coffee and sure food crops (rice, maize). The real price 271 272 of seed cotton is now about half the 1960 price. --Cotton producer prices are uniform throughout Zaire resulting in allocative inefficiencies , a dispersion of cotton production and high per unit assembly and transport costs. -Cotton lint is presently being sold to domestic spinners at a price well below the export price. -Cotton growing imposition is still widespread in the North; the location and the size of cotton fields are largely deter- mined by cotton extension agents and to a lesser extent, the territorial agronomist and village or clan authorities. The cotton extension service performs the role of a regulatory agency. --The cotton extension service is uncoordinated, poorly trained and largely ineffective in educating farmers. "Recommended agronomic practices are not followed by the majority of the cotton farmers . Seed cotton yields are lcw-an average of 399 kg. per ha., while experiment stations in the North easily obtain 1,000 to 1,500 kg. per ha., without using chemical fertilizers. —-Improved inputs such as chemmal fertilizers, pesticides, creditandmechanizationarenotusedbythefarmers andthere is almost no information on the economics of using these inputs at the farm level. --The present cotton variety in the North is degenerating; no Significant cotton breeding and cotton selection is underway to rorthern Zaire. 273 --Famers lack information on the seed cotton price , time of collectim and weight of their crop . --Per unit cotton marketing and processing costs are very high. -Seed cotton collection starts too late and is spread over an extended period of time. --Seed cotton collection is inefficient in terms of total col- lection per day and per collection team. “Transport costs are high because of the long distances involved and poor condition of the roads. -Cotton lint is charged the most expensive fare category of all agricultural products transported by public and semi-public transporters . --The coordination of cotton marketing and processing operations is weak. --Ginning costs are high because of underutilization of available ginning capacity and poor organization of processing operations. -New ginning equipment is being installed which will ‘result in an even larger underutilization of the available ginning capacity. Based on the diagnosis of these different barriers on improved performance, we will now analyze a number of different strategies for overcoming these barriers. We will first show that these strategies must be viewed within an overall political economy framework for deve10ping rural areas in rorthern Zaire. 274 Strategies for Developing Northern Zaire Northern Zaire is an agrarian dominated region. The government of Zaire has now committed itself to develop the North, based on indus- trial developmeht of the third growth pole centered in Kisangani and the development of rural areas. In fact, the southern part of the country is leading the country through expansion of the mining industry and the development of the major food crop subsectors, particularly maize, but also the cotton and livestock subsectors. The Zaire govern- ment favors a balanced growth of all the regions according to their natural resource endowments . However, a cotprehens ive strategy for developing the North has not yet been announced although baseline studies have been corpleted or are underway by private firms . One of the main problems of developing the North is the govern- ment' 3 adherence to uniform pricing policies for most agricultural pro- ducts. These policies result in low production densities, and, as a consequence, high assembly and transport costs, lack of regional specialization and lack of an adequate infrastructure in areas with a corparative advantage for the development of a particular subsector . Thus , there are substantial real economic costs involved in maintaining uniform prices on social-political grounds. This problem will be discussed in the next section. Strategies for developing rorthern Zaire will require pi lot development projects and policy guidelines to move towards regional specialization of farming systems. Major shifts in location of agri- cultural production are likely to occur with improved transport and cotmunications and with producer prices based on cotparative advantage . 275 Until more research on farming systems in the North is corpleted, it will be difficult to indicate the areas for regional specialization and to identify viable pilot development projects for the crops other than cotton. This cotton research will serve as a model for undertaking studies on other comrodities to assist policy makers in furthering our understanding of potential regional specialization of production in rorthern Zaire. Strategies for Ebpanding Cotton Production in Northern Zaire The first priority for improving the performance of the total cotton subsector in Northern Zaire is to expand cotton production within a framework of regional. specialization in order to reduce assembly and transport costs and to utilize the excess ginning capacity. There is a colonial legacy in Zaire of assuming that farmers are unresponsive to production incentives , to new technologies of production and to economic opportunities. This is reflected in low producer prices paid to farmers, in requiring farmers to grow certain crops and to observe certain agronomic practices and in treating the farmers ' fields with pesticides. Instead of performing an educational role, the cotton extension service played the role of a regulatory agency during the colonial period and it is still playing this role today. This author's field research has clearly shown that farmers do respond to production incentives and economic Opportunities and, there- fore, strategies for expanding cotton production need to focus on improving the incentives for small farmers. These strategies should provide cotton farmers with improved production incentives , extension education for improving cotton yields and for expanding acreage per farm. 276 Revis inLPricing Policies and Removing the Impps ition System Cotton farmers definitely need stronger incentives to increase cotton production. Our field research showed that small price changes will have little effect and, thus, a substantial increase is called for. Farmers will also adjust their production rather slowly to modest changes in producer prices . A substantial increase in the seed cotton price will stimulate cotton production and will speed up the diffusion of innovations, the absorption of new inputs and the utilization of idle processing capacity. A substantial increase in the seed cotton price will reduce the need for the imposition system as farmers will be motivated to produce cotton without corpulsion. By itself, the imposition system does no ham but it has a high administrative cost as it detracts extension agents from their primary role which is the education of farmers. Thus, an increased producer price will allow the imposition system to be reroved and will free cotton extension agents from regulatory tasks. Vhen cotton production becoTes profitable, farmers will also be inter- ested in protecting their crop from insects and pests. Thus, the government should drop spraying and dusting the f armers' fields and instead should concentrate on teaching farmers how to treat their fields with pesticides. The price increase could be both in cash and in kind. For example, many farmers cannot acquire small farm tools in their community and trey could be made available at subsidized rates. Since coffee production already exceeds Zaire ' s quota, there are sound arguments for increasing the export duty on coffee and channeling 277 part of this additional revenue into crop diversification in the major coffee growing areas of the North where cotton could replace some excess coffee production. However, as long as the returns to labor in coffee production are well above those for cotton, government attempts to replace coffee with cotton will be futile. Hence, a substantial increase in the seed cotton price is needed to encourage farmers to shift from coffee into cotton production. In terms of economic efficiency, uniform pricing--paying all farmers the same producer price-~should be abandoned. A shift to vari— able producer prices based on transport costs would induce a concentra- tion of cotton production in a few areas and would substantially reduce per unit assembly, transport and ginning costs. Pilot cotton development projects could then be located in the areas with natural resource endowments favorable to cotton. However, regional pricing policies are probably unacceptable to the Zairean government for political reasons as it would condemn the rerotest regions to near subsistence existence. The real cost of a uniform pricing policy is thus accepting low cotton production densities, high per unit costs of assembly and transport and the need to develop a better feeder road system for the entire North. This study has shown that COGERCO and then ONAFITEX have stabilized the'cotton lint price for dorestic mills in order to maintain low textile prices for Zairean consumers. For example, in 1973 and 1974 when the world market price of cotton more than doubled producer prices remained unchanged and domestic lint prices were only half or even less than half of export prices. We see no reason why producer prices and domestic lint prices should not be set in relation to world market 278 conditions. Since only a small quantity of Zairean textiles are exported, there is no valid argument for providing cheap lint to the mills. Even during the years when the domestic lint price was 25 percent above the export price FOB Matadi, the domestic price was still below what domestic spinners would have paid for imported cotton because of import duties, taxes, transport and handling costs, etc. Basically, the question of domestic cotton pricing is one of transferring consumer's surplus into a producer's surplus in order to stimulate cotton production. This author believes that a temporary increase in the domestic lint price of 15 to 20K. per kg. will generate enough revenue to enable ONAFITEX to raise the producer price from 6.5K. to 10 K. per kg. of first grade seed cotton. Such a price would sub- stantially increase cotton production, reduce per unit marketing and processing costs and provide ONAFITEX with substantial export earnings. A price increase of 15 to 20K. per kg. of cotton lint for domestic spinners would raise the price of domestic lint to about 60 to 65K. per kg. of lint, still well below the price which they would have to pay for imports of the same grade and quality. The uncertainties associated with such a scheme are the assumptions that cotton lint export prices will remain high and that farmers would react fairly quickly to a 50 percent increase in the producer's price. Cotton pricing policies cannot be divorced from price policies for other crops, especially since many of the other crops are grown either in a rotation with cotton or are in competition for labor with cotton. Thus, pricing policies need to be designed for a cotton-foodrexport crop production system. 279 In Zaire, low producer prices are fixed for all basic food and fiber crops in order to protect the real incotes of urban consumers. In fact, this is the major contradiction in agricultural policy in Zaire / and in developing northern Zaire . Since agriculture has been declared the priority of the priorities, the time is ripe to shift from negative price policies which depress agricultural prices to positive price policies which provide adequate incentives to farmers to raise agri- cultural production and income and to assist Zaire in becoming self- sufficient for the basic food staples. Diffusion of Improved Agronomic Practices Although experimental plots with improved agronomic practices and no chemical fertilizers consistently produce more than 1,000 kg. of seed cotton per ha., we find that farmers'yields average around 400 kg. Thus, there is a large opportunity for increasing yields by introducing improved agronomic practices . There is a body of improved cultural practices which were recomended by INEAC before 1960, INERA after 1960 and the CFDT-mission in Ubangi-Mongala. These improved agronomic practices include recom- / mendations on: (a) duration of fallow, (b) crop rotation and inter- ‘ cropping, (c) use of improved hand tools (d) choice of field plots, (e) land preparation, (f) time and method of planting, including plant density, (g) timing and method of thinning and weeding, (h) disease and insect control, (i) time and method of harvest and (j) sorting of cotton, storage of seed cotton and storage of cotton seeds. These improved agronomic practices could be extended immediately to farmers. The cotton extension service should concentrate on the 280 introduction and diffusion of improved agronomic practices. Demonstra- tion.plots in selected villages would be very useful in showing farmers what could be achieved. For its part, ONAFITEX should sort cotton seeds in the ginneries and disinfect before distribution to farmers in cmder to improve germination rates and to assure adequate plant densities. In the next section, we will advocate an overhaul of the cotton extension service to facilitate the introduction and diffusion of improved agronomic practices possible. In particular, we believe that the creation of regional cotton extension committees could be very useful for launching a cotton extension campaign to diffuse improved agronomic practices. Overhauling the Cotton Extension Service Unless the agricultural extension service is reformed and retrained, there is little hope that better agronomic practices will be observed by farmers or that pilot cotton development projects will succeed. Several actions need to be taken to reform the cotton extension service in order to shift it from a regulatory to an educational agency. The administrative structure of agricultural extension should be reformed in order to coordinate and integrate agricultural extension efforts. Agricultural monitors responsible for cotton extension should be placed under a single authority instead of three different ones as is presently the case. This authority does not necessarily need to be ONAFITEX although this would facilitate the organization of cotton extension. ONAFITEX which by decree is responsible for providing technical services to cotton farmers needs a certain degree of control 281 over the cotton extension service. ONAFITEX will need a cadre of well trained cotton extension officers who will be in charge of developing and implementing an overall cotton extension program. Agricultural extension workers need to be retrained continuously through formal training sessions and by means of extension bulletins, extension seminars, etc. Their performance should be reviewed periodically. Prorotions, transfers, etc. should be a function of their effectiveness. Training should include recommended agricultural practices , the design of demonstration plots,cormunication skills and elementary farm manage- ment, particularly partial budgeting techniques. The function of extension agents also needs to be reviewed and they should be provided with adequate transportation, facilities for field work and demonstration plots, opportunities for professional advancement and supportive personnel and services. They should have the opportunity to provide feedback information to the officer in charge of developing and implementing the cotton extension program. In order to better organize and coordinate cotton extension efforts, a cotton extension committee should be appointed in each regional branch of ONAFITEX and charged with drawing up a campaign for the next crop. The committee should be chaired by the regional or national director of the agricultural service of ONAFITEX. The committee should include representatives from the Department of Agriculture, INERA, CFDT, the extension service of ONAFITEX and the regional administration. Members who may be co-opted would be cooperative leaders , customary chiefs, cotton production specialists, etc. This committee should be responsible for the development of 282 general guidelines for the expansion of cotton production. These guide- lines should constitute the terms of reference for all cotton extension agents. Immediate steps should be taken to develop a cotton handbook which summarizes and translates the recormended cultural practices into local languages for all cotton extension agents. Each year, new cotton research results should be added to this handbook. Presently university trained agronomists at the "graduat" level and the "ingenieur" level have acquired specialized skills for agri— cultural research but are not well prepared for agricultural extension responsibilities. At the National University of Zaire, there is no department concerned with the problems of diffusion of innovations and agricultural extension. The department closest to this erphasis is the department of communications which is mainly concerned with the mass media: press, radio, TV, film and their impact in urban areas. Thus, there is an urgent need for a department of agricultural extension. In order to create a capacity for agricultural extension education in Zaire and to focus more attention on the vital role which the agri- cultural extension service plays in increasing production, incores and employment, a department of agricultural extension stould be set up at the National University of Zaire. This department should develop a curriculum in agricultural extension education including the following fields of study: communications, rural sociology, fanm management, agronomy and agricultural development. It is logical that such a department would be created in the Faculty of Agronomy. It could be added to the existing Department of Agricultural Economics or it could be established as a separate department. The major constraint for 283 creating such a department is the shortage of qualified staff in this field, either Zairean or expatriate. Graduates in agricultural extension will assure leadership in extension positions in the Department of Agriculture, in ONAFITEX and in other product marketing offices. They will be responsible for the training and upgrading of agricultural extension workers. An extension liaison department should be created in INERA to ensure that research findings are diffused via the agricultural exten- sion service and that there is constant feedback into the research system. This unit should provide liaison between extension workers and researchers which at present appears extremely weak. A major responsi- bility of this department would be the preparation and distribution of extension bulletins. This unit could also provide the needed link between INERA and ONAFITEX, between INERA and the CFDI‘ mission, and between INERA and the Faculty of Agronomy. A Pilot