’ ImamWItwimumiwuwl MSU 104 RETURNING MATERIALS: Place 1n book drop to remove this checkout from LIBRARIES ‘ _ , ”- your record. HNES w111 be charged if Book is returned after the date Stamped be1ow. U " i ‘5 / A ‘ I It" 9 mm. Lg‘ fijfibeAiVN-V’ EJ5N‘.’ fl”! ‘4 AN ANALYSIS OF SMALLHOLDER RAINFED CROP PRODUCTION SYSTEMS: A CASE STUDY OF THE NUBA MOUNTAINS AREA, WESTERN SUDAN By Gaafar Bashir Mohammed A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultural Economics 1982 ABSTRACT AN ANALYSIS OF SMALLHOLDER RAINFED CROP PRODUCTION SYSTEMS: A CASE STUDY OF THE NUBA MOUNTAINS AREA, WESTERN SUDAN By Gaafar Bashir Mohammed The focus of this study is on rainfed crop production systems in the Nuba Mountains area of Western Sudan. The two smallholder farming systems in the area--traditional farming and the Nuba Mountains Agricul- tural Production Corporation (NMAPC) Modernization schemes--were consider- ed. The objectives of the study were: to identify the present input- output relations and constraints of the two smallholder production sys- tems; and to assess the impact of policies and management alternatives aimed at improving performance in the two production systems. The general research approach employed representative models to fo- cus on the production system at the farm level. Primary data were gener- ated from two field surveys carried out in the study area. The FAO sur- vey (l978/79) data were combined with data from the researcher's survey (1979/80) to provide a descriptive analysis of the smallholders' environ- ment and production practices. Building on this foundation, the approach utilized descriptive statistics to derive three representative farm pro- duction categories mainly on the basis of farm level resource differences. Gaafar Bashir Mohammed These three categories were then represented as sub-models in each of the two linear programming production models (traditional and NMAPC) that were constructed. In the LP models, the objective function was to maximize net farm income subject to satisfying the minimum consumption requirements of the farming household. To account for the seasonality of production, the activities and resources of the LP models were disaggregated by monthly periods. The traditional and NMAPC LP models differ from each other in activities, constraints, and input—output coefficients. In particular, the NMAPC LP model incorporated the mechanized cultivation activities and the institutional features (tenancy size and crop mix) that were introduced by the NMAPC program. The analysis and results of the base production plans of the two LP models were used as departure points for later model experiments. Analysis of the traditional farm model showed the cropping pattern of the smallholder to be dominated by sorghum. Net returns are very low, a product of the low productivity of land and labor. Low productivity_ and returns are the result of low crop yields and seasonal labor con- straints. Further experiments revealed that: (l) farming returns are highly sensitive to crop yield levels; (2) short-term credit can help smallholders augment their labor resources and expand the area cultivated, resulting in substantial improvement in returns; (3) smallholders cur- rently grow late planted crops to smooth out labor bottlenecks, but through the use of credit for hired labor they can plant earlier and Gaafar Bashir Mohammed realize higher crop yields; and (4) the current cotton prices would need to be raised substantially to induce traditional farmers to grow cotton. Analysis of the NMAPC farm model showed that NMAPC participants earn low returns from their scheme plots, due to the relatively small cultivation size and the low productivities of the two crops (cotton and dura) grown. Cotton is especially unprofitable. Several other findings emerged. First, the contemplated expansion of NMAPC tenancy size can be expected to increase participants' returns significantly. However, the increase in cultivation size can also be expected to intensify the labor and operating capital bottlenecks. Second, despite the effect of mechani- cal ploughing in reducing weeds, the NMAPC participants (especially under the expanded tenancy size) face a labor constraint in weeding. Third, credit is especially needed to finance cotton picking operations. Fourth, for the NMAPC participants, an unrestricted crop mix which includes the introduction of sesame as a third crop is advantageous and more rewarding than the present standard two-crop mix. The policy implications of the study indicate a need for applied research to improve smallholders' farming; a need to reduce costs of NMAPC and improve its services; and a need for changing the NMAPC's pre- sent fixed tenancy size and crop mix policies. Dedicated with gratitude and appreciation to my dear parents ACKNOWLEDGMENTS The author wishes to thank Professor Vernon Sorenson, my major pro- fessor. To him I am indebted for much direction, help and encouragement. I am also indebted to Dr. Eric Crawford, my research supervisor, who has shown deep interest and given valuable insights, direction and encourage- ment, especially when it was most needed. I also greatly appreciate the help and encouragement given by Professor A. Allan Schmid who acted as my major professor earlier in my program. I greatly appreciate the help and encouragement by Professors Carl K. Eicher and Carl Liedholm as mem- bers of my guidance committee, throughout my program. I wish to express my appreciation to all who have contributed, in many ways, to make my study program and this work possible. My special appreciation is extended to the African-American Institute for financial support and for funding of this research; the Government of Sudan for financial support; Dr. Carl Gotsch for recommending me to this department; the Ford Foundation Khartoum Office for financial assistance with my field work; the Department of Rural Economy, University of Khartoum, for help and assistance with my field work; the FAQ for providing some of my research data; the officials of the Nuba Mountains Agricultural Produc- tion Corporation, especially Sayed/Farah Mustafa, for much help and assis- tance during my field work; the farmers in the Nuba Mountains for their invaluable help and cooperation during my field work; Fatehi Bashir, ii Ibrahim A. Rahman, and Ibrahim Sharif for help in collecting the field data; my friend and colleague Mohamed A. 0. Obnouf for help and encourage- ment; the staff of the Agricultural Economics Programming Unit, Michigan State University, for help and assistance in computer work; Debbie Greer for ably typing and assembling the final manuscript. I am deeply grateful for the support, patience and understanding of my wife Nour. I am also greatly indebted to all my family and friends for their invaluable support and encouragement. ' Above all, my special praise and thanks be to Allah "God," Cherisher and Sustainer of the Worlds, Most Gracious, Most Merciful, for his innu- merable bounties. TABLE OF CONTENTS Page LIST OF TABLES .......................... viii LIST OF FIGURES .......................... xii CHAPTER ' I. INTRODUCTION ....................... l ‘ Background: Traditional Agriculture in Sudan ...... l The Setting: Policy Issues and Programs ........ 4 . Objectives of the Study ................. 9 Organization of the Study ................ 10 II. RESEARCH METHODOLOGY ................... 12 The Study Area ..................... l2 Summary of the Research Approach ............ l4 The Analytic Approach .................. l5 Limitations of the Analytic Approach .......... l7 Sources of Data ..................... l9 FAO Survey ...................... 19 The Researcher's Survey ................ 20 The Survey ..................... 2l The Population ................... 2l Sampling Design and Method ............. 22 Sample Size ..................... 23 Data Collection ................... 24 Other Sources of Data ..... * ............ 24 III. CHARACTERISTICS OF SMALLHOLDER AGRICULTURAL PRODUCTION SYSTEMS IN HUBA MOUNTAINS AREA . . . .2. . .-— . , 26 Climatological Environment . . ............. 25 Boundary ....................... 25 Climate ........................ 2 Rainfall ....................... 27 Soils ......................... 3o Vegetation ...................... 32 Traditional Smallholder Agriculture ........... 32 Household Characteristics ............... 32 Household Head Characteristics ........... 33 iv Page CHAPTER III. (continued) Household Composition and Demographic Characteristics ......... . ......... 36 Household Grain Consumption ............. 40 Land Ownership and Cultivation Characteristics . . . . 43 Land Ownership ................... 43 Cultivated Land ................... 44 Agricultural Technology and Crop Practices ...... 50 Farm Capital .................... 50 Rotations and.Dominant Crops ............ 54 Agricultural Practices ............... 55 Crop Varieties . . . .2 .............. 55 Agricultural Operations .......... . . . . 58 Agricultural Operation's Calendar ......... 6l Crop Yield Levels ................. 64 Farm Labor and Organization ............. 64 Family Labor ................... 66 Hired Labor .................... 66 Nafir Labor .................... 67 Prices and Market Organization ............ 7l- Contrasting Features of NMAPC Agriculture ........ 75 Land Ownership .................... 75 Rotations and Crops .................. 76 Agricultural Services ................. 78 Mechanized Cultivation ............... 78 Provision of Seeds ................. 80 Cotton Pests Control ................ 80 Tenant's Credit ................... 8l IV. REPRESENTATIVE PRODUCTION MODELS AND THEIR LINEAR PROGRAMMING STRUCTURE .................. 82 Smallholder Farming Systems and the Representative Production Models ........... 82 Structure of the LP Models ............... 895 Introduction ..................... 89 The Objective Function ................ 9l The Activity Set ................... 96 Crop Production Activities ............. 96 Labor Hiring Activities ............... l06 Dura Buying and Consumption Activities ....... l09 Selling Activities ................. llO Transfer Activities ................. llO The Constraint Set , , ................ lll Farm Level Resources . --------------- 1‘1 Land Restrictions. . . -------------- ‘11 Labor Restrictions . . . . ------------ 112 Operating Capital Restrictions .......... ll7 CHAPTER IV. (continued) Minimum Dura Consumption ............... Agricultural Operations Balance Constraints . . . Non- -Negativity Constraints .............. Experiments and Changes Made in the Basic LP Model ........................ NMAPC Model ...................... Credit Experiments .................. Planting Time Experiments ............... V. TRADITIONAL SMALLHOLDER PRODUCTION: A LINEAR PROGRAMMING ANALYSIS .................. Basic Solutions and Optimal Production Plans for The Traditional Smallholder Categories ....... . . Cropping Pattern ................... Resource Use ........ . ............. Returns and Average Productivities .......... Seasonal Constraints and Marginal Productivities . . . Rainfall-Induced Yield Variability .......... Credit and Land Expansion Experiments: Traditional Model .......................... Cropping Pattern ........ * ........... Resource Use ..................... Effects on Seasonal Constraints ............ Returns with Credit .................. Smallholder Credit Situation and Possibilities . . . . Planting-Time Experiments and Model Results ....... The Experiment .................... Results of LP Analysis ................ Optimal Production Plans .............. Effects on Seasonal Constraints ........... Credit with Planting- -Time Experiment Results . . . . Cotton Price Variation Experiment ............ Background on Cotton Problems In the Area. . . . Experiment Results .................. Summary ......................... VI. NMAPC AND THE MODERNIZATION OF TRADITIONAL AGRICULTURE IN THE NUBA MOUNTAINS .................. The NMAPC: Background, Present Organization and Current Performance Record ............ Background ..................... The Present Organization and Production Record of The NMAPC ....................... Production Relations . ... ............. The Modernization Program .............. vi Page 118 118 119 119 119 121 122 125 125 128 129 133 138 143 143 145 146 148 148 152 152 155 155 158 158 161 161 164 168 173 173 173 CHAPTER VI. (continued) Mechanized Cultivation in the NMAPC .......... 182 NMAPC Joint Account and Cotton Pricing Policy ..... 187 The Joint Account System .............. 187 NMAPC Cotton Price Policy .............. 190 Concluding Remarks ................. . 194 Analysis of the Current NMAPC Farm Model ........ 195 Optimum Production Plans and Net Returns ....... 195 Seasonal Constraints Under the Current NMAPC Model . . 201 Cotton Price Response Analysis in the NMAPC Model. . . 204 Analysis of the Future Full-Scale NMAPC Model ...... 207 The Future Full-Scale NMAPC Model ........... 207 Optimum Production Plans and Net Returns ....... 209 Seasonal Constraints Under the Full-Scale NMAPC Model ........................ 212 An Alternative to the Full -Scale Model ........ 214 Summary ......................... 218 VII. SUMMARY, POLICY IMPLICATIONS, LIMITATIONS AND SUGGESTIONS’ FOR FUTURE RESEARCH .................. 224 Summary ......................... 224 Policy Implications ................... 230 Limitations and Suggestions for Future Research ..... 232 APPENDICES I. Maps of Sudan and South Kordofan Province ......... 