Cotton Development Project in Bas-Uélé The dispersion of cotton production in northern Zaire, fostered by uniform seed cotton prices, is largely responsible for the high asserbly and transport costs , particularly in the Uélés where produc- tion densities are the lowest. A strategy to increase cotton produc- tion will require specialization of cotton production in a particular area and concentration in a pilot cotton development project. In this respect, the potemtial for increasing cotton production in Bas-Ue’le’ is particularly high. Yields are already relatively high, on the average 687 kg. per ha. , but the variation in yields between farms is also large. Sore fanrers obtained yields of 1,200 to 1,600 kg. per ha. 284 indicating that substantial improverents in yields could be achieved by the majority of the farmers. Corpared to the 1959 situation, acre- ages per farm are now much smaller and the nmber of cotton farmers is now much lower. A pilot cotton developrent project, centered in Bambesa and focusing on extension of improved agronomic practicesmould have a great potential for increasing cotton production. This area is one of the best cotton producing areas in Zaire. The primary objectives of such a project would be threefold: an increase in yields to an average of 1,000 kg. per ha.; an increase in average cotton acreage per farm to 50 ares and an increase in the number of cotton farmers. The feeder roads should be improved and provisions should be made that the ginning mills located in the area are able to handle the increased production. Farmers should also receive guarantees that their produc- tion will be bought up early. A project similar to the CFUI‘ project in Ubangi-Mongala should test the potential of improved agronomic practices under actual farm conditions in the Bambesa area. As the ecological conditions for cotton production in Baa-Uélé are much better than in Ubangi-Mongala, the payoffs of such a project would largely surpass those for Ubangi- angala. The project should start in the paysannat of Bambsea, also known as the paysannat of the Babua, which was one of the showpiece settle- ments before independence [Young, 1965] . Farmers in paysannats have been exposed to intensive agricultural extension efforts and they are generally regarded as the most progressive farmers. However, in the 285 present political context of Zaire the paysannat formula should be adapted to suit present objectives of social organization and authen- ticity. Paysannats will have to conform to traditional forms of organization and control, and they need to be integrated into the local political party structure in order to survive. Accelerated Research on New varieties Improved cotton seeds can be very effectively distributed to farmers as all farmers receive seeds from the ginneries each.year. In fact, the distribution of seeds requires very little administrative capacity since the seeds are distributed to the fanmers during the collection of second grade seed cotton. The variety presently cul- tivated in the North, Reba B 50, is well adapted to local ecological conditions. Its resistance t0'wilt and bacteriosis is a major asset in this humid climate. Since 300 tons of Reba B 50 seeds were imported from.the Central African Republic in 1967, no new seeds were introduced and there is a serious degradation in varietal purity. This is evidenced by the reduction of average fiber length from 1 3/32" in 1967 to l l/l6" in 1974, the heterogeneity of fibers, the decrease in ginning percentage and the appearance of wilt and bacteriosis in certain areas. In 1974, according to ONAFITEX agents in Gemena, the blue disease appeared in cotton fields along the Ubangi river. If this decrease spreads, it will be necessary to introduce a new variety resistant to this pest. Research on this disease is presently underway in the I.R.C.T. researdh station in Bambari, Central African Republic.1 lSource: CFDT-mission in Gemena. 286 Since 1971 the INERA station in Bambesa has been multiplying purebred seeds of Reba B 50 from the Gandajika station in the South. These seeds have then been further multiplied in paysannats (the Babua paysannat in Bambesa) before distribution to farmers. In 1971 the Bambesa station multiplied 5 hectares and in‘1972-1973 15 hectares.2 Different cotton varieties are being corpared in the Bambesa station as well as in the Boketa station. The CFDT mission in Ubangi-Mongala now plans to multiply 2 tons of seeds from the Bambesa station for distribution in Ubangi-Mongala. Clearly, research for improved varieties yields high payoffs. As new physiological strains of parasites (mutants-biotypes) develop spontaneously, breeding and selection of pest and disease resistant varieties need to be a continuing activity. 3 The major challenge is to corbine resistance genes with others having high productivity or commercial qualities. When cheap fertilizers becore available in Zaire, it may be ecoromical to breed for high yielding fertilizer responsive cotton varieties. Presently, cotton breeding and selection in Zaire is concentrated in the well equipped INERA station of Gandajika in the South, in a savannah area. Since 1964 they have developed improved varieties for the southern cotton belt but not for the rorthern. Present staff 2In 1972 they obtained yields of 600 kg. per ha. with late plant- ing and as a first crop after fallow. In 1973 yields were 1,500 kg. per ha. with planting between 15 and 20 July, Spacing of 100 cm x 30 on, mechanized tilling and weeding and treatment with DDT cotton dust. 3Cotton resistance to insects is characterized by abundant pro- duction of gossypol in the vegetable parts, absence of nectaries and bractless bolls [National Academy of Sciences, 1974] . 287 availability does not permit the development of separate cotton.breeding and selection stations fro the North and South. As soon as adequate and corpetent staff becore available, the Bambesa station should be geared up for cotton improvement research for the northern cotton belt. The NOrth badly needs its own research program, as ecological conditions in the South are different from.the North. In the meantime, better cooperation between the French IRCT stations in west and Central Africa, INERA in Bambesa, Boketa and Gandajika, the agricultural service of ONAFITEX and.the CFDTHmission in UbangieMongala is desirable fer restoring and.maintaining the varietal purity of Reba B 50, for the identification of a better variety and for gaining access to improved varieties developed in other countries. Chemical Fertilizers The farm business survey has shown that chemical fertilizers are not being used in the North by any of the 129 farmers in this sample. In this section, we will analyze a fertilizer strategy for expanding cotton production. First, we will review the experiences with chemical fertilizers in Zaire, then we will discuss the major constraints on using fertilizers and finally, we will evaluate a fertilizer strategy for expanding cotton production in northern Zaire. Since chemical fertilizers have not been discussed elsewhere in this study and since the government has decided to create a domestic fertilizer industry, more attention will be devoted to this strategy than to the other ones discussed in this chapter. 288 Review of Experiences with Fertilizers in Zaire The first controlled experiments with fertilizers on cotton in Zaire were set up in 1935 in the Bambesa experiment station of INEAC [FOCAN, 1950]. It was concluded that nitrogen had a weak effect, that phosphates had the most pronounced effect and that potash also had a real positive effect. Increasing doses of N, P, K increased yields although at a decreasing rate. However, there were only a few repeti- tions in the experiment and the soil was far frcmlhomogeneous. INEAC also experimented in Bambesa with organic fertilizers and found that ten tons of cotton seed compost per hectare increased yields by 57 percent. Cotton seed ashes showed a pronounced effect as 800 kg. of ashes per hectare increased yields by 73 percent. Fresh cotton seed had a positive although weak effect, while half—decomposed banana debris showed a doubtful effect. Contrary to the Bambesa results experiments with chemical fer- tilizers on cotton in the INEAC experiment station of Gandajika in 1940 demonstrated a significant effect on nigrogen and a less pro- nounced although positive effect on phosphates and potash. And 450 kg. of ammonium sulphate per hectare increased yields by 41 percent.4 After WOrld War II INEAC undertook several experiments in continuous cultivation with chemical fertilizers in forest areas and savannahs in the northern cotton belt. These experiments were set up 4Too much nitrogen stimulates vegetative development of cotton and reduces fiber production. Forest soils are often nitrogen rich while savannah soils, more exposed to the sun and biochemical decompo- sition, often lack nitrogen. This may explain the difference between the results of Bambesa, in the forest area, and Gandajika, in the savannah area. 289 in the INEAC research stations of Bambesa, Boketa, Kutubongo and Magcmbo and throughout the bio—climatic regions of the North. The effects of types of fertilizer and rates of application on yields of various crops in a rotation and on the regeneration of soil fertility and the need for fallow were studied. Jurion and Henry [1969] discussed these experiments in detail and.we will limit ourselves to the major conclusions. In Bambesa, a fOrest area, an annual application of 100 kg. triple superphosphate per hectare and 100 kg. ammonium sulphate per hectare produced two weeded crops per year (for instance, groundnuts and maize or cotton) for at least six years without evidence of any serious deterioration in soil fertility. Average seed cotton yields between 1,600 and 1,700 kg. were obtained with favorable weather and 1,350 kg. with average weather. Without chemical fertilizers, pro— duction in the sixth year was 60 percent lower than in the first year. All trials carried out at Bambesa and throughout the same bio~ climatic zone showed the importance of phosphorous on these iron rich soils (ferrisols). The recommended doses for cotton were 100 kg. triple superphOSphate per hectare and per year and a light dressing of ammonium sulphate from the third year onwards at a rate of 100 kg. per hectare. Since 1958 the use of fertilizers has been extended to farmers in paysannats in order to prolong the cultivation cycle by one year and to cut the effort needed for land clearance by one-third. In the Ubangi forest area the Boketa station with a network of local trials found that soils did not react well to chemical 290 fertilizers and that maintaining the soil's fertility was complicated. Here organic matter played a much more important role than it did in the Ueles. Nitrogen was identified as a principal influence and phosphorous as a secondary one, with the action of these elements being relatively weak (a 24 percent and a 19 percent increase in cotton yields, respectively), often irregular and always uneconomical. At the Kutubongo station in northern Ubangi, located in a savannah zone and dominated by Imperata Cylindrica, trials brought out the importance of nitrogen and phosphorous and the inefficacy of potash. However, the natural productivity of these soils is very low and even production increases of 50 to 100 percent are not likely to make fertilizer applications economical. Similar results occurred in studies conducted in the Magombo station in northern Uélé in the Guinean savannah: a significant effect on cotton yields from nitrogen and phosphorous, but not enough to make application economically justified. In the southern cotton belt ferralsols fromlquartzites reacted best to chemical fertilizers, but heavy fertile ferralsols did not react at all for some unexplained reason. If the effect.on the whole rotation.was taken into account for ferralsols fromlquartzites, the results proved that chemical fertilizer use was economically sound. Moreover, there was a very strong indication thatcxmujmmxmmscultiva- tion of land was feasible. The CFDT.Mission started small scale fertilizer trials for cotton in Ubangi-Mongala in 1973. The results showed that application of 100 kg. fertilizer 20—20—0—6 (N-P—K—S) per hectare in side dressing 291 me week after planting does not autoratically return a net profit (Table 9.1) .5 The most fertile soils, those with highest yields on the control fields, responded best to fertilizers. Only one dose, 100 kg. per hectare, was applied and the results were corpared with nonfertilized experimental plots. Good cultural practices were maintained on the fields. Table 9.1. Republic of Zaire: Effect of 100 kg. 20-20-0-6 Composite Fertilizer per ha. in Side Dressing, One Week After Plant- ing, on Experimental Cotton Fields in Ubangi—Mongala, 1973-19741 . T . . ' . . Experimental Yield in kg. /ha. ~ertilizer Response Net Profit of Fields on Control Fields in kg. Seed in Zaires Net loss in iCotton per Z /Hectare2 Iectare L 1 1,433 7 517 34 26 2 1,015 3 97 6 — 2 3 438 E 62 4 - 4 4 1,180 346 22 +14 5 635 , 166 11 + 3 .1 l ONAFITEX, Direction Régionale Nord-Ouest, May 1974. "Essais Agronomiques 1973." Gemena: Mission CFDT. 2Fertilizer 20-20-0—6 cost 8K/kg.; seed cotton price was 6.5K/kg. In conclusion, the INEAC and CFDT experiments showed that the use of chemical fertilizers for cotton production could be economically justified, depending on the soil type and whether or not the effect on the whole rotation was taken into account. Yields could be raised 5The fertilizer was flown in from the FED—CAK project in Eastern Kasai; the bags were wet and the fertilizer had hardened. 292 significantly and labor saved for the farmer because of a prolongation of the cultivation cycle and a less frequent need to clear land. However, the optimum.doses for financial profitability were not studied properly [Tshibaka, 1974]. Mbst studies indicated either that they had an economic effect or that an economic effect could not be proved. In fact, the rates of application (0 kg:, 100 kg., 200 kg., etc.) usually did not permit a detailed study of the economics of fertilizer use. Major Constraints on Using Fertilizers Fertilizers have been used in Zaire for many years mainly by cxmmercial plantations growing export crops such as oil palm, coffee, rubber, cocoa, tea and.by plantations growing sugar cane. Small quantities have also been used for vegetable production around the major cities but practically none for the basic food crops and cotton, except in some isolated development projects. Imports of fertilizers reached 13,222 tons in 1959 and declined sharply in the 19605 because of internal disorders and general import quotas due to a lack of foreign exchange. They readhed the pre- independence level again in 1969 and are now increasing. Ckmpared to west African countries, Zaire is a very small consumer of fertilizers with less than 1 kg. per head in Zaire and over 10 kg. per head in.most west African countries [verhaegen, 1970b]. Mbreover, consumption in.most of the‘West African countries is increasing at a rapid pace. In this section we will diagnose the major constraints to using chemical fertilizers in Ziare, constraints which explain the low level of present fertilizer consumption. 293 The High Cost of Fertilizers Fertilizer prices in Zaire are very high. Table 9.2 presents prices in K per kg. for different types of fertilizer shipped to various ports and unloading stations in the northern cotton belt. Several factors explain these high prices. m. Zaire does not produce any chemical fertilizers and does not have a granulation, bulk blending or bagging plant. Fertilizers are obtained mainly from the common market countries in Western Europe,6 a fact which accounts for expensive ocean freight charges. Fertilizers could be obtained at a cheaper rate from African countries. South Africa, Rhodesia and szambique could export fertil- izers to Zaire at a price c.i.f. Zaire well below the cost of fertilizers obtained from Europe [Hedley and Good, 1971] . Yet trade between Zaire and these countries raises a political problem although there is already a brisk trade. Angola, Kenya, Uganda and Tanzania are also producing fertilizers or fertilizer nutrient concentrates and could probably export to Zaire at a fraction of the prices for fertilizers bought from Europe. The major problem seems to be the establishment of the proper commercial channels. Another alternative for Zaire is to produce its own fertilizers, particularly nitrogen from liquified air. The political bureau of the M.P.R. party in Zaire decided in its meetings of Decerber 28—30, 1974 to create a dorestic fertilizer production industry [Zaire, 1975] . Pre—investment studies have been cmpleted for a nitrogen fertilizer 6Recent devaluations of the U.S. dollar to which the Zaire is pegged have made imports from mrcpe even more expensive. 294 Table 9.2. Prices for Different Types of Fertilizers and Different locations in the Northern Cotton Region, 1971 (in K. per Ton, from the Colimpex Cotpany in Kinshasa) Locations Urea Ammonium Potash Triple Super Corposite “820 me: was 222385 lam-:7 4 2 4 2 4 4 4 2 Kinshasa 8,200 4,700 7,300 7,500 8,800 Akula 8,625 5,125 7,725 7,925 9,225 Bumba 8,637 5,137 7,737 7,937 9,137 Aketi 8,702 5,502 7,802 8,002 9,202 Mungbere 9,779 6,279 8,879 9,079 10,379 Kisangani 8,691 5,191 7,791 7,791 9,291 l[Hedley and Good, 1971] . plant to be located at Inga where cheap hydro-electric power can facilitate its establishment. 7 Reported resources of natural gas in the Kivu and of phosphate rock could also play an important role in establishing a national fertilizer industry [McCune, 1972] . In fact, the Republic of Gabon, close to Zaire and to Inga, has large phosphate reserves . Transport Costs. Half or more of the cost of fertilizer in Zaire is attributable to transport costs. AMIZA, the national Zairean 7Inga is located on the Zaire river rapids between Kinshasa and Matadi, about 30 km. upstream from the port of Matadi. The hydro- electric potential of these rapids is enormous, estimated at 30 million kw. for year round operation. The first power station, Inga I, with a capacity of 300,000 kw., started operating in 1972. Inga II is now under construction. By 1980 the total operational capacity of Inga is planned at 3 million kw. or one-tenth of the potential capacity of the Inga site. A high power line now under construction will transport electricity from Inga to the rich mining regions in Shaba. 