235 II. Chapter V's Planting Time Experiments: Research Experiments and Data .................. 237 III. Article 10 of the NMAPC Basic Charter: The Joint Account System ..................... 249 IV. Foreign-Aid Research Projects in the Nuba Mountains: Objectives and Relation to Smallholder and NMAPC Agriculture ....................... 254 BIBLIOGRAPHY ........................... 260 vii LIST OF TABLES Table Page 1.1 Share of Different Production Sectors in Area and Production of Major Crops ................ 2 1.2 NMAPC Six-Year Plan Proposal for Modernization of Agriculture: Target Area, Phasing of Capital Equipment, and Total Costs, 1977/78 to 1982/83 ..... 5 2.1 Smallholders Sample Distribution by Farm Type and Region: FAO Survey, 1979 ................ 20 2.2 I Smallholders Sample Distribution by Farm Type and Region: Researcher's Survey, 1980 ........... 23 3.1 Off- Farm Occupations: Distribution by Status and Season ......................... 35 3.2 Average Household Size and Composition ......... 37 3.3 Household Demographic Characteristics .......... 38 3.4 Household Annual Dura Consumption and Initially Stored Quantities .................... 42 3.5 Average Area of Jubraka (House Plot) for the Different Regions .................... 45 3.6 Size Distribution of Plots Cultivated by the Smallholders in the Area: 1979/80 Season ........ 47 3.7 Number and Percent of Farmers with Multiple Plot Cultivation by Region .................. 48 3.8 Distance of Cultivated Plots from Villages for’the Different Regions .................... 49 3.9 Average Expenses Per Feddan for the Different ' Crops by Operation ................... 53 3.10 Crop Area in Feddan and Average Production by Region (1974/75-1978/79) ................ 56 viii Table Page 3.11 Average Labor Requirements Per Feddan by Crop Operation ........................ 62 3.12 Agricultural Calendar Time of Operations by Crop . . . . 63 3.13 Average Yield Levels for the Traditional . Smallholders by Region ................. 65 3.14 Nage Rates Per Man-Day by Region, Crop and Operation ........................ 68 3.15 Labor Use Distribution of Type of Labor, Crop and Operation ...................... 70 3.16 Average Prices Received by the Smallholders in ' the Area for the 1979/80 Season ............. 76 3.17 Size Distribution of Land Cultivated by NMAPC Tenants: 1979/80 Season ................ 77 3.18 Average Yield Per Feddan for NMAPC Participants: . 1979/80 Season ..................... 79 4.1 Production Resources and Characteristics: Distribution by Farm Size Category, for Farms Surveyed ........ 86 4.2 Selected Household Characteristics: Distribution by Farm Size Category, for Farm Surveyed .......... 88 4.3 Distribution of Cultivation Operations by Crop and Model Period ...................... 92 4.4 Per Feddan Crop Budgets for Traditional and NMAPC Systems ......................... 100 4.5 Cotton Production Activities .............. 101 4.6 Dura Production Activities ............... 102 4.7 Groundnut Production Activities ............ 103 4.8 Sesame Production Activities . . .- ..... . . . . . . 104 4.9 Comparison of Per Feddan Labor Coefficients from Three Sources ...................... 107 4.10 Labor Activities .................... 108 ix Table Page 4.11 Land Restrictions Limits in the Basic LP Models ..... 112 4.12 Monthly Distribution of Family Labor Use by Production Category (in Man-Days): Sample Averages ........ 113 4.13 Family Labor Supply (in Man-Days) in the LP Models: by Production Category and Model Period ......... 116 4.14 Nafir Labor Availability (in Man-Days) in the LP Models: by Production Category and Model Period . . . . 117 4.15 Agricultural Operations Calendar in the NMAPC LP Models ......................... 120 4.16 Labor and Yield Coefficients in the NMAPC LP Models. . . 121 4.17 Borrowing and Capital Transfer Activities.’ ..... . . 123 5.1 Basic Optimal Production for the Three Categories of Traditional Smallholders . . . . 126 5.2 Shadow Prices for the Limiting Resources: Basic Production Plans of Traditional Smallholders ...... 134 5.3 Optimal Production Plans for the Three Categories of Traditional Smallholders: Under Twenty-Five Percent Yield Reduction ................. 140 5.4 Optimal Production Plans for the Three Categories of Traditional Smallholders: Under Fifty Percent Yield Reduction ..................... 141 5.5 Optimal Production Plans for the Credit and Land Expansion Experiment: Traditional Model ........ 144 5.6 Shadow Prices for the Limiting Resources: Credit and Land Expansion Experiment, Traditional Model . . . . 147 5.7 Expected Yield of Crops by Time of Planting ....... 154 5.8 Optimal Solution of the Traditional Model in the Planting Time Experiment ................ 156 5.9 Shadow Prices for the Limiting Resources: The Planting Time Experiment ........... . . . . . 159 5.10 Optimum Production Plans: The Planting Time Experiment with Credit Option .............. 160 Table 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 .6.9 6.10 NMAPC: Areas of Cotton and Dura (in Feddans) Planted in the Modernization Schemes, 1970/71 to 1976/77 ...................... Lagaw Station, NMAPC: Total Plough Area of the Modernization Schemes, Season 1979/8O ......... NMAPC: Joint Account Expenses Based on Expected Production for the Season 1979/8O ........... NMAPC: Determination of Farm Cotton Price ...... NMAPC: Total Cotton Revenue, Joint Account, and Farmers' Prices, 1976/77 to 1978/79 .......... Optimal Production Plans for NMAPC Participants Farm Model ...................... Shadow Prices for the Limiting ResoUrces: Optimal Production Plans for the Current NMAPC Farm Model. . . . Optimal Production Plans for NMAPC Future Full- Scale Model. ..................... Shadow Prices for the Limiting Resources in the Full-Scale NMAPC Model ................ Optimal Production Plans for the NMAPC Full-Scale Model with Unrestricted Crop Mix ........... xi Page 181 184 188 192 193 196 202 210 213 216 Figure 2.1 3.1 4.1 4.2 5.1 6.1 6.2 LIST OF FIGURES Diagramatic Representation of the Farm Production Model ........ , ................. Rainfall Monthly Distribution for Four Stations in the Nuba Mountains Area ................. Shematic Representation of the Basic LP Model ...... Graphical Representation of Average Observed Family Labor Use .................... Cotton Area Response to Cotton Price Changes: Traditional Farm Model ................. Main Managerial Structure of the NMAPC ......... Cotton Area Response to Cotton Price Changes: NMAPC Farm Model .................... xii 29 90 115 166 178 206 CHAPTER I INTRODUCTION Background: Traditional Agriculture in Sudan Agriculture in the Sudan contributes nearly 40 percent of the Gross Domestic Product, and 80 percent of the population depends on its sub- sistence on agriculture and related activities. The sector is the major source of exportable commodities accounting for over 90 percent of the ‘ country's foreign exchange earnings. Economic activities of other sec- tors in the economy, especially in transportation and industry, are cri- tically linked with those of agriculture.1 ’ The Sudanese record of agricultural development has been character- ized by a marked dualism between relatively high income irrigated and mechanized agriculture on the one hand and low income traditional agri- culture and livestock raising on the other hand. An evolving consequence of this dualism is the creation of an unbalanced regional growth, with 2 3 its related social and political problems. The place of traditional agriculture, and its relative share in area and-production of major crops in Sudan is shown in Table 1.1. 1Ministry of National Planning, "The Six-Year Plan for Social and Economic Development, 1977/78-1982/83,“ Sudan, Vol. II [44]. 2ILO/UNDP Employment Mission, "Growth, Employment and Equity: A Comprehensive Strategy fOr the Sudan" (Geneva: ILO, 1976) [30]. 3The term "traditional" is used within Sudan a riculture to denote the sector of small producers (mostly under rainfed outside the domain of the "modern" (irrigated and/or mechanized) sector. In the context of this study, the term will be used more specifically to distinguish the small producers in the Nuba Mountains area from the Nuba Mountains Agri- cultural Production Corporation (NMAPC) schemes' participants. 1 Table 1.1 Share of Different Production Sectors in Area and Production of Major Crops 3 Year Average 3 Year Average 1966/67-1968169 1973/74-1975/76 Area Production Area Production Production Sector (z) (%) (z) (%) Irrigated 22.4 53.8 18.5 50.3 Unirrigated 77.6 46.2 81.5 49.7 Public 27.6 54.9 22.3 51.6 Private 72.4 45.1 77.7 48.4 Mechanized 47.2 69.2 45.6 71.2 Traditional 52.8 30.8 54.4 28.8 Source: Six-Year Plan, Table 6 [44, p. 20]. More than 50 percent of the total cropped area is under traditional agriculture, and 6.6 million persons derive their livelihood from crop production in the sector [1]. Income and productivity of traditional farmers are characteristically low. Adam and Khidir report that "the average per capita income of traditional agriculture is about one-third of the level of per capita income in modern agriculture and is only about one-fifth of the aggregate average" [1, p. 3]. Not only is pro- ductivity of traditional agriculture low, but more seriously, it has exhibited declining trends in recent years. The current six-year plan notes that "during the last decade certain important structural changes have come about in the crop production subsector..., the most notable change is in productivity of the traditional sector, which went down in relation to the mechanized sector" [44, p. 20]. The need for developing traditional agriculture was indicated as early as the 1960s (Osman [51]), and by many others since then. The ILO in its recent report (1977) argues for this need as a first priority and it emphasizes that "development of traditional agriculture and animal husbandry is vital. This is rooted in sound efficiency criteria and is also a priority on equity grounds" [30, p. 53]. Adams and Howel [3] have questioned the priority issue, and in particular have cautioned against ILO's overoptimism for development of traditional agriculture: The western Sudan and traditional agriculture have been neglected not simply because of the determination of govern- ments to promote the modern sector within easy reach of Khartoum, but because of inherent difficulties in doing something ef- fective in areas of low fertility, meager and uncertain rain- fall, scattered population, nomadism and shifting agriculture [3, p. 508]. They, nevertheless, agree with the ILO in the need for a comprehensive "all-or-nothing" approach for developing traditional agriculture, to take the form of integrated rural development programs.1 For the government, however, the approach and means for developing traditional agriculture, were seen primarily as an extension of the “modern" sector's approach and programs. The accumulated experiences of agricultural development in Sudan have been the creation and development of an institutional and organiZational system and expertise that is re- latively effective in carrying out and executing prdgrams and projects in the modern subsector (large-scale irrigation and mechanization pro- jects). In particular, these developments in agriculture followed close- ly the‘original Gezira "model."2 This fact has important implications for the development of tradi- tional agriculture in general, and for this study in particular. For, as will be discussed below, it is against this background that the Sudanese planning machinery has framed and launched programs for the "modernization" of traditional agriculture. The Setting: Policy Issues and Programs The current Sudanese six-year plan (1977/83) for social and econ- omic development gives explicit recognition, among the stated objectives for agriculture, to the "development and modernization of traditional farming, improvement of conditions for nomads, and the modernization of 1Examples of programs endorsed by ILO [30], and Adams and Howel [3], include: Hunting Technical Services' Southern Darfour, Savanna, and Jebel Marra Development Plans, for western Sudan. 2Large-scale irrigated agriculture was started by the Gezira scheme (1925). The "model" was closely replicated in subsequent developments such as Managil Extension (1956), Guneid (1967), New Halfa (1964), Suki (1971) and Rahad (1977). All of the schemes follow, more or less, the original Gezira in.the design of their organizational set-ups, tenant- management product1on relat1ons and cropping patterns. pastorial activities" [44, p. 6]. The strategy adopted in the plan for modernization of traditional agriculture includes: (i) Consolidating the studies and researches already done or underway, to determine suitable projects for mechan- ization of traditional agriculture. (ii) Establishing agriculture complexes and a network of research stations in all rainfed crop areas. (iii) Establishing modern ranches in savanna region. (iv) Encouraging and assisting the establishment of large agricultural cooperatives. (v) Encouraging the development of close relations between modern agricultural schemes (like Rahad) and the neighbor- ing traditional agriculture areas, so that the latter will benefit from the production systems used in these schemes [44, p. 11]. A major program that was implemented in the spirit of the above strategy is the modernization schemes of the Nuba Mountains Agricultural Production Corporation (NMAPC).1 Started in the early 1970s in the Nuba Mountains area of South Kordofan Province, the program was directed towards modernizing traditional agriculture in the area. Taking es- sentially the public irrigated model format, the main component of the modernization program is provision of mechanized cultivation to the smallholder participants. The NMAPC's six-year plan proposal for the modernization program is shown in Table 1.2. The biggest item in the proposed capital costs (66.7 percent) was fOr agricultural machinery. Sixty tractors/discs were to be added to the corporation's stock an- nually, to reach a total capacity of 360 tractor/discs by the end of the plan period in 1982/83, at which time a total of 480,000 feddans were expected to be under cultivation. Although mechanized cultivation was being extended to new NMAPC schemes, the realized expansion of the 1A brief history and background note on the NMAPC, together with the specific modernization objectives, are discussed in the first part of Chapter VI. .noa-n cc. no.c-—au vet .ouemeoa¢.-I sea s-gaos .no—u_so. .nsoauosu oc.a-coao so «anon .oaa.o:—u . noun. ono.p - sauna. econ .ac.~n - .4 occ- umtzz "oucaaw e.oo..~ . ~.~a... ¢.ae~._ .._a=._ o.no.._ ~.a~. a..oo.. .oco.ms. page S co_uuoaocq ace—m uo—u.go» aeguaco: o._nox page. acoxcap Mn..om Na. as. . ne=-xu.. “menu 5.». usage “Lanna-c» «Lop—ag— .33suao8 pono.v .— cam.— oc.~:oa ._au occ.c A_o:.v "caucu— s acmxacnm nun.s _o>u— ou.a .,acu—:aeu macaque». muscle—au— _os=u_=u.ea¢ ace.oo. coo.cu‘ coo.ao. eoe.a~n. ace.°.~ eoo.c.. ooe.aa a..=.ee~.. ..c¢ ashore 'm'——— I "Evan—No— ”MVP—NI ”MC—n.1,”.- va—o-I-U Mm'QNN—u— o.n~m M NO 0—. N9. IN—l Nina-o- '— 0-0 tan—o D O I NNN I NNN INNN‘ INF-N 0 0-0—0- c-o-o-o- §. a. .2 e'eg '88 8'88 s'es 8°88 can :0 cu ..ooa.m.. . “.99 _.bo_ ...ep na\~ea. ~¢\.ea_ .o\a¢a. ea\.~._ osxoeo. ohxeso. mm\~mm— o» m~\-m— .mumou pouch new .acosnvscu _~a_amu mo mcpmmca mms< “awash."msau—zuvsm< mo copumuvcsocoz soy penance; cap; som>1xvm ua.uuo~uuu 42.5 233 as... .395 o..- : 23 ~ 9.262.. .32. 9.593 3.- Ianom .5333 £33... "ova—u... 389. 16 ......... 4am------..-1%.38:-------::3--------1-1-1-»... m m d u H. mm mm w. w M m m w w n .aaaaaazu Millmm Mimi AI 1 All m AIL“ A] M Aim All...» A] .w TWW «28.33 , a um n 1 u z a u x . 22:3. . A #3.. 2328.43 .uuo .muo— .353 .3... \7 .533 , 53... i. . 39:: 3:296. LC... - 5.2.938 .98 .342 _ j. . W 22.-.: m f N a, 2233.. \/ rV _1a=39_..8x I E3§>wxouucounn 1 \ a 33-- 3553 231:9: .aaaa.,._-- m 1’ unauam «mam—=3 \ lWfi 8:8... — 17 through the monthly distribution and availability of labor and operating capital. 3. Subsistence dura (sorghum) consumption needs are represented by an explicit constraint in the model. 4. Farmers are assumed to be risk-averse. However, the only risk considered is the weather-induced effect on yield variability. The main representation of this issue is through the safety- first requirement implied by the dura consumption constraint. This strategy which is based on an empirical foundation is a realistic specification. In addition to the analysis of basic models of smallholder produc- tion, the LP models were also used in a number of experiments to analyze selected management and/or policy intervention alternatives.1 The de- tailed features of the basic LP models and these experiments are dis- cussed in Chapter IV. Limitations of the Analytic Approach In general the limitations of linear programming are related to the validity of the assumptions incorporated in the LP model.2 An important limitation of the standard LP model is that it does not include any 1With LP it is relatively easy to vary available prices and resources as well as input coefficients in order to simulate various management and/ or policy alternatives. 21n the standard LP model which maXimizes (minimizes) a linear-objec- tive-function subject to some linear constraints, several basic assumptions are made: (1) additivity and linearity of activities; (2) divisibility of activities and resources; (3)finiteness of alternative activities and resource restrictions; and (4) single-value expectations; i.e., resource supplies, input coefficients, and prices are known with certainty [27]. 18 allowance for risk, which is central to decision making among smallholder farmers [11, 53].. In this study, the objective function which maximizes net returns from crops subject to satisfying sorghum consumption require- ments, closely simulates the decision behavior of the smallholders in the area} In other words, the objective function maximizes net returns sub- ject to a safety-first constraint. This formulation has been found to be appropriate in a number of similar empirical studies [11, 53]. Another related issue concerns the type of technology that is re- presented by the LP assumptions.2 Given the small scale of traditional smallholders in the area and the dominance of hand-hoe cultivation, the linear and additive scale of production seems an appropriate representa- tion of the technology. In other words, the physical returns and re- sources required do change proportionally as the cultivation size is in- creased or decreased. Perhaps an important limitation, though not direct- ly related to the LP format, is the limited representation of the weather- induced yield variability in the model. This could have been improved by incorporation in the model of a simulation component.3 However, lack of data needed for such an approach made this alternative infeasible. Empha- sis was given instead to management factors, together with sensitivity analysis of yield levels in relation to variations in annual rainfall. 1The revenue maximization and security objectives of the smallholder production are discussed in somewhat more detail in Chapter IV. 2Namely the constancy and additivity in production technology. How- ever, the LP technique and its different versions offers some means of treating some of the nonlinear/nonconstant implied technologies. 3For an example of studies using such an approach see Crawford [18] and Lynam [35]. 19 Sources of Data Two primary sources of survey data are used in this study. One survey was conducted by the FAO in the 1978-79 season; another supplemen- tary survey was carried by the researcher during the 1979-80 season. The following is a detailed description of the two data sources: FAO Survgy The survey was carried out during February through March, 1979. Interviews were conducted by students of the Department of Rural Economy, University of Khartoum and members of the Statistics Department, Ministry of Agriculture, under the supervision of the FAO and staff of the Rural Economy Department. The survey obtained basic farm management data, using a recently developed FAO "Farm Management Data Collection and Analysis System" [23]. A single-visit survey method was used. The two main categories of data sought in the survey pertain to resource inventory data (farm labor, land, livestock, tools, etc.) and resource utilization data (input-output of labor, crops, livestock, etc.). The data on the latter category was dis- aggregated on monthly basis. The two main classes of farmers surveyed were: traditional smallholders and NMAPC schemes participants. The eight regions of the Nuba Mountains area were covered in the survey. The sample distribution of smallholders by farm type and region is shown in Table 2.1. 20 Table 2.1 Smallholders Sample Distribution by Farm Type and Region: FAO Survey, 1979 Traditional MNAPC Region Farmers. . .Participants . .Total. Kadugli 34 21 55 Lagawa 29 10 39 Talodi 18 ll 29 Kalogi 12 11 23 Abu Gebha 24 16 4O Abassya 4 14 18 Um Brembita 24 ll 35 Dilling 4 4 3 Total ,. . 149.. ‘ 98 , .247 The Researcher's Survey With the FAO survey data in hand,1 the researcher conducted another survey for the following purposes: 1. To obtain a first-hand familiarity with the area and its dif- ferent regions. 2. To collect additional data for the study. Emphasis was given to the following: a) Input-output relations for only the four main crops con- sidered in this study (i.e., sorghum, cotton, sesame, and groundnuts). b) Data pertaining to mechanization, available through regional offices of the NMAPC. 1Through his association with the Department of Economy, University of Khartoum, the researcher had access to this data before his own survey was carried. This enabled him to meet and discuss with some members of the survey team the nature of the information collected. This discussion was helpful for the researcher in designing and conducting his own sur- vey (i.e., sampling frameworks, locations, route of travel as well as other logistical matters). 21 c) Updated marketing information for the above four crops. d) Food and in particular grain (sorghum) consumption of house- hold units. The Survey Assisted by two interviewers (one university graduate and one high secondary school graduate), the researcher conducted the survey in the Nuba Mountains area from May 1, 1980 to July 1, 1980. The eight regions of the area were covered one at a time, in the following order: Abassya, Um Brembita, Abu Gebha, Kalogi, Talodi, Kadugli, Lagawa, and Dilling. Relevant information from government offices and departments (especially those of NMAPC) was obtained while at each region. The Population Administratively, the Nuba Mountains area is part of South Kordofan Province. The latter is composed of four administrative districts; Eastern, Western, Southern and Northern (see Map 2, Appendix I). The eight regions of the Nuba Mountains are distributed by district as follows: South Kordofan Province District . Eastern, . Western . Southern w Northern Abassya Lagawa Kalogi Dilling Um Brembita Talodi . Abu Gebha Kadugli The two sub-classes of smallholders of interest in this study are traditional farmers and NMAPC schemes participants. The exact number of' smallholders in each of the eight regions is not known. An estimate of 22 traditional farmers in the province according to the South Kordofan Farmers Union (SKFU) is 380,000. The Nuba Mountains Farmer's Union com- prises the majority of the SKFU's total members. The NMAPC schemes participants on the other hand are known and recorded for each scheme in the eight regional stations of the NMAPC. Sampling Design and Method Geographic location was used as the first stratifying variable. Samples of traditional and NMAPC participants were taken from each region separately. The sampling method was as follows. After discussion with the agricultural officers of the regional station, a purposive selection of one or two NMAPC schemes was made. Criteria for selection of these schemes included factors such as near- ness, conditions and access to the fields (the survey was at the begin- ning of the rainy season), and scheme crop rotation.1 A simple random sample was then taken from the list of participants in each scheme. For the traditional smallholders, a purposive selection of one or two villages was taken from each region. Criteria for selection of villages were similar to those of NMAPC schemes. After discussion with the village head a list of the farmers was prepared and a random sample drawn. Smallholders who held plots in the NMAPC schemes were not interviewed; only farmers with traditional plots were included. 1Biases resulting from this purposive selection were judged to be minimal. Diversity in agricultural and smallholders conditions with- in each region is generally small. This is especially true for the NMAPC schemes which are operated under many standard features and similar conditions (see discussion in Chapter VI). 23 Sample Size Statistical theory can help determine the desired sample size from a given population, based on information about the criterion variable used for selection. Ideally, the distribution of this variable in the population in question, together with desired levels of accuracy of re- sults and analysis should be known in advance. In most situations, as in this case, it is difficult to obtain this information in advance. Further, and most likely, a multiplicity of variables (household size, cropping pattern, domestic organization, etc.) rather than only one vari- able are of interest in investigating farming and production systems. In this study, practical considerations such as budget, time, conditions of the fields in the rainy season, and judgment, were the main factors that determined the sample size. The distribution of farmers by farm type and region in the final sample was as shown in Table 2.2. Table 2.2 Smallholders Sample Distribution by Farm Type and Region: Researcher's Survey, 1980 Traditional MNAPC Region Farmers Participants Total Abassya 7 l 8 Um Brembita 3 4 7 Abu Gebha 9 9 l8 Kalogi 10 10 20 Talodi 9 8 l7 Kadugli 36 31 67 Lagawa 9 10 17 Dilling 16 16 32 Tota1 99. 89 188 24 Data Collection Discussions with some of the members who participated in the previous FAO survey were helpful in developing questionnaires for this survey. Interviewers were introduced to the questionnaires first in Khartoum, where it was explained and discussed with them. Upon arrival in the field, during the first days at Abassya, questionnares were again tested and discussed with interviewers. Farmers were interviewed at their farms or in the villages. To facilitate communication, the purpose of the survey was first explained to the head of the village (Shiekh), when first preparing the list of farmers and choosing the sample. The Shiekh in turn introduced the in- terviewers to the farmers and briefly explained the purpose of the survey to them. More explanation were then made by the interviewers to each individual farmer before recording the information. For the NMAPC participants, the agricultural inspector of the sta- tion or the agricultural officer resident in the field (Khabir) took the role of the village head in explaining and introducing the interviewers. Other Sources of Data Besides the two primary sources of data just described, numerous other sources of data were used in this study. These include: 1. NMAPC official records at the regional field stations and head- quarters. 