295 shipping agency, reported that shipping fertilizers in 1971 from Antwerp via.Matadi to Kisangani would cost 35.73Z per ton; to Isiro 44.962; to Iibenge 36.222,prices which included ocean freight, handling costs, duties and taxes and shipping costs in the interior. Thus, transport costs and taxes and duties amounted to between 3 and 4 K per kg. of fertilizer. Most fertilizers are imported in small lots, although buying in large quantities could possibly reduce freight costs and purchase price [MCCune, 1972]. Another possible means of cost savings would be the importation of concentrated fertilizers or fertilizer nutrients and blending them with inert fill in Zaire. Presently, low analysis fertilizers such as ammonium sulfate (21 percent N) are comronly imported, together with some urea (46 percent.N). The cost per unit of nitrogen in urea is much lower than in ammonium sulfate (see Table 9.3). If only high analysis fertilizers were imported, substantial savings could be realized. [A local bulk blending and.bagging plant would then be necessary to arrive at the desired grades. Such a plant does not require a lot of capital or managerial skills and could be set up quidkly. Shipment to interior ports in Zaire by CNATRA, the national river transport agency, is expensive. It takes up about 25 to 35 percent of the total transport costs from.European ports to interior ports or unloading points in northern Zaire. Fertilizer is classified as general merdhandise, the most expensive fare category, although fertilizers do benefit from a 25 percent reduction in total fare. CNerland transportation is also very expensive, on the average 10 K per ton and per km. for general merchandise if no return freight is 296 15:: sun woman 583.? one 85m Jogger umuflfifiom 3.80950 mo umoo Hospoo ecu cg “93E mun mag. 6836328 so 6.858 mas... o8 Ema 68o o8 88%: £3.33qu 51515 .06 once amount 05 908 :63 £8 moo M 123.2 ma 86098 8o moo x mass ma 825 3238 529:6 mo 83o unmanned 500.5? on“. us ~333qu 330950 mo made, ucofluso ounce—5 05 .ouoonwmsm eBHcQEB ma 2 .uooflm mun cocouommoqo sou swoon»? Eon Moo M msmsd .:B moo M m.~$.3 B o: moon 888 L50 sou mom M m.m3.mv Boone—”om ganged mo coed ucofiflsc Sonya on» can $38960 sou Hoe M o.hmm.mv mumfimooojnwgm 333 m0 ooeuo unmanned: ougmooo won . finance—50 sou Hoe M mess; nous mo 83o afiefisc comouuflc o5 um Eugene 3151: Homage 3.30960 mo 33> pooh—us: och. cc 3 www.mm omega gram ommém humans: mo s8. moo M SAT: momo woe omM wow 2 we 2 woe moonstone an osmosoo osoflusz mag: macs memo mums mafia cos woo .M ca 88 H38. 03; 034 024 03; SH; res on... :8? 3 x .5 m: ogoasvfi phoemcmuu pond afiazo 8? 8302 to too mfi B m-.a Rae. mmas Rim mass igneous 8289 88o 5:8 Steamed 8238 8298 330950 Hoodm campus. gemmmuom :BecQEa mono HBO.» Mom .M :3 Home .9550 CH umou ucoeoflsz “madame “mHHoN mo 038.6% .m.m canoe 297 available. With return freight, the rate is half. In order to reduce the effect of transport costs on fertilizer prices, only highly con- centrated fertilizers should be imported, preferably from suppliers within Africa. Granulation, blending with inert fill and bagging could be done in Zaire at low cost. Fertilizers should also be assessed a lower fare rate by CNATRA. Custom Duties and Taxes. The following duties and taxes were imposed on imported fertilizers until January 5, 1973: import duty 5% statistical tax 3% turnover tax 9% conjuncture tax (temporarily) 5% Total .22% Duties and taxes raised fertilizer prices by 22 percent and generated an estimated 90,000 Z in the government revenue [McCune, 1972]. Clearly, government policies were not favoring the importation and utilization of fertilizers. Most developing countries subsidize rather heavily the import of fertilizers in order to stimulate domestic agricultural pro— duction. In its meetings of December 28—30, 1974 the political bureau of the M.P.R. party decided to drop all remaining import taxes on agricultural inputs in order ’00 stimulate agricultural production [Zaire, 1975]. Inadequate Delivery Systems. Fertilizers are distributed by private companies in Zaire, a distribution system is geared to carmercial plantations. The distributors have offices in the major cities but virtually no sales agents in rural areas. This makes it very difficult 298 for smallholders to acquire small quantities for their own use. Moreover, large buyers receive considerable price discounts. Most dealers have at least two prices, one for buying 'a bag or a few bags and another for buying 50 tons or more. The price may differ as much as 20 Z per ton [McCune, 1972]. Thus, small farmers have an economic disadvantage in buying fertilizers vis-a—vis the large plantations, unless they might pool their orders . 1m Product Prices Presently, practically no fertilizer is used on food crops nor on cotton except in development projects such as the FED—OAK project in Eastern Kasai and the CFDI' project in Ubangi-Mongala. Yet the results of fertilizer applications in these projects are promising if cost-benefit ratios can be improved. Fertilizer prices are high, on the average about 7 K per kg. of fertilizer, and product prices are relatively low. 8 We have already indicated that seed cotton prices are low. Most food crop prices were raised in 1973 or 1974 and this should have a favorable effect on the cost-benefit ratio of using fertilizers on cotton and on food crops in the rotation (direct effect and residual effects). If one uses the marginal principle, the last kg. of fertilizer applied on a cotton field must produce at least one to one and a half kg. of seed cotton, depending on the type of fertilizer used--one kg. for ammonium sulphate to one and a half kg. for urea, potassium sulphate 8Fertilizer prices at Colimpex, Kinshasa ranged from 6.3 10.3 K per kg. and nutrient costs ranged from 18.5 K to 22.7 K in 1971. The seed cotton price for first grade was 6.5 K per k 1972, 1973 and 1974. K to per kg. g. in 299 triple super phOSphate or corposite fertilizer. On a fertilizer nutrient basis, the last kg. of nutrient applied must increase cotton production at least by 3 kg. of seed cotton, depending on the type of fertilizer used. Thus, the average productivity of fertilizers must even be higher, in the order of 4 kg. or more per kg. of nutrient. This requires fertilizer responsive cotton varieties and adequate plant protection, a point which brings us to the next constraint. Lack of Fertilizer Response Data Presently, very little research is done in Zaire on fertilizer response except in the South for maize in the National Maize Program and in the Kasese-Kaniama project. Although fertilizers are used for cotton production in the FED—CAK project in Eastern Kasai, it is difficult, if not impossible to derive accurate benefit-cost ratios for fertilizer application in this project. Soils in Eastern Kasai are also quite different from soils in either Ibangi-angala or in the Uélés. INERA ccncentrates on genetic cotton improvements and does not carry out fertilizer trials on cotton.9 CFDT in Ubangi—Mongala is engaged in some interesting experiments but, since only one rate of application is applied, no production functions and, thus, no Optimal doses can be derived. Research on fertilizer use for cotton production by INEAC before 1960 showed that soils in the Uélés gave the most favor— able response; however no experiments are now organized in this region 9The 1973 INERA program provided for a series of fertilizer trials on cotton in Gandajika but they were not implemented. The 1974 program repeated the trials planned for 1973. 300 although the potential for increasing cotton production in the North is the greatest for the Uélés. Sociological Factors Cotton farmers and cotton extension workers have no experience with chemical fertilizers. Thus, the introduction of chemical fertil- izers into traditional farming will encounter many obstacles. Fertil- izers will have to be quite profitable and have low risk for farmers 10 Asfarmers in order to overcome traditional resistance to change. will not be able to pay for this new input, credit institutions will have to be developed. Pesticides will also be needed as research has clearly demonstrated the need for pest control in tropical agriculture when fertilizer is used.11 Thus , the problem of introducing chemical fertilizers is one of introducing a package of innovations along with a rather drastic change in cultural methods . Communications research will be necessary before launching sud': a new technological package. In the South of Zaire demonstration plots with high yielding hybrid maize and fertilizers proved to be very effective in influencing the adoption of these new inputs. But this adoption was facilitated by substantial government or external financing, supports which will be necessary to introduce this nev technological package throughout Zaire and to overcote the initial reluctance of farmers to change their loAccording to Verhaegen [1970b] the ratio of increased revenues to increased costs as a result of fertilizer applications must reach at least two to assure adcption. l:I’Cotton acaridosis reaches much more dangerous proportions where cultivaticn is intensive; nitrogenore fertilizers make cotton plants more attractive to Lygus Vosseleri [Jurion and Henry, 1969] . 301 time-worn agricultural methods . New technologies are unlikely to become adopted unless they fit into a production system adapted to the farmer's goals , opportunities and resources . The Agricultural Extension Service Unless the agricultural extension service is reformed and retrained, there is little hope for proroting fertilizer use in traditional agri- culture. If the government decides to introduce chemical fertilizer on a large scale, then major efforts need to be made in the development, training and support of extension staff. Conclusion The use of chemical fertilizers for cotton production shows promise but cannot be recommended as a major strategy for increasing cotton production. This study has shown that some major constraints limit the use of fertilizers, that the economics of fertilizer use on cotton production is questionable and that more research is needed on the economic effects of different types of fertilizers and different rates of application. If seed cotton prices are increased and if fertilizer prices are decreased, then it might be profitable for cotton farmers to use fertilizer. The fertilizer price could be lowered by importing highly con- centrated fertilizers, preferably from suppliers within Africa, by removing all remaining import duties and taxes and by reducing the fertilizer transport rate schedules of CNATRA and other public and semi-public transport agencies. Granulation, blending with inert fill and bagging should be done in Zaire. 302 Zaire has recently decided to create its own fertilizer industry; Thus, efforts should be undertaken to introduce chemical fertilizers in pilot cotton development projects and then later among smallholders if it proves profitable. The government should seriously consider the subsidization of fertilizers in order to promote their introduction among smallholders. At the same time, an input delivery system should be set up which focuses on smallholders. Pesticides Basically, there are three kinds of chemicals which can be used to control cotton pests: insecticides to combat insects and spiders and the parasites (bacteria, fungi, viruses) they transmit, fungicides to protect plants against fungi and herbicides to combat weeds. Birds and rodents inflict virtually no damage on cotton but are serious predators of food crops. In this section we will concentrate on the application of insecticides and fungicides on cotton. Herbicides may hold a bright future for weed control in cotton mhen.labor becomes scarce and expensive. weeding is a critical activity in cotton production and herbicides could fulfill a useful role in this respect. Presently, they are unknown in traditional agriculture in Zaire and their cost is probably prohibitive. Review of Experiences with Pesticides Cotton is undOUbtedly the plant.whose parasites have been the most closely studied. In Zaire the use of insecticides on cotton started in 1952. In the northern cotton belt insect control programs were undertaken for the first time in 1954 and were confined to places 303 infested with Helopeltis. Nearly 5,000 hectares of cotton were treated for Helopeltis in 1955 and 1956; a second pass was needed on 33 hectares in 1955, but none in 1956. In those two years the rate of infestation dropped to three insects per 1,000 plants. In Ubangi 6,000 and 10,000 hectares, respectively, were treated in the course of the 1957 and 1958 seasons. In 1959 445 hectares were treated for Helopeltis, the rate of infestation falling from 20 to 1.5 insects per 1,000 plants. Only 200 hectares had to be treated a second time [Jurion and Henry, 1969] . Insecticides are extensively used in the FED—CAK project in Eastern Kasai and in the FIWA—BILI project in northern Ubangi. Since 1972 ONAFITEX has treated cotton fields with insecticides. Only cotton fields grouped in large blocs and close to a road are treated. Farmers have shown virtually no resistance to insecticide applications. The product used was special DDT cotton dust LIMI'EX and the dusting equip- ment was (word—41 Platz, portable sprayers or dusters powered by a two- stroke engine. The recommended rate of application was 12 kg. per hectare and per treatment, but many fields received from 15 to 30 kg. per pass. The equipment, however, was not well suited for dusting. Dusting equipment powered by hand which the farmers can buy and use seems preferable to the rather complicated and expensive Platz dusters. However, over 1,100 imported Platz dusters were purchased by ONAFITEX. The CFDT mission organized insecticide trials in Ubangi-angala during the 1973-1974 campaign. A randomized block design was used. Treatments started six weeks after planting and were repeated every two weeks for a total of six treatments. Three insecticides were corpared: 304 --special cotton dust LDWFEX, the product actually used.by ONAFITEX, containing 5 percent DDT, 10 percent HCH and 40 percent sulphur. --Nuvacron Ulvair Combi A400 (CIBArGEIGY) containing 250 g./l. of DDT and 150 g./l of a systemic organic phosphate (monocrotophos). --Emulsion Azodrine/DDT (Shell) containing 100 g./l. of mono- crotophos and 200 g./l. of DDT. very few differences were observed among these three products. Cotton dust seemed more effective against Spiders (Hemitarsonemus) and fertilized the plants with sulphur. However, dusting with Platz equipment was quite impractical and would be difficult to introduce to extension agents. The best fonmula appeared to be UEV (Ultra Low velume treatment) without water, with a sprayer powered by flashlight batteries and a two to four liter solution per hectare. No water or fuel is necessary and the equipment is very lightweight. One unit is needed per 20 hectares. Seed cotton yields from 600 to 1,100 kg. on the treated fields were obtained. Since the fields treated with dust were control fields a comparison could not be made with untreated fields. Thus, data on the financial profitability of insecticide treatments is not available. However, we believe that insecticide treatment is financially and economically sound in most areas when the proper timing and rates of application are observed. However, spraying or dusting should not become a routine. When the level of infestation exceeds a minimum level treatment should be applied. Thus, control and warning stations are necessary to monitor the level of infestation. Regular insect counts following a sampling plan in selected areas, should be 305 organized, and.minimum infestation levels based on economic criteria should be established. Insecticide treatments should not substitute for indirect methods of pest control such as breeding and selection for genetic resistance to cotton diseases and insect attacks, disinfection or coating of seeds before planting with an organoemercuric fungicide, sorting of seeds to increase genmination rates, uprooting and burning of cotton debris after harvest, observation of recomrended rotations, fallow times and time of planting and harvest, and destruction of fields invaded by wilt or other fungus diseases.12 Major Constraints on Using Pesticides Most of the constraints to the use of chemical fertilizers also apply for pesticides. Presently, ONAFITEX treats all fields which can be easily reached. Dusting or spraying is becoming a routine. In order to make treatments profitable, applications need to be a function of the pest population buildup and the expected damage to yields and to cotton quality rather than a schedule of applications determined by the calendar. Thus, regular insect counts, following a sampling plan in selected areas, should.be organized and minimum infestation levels based on economic criteria should be established. The farm.business survey data shows that only 10.5 percent of the farmers in Ubangi- Mongala and none of the farmers in the Uélés had their cotton fields treated with insecticides in 1972-1973 by ONAFITEX. The Agricultural Service of ONAFITEX is unable to treat all cotton fields because they 12Time of planting and harvesting is particularly important for reducing damage fromlHelopeltis. 306 are so dispersed and often located deep in the forest. Thus, the cost of pesticide treatment of cotton fields is not equally supported by all cotton farmers. This could be avoided by shifting responsibility for plant protection to fanners by providing them with locally made hand powered dusters or Sprayers and pesticides at a subsidized price to stimulate adoption. Conclusion The present trend in agricultural policy in Zaire is for more direct government intervention at the farm level whereby new technologies or agricultural methods are imposed upon the farmer rather than using the agricultural extension service to diffuse, demonstrate, persuade and teach farmers to adopt innovations. Direct government intervention at the farm level, the model adopted under colonial rule, was very successful in the short run but failed to pay off in the long run. The present approach to cotton plant protection is quite paternalistic and is modeled on the colonial experience. For long run success it is necessary that farmers themselves ad0pt this new technology. This will require an expanded and redirected extension effort. We believe that the cotton extension service should encourage farmers to purchase locally made hand powered spraying or dusting implements and that farmers should be taught how to operate these implerents and when to apply pesticides . To favor introduction of pesticides and hand powered elements , the government should temporarily subsidize these inputs until they are in widespread use. 307 Medhanization From 1967 to 1973 the Zairean government spent approximately 17 million dollars on the importation of heavy agricultural machinery, mainly tractors and accompanying implements. Although the subsidiza- tion of imports of tractors and related equiprent has now been dis- continued, mechanization is still advanced as a strategy for increasing agricultural production in Zaire and.mechanized plowing is being sub- sidized by the government. In this section, we will review the experiences with mechanization in Zaire and we will discuss the major problems of a mechanization strategy for increasing cotton production in northern Zaire. Review of Experiences with Mechanization From 1950 to 1960 tractor mechanization was tried out on a large scale in Zaire, mainly in savannah soils. In the Imperata savannahs of northern Ubangi mechanized preparation of the soil increased.yields by 15 to 24 percent over manual preparation. In the Guinean savannahs of northern Uélé mechanical preparation produced equal or slightly higher yields than manual preparation. Mechanical land clearing and plowing in the Gandajika region (KaSai) resulted in an 18 percent increase in output for the first cotton crop after clearance with a residual effect of a 24 percent increase in maize yields grown after cotton. In.the Kongolo area (northern Shaba) mechanized land clearing helped to eliminate Imperata cylindrica and reduced the need for weeding by about 60 man-days per hectare. This made it possible to Observe the agri- cultural time table. Seed cotton production in kg. per farmer increased from 447 kg. to 885 kg. because of a combination of an increase in 308 yields per hectare and a larger acreage under cultivation. However, mechanization is reported to have hastened the deterioration of surface- soil structure. In Kasongo (Mamiéma) farmers' incomes as a result of mechanization increased in 1957 by 23 percent, in 1958 by 69 percent and in 1959 by 187 percent as corpared with hand tilling in 1956. A combination of higher yields per hectare and larger acreages planted was responsible for the increase in farmers' incomes. Farmers paid 937.50 francs (18.755) per hectare for plowing in 1958 and 350 francs (7.00$) for harrowing. The crop rotation was cotton, maize and groundnuts. With manual operations, the harvesting of one crop is often not completed before it is time to plow for the next crop. This is the bottleneck that limits the possibility of extending the acreage culti- vated by hand according to Jurion and Henry [1969] . In general, the principle of mechanization was readily accepted by the peasants, and in most cases, a few years of propaganda were enough to persuade everyone to adopt mechanized cultivation which reduced the arduousness of agri- cultural 1abor. However, in some cases, mechanization had no other result than to increase the farmer's leisure as acreages were not expanded. In 1958 it was estimated that 11,000 to 12,000 hectare of cotton were plowed with tractors. Farmers paid on the average 900 francs (18.005) per hectare for plowing and 500 francs (10.003) per hectare for Rare plowing. Yields in kg. per hectare increased on the average by 15 to 20 percent after mechanical plowing. Mechanized clearing of a savannah, no matter where it was located 309 in Zaire and however light the tree cover, was not economical. Only in exceptional circumstances was the use of heavy machinery to clear heavy forest cover economical. Mechanization of post-harvest operations (thrashing, drying, etc.) was found very important because it was estimated that about 42 percent of family labor was absorbed in post- harvest activities [Geortay, 1961] . Animal powered mechanization was successful in several regions of Zaire before independence. Draught oxen were used for plowing in western parts of Shaba province (Dilolo, Sandoa, Kapanga) and in the South of Kasai province (Luisa). local herds exist in these regions, built up outside the tsé-tsé infested areas, and soils are sandy. An oxen training center in Kabelenge (Sandoa) trained farmers in handling draught oxen. Peasants were able to buy a pair of cattle with payment facilities. There were 200 oxen teams in service in the Sandca- Dilolo territories in 1959 and about 100 teams in Luisa. In 1960 the price for custom plowing with oxen was 250 francs (5.00$) per hectare and 0.15 to 0.20 hectare could be plowed per 4—5 hour day of work. The farmers who used oxen for plowing doubled their average gross income as they could cultivate twice or even three times as much land as they could with manual tilling. Oxen were also used for transport of produce. However, on heavy soils, difficulties were encountered with animal traction. The quality of work was also below that of tractor plowing, but oxen could be used on fields with tenmite heaps, stumps, etc. Since 1969 cotton fields in FIWA have been plowed with heavy tractors (type FIAT 70 C with NARDI disc plows). In 1971 the BILI region was added to this project. Farmers had to pay 5 Z per hectare 310 for tractor plowing in 1969, half at the time of plowing, half after harvest. In 1970 the rate was increased to 62 per hectare and in 1972 to 82 per hectare. In 1973 it was estimated that tractors needed to be replaced after 4,000 hours of operation and disc plows after 2,000 hours. Operating expenses, excluding the costs of expatriate garage personnel paid by the Belgian Aid Mission, and service parts for tractors and plows received from the government, were, in Zaires per hectare:13 Gasoline 0.33 Diesel Fuel 3.07 Oil 0.54 Tractor Drivers _l_.__5_g Total 5.44214 Average cotton yields in FIWA and BILI were between 800 and 1,000 kg. per hectare with tractor plowing,strict observation of recommended cultural practices and several treatments with DDT cotton dust. Average yields in Ubangi-Mongala in 1972-1973 were only 257 kg. (Table 6.22) . The average cotton acreage per farm in FIWA-BILI was about one hectare. The farm business survey in Ubangi-Mongala showed the average cotton acreage per farm to be only 40.0 ares (Table 6.22) . Although detailed information on farmers' incomes in the FINA-BILI project could not be obtained,it is generally believed that incomes were considerably above those of farmers outside the project area. The 13In most development projects, the cost of expatriate personnel is not included in profitability calculations. 14The doubling of all fuel prices in 1974 substantially increases the variable cost of tractor plowing. 311 project staff estimated that a family could produce each.year 800 kg. of seed cotton on one hectare, resulting in gross revenues of 800 kg. x 0.062 = 48.002 and 1,400 kg. of unshelled groundnuts on another hectare, amounting to 1,400 kg. x 0.042 = 56.002 and 4,000 kg. of dry manioc Chips on a third hectare earning 4,000 x 0.0252 = 100.002. In this example the cost of tractor plowing is 2 hectare x 8.002/hectare = 16.002 and gross revenues are 48.00 + 56.00 + 100.00 = 204.002. The crop rotation is cotton as the first crop in the rotation, planting in June, harvest in November-December, then groundnuts from.March to June and, finally, manioc from.Ju1y to the end of the next year. Thus, each farmer would have each year three one hectare fields of which two hectares would be plowed mechanically. It is doubtful whether family labor would be sufficient to harvest one hectare of groundnuts in May- June, to plant one hectare of cotton seeds in June-July and to plant manioc cuttings in July-August. More research is needed to find out if this scheme is feasible. Several farmers in the FIWArBILI project did not harvest all cotton on their fields which indicated that harvesting can become a major labor bottleneCk when fields are tractor plowed. In July 1972 a Ministry of Agriculture conference on tractor mechanization was held in Kinshasa and experience with tractor medhan- ization in FIWArBILI, MBANZArNGUNGU, RUZIZI valley and GANDAJIKA.were reviewed. All these projects were using the same type of equipment which had been supplied by the Zairean government. Total costs for tractor plowing were between 15 and 19.52 per hectare. Farmers had to pay 19.52 per hectare in MBANZA-NGUNGU, 8.002 in FIWArBILI, 6.002 in 312 the RUZIZI valley and 8.002 in Gandajika. All these projects employ expatriate personnel which is not accounted for in the cost of tractor mechanization. Major Problems of Mechanization Many agricultural developrent projects in Zaire rely heavily on tractor mechanization, yet the financial. and economic impacts of mechan- ization have not been carefully analyzed. Thus, a careful study should be made of tte potential of selective mechanization policies for the development goals of Zaire, such as attaining self-sufficiency in basic foodstuffs, increasing rural employment and income generation and promoting rural development in general. Animal powered mechanization, abandoned since 1960, should be considered and studied for its potential contribution to rural developrent, particularly in the savannahs of the North and South which are sandy and are relatively free of tséF-tsé flies. It has been shown that tractor mechanization can improve the peasant' s incore, provided he cultivates a larger area and adopts the recommended cultivation techniques to ensure high yields from his crops. These conditions, however, are not always met. Many tractor and tractor implements are misused or badly maintaired. This stems from inexperience, poor training of mechanics and drivers, lack of service parts, ignorance of the actual cost of these machines and, generally, machinery that is poorly adapted to tropical working conditions. Most tractors accumulate far less total working hours in a tropical environment than in terperate regions. 313 The main advantage of tractor mechanization is that it enables farmers to increase the acreage under cultivation. The time needed for land preparation is drastically reduced, weeds are less of a pr0blem and, thus, the agricultural calendar can be more easily observed. Nevertheless, there is not always a yield per hectare increase associated with tractor plowing , and tractor mechanization may damage physical soil structure, particularly when the soil is wet. And yields per hectare do have a decisive effect on the profitability of mechanized cultivation. Many fields cannot be mechanically tilled, particularly in forest areas. Termite heaps and tree stumps make tractor cultivation.impos— sible. Many fields are located deep in the forest and can only be reached by a narrow, winding path, crossing rivers and creeks . This explains why mechanization has been mainly restricted to savannahs. Farmers also have to group their fields in large blocks close to a road for tractor mechanization. Mechanical picking of cotton in Zaire at present wage rates is uneconomical. In the U.SLA. a skilled hand picker can gather around 300 pounds (136.1 kg.) of seed cotton per day under ideal conditions. A modern mechanical picker under similar conditions can gather cotton at the rate of 300 pounds per quarter hour. The break-even point for mechanical picking even on U.S. small farms as long ago as 1954 occurred at a wage rate of $9.00 per day [Sinclair, 1968] . In Zaire with wage rates at $1.00 per day or less it is obvious that mechanical picking is not feasible for many years to core. There is still room for substantial improvement in the basic 314 tools and equipment used by farmers. Improvetents could be made in the design as well as in the quality. In particular, the supply of these implerents in rural areas at reasonable prices needs urgent government attention. Conclusion Presently, tractor mechanization of cotton production in the North cannot be recomended. All mechanization projects in Zaire rely heavily on expatriate manath and the economic effects of tractor mechanization have not been studied properly to justify tractor mechani zation. However, the government should take steps to improve the supply of improved hand tools and equipment in rural areas of tie North. Agricultural Credit Lack of agricultural credit is often seen as a constraint on agricultural production in developing countries . In this section, we will review the experiences with agricultural credit in Zaire and dis— cuss the administration of agricultural credit. In that context, cooperatives will be reviewed briefly for their potential in channelling credit to smallholders. We will conclude by evaluating a credit strategy for expanding cotton production in northern Zaire. Experiences with Agricultural Credit Agricultural credit was given to white settlers (called Colons) in southern Shaba as early as 1911. The primary purpose of these loans was to attract settlers, to protect the southern borders and to supply 315 the rapidly developing urban centers with food. These loans eventually became grants. Several credit institutions were set up during the colonial period [Mabi, 1974; Mbuyi, 1970]. Their main purpose was to attract white settlers. The loans given were practically grants. Gradually, credit institutions evolved into true credit granting agencies, but the credit terms were too strict in relation to guarantees or collateral to appeal to indigenous farmers and, thus, they served primarily the white cxmmmnity. In 1941 the "Fonds Spécial du Credit Agricole Indigéne" was set up specifically for indigenous farmers and their local collectivities. Other governmental financial institutions were created later for the same purpose but they had only a limited success. They granted.credit only to those who could offer material guarantees and.many recipients confused a loan with a grant. Private banks catered only to capital intensive, commercial plantations. The amount of agricultural credit given to smallholders from 1960 to 1970 was insigificant. USAID helped to constitute a "Fonds Congo" which worked through private banks, primarily for the rehabilitation of abandoned plantations. On March 2, 1967 the "Fonds National de Credit Agricole et Artisanal" (FNCA) was created, absorbing "Fonds Congo" and other credit institutions established during the colonial period. These institutions had a 1imited.impact on agricultural production; lack of loan supervision, political interference and poor management greatly reduced their efficiency. CADEZA (Caisse Generale d'Epargne du Zaire), the national savings and loan institution, had on December 31, 1969 316 only 0.26 percent of its outstanding loans in agriculture [vasos, 1973]. In 1970 two new financial institutions were created, SOFIDE and Supervised Agricultural Credit (Crédit Agricole Contrélé). SOFIDE (Société Financiere de Développement) was set up with aid fitmtthe Wbrld.Bank. From the date of its creation to June 30, 1972, only 2.2 percent of its credit was channeled into agriculture although this institution accounted for 70 percent of all credit granted over this period.by specialized financial institutions [Mabi, 1974]. The Supervised Agricultural Credit project was financed largely by counterpart funds from USAID. The long term objective of this project was to prepare the ground for the establishment of a national agricul- tural credit bank. Originally, 500,0002 was budgeted for this project. It was decided to grant loans to individual farmers, to cooperatives, paysannats, small agricultural businesses and small agricultural market— ing firms. In fact, a consultant for the project recommended lending fer marketing projects as they were the most creditworthy, and he strongly urged lending £9 cooperatives rather than lending through cooperatives for production credit [Hirsch, 1970] . Moreover, he recom- mended that lending to individual borrowers only be allowed in exceptional circumstances. Three types of loans were made available: short-term loans, up to 24 months, interest rate 11 percent; mediumrterm.1oans, from 25 to 60 months, rate 8 percent; and long-tenm loans, from 61 to 120 months, rate 6 percent. The funds were disbursed in installments. The short-term loans were to be repaid.with the proceeds from the sale of agricultural products and the medium? and long-term loans were repaid in annual annuities. The recipients needed to have 20 percent of the required capital on hand in order to become eligible for a loan. 317 A central loan committee decided on all loan applications. However, cooperation from private banks in administering the loans was not forthcoming. Originally, four areas were selected for inclusion in the project: tl‘e Kinshasa area, the Mbanza-Ngmgu area (ex-’I‘hysville) , the Bumba area and tie Gandajika area. The Bunba area, the major rice production center in Zaire, was dropped before the project started pending a change in the pricing structure of rice. The Gandaj ika area was also dropped terporarily. The project management was composed of an expatriate project leader, sponsored by USAID, eleven expatriate supervisors from I.V.S. (International Voluntary Services , Inc.) and an equivalent number of Zairean counterparts. Technical assistance to farmers was an integral part of the project and was the major responsiblity of the I.V.S. people and their counterparts. The project was fairly successful as compared with previous agricultural credit schemes. Of the 71 loan applications received by December 31, 1971, 41 were approved and 83, 6262 was disbursed, on the average 2,2602 per recipient [IVS, 1972]. About 55 percent of tie loans were for operating expenses and 45 percent were for investments. All production loans were made to middle class farmers, rather than the small farmers for which the program was originally designed. In December 1971 USAID discontinued technical assistance to the project and reduced by one-third the amount of loan funds for the project.15 Hence, the project had not been operational long enough to 15This coincided with the arrival of a new food and agriculture officer in USAID/Kinshasa. 