2. Agricultural Research Corporation ia) Kadugli Research Station b) Headquarters at Nad Madni 10. 25 Agricultural Bank of Sudan a) Dilling Branch Meteorology Department a) Headquarters at Khartoum Mechanized Farming Corporation a) Dilling Branch b) State Farm at Habila (Dilling area) South Kordofan Province, Commissioner's Office a) Agricultural Service Department b) Cooperative Department c) Planning and Development Department Proceedings and Discussions of NMAPC Norking Agricultural Con- ference (June, 1980). Informal interviews with officials at NMAPC and other government departments. Reports, working papers, and discussions with members of foreign research and development projects in the region. a) Hunting Technical Services Ltd. b) German Technical Aid Project c) The EEC (SATEC) project Other miscellaneous reports and documents. CHAPTER III CHARACTERISTICS OF SMALLHOLDER AGRICULTURAL PRODUCTION SYSTEMS IN NUBA MOUNTAINS AREA This chapter is intended to serve two primary purposes: 1. to give an understanding of the general context within which the farming units operate by describing the nature, amounts and variability of its resources, agricultural activities and production; and 2. to serve as a base for the design of the LP production models and the following LP analysis presented in Chapter V and VI. The first part describes the climatological environment (climate, rainfall, soils and vegetation) of the area. Next, both traditional and NMAPC agricultural production systems are discussed. Climatological Environment Boundary The Nuba Mountains area is a hilly area in the north central part of South Kordofan Province (see Map 1, Appendix 1). It is part of the central clay plains of the Sudan which extends from the east to the west of the country. Latitude 120 10' N forms the dividing line between the sandy steppes of North Kordofan and the clay plains of South Kordofan. The southern 26 27 boundary of the latter province is Bahr-el-Arab, which is at latitude 10° 25 n. Climate The area is located in the savanna belt of the northern hemisphere tropics, having a continental climate classified by Meigis [37] as "hot semi-arid." The main influence in the climate of the area is brought about by the migratory movement of the Intertropical Convergence Zone (ITCZ), [41]. The ITCZ moves from north to south and back again each year. This movement of the ITCZ is associated with a shift of wind direc- tion from north to south, carrying moist air over the area. This occurs around mid-April and brinQS'in the first erratic showers, signaling the beginning of the rainy season, which continues until the end of October. By the end of October the wind changes from a southerly to a norther- ly direction, bringing dry air to the area. This is the beginning of the dry period which extends fron November through mid-April, where again the area falls under the influence of the ITCZ and the cycle is repeated. Rainfall Amount and distribution of rainfall are the most important factors influencing economic activity and social life in the region. In particu- lar, important aspects of agricultural production (activities, operation timing, yields, etc.) are determined to a large extent by these factors. Although the rainy season extends from May until October, most of the rainfall occurs between July and September. The amount of rain in- creases southward. In the north of the region average rainfall is 500 mm, it increases to 800 mm in the Southern Jebels. Within the Nuba Mountains 28 area, there is less variation in amount of annual rainfall. Annual rain- fall data for four stations in the region are shown in Figure 3.1. The - stations: Dilling, Abassya, Lagawa, and Kadugli are located in the North, East, Nest, and South divisions respectively. Except for Kadugli (South), which has an average annual rainfall of 720 mm, the rest of the stations have an average rainfall slightly above 600 mm. A study (HTS [38]) into the variability of annual rainfall in the area showed that in the north, where the mean annual rainfall is 500 mm, it can be expected to be less than 365 mm in 20 percent of years, and less than 640 in 80 percent of the years. In the south where mean annual rainfall is 800 mm, annual rainfall of 695 mm or less can be expected in 20 percent of years, and annual rainfall of 895 mm or less in 80 percent of years. The study concluded that "there is no firm evidence of any long-term cyclicity in the rainfall fluctuations occurring in the savanna regions" [38, p. 23]. However, despite this relative stability of long-term and annual distribution of rainfall, the monthly rainfall distribution is highly variable. For example, HTS [38] estimated that for the Kadugli area, on average 11 percent of the annual rainfall falls in May (with coeffi- cient of variation (C.V.) of 0.75); 14 percent falls in June (C.V. 0.4), and 62.5 percent in the months of July, August and September (C.V. 0.35 - 0.4), and the remainder falls in October (C.V. = 0.5). It is worth mentioning here that these patterns of rainfall distri- bution and characteristics has important implications for crop yield levels and husbandry practices. First, although annual rainfall has a low probability of declining to levels that would result in crop failure, annual rainfall variations results in substantial variation in yields. 29 Esoueegx .ucoEuemaoo Amopoeompo: an uwuw>oem mama soot umuuzsumcou .mw..< 2.3.2.20: was: 2: .... meet—mum ..son. .8» 53323.5 355: 2353. Tm 9.sz .. x < .. a .. ////é - R 1 3 . 8 1S . . 2 1 m. - 8. 1 3. . a. r 2.. - ...... 1 S. 3. NS . .... . .... .... 2...; . .... 1m: .22- .3: 52-..... Au o— 0:“. .. ..ugod I .0— 0: Jo: u>mmaa.< Au .8 can .93; .I .~O cup 4.: 2:2:— USN ...... ..l. :35... :85... a e m ... .. .. .. < x . /é . a 1 ...... . m. 1 8 W, - 8. - 2 . ...: - 8. . S. 1 m2 . ...: 1 S. . 8~ . .... . I M .8N Iti‘ . . e #37:... w an Jenna..." ..l. C .2 as .23 ... .8 a: J... :93. :nu"... a .3 on $9.... ... .3 o: ...: 2.3: .35... 30 Second, aside from these between-years variations, there is high vari- ability in the monthly rainfall distributions. This greatly influences the pattern and timing of agricultural operations, especially those tak- ing place at the beginning of the growing season (i.e., land preparation and planting). Soils The Nuba Mountains are outcrops of resistant Basement Complex rocks, mainly granites, mica schists and quartzites. The topography varies from undulating to rugged rising to a height of a few hundred feed up to 3,000 feet above the plain. The jebels and associated foothills occupy 40 percent of the area. These jebels are separated by a series of gently undulating or almost flat intermontane plains occupying the remaining 60 percent of the area. The distribution of soils in the region is complex. The nature of these soils follow more or less their location with respect to the jebels, footslopes and planes.1 Six classes of soil types has been defined for the area (HTS [38], AHT [4l]), with the aim of considering their agricul- tural potential. All of these subsoils except the dark cracking clays has limiting physical and/or chemical properties that render them of low agricultural potential in crop farming.2 They are suited and used for 1This distribution, in a form of catenary sequence, determines their texture and clay content (for a diagramatic representation of this, see HTS [38], p. 16). 2The dark cracking clays are the Nb3 soils according to the HTS [38] classification; or the $2 soils according to AHT [4l], which follows the Sudanese suitability classification. 3l dewlings, limited grazing sites and to a lesser extent for fruit and vegetable production, the latter being in and around flood plains, which comprises 5 percent of the area. The soils with the greatest potential and used for crop production, are the dark cracking clays also kndwn as vertisols or black cotton soils. Covering 40 percent of the area and more than half of the intermontane plain, these soils occur mostly in the middle and lower slopes and in the valley bottoms. These soils support 80 percent of cropping in the region. Almost all the smallholder agriculture and NMAPC schemes are located in these soils. This class of soils with a high clay content (60 percent), generally has no limiting chemical properties (i.e., nonsaline and nonsodic), al- though nitrogen and phosphorous are low. The main difficulty in managing these soils is due to their physical structure. Quick ceiling of the cracks leads to run-off and serious erosion caused by additional accumu- lating water. The greatest difficulty however is attributed to the ex- tremes of consistency exhibited by these soils. They are very hard when dry and very plastic when wet. This has an important implication for land preparation and tillage operations, whether performed by man or machines and leaves a short time to perform tillage_under optimal soil condition. Under existing farming practiced in the «area. the fertility of these soils depends primarily on the management practices. In the present system no fertilizers are used, instead the land is cultivated for 2-4 years and then allowed to rest for roughly an equivalent period of time before it is brought back for cultivation again. 32 Vegetation Vegetation growth in the area follows the savanna pattern. These are mostly of the accaia species. AHT [41] contains a concise account of the exploitable vegetation resources in the Nuba Mountain area.I These resources also offer substantial off-season employment to farmers in the region. The most important of these: firewood and charcoal (mostly Accacia seyal), timber for building (various species including Boragrus aethiopiam (daleib), bark for ropes (Accacia Senegal, Adonsonia digitata as others), A. senegal is also used for gun, bamboo poles (Oxythenantera abyssinica, grass fencing (sheragnia) (namely Hyparrhenia spp) [38]. The vegetative growth of shrubs and trees also constitute an impor- tant investment by the farmer in the form of land clearance, when the land is brought into cultivation for the first time. Traditional Smallholder Agriculture Household Characteristics Household, for the purpose of this study, refers to the unit of fam- ily members who live and eat together. This unit includes a "household head" who is the decision maker in all aspects pertaining to the well being of the family. In particular the household head has the responsi- bility of making decisions concerning the agricultural production and its related activities. .ISee Ministry of Agriculture Food and Natural Resources, "Nuba Hountains Region, DeVelopment POtential Survey, Annex l-5," prepared by AGRAR-UHD HYDROTECHNIK ESSN. FRG, Sudan, l977, :41}. 33 Household Head Characteristics: 1. Age, All of the household heads interviewed in the researcher's survey (1979/80) were males,1 93 percent of whom were in the age bracket of l6-65 years. The remaining 7 percent were above 65 years old. Educational Achievement. Analysis of the educational achieve- ment of the household heads interviewed shows that 80 percent of them have no formal education. Eight percent had “Khalwa" education,2 and 12 percent had some primary school education. Such a distribution is typical of traditional farmers in the developing countries. Education, being an important form of investment in human capital, is thought to be an important fac- tor because it increases the individual's awareness of alterna- tives, facilitates learning and adoption of new ideas and in general tends to increase the productivity of human capital. In a number of farm and production studies, educational achieve- ment has been used as a proxy for managerial ability. Because of the difficulty of equating school years to managerial ability, such a formulation is likely to result in an insignificant co- 3 efficient. In this study, it is argued that production 1 Very few of the NMAPC tenancies were registered in the name of women, in which case they are either managed jointly or exclusively by the husband. ' 2 Khalwa is a l-3 year religious education in basic reading and writing. 3Other studies (e.g., Massel and Johnson [ 36 ]) using farmer's experience and skills, as proxies for management reported highly signi- ficant coefficients. 34 decisions are related not only to education, but rather to the resource endowment, constraints and opportunities that face the farmer. It is better to address the impact of these factors on managerial decisions directly. 3. Off-Farm Occupations. Table 3.1 shows the distribution of off- farm occupation by status and season for the household heads. The nature of the seasonal crop production in Nuba Mountains has two important implications. First, many persons (13 percent of those interviewed) whose main occupation is not agriculture,l are involved in agricultural production during the season. Second, and most importantly, is the fact that smallholders are involved in a number of off-farm activities when the growing season is over. Although the off-farm employment is concentrated in the off-season, a number of smallholder farmers (15 percent) work as hired laborers in agriculture during the growing season. This situation arises mostly because of cash requirements for consumption or production developing early in the season (early to mid-August). This issue is discussed later in conjunction with the credit situation. The type of off-farm/off-season employment in the region is mostly influenced by the savanna climate of the area. Cutting wood/hay, fencing, charcoal burning, honey collection, gum arabic collection,-rope marking, etc., are the most common activities. At present, very few farmers (7 percent) migrate from their localities to work as hired laborers in 1Those reporting a permanent job (mostly trade and government jobs) during the season (see Table 3.1). 35 .uuo .o:_am.u ..m: .uoxcus man :. Susana. ..osasaozm .m:.:can —uou.uzu .a=_ep—:n «mac: .=o_uuu__ou xacoz .ma seam u .comamm m:_xc.m nose 0:» acpgae no: so cucumczu .macoucoamo. uo: mamcoamog co comma .gueos a a. xuoz a cagu «mo. so m_mom —u:=_u:ou :_ Ho: 5. u .comaum ac—zocm couea so .ou uuuupgumm. »u_>—uuo N. 1 ..ou» apes: oz» a. c. ua>po>=¢ up 1 “edemELoa a “ma ewcau.m=ou maumumm .Acm\mNmpv name xm>czm ”ougzom N .p _ Nm N we o.mm umaoocoppmomvz .m N m c. m. m.o_ aoa u=w5:.m>ou .e N m N c.m Acm>_—ma swam: .m m a c. N... xa:\cooz m=_uu:u .N c m_ — m. o.N— munch .— .—ug=u_:u..mo:oz m . e ma .aezh .N m m ..N acueaamzz .p .xuoumm>_. N .p a. m.N_ .mco.u.uach .n m 2 MN To. “is: .N c _ p a c. m.m .ouoom1azm coupcusuuz .. "puczu—aupcm< :omumm camaom comoom :omoom :omaom comaom memo: memo: m:_c:o maven: uoz m=_.:o mcpcaa uoz omcvgao nmwgaa uoz . 1 _a:o_mauoc 1 _a=omamm acocascma u_ouwmwoz u_ouwm“oz .copunaauuo scum who n acopuammUoc 5o magnum commmm new magnum xn :owuzawcpm_o ”meowumazooo Ecmuuewo F.m «Pomp 36 the larger—scale mechanized farming of Mechanized Farming Corporation (MFC) in the North and Eastern parts of South Kordofan Province.1 Household Composition and Demographic Characteristics Table 3.2 shows the distribution of hoUsehold members in relation to the head. Although the average household head is shown to have more than one wife, this statistic is influenced by a few (23 percent) house- hold heads reporting more than one wife.2 The wife plays an important role in agricultural production, not only by helping the husband in the work of the agricultural plots, but also by undertaking the responsibility of cultivating the "Jubraka" (small home plot). Another important aspect in the role of the wife is that of her kinship relation in regard to the availability of the "Nafir" (communal or exchange labor).3 The demographic characteristics of the survey (l979/80) sample are shown in Table 3.3. The sex ratio indicates a marked dominance of males in the household.4 It should be noted that this statistic is influenced mainly by the dominance of sons over daughters in the household membership (see Table 3.2). 1Dilling, Habila and El-Beida mechanized farms. Affan [4] discusses some of the socioeconomic aspects of traditional farmers' employment in Habila mechanized areas. 2Those reporting two, three, or four wives were 20, 2, and 1 percent, respectively. 3 \. Especially among the Nuba ethnic groups [38]. 4This was also found to be true at the province (South Kordofan) level in Sudan's second population census (1977). 37 Table 3.2 Average Household Size and Composition # of ' ‘ Relation to HouSehold Head: Household Region Households Hife Son Daughter Relative Size Abassya 8 1.13 2.50 1.75 0.50 6.88 Um Brembeta 7 1.57 2.57 2.57 0.43 8.14 Abu Gebha 18 1.17 2.56 2.17 1.04 7.61 Kalogi 20 1.20 2.35 2.00 0.60 7.15 Talodi 17 1.29 2.18 2.17 0.35 6.94 Kadugli 65 1.23 2.39 2.20 0.35 7.17 Lagawa 19 1.16 1.68 1.00 1.00 5.86 Dilling 31 A 1.36 2.68 2.61 0.58 8.23 All Regions 188 1.23 2.33 A 2.07 0.52 7.18 Source: Computed from survey data (1979/80). 138 ..882828.. 8888 xo>288 582. wouaasou "mucaom 82.8 88.8 2..2 88.8 8..2 .8.2 8..8 28.8 82.8 2..e82 888.882 .2 .8 . .8+8.. 2.. 22. 2... 88.. 88.. 8.. 8... 82. 8.888 2888888888 .8 28. 88. 88. 88. 88. 28. 28. .8. 2 .8288» 88-8.. 28. 28 882 88 22 88 82 .8 2 88.88.8888 88.882 .8 88. 88. 88. 28. 88. 88. 8.. 28. 2 ......8 88 x. 8 8 .. 2 8 8 . 8 2 8888 888888888 .8 .8. 88. 28. 88. .8. 88. 88. .8. 2 .8288» v.1c. 88. 88 8.2 28 .8 88 82 .2 2 8..88 888888888 .8 .88. x .mwmw. 28. 28. .8. .8. 28. .8. 88. 88. x 88 88.8.8882. .2 .88. x.m«. .... 28.. .... 28.. 88.. 2.. 88. 88.. 8.888 888 .. 88....8 838888 ..88882 .88.8. .88.8. 88888 88¢ 88888888 8: 828888< 8.88.88.88.888 88.28.28888285u upcamgmoEma 8.888888: m.m 8.888 39 An interesting characteristic, as can be seen from the table, is the high proportion of dependent children (0-16 years old), which goes as high as 50 percent in Dilling region. This feature is more or leSs typical of developing country populations. An important implication of this feature is that it results in a high dependency ratio. This would be especially true if coupled with a high dependent-aged ratio. The latter ratio was found to be small in this case (only around 4 percent). Still the figures for the dependency ratio were markedly high for all regions. The high proportion of children also has a major influence on the remaining two important characteristics, i.e., the family size and the active population statistics. Active population ratio (16-65 years) or what is sometimes referred to as economically active population ratio, gains its importance from . the fact that it is commonly used to make inferences about the productive capacity of the household.1 In traditional agricultural settings, this is deficient in many aspects. Most important criticisms include such as: it includes individuals (such as women) who participate minimally in the economic activity(s) under consideration and others (such as students) who participate, if any, only on part time basis. Also it excludes individuals less than fifteen years (and less importantly those above sixty-five years) which is not uncommon in developing coun- tries to start working. Finally, this index (economic active population 1Sometimes used for a region or a country likewise, as was the case in Sudan's second population census (1977). 40 16-65) is deficient in that it does not give consideration to domestic organization and other labor recruiting institutions relevant to the household (e.g. hired and communal “Nafir" labor, which are discussed in detail later in this chapter). The alternative adopted in this study is to make use of the observed family labor utilization and distribution in the course of the analysis. The active population ratio in this case, and as mentioned above, is greatly influenced by the high dependent children ratio. The highest active population ratio was observed in Lagawa (65 percent) and lowest (42 percent) in Dilling region. We note that this is consistent with the inverse effect of dependent children; the latter ratio was lowest in Lagawa (30 percent) and highest in Dilling (51 percent). Lastly, the observed high proportion of dependent children, in- fluences in a straight forward fashion the overall family size of the household. As can be seen from Table 3.3, the average family size is around seven members. Household Grain Consumption Consumption is the goal and center of the household unit activities. In traditional agricultural settings consumption decisions and activities are interwoven in a complex fashion. In this case consumption require- ments and preferences are reflected directly in the crops grown and areas devoted to them in the production sphere. Almost all crops grown by , smallholders in the Nuba Mountains area are partially used in consumption (e.g., sorghum, sesame, groundnuts, lubia, etc.). However, sorghum (dura), holds a special place in the production/consumption complex of the house- hold. 41 More than 80 percent of the cultivated area in the Nuba Mountains area is under dura. Beside its varied consumption uses as the basic staple food, dura has other important uses for the household. It~ is used to meet certain social and ceremonial obligations, in-kind payment of hired labor, and is sometimes used (in a form of barter exchange) to pay for goods bought at village shops. This background note is intended to provide understanding of the special place that dura has in the smallholder's cropping plans and rotation. This will be discussed in more detail later, but another methodological point of relevance here is the behavioral as- sumption (pointed out in the previous chapter) that farmers, as one way of risk management decide on the production of dura such that at least household requirements are secured. This assumption is based on the observed and on-going tradition of retaining portions of the produced dura.1 Table 3.4 shows the quantities initially retained at the household and their percentage of the total production. The table also shows the average annual consumption of dura by the house- hold. Resulting estimates of consumption seem large and are also extremely variable (see standard errors). These estimates are further discussed, compared and adjusted later, when developing the dura con- sumption constraints in the LP models (Chapter IV). 1Farmers store dura at home in a structure locally known as "Seiba." 42 Table 3.4 Household Annual Dura Consumption and Initially Stored Quantities Household Stored Duraa 'Annual Consumption Region Quantity %'of Quantity (90-kg Sacks) Production (Malwas)b Abassya 12.14 86.74 409.5 (2.42) (2.79) ((89.36) Um Brembita 21.25 52.97 702.00 (15.34) (16.72) (134.26) Abu Gebha 25.28 68.84 697.67 (5.25) (5.77) (62.86) Kalogi 13.37 64.66 536.90 (1.84) (4.93) (46.64) Talodi 30.30 59.10 434.35 (11.09) 3 (9.26) (40.55) Kadugli 8.48 87.84 513.80 (6.82) (2.57) (35.70) Lagawa 12.19 58.70 403.00 (3.31) (6.77) (42.14) Dilling . 13.25 63.80 556.12 (1.89) (5.8) (55.51) ( ) = Standard error. Source: Computed, survey data (1979/80). aRefers to quantities initially retained from the harvest, but not necessarily consumed at the household. bMalwa is a local volume measure = 1.4 liters; approximately, the 90-kg dura sack has 30 malwas. 43 Land Ownership and Cultivation Characteristics Historically there has been no land shortage in the Nuba Mountains regions. Recently, however, a number of developments are just beginning to alter this situation. Notably of these are: the continuous expansion of large scale mechanization (especially in the northern and eastern parts of the region),1 introduction and expansion of NMAPC schemes, small but consistent rise in settlement rate of nomadic tribes, and finally the internal pressures and dynamics of population under these difficulty- managed savanna land. All these contribute to change the land availability situation, but it has not reached a problematic or limiting extent yet. Land Ownership Ownership of land in the area is on a noncontractual basis. This is accomplished within the sphere of the village by either inherit- . ing the land or clearing it (from trees and bushes) [21]. Authority over land is also recognized by the different villages in a traditional way.2 Another issue relating to the rights of land concerns the nomadic groups in the iarea, as they pass and enter cultivated land. This has 1These include not only official MFC schemes, but also the continual- ly expanding numbers of private schemes under the so-called "undemarkated schemes." As of June, 1980, the number of registered such scheme in South Kordofan Province reached 209 (each with an area of 1,000 or 1,500 feddans). 2This among other things created a big problem for NMAPC in exer- cising authority over land and its allocation between different individuals and different villages (see discussion in Chapter VI). Land rights of a certain tribe and/or village can not easily be allocated to farmers of another tribe or village. 44 resulted in frequent and rising conflicts between the pastorialits and cultivators [30]. At present this still remains a problem and no rules are set to deal with it other than the traditional attempts of tribal recognition of rights of those involved. NMAPC as will be discussed later, also has not achieved a reasonable solution to this problem. Cultivated Land 1. Land Size. Basically there are two types of cultivated plots for the smallholder in the area. a. Jubraka: This is a backyard garden or small plot, usually situated at or very near to the home. Table 3.5 shows the average size of these plots for the different regions. As can be seen from the table these plots are usually of small size (around one third of a feddan).1 The wife usually has the responsibility of cultivating this plot. The tradition is to growa number of crops (dura, maize, cu- cumber, okra, lubia, etc.) sown in a mixture, very early in the season. This way crops can mature and support the household for some period (5-6 weeks), before the harvest of the other field plots.2 Saraya: These field plots originate from land brought into cultivation either from fallow (under the system of shifting 1One feddan = 1.038 acres. 2 Much of the Jubraka crops are of the quick-maturing types and are also consumed in a vegetable form (i.e., before full maturation). 45 Table 3.5 Average Area of Jubraka (House Plot) for the Different Regions V Region 1 . . Area (feddans) Abassya 0.26 Um Brembita 0.48 Abu Bebha 0.21 Kalogi 0.37 Talodi 0.30 Kadugli 0.34 Lagawa 0.46 Dilling 0.31 Source: Computed from the survey data (1979/80). 