318 test the feasibility of agricultural credit. It was too early to evaluate the rate of reimburserent, but tie prospects for repayment were very good according to tie IVS staff. loan recipients did increase their agricultural production and their incomes. The project was turned over corpletely to the Zaire government on December 31, 1971. Part of the funds were transferred to the National Maize Program (P.N.M.) , and all external support was cut off. Mabi [1974] indicates that the operations were suspended because of poor repayments. Thus, what started out as a unique experience in agricultural credit in Zaire was aborted for reasons which are unknown to the author. At present, there is no specialized agricultural credit institution now functioning in Zaire. Private banks continLe to play a major role in channeling credit to comercial plantations and agricultural marketing firms. In December 1970 33 percent of the loans administered by private banks were for agricultural firms; on June 30, 1972 the percentage had dropped to 24 percent [Mabi, l974] . The national bank of Zaire regulates the credit activities of private banks. Since July 3, 1972 credit ceilings for agriculture have been abolished and loans to agricultural firms no longer need to be covered by deposits or other collateral. And since February 1, 1974 comrercial agricultural credit benefits from the lowest interest rates charged by private banks. Moreover, discount rates of commerical paper at the national bank are now fixed at 3 percent for agricultural credit, at 5.5 percent for credit for small and medium sized enterprises and at 10 percent for other firms. Each bank has a minimum quota for credit which it has to lend to agriculture and to small and medium sized 319 enterprises. Failure to reach this quota results in requirements for covering loans with reserves for at least 40 percent of the value, depending on the deficit which they run vis-a-vis the quota. Thus, the government via the national bank is pressuring private banks to lend money for agricultural purposes and for small and medium sized enterprises. An agricultural development bank, SOFIDAG (Société Financiere de Développement Agricole), was created in 1974 as a subsidiary of SOFIDE. Its operations have not yet started and it is doubtful whether this bank will cater to smallholders. Small amounts of credit are provided in kind and repayment in kind is made at harvest in several agricultural development projects such as the FED-CAK project in Eastern Kasai, the FIWArBILI project of the Belgian Aid Mission, the C.D.I. project in Bwamanda, the G.E.R. Mechan- ization project in lower Zaire, the Chinese agricultural development projects in Mawunzi, Bumba and Ruzizi valley, a.o. In financing cotton marketing operations for farmers and in providing extension services to cotton farmers, particularly the treat- ment of cotton fields with pesticides-free of Charge, ONAFITEX is indirectly providing credit to cotton farmers. This credit is covered by the difference between the final sales value of cotton lint and by-products and the price paid to the farmers after deductions of all marketing and processing costs. The long-run aim of ONAFITEX is to tmflr>develop cotton farmer cooperatives capable of assuming responsi- bility for cotton marketing operations. The role of ONAFITEX will be limited to technical assistance, auditing of accounts, agricultural 320 extension and pr0bably guaranteeing loans to cotton cooperatives and cotton fanmers. At present, no efforts are made by ONAFITEX to establish cotton c00peratives. If a cotton production development program is enacted, it is likely that ONAFITEX will assume the role of credit retailer and*will provide the cotton extension services. .ONAFTTEX is already trying to consolidate all cotton extension efforts within its organization. If 000peratives are formed at a later stage, they will probably deal directly with the specialized credit institutions and ONAFITEX‘will then provide the needed guarantees [Mabi, 1974]. The political bureau of Zaire recently decided to create an agricultural bank to encourage the development of a "Fonds Agricole" which has just been created by the Executive Council to facilitate investments in agriculture [Zaire, 1975]. Thus, agricultural credit is now receiving major policy attention. The Administration of Agricultural Credit and Cooperatives It is important that the objectives of any agricultural credit project are made clear. This was a major shortcoming of the Supervised Agricultural Credit project. This project, although operating with its own extension service, was not related to a broader development program, A new technological package was introduced to farmers. The project was a general purpose project, credit for its own sake, aimed at increasing agricultural production across the board. This was probably the main reason for termination of the project. Recipients of credit need to realize the distinction.between a loan and a grant. Indeed, there is a long colonial history of confusion 321 between loans and grants in Zaire. It is difficult to convince farmers of the exact nature of a loan, particularly when credit is given in kind and when a government institution is involved. Therefore, credit institutions need to be separated from government or pararstatal institutions. The requirement of adequate guarantees (collateral) helps to make clear the exact nature of a loan but, at the same time excludes most smallholders. One long-run aim.of extension education should be that of helping farmers use credit properly. Financial profitability of the operation for which credit is given is the key to repayment. All experts agree on this point. Therefore, adequate market outlets for the increased production must exist. Thus, it is sometimes necessary to tie credit for the develop— ment of marketing facilities to production credit. In order to assure financial profitability and repayment, it is usually necessary to include an agricultural extension service in an agricultural credit project. It is not clear whether or not the extension service should be made responsible for the supervision and administration of the loans. Some argue for a distinction between credit institutions and the extension service; others give examples where the two have been merged with success [verhaegen, 1970a]. Isolated agricultural credit.projects such as Supervised Agricul- tural Credit probably failed.because of the limited role which credit played in increasing the productivity of farmers: no new technologies of production were diffused and the performance of the marketing system was basically unaltered. Thus, credit must be part of a package of improved inputs whidh raises productivity of farmers and substantially increases their production and incomes. 322 Cooperatives are often a recommended institutional mechanism for administering agricultural credit. The historical record of farmer c00peratives in Zaire is one of failure. Cooperatives were well developed before 1960 but'were strictly supervised and controlled by the government. .After 1960 most cooperatives withered. .Most.are still officially recognized but show little or no activity.16 Sore marketing cooperatives are playing a useful role in assuring a market outlet for farmers and in supplying them with small farm tools, DDI', etc.l7 Only 21.4 percent of the farmers in the farm.business survey expressed an interest in joining a cotton cooperative and only 23.8 percent of the farmers would agree to join a collective field if asked to do so. Several conditions are necessary for the success of cooperatives. Father A. Cauwe [1969], founder of the Kisantu c00perative, has dis- cussed these conditions at length. Cooperatives must build upon native traditions [Hirsch, 1971]. This is probably the main reason why cooperatives created during the colonial period failed after indepen- dence. According to Hirsch [1971], "Africa does not need to take over European and American models of cooperatives without examination but rather a significant contribution could be made to agricultural and industrial development through modernizing the indigenous cooperative societies." This is also the main thrust of the government's "second revolution" based on the philosophy of authenticity [Zaire, 1975]. 16Hirsch [1971] listed 91 agricultural cooperatives in Zaire. 17For instance, the cooperative of the Turumbus in Yangambi. 323 In its meetings of December 28-30, 1974 the political bureau of Zaire decided to create a new type of cooperative for the production and marketing of agricultural products [Zaire, 1975]. It was also decided to create "brigades d'encadrement agricole," or brigades for agricultural extension. Thus, cooperatives will be created from the top down with the danger of creating bureaucratic institutions which are overstaffed and carry heavy overhead costs. The challenge for the Zairean government is to conceive and to promote cooperatives or other forms of organization which are socially acceptable and.which build upon the traditions of people and provide farmers with incentives and the means for increased production, incomes and employment. However, not everything should be supplied by the government as this would again lead to paternalism. Instead emphasis should be on appropriate education for farmers. The chances of cooperative success also depend very mMCh on the quality of managerial and educational staff. Individual farmers will always judge cooperatives in terms of how they will help to improve their material welfare [Hyden, 1970]. The next decade will undoubtedly see an intensive search for and selection of new forms of institutional organization for traditional farmers. Conclusion Most agricultural credit schemes in Africa have failed. In the farm.business survey, we found no indications that lack of agricultural credit at the farm level was a major constraint on cotton.production. within the present administrative capacity of Zaire and looking at the historical record of farmer cooperatives, it is doubtful that agricultural credit could be profitable for increasing cotton production, particularly 324 since there is not an improved package of inputs or the required extension service for introducing package programs. There is some scope for agricultural credit to enable farmers to buy insecticides and hand powered dusters or Sprayers but subsidization of these inputs to promote introduction is a more effective strategy than setting up a special credit program, CHAPTER X SUD/MARY, POLICY PRESCRIPTIONS FOR INCREASING PRODUCTION AND FOR IMPROVING THE MARKETING PERFORMANCE OF THE COI'ION SUB-SECTOR OF NORTHERN ZAIRE AND FUTURE RESEARCH NEEIE seer Before independence in 1959 about 800,000 families cultivated cotton in Zaire, producing 63,000 tons of cotton fiber of which 53,000 tons were exported and 10,000 tons were absorbed by domestic mills. By 1965, however, seed cotton production amounted only to 10 percent of the 1959 level and from 1965 to 1968, Zaire had to import cotton lint to supply the expanding domestic textile industry. In 1969, with the return of security and stability, seed cotton production started increasing and is now around 22,000 tons of fiber and about 6,000 tons are exported. New mills are coming into production and will soon absorb all of the domestic production. An increase in cotton production is a high priority for the government, because of growing internal demand and good export prospects. Cotton and coffee are important cash crops produced by smallholders. However, the coffee export quota has been reached and the government now wants to promote crop diversification in the main coffee growing areas which also overlap with the main cotton producing regions. In order to provide the cotton marketing office and tle Department of Agriculture with relevant data and with prescriptions for increasing 325 326 cotton production and for improving marketing efficiency, an economic analysis was undertaken of the rmuthern cotton belt. In Chapter II, the agricultural economy of Zaire was reviewed. Agriculture in Zaire is dualistic and traditional agriculture has been neglected. Shortage of qualified staff and lack of an adequate trans- port infrastructure for both the supply of inputs and the evacuation of produce are major constraints on a rapid increase in production. .Market- ing margins are high and production incentives for farmers are low. The government has recently created parastatal marketing offices for most agricultural products and has taken over all plantations, ranches, farms and agro-industrial complexes previously held by foreigners. The Zaire government has now declared agriculture the "priority of the priorities" and is focusing major attention on agricultural development although this is not yet reflected in the government budget. For example, in 1972 less than 3 percent of total government expenditures were allocated to agriculture. A description of cotton production in northern Zaire is covered in Chapter III. Cotton was introduced in Zaire in 1918 as a required crop and it remained an imposed crop ever since. During the colonial period an intensive agricultural extension network promoted cotton production, particularly in paysannats where cotton was the principal cash crop. Since independence, extension agents have not been retrained; they lack travel facilities, and are insufficiently motivated. Presently, the extension service is largely ineffective. The agricultural calendar is poorly observed, seeding is retarded, fallow periods are too short and many cotton fields are poorly maintained. Three different agencies 327 are now responsible for cotton extension work. The agricultural service of (NAFITEX now has its own extension agents who introduced large scale pesticide treatments on farmers' fields in 1972. Since colonial times there has always been a uniform seed cotton price paid to farmers in Zaire. In 1970 the price of seed cotton paid to producers was the lowest of all African countries for which price data could be obtained. Presently, three cotton production projects are underway in Zaire: (l) the FED—CAK project in eastern Kasai, an integrated "paysannat" development scheme using improved seeds, fertili- zers, pesticides, credit, tractor mechanization and improved hand tools. The project is exceeding its production objectives for cotton and food crops. It is not clear whether the project will becore self sustaining by 1978, the target date. ( 2) The CFDI‘ project in Ubangi-Mongala which has as its broad objective the developrent of cotton production mainly through improved agricultural extension services to small farmers. Presently, this project is engaged in varying research activities which are basically the domain of INERA. (3) the FIWA-BILI project of the Belgian Aid Mission in northern Ubangi which is concentrating on tractor mechanization and pesticide applications in cotton fields. The main problem with this project is its lack of focus. Ecological conditions in the North are well adapted to cotton production. Soils in the Uélés are generally better than those in Ubangi-Mongala. Land tenure is no constraint on cotton production. The cultural practices used by farmers are primitive and very labor intensive. The only source of power is human labor. The population in Ubangi and Mongala sub-regions is increasing at 328 a rapid pace but several zones in the Haut—Zaire region have low to negative rates of population growth. The northern cotton belt has gained in importance relative to the southern zone since 1960. The Ubangi subPregion now contributes 19.0 percent of the total cotton production in Zaire as contrasted with 9.6 percent in 1959. The decrease in cotton production in the Equator region since 1959 is almost completely attributable to a reduction in the Mengala sub-region where production is now only about one-fourth of the 1959 level. Cotton production fell relatively more in the Uélés than in UbangieMongala. The potential for increasing cotton production in the Uélés is much larger than that in Ubangi-Mongala. Cotton marketing and processing are discussed in Chapter IV. In 1971, a national cotton marketing office (ONAFITEX) was created as a.major step toward rationalizing the cotton marketing structure in Zaire. Until then, conventions with foreign controlled private cotton ginning firms regulated cotton processing and marketing on a custom work basis. This approach resulted in generous profit.margins being accrued by the ginners. By 1970 the reserves of COGERCO, the price stabilization fund and regulatory agency, were eXhausted and this prompted the government to intervene and establish ONAFITEX. In 1970 half of the 115 ginning plants in Zaire were closed. Although sufficient ginning capacity is presently available, thirty new ginneries are now'being installed in Zaire. Technically ginning has not been a bottleneck in cotton processing and there does not appear to be an economic justification for the government's purchase of the new ginneries. 329 Cotton marketing costs are very high. Transport costs repre- sented one-fourth of total marketing costs, including the purchase of seed cotton in 1969 and they were much higher for the Uélés than for Ubangi—Mongala. Transport costs are very high because of the poor roads , the long distances involved in hauling seed cotton and the luxury goods rate charged by ONATRA and VICI-ZAIRE for the transport of cotton bales. ONAFITEX has reduced marketing costs by laying off expatriate persomel, and by reducing transport rates, financing and baling costs. However, the collection of seed cotton in 1973 and 1974 started late and was extended over much of the year. This reduced the farmer's willingness to plant cotton; it occupied collection agents, many of whom are cotton extension workers , for a long time on collection; and it increased transport costs. The Zairean textile industry is well developed. A large textile mill is now going into production in Kisangani and this will consider- ably improve the marketing efficiency of cotton from the Uélés. The average sale prices of cotton fiber to dotestic textile factories retained remarkably stable over the 1963-1973 decade. Presently, local fiber prices are kept low to protect the Zairean textile industry and Zairean consumers. In 1973-1974 fiber prices were about half of the export price. The prospects for expansion of the domestic market for locally made textiles are bright because of the high income elasticity of demand for cotton. The cottonseed crushing industry is well deve10ped but suffers from outmoded equiprent and high transport costs . Since 1960 most 330 mills have been operating in the red. In 1973 most cotton seeds in the North were either burned or dumped; sore were exported. A high capacity oil extraction mill is being opened in Gemena in Ubangi and will process all the seeds from Ubangi-angala and probably some iitmtthe Uélés. Chapter V focuses on the methods of primary data collection on cotton farms in northern Zaire. A.farmlbusiness survey was conducted of 160 cotton farmers selected at random, with two visits per farm, once before harvest and once after. Four field enumerators were hired to interview farmers. This survey was supplemented with visits to the ginning mills, to the regional branch offices of ONAFITEX and to the central headquarters in Kinshasa. The farm business survey data are presented and analyzed in Chapter VI. The average age of the farmers was 44 years. The average years of formal schooling was 1.4. Almost half of the farmers had been employed off their fanm in some way in the past. Each family was composed of 5.05 persons. Maize and manioc were the most important food staples in the Northéwest and manioc and bananas in the North-East. land tenure did not appear to be a prdblem. On the average, 86.0 percent of the farmers were under obligation to grow cotton. The average acreage imposed was about half a hectare. Stiff fines and/or jail terms were given to farmers who ignored cotton growing requirements. Since farmers were only interviewed twice it was not possible to collect accurate labor data. Communal labor was fairly common. Most farmers had to help repair the roads which pass through their villages, for an average duration of 6.7 weeks per year. About half 331 of the farmers also had to work for the chief of the village or head of the clan for an average of 10.5 weeks. These activities interfered with their cotton operations . Most farmers possessed only a few elerentary hand tools such as a machete, hoe and axe. Chemical fertilizers were miknown to all cotton farmers interviewed. Some farmers (12.9 percent) complained that they had received cotton seeds too late. Fields were planted one to three weeks after the optimal time of planting. The recomended crop rota- tions were generally observed although cotton was almost never grown in a pure stand. Fallow time was on the average 6.5 years, far below the recommended period of 10-20 years. Cotton fields were located on an average of 1.9 km. from the cotpound of the farmer. Only 19.2 per- cent of the cotton fields were grouped in large blocs. Only about one-third of the farmers decided where to locate their cotton fields and only one-tenth determined tie size of their cotton fields. Agri- cultural monitors and to a lesser degree territorial agronorists and village or clan authorities were primarily responsible for determining the location and size of cotton fields. The average cotton acreage per farm was small: 35.7 ares, less than one acre. There was a large variation in the size of cotton fields between farms. The average farm size was 83.0 ares; 113.6 ares in Ubangi-Mongala, 34.3 ares in Bas-Uélé and 53.7 ares in Haut-Uélé. The largest cotton fields and the largest number of f relds per farm were in Ubangi-Mongala. Cotton plant densities were fairly uniform but well below the recormended density. Cotton maintenance was general 1y poor and farmers 332 harvested too late. The average yield per farm in kg. per hectare was 399 kg. (257 kg.in Ubangi-Mongala, 687 kg. in Bas-Uélé and 436 kg. in Haut-Uélé) . Yields differed widely between farms and between sub- regions. Yields in Bas-Uélé were more than double those in Ubangi-Mongala. Gross revenues from cotton in Bas-Uélé were also more than double those in Ubangi-Mongala or Haut-Uélé . On the average , farmers earned a gross revenue of 8.792 (l7.58$) from cotton production. Most famers leaned about the cotton price on the day of sale or only a few days before. Farmers interviewed did not know the weight of seed cotton they brought to the collection center. The majority of the farmers (67.7 percent) affirmed that the producer price over the last five years had influenced their decision to plant more or less cotton. Farmers thought that the present cotton prices were too low as fanmers reported that the price would have to be raised substantially before they would increase their acreage; on tle other hand, farmers reported that prices would have to fall con- siderably before they would reduce their cotton acreage . At present cotton prices farmers are insensitive to small price changes. Farmers in the Uélés were much more seisitive to price changes than those in Ubangi-Mongala. About half of the farmers believed that cotton pro- duction earned them a lot of money but 96 . 