46 cultivation), or a new cleared land. In the first year of cultivation, the land is referred to as "Harig,"1 in the second and subsequent years it is called Saraya. All poten- tial for expanded land base as referred to in this study comes from these lands. Table 3.6 shows the size distribu- tion of the smallholders cultivated plots in the area. Eighty-three percent of the farmers cultivate less than 15 feddans, and more than two-thirds of the farmers with less than 10 feddan cultivation size. Further, the picture for each of the eight regions is more or less the same. 2. Multiplicity and Distance of Plots. An additional feature, which is common to savanna agricultural systems is that farmers tend to cultivate more than one plot of land. This is primarily because of the fallowing and the search for suitable soil areas. Table 3.7 shows that 40 percent of the farmers cultivate two separate plots, 21 percent have three plots, and 6 percent cul- tivate four or more plots. With the exception of the Jubraka which as mentioned above is usual- ly situated right or near the house, these plots are all situated out— side the village. Table 3.8 shows the distance of these plots from the residence. As can be seen from the table more than half of these plots are far, with the majority (48 percent) being very far. This has an im- portant implication for access to these lands and for time spent in travelling to and from fields. 1The name Harig (meaning burning) refers to the land preparation method, which is through burning of the grasses. Z17 .8383: 38.8 882,288 .3... 288888.88 "3.58 ..8 8 2.. 2 . 8 8 8 8 8. 8 8. 82 8.88 88 8.88 .8 888.888 ..< ..8 2 ..8 2 2.8. 8 8.8. 8 8.88 8 8.8. 8 88....8 8.8. 8 8.2. 2 8.88 2 8.82 8 8:888. 2.. . 8.8 8 2..8 8. 8.88 88 ..8888. 2.2 . 2.2 . 2.2 . 8.8. 2 8.88 8 ..82 8 2.2 . .88... 8.8 . 8.8. 2 8.8. 8 8.8. 2 ..28 8 8.8. 8 .88... 8.8 . 8.88 8 2..8 2 8.2. 8 88888 888 8.88 2 2.8. . 2.8. . 2.8. . 2.8. . 88888888 88 8.82 2 8.28 8 8.2. . 8888888 8 8.29828... 8 8298.288. 8 828.58.. a 838:8... u 8.28.582. . s 8.28.28“. 0 .02 we .82 .538.— .1112288m12811 1112848888111 ”8888882 8. 888.8 8888>.8.88 .8 8888 .888. 888888 882828. "8888 888 8. 8.88.88..888 888 88 8888>.8.88 888.8 .8 88.888.888.8 82.8 8.8 8.888 48 ..om\m2m.v 8288 88>288 88.8 88288880 ”882888 8.. 8 8.8 8 8..2 88 8.88 82 8..8 88 888.888 ..8 8.8 2 ..8 . 8.88 8. 8.88 8. 88....8 8.8. 2 8.8. 2 ..28 8 8.88 2 838888 8.. . 8.8. 8 8..8 82 8..8 82 ..88888 8.8 . 8... 2 8... 2 ..28 8 8.82 8 .88.8. 8.82 8 8.88 8 8.88 2 8.8 . 288.88 ...8 .. 8.22 8 .... 2 88888 888 8..2 8 8.8. . 8.8. . 88888888 88 8.82 2 8.28 8 8.2. . 8888888 2 8285282 2 8288288 2 8288288 2 8288288 2 8288288 88 .oz 88 .8: . .8 .oz 88 .oz .8 .oz 88.88m 888.8 8 888.8 8 888.8 8 888.8 2 88.8 . "828.8 mumcmamm m8228>.2.=o 8285288 88 288882 88.888 88 88.888.8.88 88.8 8.8.8.82 88.8 8.8888. 88 8888888 888 888882 2.8 8.88. .282888 28 88888.2 8828888 28 88888.5 88 88:8 8885 n 888 .>8 .282888 28 88888.2 omnoN\888. 28 88888.5 88-88 n 2888 .282888 28 88888.5 omuo.\888. 28 88888.8 omum. u 88828 .882888 88 88888.5 8.28888 28 88888.5 8. 8888.888. n 8882 .>8 ..om\m2m.v 8888 28>288 2888 88888588 "mugzom 8.28 88 ..8 8. 8.88 88 ..8. 82 888.888 ..8 ..88 2. 8.8 2 8.82 8 8.8. 8 88....8 8.28 8 8.8 . ..28 8 8.8 . 888888 8.88 88 ..2. 8 8.88 82 ..2. 8 8.88888 8.88 8. 8.82 8 8.2. 8 888.8. ...8 .. 8.8 . 8.88 8 .88.88 2.8. 8 .... 2 8.88 8 8.22 8 88888 888 8..2 8 8.8. . 8.8. . 88888888 88 8.28 8 8.2. . 8.88 8 8888888 8 8882.88 2 8885888 2 8885888 2 8.85888 88 .82 88 .82 88 .82 88 .82 88.882 8888 .> 8288 88882 88882 .> 8888.8 m:.a8>.8.:o 8885888 88 8888888 888 888582 888.882 8888888.a 828 88. 8888...> 8888 888.8 8888>.8.=u 88 88888888 m.m 8.88. 50 Another implication of this is the difficulty and cost entailed in V the transport of food and drinks to these fields and transport of the harvest to the home or to the market.' Agricultural Technology and Crop Practices Traditional smallholder agriculture can be described as consisting of small operating units, low level of capital inputs, dominance of hand labor, and generally low and unstable production. What follows is a brief description of smallholders' farm capital outlays, rotations and dominant crops, the set of agricultural operations adopted, and the farm labor organization. Farm Capital The level of capital investment and usage for smallholders in the Nuba Mountains area is very low. The bulk of capital outlays are operat- ing capital (cash expenses), used mainly to pay for hired labor and ex- penses of the Nafir (communal or exchange) labor. Fixed capital items include primarily the hand tools with which all field operations are l carried out. Almost all farmers also have a "Seiba" (a cyiinderica] grass and mud structure) for storage of sorghum at home. 1Only few hand tools are used. These include: long-handled flat hoes (Siluka) for checkrow planting, short-handled flat hoes for weeding (Jerraya), short-handled hoes with a strong angled blade (Sakaba) for cutting down sorghum stalks and light bushes, a range of knives, axes, blades or sickles for harvest, flat wooden paddles for threshing sorghum and long two-pronged rakes for cleaning fields of crop residues [38]. 51 Operating capital is one of the most important factors in the per- formance of smallholder agriculture in the area” . Its level and timely availability, greatly affects the siae and efficiency of the operating units. Its special importance is not only for meeting productive expenses, but also for meeting urgent consumption needs. Consumption needs tend to coincide with the seasonal needs to cover production expenses. For the most part farmers meet these demands for capital through borrowing from local money lenders, under what is locally known as the "shiel" system. Under this system, money lenders who are mostly village merchants and shopkeepers, advance cash or goods to the farmer. The latter in turn has to pledge his prospective crop at a value substantially lower than market price.1 The implied interest rate differs according to location 2 and time of borrowing but is usually substantial. When the farmers do not get involved in sheil, they often are forced to work as hired laborers for others in order to obtain their needed cash.3 Operating capital needs of traditional smallholders are predominantly for meeting labor expenses. Hired labor is paid in cash with or without 1"The surveys carried out by the Agricultural Bank of Sudan in S. Kordofan Province showed that about 50 percent of the cultivators re- ceived sheil and realized only 50 percent of the market price for their produce" [8, pp. 95-96]. 2The implied interest rate usually declines as the harvest time approaches, but it can reach as high as 300 percent for loans made at the peak of the season [38]. 3Sometimes the two go together when “the sheil merchant who is often a large farmer, gives no loans before his own farm work is done. The smallholder may have to work for the large farmers before he can get any loans" [8, p. 95]. 52 additional pay in-kind. Nafir labor is paid solely in-kind (food and drinks). Occasional expenditures are made on items such as seeds, chemi- ‘cal insecticides, sacks, and transport of crops to home or markets. The dominance of operating capital use in labor expenses reflects the fact that at present, hand labor is the principal production input for small- holders in the (area. ‘Table 3.9 shows the average expenses per feddan for the different crops disaggregated by operation. Weeding and harvest, which are the most labor demanding operations, are also the most expen- sive. The variability in the average expenses (see ranges in Table 3.9), is related to the composition of the farm labor force; which at one ex- treme could be composed of entirely hired labor, and on the other ex- treme of entirely (unpaid) family labor.1 Formal sources of credit to smallholders in the area are virtually nonexistant. However, the Dilling Branch of the Agricultural Bank of the Sudan (ABS), has recently conducted a trial experiment to extend cre- dit to smallholders through cooperative societies.2 Results of the ex- periment were encouraging for expansion of the program [8]. Another po- tential source of formal credit exists for the NMAPC schemes participants. At present however, only mechanized cultivation is done for the partici- pants on credit basis. The issue of NMAPC participant's credit will be discussed more in Chapter VI. 1Family labor was not costed in the calculation of the average ex- penses shown in Table 3.9. 2The issue of credit and the ABS trial experiments are discussed in more detail in Chapter V. 53 Table 3.9 Average Expenses Per Feddan for the Different Crops by Operation (1n Ls: 1 Ls = $2.0) Operation Sorghum Sesame Groundnuts Cottona Land Preparation 0.45 0.07 0.90 --- (3.0) (1.11) (4.76) Planting 0.29 0.39 0.55 0.29 (2.0) (4.0) (2.25) (2.0) Weeding lst 1.14 1.18 1.57 0.95 (5.46) (7.27) (4.48) (8.0) Weeding 2nd 0.59 0.18 0.99 0.43 (3.29) (1.40) (3.67) (3.75) Harvest lst 1.11 0.51 0.71 1.31 (6.67) (2.8) (4.0) (7.2) Harvest 2nd 1.01 0.36 0.53 -- (5.16) (5.0) (3.68) Collection and 1.0 0.57 0.66 0.95 Transport (6.67) (3.33) (3.0) (5.0) Clean up (Al-awdi) -- -- -- 0.83 (3.10) ) = Range of the estimate, i.e., difference between largest and smallest value reported. The ranges in the above table indicate a highly skewed distribution with a few high values. The reason for this as indicated in the text, is because of the differences in labor com- position of the farming households. Source} Computed from survey data (1979/80). aEstimates from the NMAPC participants' sub-sample. 54 Rotations and Dominant Crops Rotations and crop composition are influenced by physical environ- ment, the state of technology, market opportunities, dietary preferences and resource base at the farm level. Rotations The agricultural system of smallholders in the area is characterized by shifting cultivation. This system is forced to adapt to the conditions of: lack of animal manure or use of commercial fertilizers; absence of appropriate rotation that can sustain yields for longer periods; and) absence of pressure on land. This last factor, has been undergoing changes recently. Although it has not reached limiting magnitudes yet, its effects are beginning to show by an increase in the cropping periods and a decrease in the fallowing. No definite rotation is practiced by the farmers in the area. The most common rotations that are followed loosely by the farmers include: sorghum-sesame-fallow-fallow; or a modification of this, sorghum-sesame/ sorghum-sesame/sorghum, followed by a fallow. Cotton used to be grown by farmers in the area as a third crap in the rotation. Recently the relative decline of cotton growing by traditional farmers has resulted in an increase in the proportion of sorghum on land presently under cul- tivation [38]. When soil conditions permit, groundnuts may be introduced into the rotation. At present however, few farmers grow groundnutsand only in a limited scale.1 1As indicated in the next section, groundnuts are grown relatively more extensively by the farmers in the northern part of the area (Dilling and Abassya regions). 55 Dominant Crops The dominant crops of the Nuba Mountains regions include: cotton, sorghum, sesame, and to a relatively less extent groundnuts. Other minor crops include: millet, lubia (vigna), and maize. Horticultural crops are limited, more or less, to flood plain areas, and are of even less importance. Table 3.10 shows the area and production trends of the four main crops in the regions of the area. Noticeable from the table is the de- cline of cotton feddanage in all regions; to such an extent that cotton growing is almost disappearing from traditional smallholder agriculture. The primary reason for this, is the decline in the farmer's price of cotton, in real terms and relative to other crops prices [39]. This issue, being of current policy concern, will be discussed in more detail in Chapters V and VI. As can be seen from the table, the crop feddanage is dominated by sorghum, followed by sesame. Groundnuts, although grown in all eight regions, is concentrated in the northern regions (Dilling and Abassya). The light loamy soils in these northern regions are more suited to the production of groundnuts. Agricultural Practices The following is a brief description of the varieties grown, the set ofagricultural operations, their labor requirements, the agricultural calendar, and the level of crop yields attained. Crop Varieties With the exception of cotton, there is no certified crop variety grown by the smallholders in the area. In addition there is considerable Table 3.10 Crop Area in Feddans and Average Production by Region 565 d (1974/75-1978/79) Region Season _____§gt§gg. a Dura b Sesame c Groundnut: Are; Avg. Prod. Ari Avg. Prod. Are_a_ Avg. Prod. Ara Avg. Prod. Kadugli 1974/75 31917 4 122966 7 23098 3 8208 6 1975/76 57287 3.5 17960 8.5 25590 2 7532 6.2 1976/77 35200 3 96876 8 8141 8 4505 8 1977/78 22401 2.8 69729 10 14377 2 7955 6 1978/79 14419 2.5 99299 10.8 11432 6.5 7923 6 Lagawa 1974/75 4650 - 18621 8 7707 3.5 4187 8 1975/76 6400 - 20100 8 8800 3.5 7500 8 1976/77 2686 - 23200 6 8356 4 6926 6 1977/78 1450 1.7 20400 8 8400 3 8000 8 1978/79 800 3.5 21500 10 8000 3.5 7000 9 Dilling 1974/75 4124 2.4 38636 9 19615 4 36082 10 1975/76 4990 2.6 38752 6 12549 2.5 22157 7 1976/77 3294 2.9 20264 4 8150 2.5 11010 6 1977/78 1284 3 20193 6 10375 3.6 13813 7.5 1978/79 1733 3 20088 10 8137 2.5 8224 7.5 Talodi 1974/75 20000 2.2 30112 - 2543 3 5899 7 1975/76 20966 2.3 36903 7 1754 3 5688 7 1976/77 11551 3.7 32080 8 616 5 3087 5 1977/78 5129 3.5 26003 - 871 5.4 1847 5.8 1978/79 7450 5.5 23780 9.3 930 3.2 2110 7.9 Kalogi 1974/75 16163 2.5 49797 7 13646 4 1381 7 1975/76 12976 3 38937 6 6783 4 1830 4 1976/77 13220 - 37295 - 9326 - 1170 - 1977/78 7386 2.9 37051 6 11662 5 837 7 1978/79 8029 3.5 41965 8 15619 4.5 1224 7.5 Abu Gebha 1974/75 21918 4 30565 14.5 25310 5.5 1354 12 1975/76- 2733 3 42941 10 31217 5 1779 15 1976/77 4962 3 41128 12 26901 4 278 4 1977/78 7532 3 34305 7 18795 4 2299 5.5 1978/79 4314 2 58200 11 35186 4.4 2667 9 Um Brembita 1974/75 2143 1.11 35647 10 17966 3 2364 10 1975/76 1532 2.7 35490 10 18969 2 3560 12 1976/77 1700 2.4 37617 8 14777 3 3719 10 1977/78 1480 1.2 31199 8 14596 2 3473 8 1978/79 350 2.5 36864 12 12122 3.5 4473 10 Abassya 1974/75 2300 1.5 60454 7 31724 3 120585 7 1975/76 164 3 58516 11.5 20418 3.5 20025 10 1976/77 119 4 53230 11 16049 . 