1 percent also considered cotton a risky crqa. Oily one-fifth of the farmers expressed an interest in joining a sales cooperative for cotton. ‘ Of the farmers interviewed, 87.5 percent knew an agricultural monitor (extension agent). The territorial agronomist was less known by the farmers. Generally, the farmers believed that the help of 333 agricultural monitors and the territorial agronomist was needed to produce more cotton . Cotton was ranked as the first source of sales revenue by 42 per- cent of the farmers; 23 percent also ranked cottonas the second most important source. For 18 percent of the farmers, coffee was the first source. Cotton was by far the primary source in Bas-Uélé and coffee was more important in Ubangi-Mongala. Farmers paid an average head tax of 2.472 per year as well as other taxes. The dowry represented a considerable outlay of money relative to the farmer's incore; the average cash portion of the dowry was 34. 82. Most farmers like to grow cotton for the cash incore it provides. However, they complained about the low seed cotton price and the small amount of money they earn with cotton. They consider poor health as a major constraint on their ability to produce more cotton. The lack of incentives and the problers connected with the delayed purchase of their cotton are given as other important factors for giving up cotton production. They expect the government to raise cotton prices and to provide farm tools, machinery and gifts as incentives to expand production. In Chapter VII we analyzed the returns to labor in cotton produc- tion using labor data derived from agricultural experiment stations. We found that the returns to labor in cotton production are comparable with trose for food crops. These returns are below the legal minimum wage of unskilled workers in rural areas. Perennial crops earn much higher returns to labor than cotton, with cocoa the most profitable crcp followed by coffee and oil palm. These returns are well above the legal minimum wage. _334 In December 1973 the real purchasing power of seed cotton was about half of the June 1960 level. Thus, other things being equal, the real returns to labor for cotton production are now much below those of 1960. This author strongly believes that erosion of the farmers' real purchasing power of seed cotton explains a large part of the lack of interest in cotton production, low yields and the present low level of production. Presently, cotton farmers can derive much higher labor earnings per day from other crops than cotton, particularly coffee, which has expanded greatly among smallholdersgsince 1965. Maize and paddy can also be more profitable if there is an assured market outlet at the official minimum purchase price. Chapter VIII contairs an analysis of the factors explaining cotton production. A multiple linear regression equation model was used to explain the variation in yields beoween farms. For the northern cotton belt, a coefficient of multiple determination (R2) of 0.58 was obtained and the dummy variable for Bas-Uélé turned out to be the most important factor. Other important variables with a positive effect were cotton revenues, the distance from the house to tie field, cotton plant density, imposition of other crops than cotton, labor performed on other activities than maintenance apd labor units available on the farm. The only variable with a negative effect was the dummy variable for the Mongala sub-region. In the analysis of factors explaining acreages, an atterpt was made to determine the price elasticity of supply in terms of acreages planted. The results do not confirm our hypothesis of a positive nominal price effect. The following factors probably influenced the 335 unexpected outcome of our supply response analysis: the real cotton price remained almost constant over the 1970-1972 period, the 1970 acreage data could have been inaccurate and other things may not have remained equal over the time period considered. A multiple linear regression statistical analysis was carried out to explain the variations in cotton acreages between farms. The coeffi- cients of multiple determination obtained in this analysis were below those for the analysis of yields. For the North, one-half of the variance in cotton acreages was explained in terms of the variance of eight other variables (R2 = 0.50) . TWO dummy variables are included in the equation, one for Ubangi and one for Mongala, reflecting the major differences between sub-regions. The most important variables were acreage in other crops, particularly in Ubangi-Mongala, and years of fallow, particularly in the Uélés. Both variables had a positive effect on cotton acreages. Another variable with a positive effect was work for chiefs and other authorities . Variables with a negative influence on cotton acreages were the labor inputs per hectare in cotton production with labor for cotton maintenance having a stronger negative effect than labor for other cotton activities. A comparison of the cotton production determinants between 1958- 1959 and 1970-1971 using Official government data revealed that pro- duction in 1971 in Ubangi was at the 1959 level but that the increase in the number of cotton farmers and in total acreage were offset by a large decrease in yields per farm in kg. per hectare. For Mongala, production dropped 68 percent as a result of a decrease in number of cotton farmers, a large decrease in the acreage per farm and an even 336 larger decrease in yields per farm in kg. per hectare. In Bas-Uélé the number of cotton farmers , the average acreage per farm and the yield per farm in kg. per hectare decreased all in the same degree, about 40 percent and accounted for the large drop in production, 77. 7 percent. In Haut-Uélé the same picture was repeated but the decline in yields was much larger, 73.1 percent,while production fell by 81.5 percent. A corparison of the cotton production determinants derived from the farm business survey with the 1970-1971 Official data gave almost identical data for Ubangi-Mongala but marked differences for the Uélés, except for the acreage data. Yields in kg. per hectare for the Uélés in 1973 were at the 1959 level, sorewhat higher for Bas-Uélé and sam- what lower for Haut-Uélé. This contrasts with the official 1971 data which showed a sharp decline in yields since 1959. As a result, this author questions the 1970-1971 yield data for the Uélés. According to the 1972-1973 farm business survey data, the low level of cotton production relative to 1959 can be explained only in terms of a decrease in acreage per farm and a large decrease in the number of cotton farmers in the Uélés. In Chapter IX alternative strategies for increasing cotton pro- duction in the North were analyzed and evaluated. First, the barriers to improving the performance of the cotton subsector in northern Zaire were listed. Then, the strategies for increasing cotton production were examined within the context of increasing regional commodity specialization within northern Zaire and in a political economy framework of developing the North. _337 In the immediate future, the most promising avenues for expanding cotton production are a revision of cotton pricing policies, diffusion of improved agronomic practices , an overhaul of the cotton extension service, a pilot cotton developlent project in Bas—Uélé and accelerated research on new varieties. A substantial increase in the seed cotton price from 6.5 Makuta to 10.0 Makuta per kg. for the first grade seed cotton is needed to provide adequate incentives for farmers to expand cotton production. Such an increase in the producer's price could be made possible by raising domestic cotton lint prices in relation to world market conditions. An expansion in production will, in turn, reduce per unit cotton marketing and processing costs and will provide ONAFITEX with substantial export earnings. When more favorable producer pricing policies are adopted, the cotton imposition system can be reIoved and the cotton extension service can gradually shift from its traditional regulatory role to an educational role. There is a large opportunity for increasing cotton yields in northern Zaire by introducing improved agronomic practices such as timely planting and harvesting, correct Spacing, frequent thinning and weeding, etc. However, a major overhaul of the cotton extension service is needed in order to create the administrative structure and leadership for effective small farmer extension programs. The following extension refonms were recomended: retraining and \pgrading of extension workers , creation of regional cotton extension comittees and the establishment of an extension department at the National University of Zaire and at the National Institute for Agronomic Studies and Research . 338 A.pilot cotton development project in Bas-Uélé, centered in Bambesa and focusing on an extension of improved agronomic practices, wouid have a great potential for increasing cotton production. More- over, specialization of production in this area.would substantially reduce per unit assembly, transport and ginning costs. Research on new varieties for the North needs to be accelerated and improved varieties should be imported from.neighboring countries as the present variety is degenerating. Chemical fertilizers, tractor mechanization and agricultural credit are presently less promising strategies for increasing cotton production in Northern Zaire. More technical and economic research at the farm level is needed to test alternative rates of application of fertilizers, selective mechanization, and the need for agricultural credit. When the cotton producer's price is raised and when the extension service is upgraded, these strategies may becoxe feasible. Zaire is a very small consumer of fertilizers with less than 1 kg. per head in Zaire as compared with over 10 kg. in most.Western African countries. Fertilizer prices are very high in Zaire because of the predominately western European origin, ocean freight, high transport charges to interior parts of Zaire at the general merchandise fare and poor delivery systems. Low product prices, lack of fertilizer response data, sociological factors and a rudimentary agricultural extension service are the other major constraints on fertilizer utilization. Zaire has recently decided to create its own fertilizer industry. The government should seriously consider the subsidization of fertilizer in order to promote its introduction among smallholders, first in pilot 339 cotton development projects and then later among smallholders if it proves profitable. The present approach to cotton plant protection is quite pater- nalistic and is modeled.on the colonial experience. For long run success it is necessary that farmers therselves adopt spraying or dusting with pesticides. To favor introduction of pesticides and hand powered elements, the government should temporarily subsidize these inputs until they are in.widespread use. From 1964 to 1973, the Zairean government spent approximately 17 million dollars on the importation of heavy agricultural machinery. Tractor mechanization was tried out on a large scale in Zaire froml 1950 to 1960, mainly in savannahs. The use of machinery was only justified when a substantial increase in the area under cultivation occurred. Usually cotton yields increased by 15 to 25 percent over manual land preparation. Animal powered mechanization was tried with success before independence in savannah areas of southern Zaire. Presently, animal powered mechanization has almost completely dis- appeared in Zaire. Tractor mechanization of cotton production in the North cannot be remanded. All mechanization projects in Zaire rely heavily on expatriate management and.the economic effects of tractor mechanization have not been properly studied to justify tractor mechanization. However, the government should take steps to improve the supply of improved hand tools and equipment in rural areas of the NOrth. Agricultural credit schemes for smallholders in Zaire have failed because of unrealistic demands for guarantees, lack of loan supervision, 340 political interferences, poor loan managetent and lack of integration with otter agricultural projects. The government is now pressuring private banks to lend money for agricultural purposes . In northern Zaire, we found no indications that lack of agricultural credit at the farm level was a major constraint on cotton production. Within the present administrative capacity of Zaire and the historical record of farmer cooperatives, it is doubtful whether agricultural credit would be profitable for increasing cotton production, particularly since there is not an improved package of inputs or the required extension service for introducing package programs . Subsidization of inputs such as insecticides, hand poered dusters or Sprayers, etc. is likely to be a more effective strategy to prolote cotton production than setting up a special credit program. In the remainder of this chapter, a series of recommendations are made for increasing cotton production and for improving the market- ing performance of the cotton subsector of norther Zaire. Policy Prescriptions for IncreasinELProduction Improvinthhe Marketing Performance of the Cotton Subsector of Northern Zire Revising Pricing Policies and Reroving the Imposition System Cotton pricing policies cannot be divorced from price policies for other crops and thus, pricing policies need to be designed for a cotton— food-export crOp production system in relation to the produc- tion objectives for each comodity. For example, since coffee production already exceeds Zaire's quota, trere are somd arguments for reducing the coffee price . 341 Cotton farmers definitely need stronger incentives to increase cotton production. We, therefore , recommend an increase in the dorestic lint price of 15 to 20K. per kg. ‘which will generate enough revenue to enable ONAFITEX to raise the producer price from 6.5K. to 10K. per kg. of first grade seed cotton. A price increase of 15 to 20K. per kg. of cotton lint for domestic spinners would.raise the price of domestic lint to about 60 to 65K. per kg. of lint, still well below the price whidh they would have to pay for imports of the same grade and.quality. An increase in producer prices would substantially increase cotton production, reduce per unit.marketing and processing costs, provide ONAFITEX with substantial export earnings and speed up the diffusion of innovations. .A substantial increase in the seed cotton price will also reduce the need for the imposition.system.as farmers will be motivated to produce cotton without compulsion. This will free cotton extension.agents from regulatory tasks and allow them to pursue an educational function. The price increase could be both in cash and in kind. For example,:many farmers cannot purchase small tools in their community and these could be made available at subsidized rates. Diffusion of Improved.Agronomic Practices and.New Inputs There is a large opportunity for increasing yields by introducing improved agronomic practices such as timely planting, thinning, weeding and harvesting, correct spacing, use of improved hand tools, etc. The cotton extension service should concentrate on the introduction and diffusion of improved agronomic practices. Demonstration plots in 342 selected villages would.be very useful in showing farmers what could be achieved. Cotton seeds should be sorted in the ginneries and disinfected before distribution to fanmers in order to improve germination rates and to assure adequate plant densities. When cotton production becomes profitable for farmer, the govern- ment should drop spraying and dusting the farmers' fields and instead should concentrate on teaching farmers how to treat their fields with pesticides. The government should provide farmers with locally made hand powered dusters or Sprayers and pesticides at a subsidized price to stimulate adoption. Since Zaire has decided to create its own fertilizer industry, efforts should be undertaken to introduce chemical fertilizers on pilot projects and then later among smallholders if it proves profit- able. However, the costebenefit ratio of fertilizer should first be improved considerably by increasing the seed cotton price, by elimina- ting all import duties and taxes and by reducing the fertilizer transport rate schedules of ONATRA and other public and semi—public transport agencies. Only highly concentrated fertilizers should be imported, preferably from suppliers within Africa. Granulation, blending with inert fill and bagging should be done in Zaire. The government should seriously consider the subsidization of fertilizers in order tijromote their introduction.among smallholders. At the same time, an input delivery system should be set up'which focuses on smallholders. 