3 19450 4 1977/78 1089 2.4 40521 9 11463 3 17705 4 1978/79 518 3 85224 8.5 23605 3 32129 6 Source: NMAPC records. “In small kantars (45 kg). b In 90 kg sacks. cIn 75 kg sacks. dIn small kantars (45 kg). 57 diversity in the agronomic characteristics and genetic nature of the varieties grown. Most are of indigenous or local types commonly known as "Baladi" varieties. 1. Cotton, Cotton grown in the Nuba Mountains area is the short staple, American type cotton. The variety in use at present (Albar A(57)12)], was developed by Kadugli Research Station (KRS), which has operated as a breeding station for cotton under rainfed conditions. With optimum planting time (recommended as July 1), the currently used variety has a maturity period of 150-155 days. 2. Sorghum. There is considerable genetic and agronomic diversity of sorghum varieties grown in the area. Classified according to the maturity period, these varieties are grouped into: a) Early-maturing varieties (100-110 days). These are dwarf white-seeded varieties. They were first introduced in the large scale mechanized sub-sector; their uniformity in ma- turity and height make them suitable for mechanized harvest- ing. "Um Brenien" and "Gadum-el-Hamam" are the leading varieties in this category. Very few smallholders in the area grow these varieties, but under NMAPC schemes, officials encourage their adoption and use. The quality of dietary 1 Variety development and improvement research at KRS, has already succeeded in developing a new variety (BAR 24/4H), which in limited tests has outperformed the variety in use (Albar A(57)12) in a number of re- spects (yield, ginning percentage, lint and fiber qualities) and pending future quality tests might substitute the existing variety [6]. 58 characteristics of these varieties is widely believed by farmers to be inferior to other varieties, a factor that has prevented its widespread use. b) Medium-maturing varieties (120-130 days). Popular varieties in this category include the easily threshed "Karamuka," "Marig," and "Bakhit." These are most common in the north. c) Late-maturing varieties (150-160 days). Important varieties of this type are "Kurgi," "Kulum," and "Magoy." These are mostly grown in western and southern parts of the area. 3. Sesame, Considerable genetic diversity also exists in sesame. This is more manifested in other agronomic characteristics (pod shape, seed color, and branching style), rather than in height or period of maturity. Most varieties mature between 90 and 110 days. All varieties grown are dehiscent (opening of pods) and quick in shattering seeds upon maturation, a feature thus, with important implications for timeliness of the harvest operations. 4. Groundnuts. Only two main types are grown in the area. A local spreading variety known as "Baladi," and the newly intro- duced, erect type, "Barberton" [38]. Agricultural Operations 1. Land Preparation. Land preparation methods depend on whether the cultivated plot is “Harig” or "Saraya." a) Harig: refers to first cultivation year (either from fallow or new land). Land preparation in these plots involve set- ting of fire to kill the flush of weeds and grasses that emerge after early showers of the rainy season. 2. b) 59 Saraya: denotes plots in second and subsequent years of cultivation. If the land was well weeded in the previous seasons, limited land preparation is required. Sweeping and burning of stalks and crop residues is undertaken either after the harvest or prior to planting. This operation generally is not very laborious, except for cotton. The clean up operation after cotton (locally known as "Al-awdi), involves pulling the cotton stalks and burning the residue. NMAPC administration which supervises the operation, sets May 31 for finishing the clean up. Allowing at least one month of open land, reduces the transmittal of cotton diseases from one season to the next. Another reason mentioned by the administration is that cotton stalks on poorly cleaned schemes greatly hamper the mechanical cultivation operations in the coming season. Farmers in the NMAPC schemes are very reluctant to carry out this operation, and the administration often resorts to judiciary powers to enforce it.1 Planting. Crops are planted usually in a single stand and crop mixing is seldom practiced.2 Checkrow planting is advised for all crops. However, at present this is followed only for sorghum and groundnuts; sesame and cotton are broadcasted. Checkrow plant- ing is especially recommended for cotton, and the current practice of 1 Through local and traditional courts which fine violating farmers. 2Crop mixing, however, is the standard in planting of the "Jubraka." 3. 60 broadcasting is very much blamed, among other things, for the observed poor crops. Also, another important aspect of the planting operation is its~ timing. Although, this is dependent on rains, recommended sow- ing dates are: July 1 for cotton, early-June for sorghum and sesame, and late-June to mid-July for groundnuts [6, 7]. Weeding. This is the most labor intensive operation of the cul- tivation process. When land is first brought into cultivation (fallow or new land) burning is used as a weeding method. In subsequent years, this method would not be effective because density of weeds and grasses is too low, rendering the burning control method ineffective and hand weeding as the only alterna- tive. In general weeding levels and timingS'are related to a number of factors. These include: previous season management, current season rainfall, timing of the operation, availability of family labor and/or operating capital, crop establishment level and anticipated returns from the operation.T At present up to two weedings are followed. Very few farmers in the sample reported carrying out a third weeding operation. Harvest. Except for cotton, harvest is usually composed of two operations. The first is cutting of sorghum and sesame, and pulling of groundnuts. After being left to dry for sometime Failing or poorly established crops are customarily left unweeded. 61 (around a month for sorghum,1 and 1-2 weeks for sesame and groundnuts), the second operation commences. It involves threshing for sorghum and sesame, and stripping of pods from groundnuts. Although, two hand pickings are recommended for cotton, at present only one pick is practiced. ' Sesame harvest presents a special problem. The dehisence (opening of pods) and quick shattering of seeds upon maturation, requires the harvest of the crop within a relatively short time, making the operation a labor intensive activity. Table 3.11 shows the average labor requirements per feddan for the different operations. Weeding is the most labor-demanding activity, followed by harvest. No separate land preparation for cotton is done other than pulling of cotton stalks and clean up (Al-awdi). Agricultural Operations Calendar An important aspect of an agricultural operation is its timing. This is of even greater importance under rainfed conditions. It is therefore, important to consider timing as an explicit management practice. Table 3.12 shows the calendar times of the different operations for the sample of the traditional smallholders. Although some operations are carried as early as April (land preparation), and some as late as March (cotton picking), the bulk of the operations are concentrated in the period from June to December. This concentrated pattern (especially in June-August), which is primarily a factor of the climate, results in in- tensifying the seasonal constraints in labor. 1Except for the variety "Karmuka" which is usually threshed within a week of cutting. Table 3.11 Average Labor Requirements Per Feddan by Crop Operation (in 7-Hour Man Days) Operation Sorghum Sesame Groundnuts Cottona Land Preparation 3.80 2.80 3.83 , -- Planting 2.92 2.53 4.77 1.15 Weeding lst 7.48 7.61 10.37 4.45 Weeding 2ndb 3.94 3.60 8.36 2.40 Harvest lst 5.60 6.05 5.32 8.67 Harvest 2nd 3.49 2.23 5.77 -- Clean up (Al-awdi) -- -- -- 3.84 Source: Computed from survey data (1979/80). aEstimates from the NMAPC participants' sub-sample. bVery few farmers practiced a third weeding. Percentage of farmers reporting second weeding: sorghum = 60%; sesame = 56%, groundnuts = 76%; cotton = 53%. €53 .ua_a gaucupau unguap a» umo.—gumu .Aucoaaosh amen ..o.—v wavy Laccapau .auo: a .o_aaam-a:m .mu:a¢.u_ugua uasam use each uoazasou "ougaom .mzucos o>_uuoammg as» :. co.uogmao as» m:_xgsou mgmsgm» so a u A V Ao.~VAo.v. Ac.am. - - - - - - .c:a-.amu .La< A_uza-_<. a: com—u - - Ao.~v.~.c_. Ac.mmv Ao.m.Ac.m_. .c.~o. .c.m.Ac.m_. Ao.~mv .coa-.auc .>oz .aou-.uuo .>oz .ga<-.>oz .uma new umu>gaz Ac.~.fic.~_v “c.5m. .a.¢mv.o.omv Ao.omv Ao._eVAo.mv .c.emv Ac.mVAo.~V Ac._mv .sa:-.uoa .caa .>oz-.uuo .uuo .>oz-.aom .uuo .nmm-.uuo .>oz um— umo>gmz Ac.~vao.-. Ac.~mv Ao.~..c.~v Ac..ov .o.mmvac.¢_v .o.~mv Ac.hv.o..v Am.~mv .uuo-.m=< .aom .aam-.:=a .m:< .nmm-._:a .m=< .uou-._=e .m:< ucm mc—uooz .o.~VAc.mv Ac.~m. Ao.nmvac.~_v .c.cmv Ac.pmVAo.~m. .c.~ov .m.~vac._mv Ac.¢¢. .uau-.—:a .a:< .m=<-.:=a ..aa .awm-.m:< ..aa .qmm-.a=< .—:a um_ ocpuom: Ao.m_.Ao.~mv Aa.oev .c.ovaa.eev .o.cmv Ao.__v.c.~m. Ao._mv A~.oVAo.c_v Ao._~. .qwm-.paa .o=< .m:<-.—:e .caa .m:<-._:a .=:a .m:<-xa= .caa mcvuca—a - - Ac.AVAo.m~V Ac.me. Ao.¢VAo... Ac.mmv Ac.~_v.c.mv .~.cmv ._:e-.ca< .cza .—:a1.:ae .ga< .c:a-.uoa .La< coFuoguaoga new; magma gaze: macs: gaamml magma gaze: ummcmw grace: acouaou muscuczogu osemom sasmcom aogu ma mcowumgmao we mewp Laucm—mu .mczapauwcm< Np.m m_nmp 64 Land preparation is split between those who carry out the operation after harvest, and those who carry it out just before next year's plant- ing. Cotton clean up (Al-awdi) in the NMAPC schemes, by law belongs to the first group and should be finished by May 31. The harvest operations are more or less streched out. This is mainly because of the relatively short growing period of sesame and groundnuts, and the delayed cotton picking. For cotton, only one (instead of the recommended two) picking is carried out. Crop Yield Levels Yield levels attained by smallholders in the Nuba Mountains area are variable. Among the important determinants of yield is the amount and distribution of rainfall [6, 21]. Table 3.13 shows the yield levels obtained by the traditional farmers in 1979/80 season.1 Cotton yield levels (missing from the table) are reported later in connection with the NMAPC participants' production. In Chapter V, sensitivity analysis on crop yield levels, is used to examine effects on farmer's returns and cropping pattern.2 Farm Labor and Organization In the Nuba Mountains area, there are three types of labor involved in performing the agricultural operations: family labor, hired labor, and "Nafir" labor. l able. 2This used to simulate 25 and 50 percent reductions in yield levels, corresponding to variations in annual rainfall levels. For this season, amount and rainfall distribution were not favor- 65 Table 3.13 Average Yield Levels for the Traditional Smallholders by Region Region Sorghum ' Sesame Groundnuts (90 kg Sacks) (75 kg Sacks) (100 lb. Kantars) Abassya 3.87 1.29 9.34 Um Brembeta 3.05 1.47 ~- Abu Gebha 6.01 2.92 -- Kalogi 2.63 1.38 10.0 Talodi ' 6.59 5.82 . -- Kadugli 4.73 3.47 ' 13.03 Lagawa 4.56 2.16 . 9.87 Dilling 3.98 1.00 3.75 (--) = Not Available Source: Computed from survey data (1979/80). 66 Family Lab0r (FL). FL is used here to define the labor resource provided by any of the household members. Historically and traditionally this is the main source of labor available to the household [21]. Labor contributed by this-category is not paid directly, but rather is compensated by returns to the whole household. Availability of this type of labor is related to the size, age and sex composition, and domestic organization of the household. As have already been pointed out, households in the area have high ratio of dependent children, and that the wife in.addition to her domestic roles is responsible for the cultivation of the "Jubraka." However, it needs to be pointed out this is not meant to imply a rigid role specification. For it is not un- common for the husband to help in weeding of the "Jubrakd'; the wife to participate in planting and weeding of field plots; or for the children to help on a part-time basis, or in delivery of food and water to the fields. Hired Labor (HL), This category refers to that part of the labor which is governed essentially by a market institution. It is this source of labor, which a farmer can rely on for needs beyond the available family labor, being constrained probably only by his operating and cash resources. Unlike "Nafir" labor, which tends to be employed mainly in the labor intensive opera- tions (e.g., weeding and harvest), HL is employed in the whole range of the agricultural operations. 