343 Overhauling the Cotton Extension Service Several actions need to be taken to shift the cotton extension service from a regulatory to an educational agency . The administrative structure of agricultural extension should be reformed in order to coordinate and integrate agricultural extension efforts. Extension monitors should be placed under a single authority. Since ONAFITEX reeds a cadre of well-trained cotton extension officers, it should have some influence and control over cotton extension monitors regardless of which agency has formal responsibility for the extension system. Agricultural extension workers need to be retrained and their performance should be reviewed periodically. The function of extension agents needs to be reviewed; they should be provided with adequate transportation, facilities for field work and demonstration plots, opportunities for professional advancement and supportive personnel and services . A cotton extension committee should be appointed in each regional branch of (NAFI'I‘EX and charged with drawing up a campaign for the next crop. Immediate steps slould be taken to develop a cotton hand- book which summarizes and translates the recommended cultural practices into local languages for all cotton extension agents. Each year, new cotton research results should be added to this handbook. A department of agricultural extension should be set Lp at the National University of Zaire, preferably in the Faculty of Agronomy, either as a new department or added to the existing department of agricultural economics . A curriculum in agricultural extension educa— tion should be developed including courses in communications, rural 344 sociology, farm management, agronomy and agricultural development. An extension liaison department should be created in INERA to ensure that research findings are diffused and that there is constant feedback into the research system. Establishing a Pilot Cotton Development Project in Bas-Uélé We recomend the establishment of a pi lot cotton development project, centered in Bambesa, which focuses on extension of improved agroromic practices. The Bambesa project can be modeled after the CFDT project in [bangi-bbngala. The CFDI‘ project has demonstrated the importance of testing the potential of improved agronomic practices under actual farm conditions. The primary objectives of such a pilot project in Bambesa would be to increase the yields to an average of 1,000 kg. per hectare, an increase in average cotton acreage per farm to 50 ares and an increase in the number of cotton farmers. The feeder roads should be improved in the project area and provisions slould be made for the ginning mills to handle the increased production. The Bambesa pilot project would foster the specialization of cotton production and would greatly reduce per unit cotton assembly, transport and ginning costs. Technical Cotton Research Cotton production in northern Zaire could benefit greatly from a good research program which is presently absent. The following points stould be considered. Research on Improved Varieties As soon as adequate and cmpetent staff can be found,the research 345 station in Bambesa should be geared up for cotton breeding and selection for the northern cotton belt. In the meantime, the Gandajika station in the South should continue its plant breeding activities , but selection of new varieties for the North should be concentrated in the northern ecological area, preferably in Bambesa. Better cooperation between the French IRCI‘ stations in West and Central Africa, INERA in Bambesa, Boketa and Gandajika, the agricultural service of ONAFITEX and the CFUI‘ mission in Ubangi-Mongala is desirable for restoring and maintaining the varietal purity of Reba B 50 and for gaining access to better varieties developed in other countries . Research on Better Agronomic Practices and Improved Input Packages Presently, INERA is concentrating almost solely on the breeding and selection of new cotton varieties. However, there are many agronomic problems which need urgent research attention before recomrendations can be drawn up for the cotton extension service. Improved input packages will most likely include chemical fertilizers, pesticides, credit and selected mechanization technologies. Research is needed to determine the optimum types and rate 5 of application of these inputs and the financial and ecmomic returns on these packages. Since cotton is always part of a food crop rotation, the residual effects of improved inputs on food crops need to be taken into account for a proper analysis of the impact of a new cotton production technology. The research efforts of the CFDI‘ mission in Ubangi-Mongala are laudable and are a step in the right direction. However, their efforts 346 are clearly not sufficient, and more rigorous trials and testing are necessary before solid recomrendations can be made. Improving the Performance of Cotton Marketing The following points should.be considered in attempting to improve the operational efficiency, vertical coordination and organe ization of the marketing and processing activities of ONAFITEX. gperational Efficiengy_ Substantial gains in Operational efficiency could be achieved by ONAFITEX through better organization and coordination. The following steps should be taken to enhance operational efficiency. Reducing Transport Costs --The organization of seed cotton collection should be better organized to reduce transport costs. Half capacity runs and onedway payloads should be avoided. --Seed cotton collection should start in December and terminate by the end of March, allowing most cotton to be transported during the dry season. Thus, collection agents would also be able to devote more time to agricultural extension as the buying season would be shortened. --Standardized canvasses and large scales should be utilized to speed up the weighing and collection of seed cotton. --Seed cotton should be kept in the storage facilities of the collection centers when storage facilities at the ginnery are inadequate. 347 -—Efforts should be made to obtain a more advantageous fare classification for cotton bales transported by public and semi—public transporters. --Marginal cost pricing should be applied for the transport of cotton seeds from ginneries to crushing mills. Reducing Ginning Costs --The overstaffing at ginning mills should be corrected. --More spare parts should be stocked for the repairing of gins, power sources and transport equipment to overcome undue delays in ginning and seed cotton collection. --During the dead season all ginning and transport equipment should be overhauled. --If necessary, two or even three work shifts should be organized in ginneries when storage facilities are inadequate to handle all incoring seed cotton. Improving Organization and Administration --Operational efficiency is hampered by serious communication problems between the central headquarters in Kinshasa, the regional branch offices and the ginning mills. Steps should be taken to improve communications . --The flow of funds in appropriate denominations and quantities for the buying of seed cotton needs to be accelerated. - the budget control and accounting system needs to be strengthened , although it should not slow down the collection and ginning of seed cotton . 348 Diffusion of Information --Producer prices should be announced prior to the planting season, preferably during the distribution of cotton seeds. --Farmers should be infonmed in advance of the collection date and the price they will receive for seed cotton. --Since 7 percent of the farmers possess a radio, this media should be used to promote better agronomic practices and to inform them about the seed cotton price and the date of collection. Rational Utilization of Cotton By-Products ways should be explored to utilize cotton seeds that are presently burned or wasted. Inoperative crushing mills should be studied to determine whether new crushing equipment is needed or whether the available equipment should be repaired. An overall evaluation of the cotton seed crushing potential in Zaire should be carried out to guide the government in its efforts to use cotton seeds more rationally. Counterpart Training ONAFITEX has hired skilled, experienced expatriates which were fermerly employed by the private cotton companies. Their primary role should be one of advising and counterpart training and not.of decision- making and execution as is often the case. This will ensure that their efforts and experience are not dissipated on minor activities and that a Zairean capacity for marketing and processing management is created. 349 Future Research Needs The present lack of knowledge about the major constraints on cotton production and the high costs of cotton processing and.marketing requires on-going farm management, marketing and macro—economic research. Ckmmmnications research for agricultural extension is needed to guide the cotton extension services. Farm Management Research on Cotton Production This study has utilized a farm business survey approach and, as a result, allocation could not be accurately measured. A cost route survey in the major regions could yield the needed data for a detailed study of the returns to labor in cotton production as well as data on labor bottlenecks. Cost route studies will help guide the introduction of improved inputs and the design of appropriate pricing policies. The results of cost route surveys would also help in developing simple extension guidelines on the economics of producing cotton and other crops. The identification of labor bottlenecks could help_identify optimal production systems, including the deve10pment of appropriate types of mechanical technology for breaking seasonal labor bottlenecks. It is reermmended that sudh a study be organized as an integral part of pilot cotton development projects. The FED—CAK project in Eastern Kasai merits a full farmlmenage- ment survey and a detailed cost-benefit study of fertilizer application, pesticide treatment, tractor mechanization and agricultural extension. However, it is uncertain whether the results obtained in Eastern Kasai are transferable to the North because population, soils, climate, cotton varieties and marketing structures vary between Eastern Kasai and.the North. 350 Tractor meChanization in FIWA—BILI should be studied as this project is currently not generating the data needed for a rigorous economic and financial analysis of mechanization. The CFDT team in Ubangi-Mongala should continue its research activities in order to develop general guidelines for the expansion of cotton production. However, they should adopt a more rigorous research design, including more repetitions per trial and more varia- tion in types and rates of fertilizers applied, in order that an economic analysis of optimum types and dosages can be carried out. Communications Research for Improving Cotton Extension and Facilitating an Information Feedback to Researchers Communications research for cotton extension could have high pay-offs in Zaire. For example, since 7 percent of the farmers possess a radio, the radio could be tested as a method of informing farmers about cotton prices, the proper time of collection and improved agricultural practices. (Xmummutetions research could direct govern- ment attention to the linkage between agricultural research, the Department of Agriculture, ONAFITEX and the cotton farmer. Marketing Research Marketing research could lead to substantial reduction in the cost of cotton marketing and processing. A study is needed to determine the optimum number, size and location of ginning facilities and the optimum transport routes. This could guide ONAFITEX in the purChase and location of new ginning equipment and for the organization of transport. 351 Macro-Economic Research Pricing of agricultural products as well as inputs is a major area requiring policy—oriented research. Increased attention must be directed to the impact of alternative pricing policies and their impact on production, imports, exports and consumption of the major food products and cotton. Since most developmental efforts are focused on specific agricul- tural projects, research is needed to identify and analyze agricultural development projects. Project analysis should draw upon the results of detailed micro research because planning from the top down has produced disappointing results. Only when a careful economic analysis of the present cotton development projects in Zaire is completed can a particular input package be recommended for introduction and diffusion by the cotton extension service. BIBLICXSRAPHY BIBLIOGRAPHY This bibliography includes all publications cited in the text, as well as important uncited material on the Zairean agricultural economy. Some works relating to the cotton subsector in other African countries are also included. American University, 1971. Area Handbook for the Democratic Repgblic of the Congo. Washington: Foreign Area Studies. Baker, B. N. and R. L. Eris, 1964. An Introduction to PERT—CPM. Homewood, Illinois: Richard D. Irwin, Inc. Bala, Ernest, 1963. "Ie Cotonnier: Sa Biologie et son Exploitation." Iéopoldville: Université Iovanium, Faculté d'Agronomie, unpublished thesis. Banneux L., 1938. Quelques Données Economuques sur le Coton au Congo Belge. Bruxelles: I.N.E.A.C., Série Technique Nr. 22. Bannink, L., 1958. "La Fumure.Minérale du Cotonnier." Bruxelles: Bulletin d'Information de l'INEAC, Vel. 7. Banque du Zaire, December 1971. Rapport Annuel 1970-1971. Kinshasa. , December 1972. Rapport Annuel 1971-1972. Kinshasa. , December 1973. Rapport Annuel 1972—1973. Kinshasa. Bartlett, Christopher D. S., 1974. "Factors Affecting the Spread of Rain-Grown Cotton in Tropical Africa." Stanford University, Food Research Institute, unpublished doctoral dissertation. Béguin, H., 1958. Geographic Humaine de la Région de Bengamisa. Bruxelles: I.N.E.A.C. Biebuyck, Daniel, 1966. Rights in Land and Its Resources Among the Nyanga. Bruxelles: Academie Royale des Sciences d'Outre-Mer, 34:2. Blalock, HUbert M., Jr., 1960. Social Statistics. New York: McGrawe Hill Book Co. Boulanger J. and A. Jarry, June 1963, "Deux Nouvelles variétés de Cotonnier en Republique centrafricaine: E40 et Reba.B 50". Paris: Coton et Fibres ngpicales, 18:2, 299-300. 352 353 choma, Reportage de YAV a Mutach, May 13, 1974. "Les Dix Jours du Commissaire de Region dans le Bas-Uélé: 3. les problémes du Bas-Uélé." Brixhe, A. , 1958. Ie Coton au Congo Belge. Bruxelles: Compagnie Cotonniére Congolaise, 3rd. édition published by the Ministére du Congo Belge et du Ruanda—Urundi . Bustin, Edouard, 1970. A Study Guide for Congo-Kinshasa. Boston: African Studies Center, Boston University. Byerlee, Derek and Carl K. Eicher, 1972. "Rural Employment, Migration and Economic Development: Theoretical Issues and Empirical Evidence from Africa," AREP No. 1, African Rural @lcyment PaEr, Department of Agricultural Economics , Michigan State University, East Iansing, Midiigan. CADICEC, 1973. Code du Travail; Mesures d'Application. Kinshasa: editions CADICEC. Cauwe, André, May 1969. "Ies Conditions de Réussite des Entreprisee Cooperatives." Kinshasa: Cogo—Afriqtie, Nr. 35, 249-264. Centurion, C.; P. Debricon; J. B. Roux; L. A. Tores Bogado, 1972. "Deux Variétés de Cotonniers Sélectionnées en Afrique Centrale Reba B 50 et Re’ba BTK 12, Se Montrent Bien Adapte’es au Paraguay," Coton et Fibres Tropicales. Paris: I.R.C.T., 27:3, 291-294. C.F.D.T., 1972. "Rapport sur la Commission Agricole du Kasai Oriental au Zaire." Paris, unpublished document. Clayton, Eric S. , 1964. Agarian Development in Peasant Economies. Iondon: Pergamon Press . CDGERCO, 1970 . "Oommentaires Concernant 1e Budget Général Présenté par le COGERCO pour la Cotpagne Cotonniére 1970." Kinshasa, unpublished document. , 1970. "Prix de Revient Industriel 1968/1969." Kinshasa: unpublished document. Collinson, Michael P. , 1972. Farm Management in Peasant Agriculture. New York: Praeger. , December 1974. "Sore Guidelines for Farm Economic Data Collection in Rural Areas of Africa," Paper prepared for the Alli/CWT/Ford Foundation Workshop on Field Collection of Socio—Economic Data in Developing Countries . Beirut: December 8-14 , 1974. DeCoene, R., 1952. "Ies Méthodes et les Progrés de la Sélection du Cotonnier a Bambesa." Bruxelles: Bulletin Agricole du Congo Belge. 354 , April 1956, "Ie Bananier dans la Rotation en Zone Cotonniére Nord." Bruxelles: Bulletin d'Information de l'I.N.E.A.C., 5:2, 113—126. De Dampierre, Eric, May 1960. "Coton Noir, Café Blane, Deux Cultures du Haut Oubangui a la Veille de la lei-Cadre." Paris: Cahiers d'Etudes Africaines. Degiorgio, Fred, Septerber 1969. Congo' 5 Agricultural Econory in Brief. Washington, D.C.: Economic Research Service, U.S.D.A., BBS-Foreign 282. Derol, J. and L. Bannink, 1961. "Méthodes Culturales Rationnelles Appliquées dans la Zone Cotonniére Septentrionale du Congo. " Bruxelles: Bulletin d'Information de l'I.N.E.A.C., 10:4, 141-261. De Plaen, G., October 1957. "Delimitation des Diverses Regions Cotonniéres de la Zone Nord." Bruxeles: Bulletin d'Information de l'I.N.E.A.C. , 6:5, 285-300. De Pooter, R. , 1960. "Ie Coton au Congo et Analyse Corptable d'une Usine Cotonniére. " Iéopoldvi lle: Université lovanium, Faculté des Sciences CoImerciales et Financiéres, unpublished thesis. Depy, G. , 1954. "Etape de la Convention avec les Sociétés Cotonniéres." Kinshasa, unpublished document. , February 1956. "La Caisse de Reserve Cotonniére." Bruxelles: Bulletin Agricole du Congo Belge. de Saint, Moulin, L., 1971. "Ies Statistiques Démographiques en République Démocratique du Congo. " Kinshasa: Congo-Afrique , 377-385. de Schlippe, P. , 1956. Shide Cultivation in Africa: The Zande System of Agriculture. London: Routledge and Reagan Paul. , 1957. Méthodes de Recherches Quantitatives dans l'Economie Rurale Coutlmuére de l'Afrique Centrale. Bruxelles: Ministére des Colonies. Deutsches Institut ffir Afrika—Forschung, 1975. "Planification et Développerent Economique au Zaire . " Hamburg : Contributions Hambourgeoises a la Connaisance de l'Afrique, Vol. 16, 168 p. Dewez, J., 1960. "La Culture Cotonniére dans la Plaine de la Ruzizi." Bruxelles: Bulletin d'Information de l'I.N.E.A.C., 9:3, 161-176. Dimandja, Berthold, July 1971. "Ie Prdolé'me de la Production Cotonniére dans le District du Sankuru." Kinshasa: Université Lovanium, Faculté des Sciences Economiques , unpublished thesis . 355 Drachoussoff, V. , 1965. "Agricultural Change in the Belgian Congo, 1945-1960." Stanford: Food Research Institute Studies, 5:2, 137-201. Du Bois, 11., August 1957. "Types d'Assolement en Culture Extensive de la Zone Cotonniere Nord." Bruxelles: Bulletin d'Information de l'I.N.E.A.C., 6:4, 227—241. Eicher, C. K.; T. Zalla; J. Kocher and F. Winch, Mai 1971. "la Creation d'Fmplois dans l'Agriculture Africaine." East Iansing: Michigan State University, Institute of International Agriculture, Research Report No. 9. F.A.O., 1962. Rapport sur les Possibilités du Développement Rural de l'Afrique en Fonction du Prgres Economique et Social. Rome. Fbcan, A., 1950. "Résultats des Essais de Fumure Minérale au Congo Belge." Bruxelles: Bulletin Agricole du Congo Belge, 41:1, 73-104. Fox, Renée C., 1970. "The Intelligence Behind the NEISk: Beliefs and Development in Contemporary Congo. " Harvard University, unpublished document. Frankart, R. , 1960. Carte des 8015 et de la Végétation du Congo Belge et du Ruanda-Urundi; Uélé. Bruxelles: I.N.E.A.C., Carte Nr. 14 A. , C. Geortay, G. , 1961. "Organisation de l'Agriculture Congolaise dans les Paysannats." Bruxelles: Bulletin d'Information de l'I.N.E.A.C., 10:1, 1-15. Gilson, A. H. , 1948. "Trente Années de Culture Ootonniére au Congo Belge." Bruxelles: Corpagnie Cotonniére Congolaise, 140 p. Harms, Robert, June 1974. "land Tenure and Agricultural Development in Zaire, 1895-1961." Madison: University of Wisconsin, land Tenure Center, L.T.C. No. 99. Heoq, J., A. Iefebvre, E. Vercruysse and A. Van Wambeke, 1963. Agriculture et Structure Economique d'une Société Traditionnelle au Kivu. Bruxelles: I.N.E.A.C. Hedley, Douglas D. and Dale Good, December 1971. "Maize Production in the Zaire Republic; An Appraisal and Policy Implications." Kinshasa: U.S.A.I. ., unpublished document. Hirsch, Hans George, March 1971. "Credit for Agriculture through Cooperatives in the Derocratic Republic of the Congo." Kinshasa: U.S.A.I.D. mission, unpublished document. 356 Houyoux, J ., 1973. Budgets Ménagers, Nutrition et Mode de vie a Kinshasa. Kinshasa: Presses Universitaires du Zaire. Hunt, K. E., 1969. Agricultural Statistics for Developing Countries. Oxford: Institute of Agrarian Affairs, University of Oxford. Huybrechts, Andre, 1970. Transports et Structures de Développement au Como. Kinshasa: I.R.E.S. et Paris: Mouton. , October-Decerber 1969. "la Politique Economique du Congo We a Travers la Tarification des Transports." Kinshasa: O.N.R.D., Etudes Congolaises, 12:4, 58-77. Hyden, Goran, Decerber 1970. "Can Co-ops Make It in Africa." Africa I.B.R.D.-I.D.A., J1me 1972. "Agricultural Sector Survey, Republic of Zaire." Washington, D.C. , Agricultural Projects Department, three volumes. I.N.E.A.C., 1955. "Ecarte1ent et Dates de Semis Préconisés pour la Culture Cotonniére en Zone Nord. " Bruxelles: Bulletin d' Information de l'I.N.E.A.C. , 1958. Normes de Main d'Oeuvre pour les Travaux Agricoles au Congo Belge. Bruxelles: I.N.E.A.C. , hors série. , 1962. "Précis des Maladies et des Insectes Nuisibles Rencontrés sur les Plantes Cultivées au Congo, au RWanda et au Burundi." Bruxelles: Publications de l'I.N.E.A.C., hors série. International Cotton Advisory Committee, January 1971. "Cotton-World Statistics," Quarterly Bulletin, 24:6 (Part II). I.R.E.S., July 1966. Etude d'Orientation_pour la Relance Agricole. Kinshasa: Bureau d'Etude du Haut Commissariat au Plan et a la Reconstruction Nationale. , 1968. Indépendance, Inflation, Développement: l'Economie Congolaise de 1960 a 1965. Paris: Editions Nbuton. , 1973. "Index des Prix de Détail des Biens de Oonsommation Courante aux Marchés de Kinshasa. " Kinshasa: Cahiers Economiqu et Sociaux, 11:3-4, p. 137. I.V.S., 1972. "End of Contract Report on the Supervised Agricultural Credit Program. " Kinshasa, International Voluntary Services , Inc. , unpublished document. Janzen, John M. , 1969. "The Cooperative in Lower Congo Economic Development, " in The Anthropology of Develppment in Sub-Saharan Africa. Iexington: University of Kentucky Press. 357 Jones, T. R., April 1966. "Comparison of Hand—(berated Machines for Cotton Pest Control in Uganda," East African Agricultural and Forestry Journal. Jones, W. 0., May 1960. "Economic Man in Africa," Stanford: Food Research Institute Studies, 1:2, 107-134. Jongen, P.; M. Van Oosten; C. Evrard, and J. M. Berce, 1960. Carte des 8015 et de la Vegetation du Congo Belge et du Ruanda-Urundi; Ubangi. Bruxelles: I.N.E.A.C., Carte Nr. 11, A.B.C.D. Jurion, F. and J. Henry, 1969. Can Primitive Farming be Modernised? Brussels: I.N.E.A.C. and Kinshasa: O.N.R.D., hors série. Kabange, Justin-Laurent, 1969. "L'Avenir des Principales Fibres Naturelles dans l'Industrie Textile." Kinshasa: Université lovanium, aculté des Sciences Commerciales et Financiéres, unpublished thesis. Kaponda, A. , 1966. "Contribution de la Culture Cotonniere au Développe- ment Economique du Maniema. " Kinshasa: Universite lovanium, Faculté des Sciences Economiques, unpublished thesis. Kinghoma, A. Malima. "The Economics of Cotton Production in Tanzania: Some Theoretical and Policy Discussion." Dar es Salaam: Economic Research Bureau, University of Dar es Salaam, E.R.B. paper 70.20. Kmenta, Jan, 1971. Elements of Econometrics. New York: Macmillan. Kombo, Timothee, 1970. "La Contribution de la Cooperation au Développe- ment Economique du Congo. " Université Catholiqte de Louvain, unpublished thesis. Kriesel, H. C. , December 1968. Cotton Marketing in Nigeria. East Lansing: Consortium for the Study of Nigerian Rural Development, CSNRD-24. Krishna, Raj, 1967. "Agricultural Price Policy and Economic Development" in Herman M. Southworth and Bruce F. Johnston, Agricultural Development and Economic Grmth. Ithaca: Cornell University Press, 497-547. Lacroix, J. L., 1967. Industrialisation au Congo. Paris: Mouton and Cie, and Kinshasa: I.R.E.S. _- Iaudelout, H., H. M. Dubois and G. De Plaen, 1955. "Ia Fumure du Cotonniere en Uélé." Bruxelles: Bulletin d'Information de l'I.N.E.A.C. Ieplae, Edmond, 1917. "Notes au Sujet du Développement de l'Agriculture du Congo Belge. " Bruxelles: Bulletin Agricole du Congo Belge, No. 8, 3-39, 172-217. 358 , 1933. "Histoire et Déve10ppement des Cultures Obligatoires de Coton et de Riz au Congo Belge de 1917 a 1933," Congo, 14th Year. , 1947. "Résultats Obtenus au Congo Belge par les Cultures Obligatoires, Alimentaires et Industriellesf' Bruxelles: Zaire, NO. 1,115-140. Iooney, Zolon M. and C. A,‘Wi1mot, October 1971. Economic Models for Cotton Ginning. washington, D. C.: U.S.D.A., E.R.S., Agricul- tural Economic Report No. 214. Lumpungu, Kamanda, 1973. "Ie Regime Foncier au Zaire et son Incidence sur le Développement.Agricole." Kinshasa: I.R.E.S., Cahiers Economiques et Sociaux, 11:3-4, 55-71. , Novembre 1970. "Culture Cotonniere et Société Rurale dans 1e Nord du Katanga (Production, Comercialisation et Pers- pectives de Développement)". Universite de Rennes, Faculté de Droit et des Sciences Economiques, unpublished doctoral thesis. , March 1974. "les Problemes Actuels de l'Economie Agricole Zairoise." Kinshasa: I.R.E.S., Séminaire sur 1e Développement Agricole au Zaire. Mabi, Mulumba, March 1974. "Les Problemes de Financement de l'Agricul- ture au Zaire." Kinshasa: I.R.E.S., Séminaire sur le Développemept Agricole au Zaire. Mac Arthur, J. D., 1968. "The Economic Study of African Small Farms: Some Kenya Experiences." Journal of Agricultural Economics, 19:1, 193—205. Magleby, R. S. and.E. Missiaen, January 1971. werld Demand Prospects for Cotton in 1980 with Emphasis on Trade by IDC's. washington, D.C.: U.S.D.A.-E.R.S., Foreign Agricultural Economic Report No. 68. Malengreau, Guy, 1949. "vers un Paysannat Indigéne." Bruxelles: ARSOM, Sciences Morales et Politiques, 19:2. Mbuyi, J. P. , December 1970. "Ie Programme de Crédit Agricole au Congo," Conference Donnée au Séminaire sur les Cooperatives Agricoles a Nganda—Kinshasa au Decembre 12. McCune, Donald L., 1972. "Trip Report, Republic of Zaire," Noverber 13-21, unpublished document. Ministere de l'Agriculture, 1968. "L'Agriculture Congolaise en Tableaux Statistiques," Kinshasa: Direction Etudes et Politique Agricole. 359 .Ministere de l'Agriculture et du Déve10ppement Rural, Avril 1970. Rapport Préliminaire 1959-1969. Kinshasa: Direction Etudes et Politique Agricole. Ministere des Colonies, 1952. Culture, Industrie et.Ctmmerce du Coton au Congo Belge: Legislation. Bruxelles: Direction de l'Agriculture. , 1957. La Situation Economique du Congo Belge et du Ruanda— Urundi en 1957. Bruxelles. Ministere du Congo Belge et du Ruanda-Urundi, 1960. Bulletin Agricole du Congo et du Ruandanrundi; volume JUbilaire 1910-1960. Bruxelles. Miracle, Marvin P., 1967. Agriculture in the Congo Basin: Tradition and Change in African Rural Economies. Madison: University of Wisconsin Press. Mekako, Kumu E. Madjo, 1973. "Etude AgrOPEconomique de Coton a 1'Ubangi." Kinshasa: UNAZA, Faculté des Sciences Agronomiques, unpublished thesis. Mubiala, Remy J., 1971. "La Reprise de la Culture du Coton au Congo." Kinshasa: Université lovanium, Faculté des Sciences Economiques et Sociales, unpublished thesis. Musée Royal de l'Afrique Centrale. (Ongoing Series). Monographies Ethnographiques. Tervuren, Belgium. Mushinz—Imana, T., 1966a. "La Commercialisation du Coton au Congo." Kinshasa: Universite lovanium, Faculté des Sciences Commerciales et Financieres, unpublished thesis. , 1966b. "Quelques Aspects de la Politique Agricole au Congo." Kinshasa: Université Lovanium, Faculté des Sciences Economiques, unpublished thesis. Mutombo, Pierre-Sylvain, 1970. Les Fibres de Coton en République Démocratique du Congo et l'Industrie Textile. Kinshasa, O.N.R.D. Nasse F., G. Chatellier and J. C. Beaudoin, 1973. L'Agriculture Zairoise de 1970 a 1972. Kinshasa: Departement de l'Agriculture Direction des Etudes et de la Politique Agricole. National Academy of Sciences, 1974. African Agricultural Research Capabilities. Washington, D.C.: Committee on African Agricul- tural Research Capabilities. Nelis M. and M,‘Wieers, July 1969. "Le Prdbléme Cotonnier." Kinshasa: Ministere de l'Agriculture, unpublished document, 18 p. 360 Norman, D. W., 1967-1972. "An Economic Study of Three Villages in Zaria Province: Parts 1, 2 and 3," Samaru Miscellaneous Papers 19, 23, 37 and 38, Ahmadu Bello University, Zaria, Nigeria. , 1971. "Intercropping of Annual Crops Under Indigenous Conditions in the Northern Part of Nigeria," Rural Economy Research Unit, Institute for Agricultural Research, Ahmadu Bello University, Samaru, Zaria, Nigeria (Mimeograph). , 1973. "Methodology and Problems of Farrm Management Inves- tigations: Experiences from.Northern.Nigeria," AREP No. 8, African Rural Emplgyment Paper, Department of Agricultural Economics, Michigan State University, East Lansing, Michigan. Nzengele, N., February 1971. "Probléme2Cotonnier au Congo et Situation du COGERCO." Kinshasa: ONAFITEX, unpublished document. Nzeza,.A., 1971. "The Demographic Transition Applied to the Democratic Republic of the Congo," in The Demographic Transition in Tropical Africa. Paris: Development Centre of O.E.C.D. ONAFITEX, Direction Regionale Nord—Ouest, May 1974. "Essais Agronomiques 1973." Gemena: .Mission C.F.D.T. Peemans, J. Ph. , July 1974. The Political Economy of Zaire in the Seventies: An Essay in Historical Perspective. Louvain: Institut d'Etude des Pays en Développement, No. 7406, 71 p. Peeters, G., 1958. "LFAgriculture Congolaise et ses Problemes de Structure," Zaire, 451-477. PERT/CPM/LOB Prgject.Management Systems for Economic Development. Washington, D.C.: Agency for International Development. Republique Démocratique du Oongo, 1970a. Rapport Annuel, Service Provincial de l'Agriculture, Province du HautrZaire. Kinshasa: Ministére de l'Agriculture. \ , 1970b. Rapport Annuel, Service Provincial de l'Agriculture, Province de l'Equateur. Kinshasa: MiniSEere de l'Agriculture. , Mai 1971. Programme Décennal de Développement Rural. Kinshasa: Ministére de l'Agriculture. Republique du Zaire, Avril 1972. Projet de Relance Cotonniere dans la Province de l'Equateur. Kinshasa: .Ministere de l'Agriculture. , 1973a. Rapport Annuel 1972, Region du Haut Zaire. Kinshasa: Division de l'Agriculture. 361 , 1973b. Résultats Provisoires de L'Agriculture: Recencement 1970. Kinshasa: Départerent de l'Agriculture. , Novembre 1973. Conjoncture Economique, Année 1972 et 1er Trimestre 1973. Kinshasa: Ministére de l'Economie Nationale, Nr. 13. , 1974. Rapport Annuel 1973, Région de L'Equateur, Division Regionale de l'Agriculture. Rideout, William M., Jr. , et a1. , August 1969. Survey of Education in the Democratic Republic of the Congo. Washington, D.C.: Overseas Liaison Committee, American Council on Education, 102 p. , 1974. "The Reorganization of Higher Education in Zaire," OLC Paper No. 5, Overseas Liaison Committee, American Council on Education, Washington, D.C. Rogers, Everett M., 1962. Diffusion of Innovations. New York: The Free Press. , in association with Lynne Svenning, 1969. Modernization Among Peasants: The Impact of Communication. New York: Holt, Rinehart and Winston, Inc. Roux, Jean-Baptiste, January-February 1974. "Vingt-cing Ans de Selection Variétale du Cotonnier en Afrique." Paris: I.R.C.T. , Technigges et DévelOppement, No. 11, 24—30. Rukandema, F. M. "Prices and Cotton Supply Analysis in Uganda." Ithaca: Cornell University, Department of Agricultural Economics, Cornell International Agriculture Mimeograph No. 43. Saidi-Mokuba, Andjoipa, 1972. "Effets Escomptés de l'Industrie Textile a Kisangani—Gas de la SOI'EXKI." Kinshasa: UNAZA, Faculté des Sciences Economques, unpublished thesis. Shaffer, James Duncan, March 30, 1970. "On the Concept of Subsector Studies" University of Florida, Department of Agricultural Economics, Technical Seminar on Subsector Modeling of Food and Agricultural Industries, unpublished document, 22 p. Shapiro, Kenneth H. , 1973. "Efficiency and Modernisation in African Agriculture: A Case Study in Geita District, Tanzania," Stanford University, unpublished Ph.D. dissertation. Sinclair, John L. , 1968. The Production, Marketing and Consumption of Cotton. New York: Praeger. Sladden, March 1949. "Note Sladden sur le Coton: Examen des Probléres Posés par 1' Economie Cotonniere, " unpublished document . 362 Spencer, Dunstan S. C. , 1972. "Micro-level Farm Managerent and Produc- tion Economics Research Among Traditional African Farmers: lessons from Sierra Ieone," AREP No. 3, African Rural Employment Paper, Department of Agricultural Economics, Michigan State University, East Lansing, Michigan. Spiegel, Murray R., 1961. Schaum's Outline of Theogy and Problems of Statistics. McGraweHill, Schaum's Outline Series. Staner, P. , June 1955. "Ies Paysannats Indigénes du Congo Belge et du Ruanda-Urundi. " Bruxelles: Bulletin Agricole du Congo Beige. Tollens, Eric F. , l974a. "L'Economie de la Culture des Fibres Textiles au Zaire." Kinshasa: I.R.E.S., Séminaire sur le Développement Agricole au Zaire. , September l974b. "Problems of Micro-Economic Data Collection on Farms in Northern Zaire. " Paper prepared for the AIXZ/CMWT/ Ford Foundation Workshop on Field Collection of Socio-Economic Data in Developipg Countries, Beirut, December 8-14, 1974. , 1975. "An Analysis of Research on Agricultural Economics in tie Republic of Zaire," African Rural Employment Paper (forth- coming), Department of Agricultural Economics, Michigan State University, East Lansing, Michigan. , and Mulumba Kamuanga, 1975. "Essai d'Analyse du Marché du Mars au Zaire," to be published in Annales de la Faculté d'Agronomie, Kisangani: UNAZA, Centre de Yangambi, Volure II. Tondeur, G. , 1947. "Utilisation de la Farine de Graines de Coton dans l'Alimentation Humaine au Congo Belge. " Bruxelles: Bulletin Agricole du Congo Belge, 38:1. , 1957. L'Agriculture Nomade au Congo Belge. Bruxelles: Ministére des Colonies. Tshibaka, Tshjkaba Bulalu, July 1974. "L'Intensification de l'Agricul- ture Traditionnelle Zairoise; la Contribution des Engrais Chimiques. " Kisangani: UNAZA, Yangambi, Faculté des Sciences Agronomiqnes, unpublislned thesis . Tshibanda-Mihuta, Mukubi, 1973. "L'CNAFITEX et la Prorotion des Cultures de Fibres Textiles dans le Développement du Milieu Rural au Zaire." Kinshasa: UNAZA, Faculté des Sciences Ecormuques, unpublished thesis. Upton, Martin, 1973. Farm Management in Africa: The Principles of Production and Planning. Teflon: Oxford Unnversrty Press . Van den Abeele, M. and R. Vandenput, 1956. Ies Principales Cultures du Congo Belge. Bruxelles: Ministére des Colonies, 3rd Edition. 363 Van De Velde, Gérard, 1969. "Cours de Teclnnologie." Kinshasa: Université lovanium, Faculté des Sciences Economiques et Comerciales . , March 1974. "Ies Transports des Produits Agricoles au Zaire." Kinshasa: I.R.E.S., Séminaire sur le Dévelth Agricole au Zaire. Van de Walle, B. , 1960. Essai d'une Planification de l'Economie Agpicole Congolaise. Bruxelles: I.N.E.A.C. , Série Technique Nr. 61. Vansina, Jan, 1965. Introduction :5 l'Ethnpgraphie du Congo. Kinshasa: Editions Universitaires du Congo. Vasos, Georges, 1973. "L'Action de la Commission Agricole du Kasai Oriental et ses Effets sur 1e Revenu et l'Emploi en Milieu Rur ." Kinshasa: I.R.E.S., Cahiers Economiques et Sociaux, 11:3-4, 37—53. Verhaegen, Guy, June 1967. "Rationalité Economique et Agriculture Traditionnelle." Kinshasa: I.R.E.S., Cahiers Economiques et Sociaux, 5:2, 189-201. , March 1968. "Le Paysan Africain, 'Homre Traditionnel' on 'Home Economique'?" Kinshasa: I.R.E.S., Cahiers Economiques et Sociaux, 6:1, 100-127. , March 1970a. "Note sur l'Organisation du Credit Agricole en R.D.C." Kinshasa: O.N.R.D. , Section de Recherches Economiques et Financieres. , 1970b. "Pour une Stratégie Nationale des Engrais." Kinshasa: I.R.E.S., Cahiers Economiques et Sociaux, 8:1, 133-180. Viokolo-Burba, B. , 1970. "La Contribution de l'Industrie Textile Congolaise a l'Economie Nationale: Periode 1958-1968." Kinshasa: Université lovanium, Faculté des Sciences Economiques, unpublished thesis. Wenbi, Antoine, 1966. La Securité Sociale au Conga. Kinshasa: UNAZA, I.R.E.S. Widstrand, Carl Gesta, 1970. Cooperatives and Rural Developrent in East Africa. Uppsala: The Scandinavian Institute of African Studies, New York: Africana Publishing Corporation. Yang, W. Y. , 1965. "Methods of Farm Management Investigations," Agricultural Development Paper No. 80 , Food and Agriculture Organization of the United Nations, Rote, Italy. 364 Young, Crawford, 1965. Politics in the Congo. Princeton: Princeton University Press. Zaire, October 23, 1972. "La Vérité sur 1e Coton: les Tribulations de l'ONAFITEX." Zaire Magazine, No. 220, 20-22. , January 6, 1975. "La Deuxiéme Revolution Mobutienne: CoTpte Rendu de la Reunion du Bureau Politique du 28 au 30 Décembre 1974." Zaire Magazine, No. 335. Zarkovich, S. S., 1966. Quality of Statistical Data. Rome: Food and Agriculture Organization of the United Nations.