67 Payment of HL is usually on a cash basis, however payment might include an in-kind component (food and drinks). Sometimes the in-kind component is paid from the crop itself (especially in harvest of sorghum and sesame). Wage rates are contracted in different ways: a) by area, usually "Mukhamas" or "Habil,"1 for weeding oper- tions; b) by production (usually per sack, or volume of cotton picked), at harvest; or c) by unit time, usually "Dahawa,"2 as in planting and weeding. Although wage rates are usually in conformity given seasons, localities or villages, due considerations are given to the variations of the same job (e.g., distance to cultivated plots from villages, intensity of weeds in weeding operations), and these are reflected in the wage rates payed. Table 3.14 shows the hired labor wage rates by region, crop, and operation. Dif- ferences in wage rates in the area is not affected so much by region, as it is by the variation of the within-region. The latter takes into account the type of crop, operation and time, and the nature of the contracting agreement. 3. Nafir Labor (ML). Nafir (cooperation or exchange labor) is one of the traditional aspects of the Nuba Mountains society. 1 2 Mukhamas = 1.75 feddans; Habil # 0.1 feddan. Dahawa. is approximately from morning till noon. 68 3.14 Wages Rates Per Man-Day by Region, Crop and Operation (In Ls: 1 L5 = 82.0) Land Weeding Weeding Harvest Harvest Preparation Planting lst 2nd lst 2nd Abassya a: 1.00 .63 b: c: d: 1.33 1.25 1.00 .50 Um Brembeta a: .50-.95 .51-.80 .40-.85 .57-1.00 .57-1.25 b: 1.75 1.00 .83 c: d: 1.00 1.66 Abu Gebha a: .35 .52-.87 .54-.87 1.74 .20-1.75 b: c: d: .24-l.00 .33-.73 .37-.75 .40-.67 .50-.87 Kalogi a: b: .40-.60 .37-.40 .44-.75 .50-.75 .43-.52 .53-1.35 c: .42 .33-.61 .56 .54-l.13 .50 d: 1.00 .18-.60 .83-.83. Talodi a: .39-.75 .19-.56 .28-.70 .30-.87 .32-.89 .42-1.67 b: 1.00 .65 c: d: .80-.83 .36-.44 .81 .16-.26 .31-.90 Kadugli - a: .35 .90 .85 .46 .50 .57 b: .50 .50 c: .50 d: .50-1.00 .50-1.33 .40-1.00 .40 .50-1.25 Lagawa a: .51-.56 .25-.66 .57 .40 .75 .74 b: 1.00 .38-.62 .60 .50-.83 c: .41-1.00 .45 .50-1.00 .50-1.00 .41 .38-.42 d: .50 .50 .45 .45-.50 Dilling \ a: .42-.60 .44-.53 .50-.93 .62-1.16 .70 .50-1.00 b: .72 .45 .50 c: .40-.70 .3l-.85 .50-.73 .44-.67 .51 d: .83 .66 .60 .42 .87 Source: Computed from survey data (1979/80). a - Sorghun b - Sesame c - Groundnuts ‘ d 8 Cotton, estimates from NMAPC sub-sample, harvest 2nd estimate is for clean up (Al-awdi) 69 This cooperative institution is prevalent in many parts of the Sudan and in many other African countries. Its cooperative nature makes it suitable to labor intensive activities, especially those with a short time span. In the area NL tends to be concentrated in weeding and harvest opera- tions. It is also concentrated in food crops, especially sorghum (see Table 3.15). The institutional nature of NL is also illus- trated in the rewarding of their services. The incentive system is not the market, but rather, is social approval, obligation, and reciprocity. That is why no direct wage is paid for the group participating in the Nafir, but a farmer provides a meal (food and drinks) for his fellow villagers who provide labor. Costs incurred by the farmer are related to his economic ability and tend to be proportional to the number of people participating in the Nafir. The availability and supply of NL is dependent on the size of the village/community, and its domestic and produc- tive organization. Table 3.15 summarizes the utilization of the three types of labor by crop and operation. As can be expected, in this smallholder case there is a marked dominance of the family labor for nearly all operations of the four crops. Second to family labor in importance and utilization is the hired labor. Hired labor use is distributed among all operations, unlike Nafir labor which tends to be concentrated in labor intensive activities for food crops. 70 .opasemuaam .ma:ua.u_uean um¢zz osu seem mmuq:_umu .A.u~3-_<. a=-:o~_u on menus; ecuuou com a .umau meo z—ecu a cones a.ca= u 4: sea "Loans 632.: u 4: «teens s_.5ae u see .Aom\m~m—v name museum as“ see; ambassau "ouezom o.mn m.m m.mm c.o ~.o_ m.om c.c m.em m.mo m.m~ ~.m~ v.9e uueu umm>euz m.~ m.cm m.—c m.~ m.a_ «.mn m.~— ~.—v m.mm o.o~ m.n~ «.mm an. umm>ca= c._ m.—~ m.- c.c m.m_ ~.om o.m o.o~ «.mo _.m m.- m.oo vew me'vmo: m.e e.o~ ~.me q... o.- m.mo ~.m ..mm «.mm o.~_ a.e~ c._o amp mcpumoz o.e mumu m.mo m.m m.o— o.m~ m.m e.m~ ~.No c.~ ~.~— ~.m~ m:_u=a—¢ c.c c.o uo.cop c.c m.m —.eo ~.__ o.mm m.cm m.o ~.m~ m.mo :o_uueeaoem use; mg: mg: «an Na: Na: NJ; as: ma: ad; «42 ad: mam neouuou maomom museueaommll. mammeom meowpmcmao use aoeo .Lonm4 so weak An copp:n_cum_o mm: Loam; m_.m m_nmh 71 Prices and Market Organization Marketing of livestock and crops in the Nuba Mountains is constrained by logistical and organizational problems. These are due in part to the 1 The remoteness of the area with its poorly developed infrastructure. other aspects of the marketing problems are related to the small scale nature of agricultural production and its scattered supply patterns. All crops in the area (with the exception of cotton) contribute to both household consumption and market. The distribution employed in some literature between cash and subsistence crops, has no relevance for the context of this study. Farmers in the. area are concious and responsive to the price structure of their crops. This is reflected in their choice of crops and the area they devote to each. Cotton is a case in point, the historic decline in real returns from cotton has lead to its virtual disappearance from the traditional sub-sector [39]. The price system is subject to control and regulation in the govern- ment's price policy. These policy measures differentiate between cotton and other crops. Cotton marketing which is intended primarily for export is undertaken by the government. Its‘transport, ginning and marketing costs are debited 1This aspect of the problem is especially important for livestock, which are marketed in large urban centers in the north of Sudan (mostly at the capital). At present, the practice is to trek the animals over these long distances, resulting in weight loss, disease and death of a significant number of animals. 72 to the so-called hjoint-account" system.1 This joint account is held between the government, represented by NMAPC and the farmers. The costs incurred under this account system are subtracted from the gross proceeds of cotton, and the remaining net returns are divided between the NMAPC and the farm- ers according to a formula. The price of cotton is calculated on f.o.b. basis at Port Sudan.2 Payment to farmers, is further, based on the qual- ity of cotton they have delivered. There are three established grades; I, II, and III, each with a fixed price. As for the other crops, the government role is confined to a‘ regu- latory function. This involves the setting of minimum prices of all main crops (it includes crops under discussion here i.e., sorghum, sesame and groundnuts) for the so-called "auction markets." (These are located at major centers, the only one in Nuba Mountains area is at Kadugli, which is the provincial capital for S. Kordofan). The prices announced each year by the Ministry of Commerce, are related to the expected world mar- ket prices expressed as f.o.b. at Port Sudan. This system of minimum prices is ensured by having the government trading companies buy quanti- ties that are offered but not bought by private traders at these prices. However, it remains to be said that the "auction markets" system has little direct relevance or impact for the smallholders in the area First, there is only one auction market in the area (in Kadugli), and transport costs from the different regions are prohibitive for the 1The joint account system is discussed in more detail in Chapter VI. The corporation law defining this system is also given in Appendix III. 2Major port of the Sudan, located at the Red Sea. 73 smallholders given their small scattered production. Second, and more important, is the fact that the smallholder's disposal and marketing of crops is not independent of their subsistence and production. As we have indicated earlier in discussing cash needs and credit, a significant num- ber of smallholders during the course of production become indebted to the local money lenders (sheil), and commit their marketable products to these lenders. A study by HTS [38] describes another form of marketing involving barter of crops, which is also common in the larea: There is another form of barter of crops which is probably even more prevalent than sheil. Most shop- keepers accept crops in lieu of cash payment for goods. The customer brings a standard measure, normally one malwa of sorghum, sesame, lubia, groundnuts or okra, and receives goods of equivalent value. The exchange value of crop bartered is invariably less than the selling price. A significant portion of the marketed production is probably transacted in this way, with the shopkeeper acting partly as a middle-man who later sells the crop at an auction market or direct to a major merchant, and partly as a speculator anti- cipating that the crop can be sold locally at a higher price later in the season [38, p. 59]. Some farmers also market a part of their crops directly to their fellow villagers, acting as retailers. This is done at home in the village, or when transport can be arranged in the nearest local market. Table 3.16 shows the prices received by the smallholders in the area for the season 1979/80. The prices received by the farmers depend on the one hand on the nature of the marketing arrangement (e.g. sheil, barter, direct retailing, etc.), and on the other hand on where and when sales take place. The best prices a smallholder can get are those pre- vailing few months after harvest, at the auction market [38]. The prices shown for cotton are from the NMAPC participants' sub-sample, and they 74 Table 3.16 Average Prices Received by the Smallholders in the Area for the 1979/80 Season (in L5: 1 L5 = $2.0) Region #18233 (32.53180 i;23”i38“ r—SiftiSS kg Sack) lb. Kantar) lb. Kantar) lb. Kantar) Abassya 8.50 12.50 5.00 3.90 Um Brembeta 4.50 11.50 4.20 Abu Gebha 3.00 9.77 3.74 Kalogi . 4.32 11.67 2.00 3.84 Talodi 4.71 12.97 3.95 Kadugli 7.60 . 7.50 6.00 3.99 Lagawa 7.56 14.00 4.85 4.10 Dilling 8.70 15.00 5.18 3.96 Source: Computed from the survey data (1979/80). aEstimates from the NMAPC participants' sub-sample. 'Prices are weighted average of three grades of cotton. The fixed price offered by NMAPC for the three grades respectively was: Grade I = 4.25; Grade II = 3.25; and Grade 111 = 2.75. These prices do not include local transport costs (i.e., from the field to the nearest collection center). 75 represent a weighted average of the three grades they have produced. The NMAPC cotton price policy is discussed in more detail in Chapter VI. Contrasting Features of NMAPC Smallholder's Agriculture The NMAPC, its role and organization is discussed in more detail in Chapter VI. What this section intends to give is a brief description of features that are particular to NMAPC smallholder participants (as opposed to traditional smallholders), with regard to their agriculture. In general, these features have not introduced (at least so far) any ma- jor change in the nature of the agricultural systems. They are designed primarily to promote the growth of cotton crop and are characterized mainly by the introduction of mechanical land preparation. Land Ownership Like all public agricultural corporations in the Sudan, NMAPC re- tains the ownership and title of all the scheme land, with farmers recog- nized as tenants. Among other aims of this policy, it is thought to en- sure running these public schemes according to specified governmental rules (especially in distribution of land), and to enforce the prescribed set of agricultural rotations and operations. Farmers of NMAPC dispute this status of land ownership and often claim the title for land individually, or collectively for different 1 schemes under NMAPC. This situation arises because of the way schemes are cleared (from bushes and trees) and registered under the NMAPC. It 1Problems of EL-garug scheme in the Kadugli region is one of many examples related to this issue. 76 is the farmers who collectively clear the land; NMAPC then supervises the scheme and divides the land (as tenancies) among the participating farmers. The net result of the disputed land ownership issue, has been reflected in an unclear production relation between the corporation and 1 the farmers. This has also caused a number of complications and problems in administration and management of many NMAPC schemes [46]. Rotations and Crops In its full development NMAPC plans to have a 15 feddan for each tenant to be grown in'a three-course rotation of cotton-sorghum-fallow. With 5 feddans for each component in the rotation, participants are ex- pected to cultivate a total of 10 feddans (divided equally between cotton and sorghum) each year. However, at present neither of these two aspects have been realized. No fallowing is practiced, instead a continuous two-course rotation of cotton-sorghum is followed; and land size cultivated per tenant is less than 10 feddans. Table 3.17 shows the tenant land size distribution at the different stations [of NMAPC. Seventy-five percent of the tenants cultivated less than 10 feddans in the season 1979/80. For the few tenants reporting a land size of more than 15 feddans, the land actually belongs to more than one individual (e.g., husband and wife and/or brothers, etc.) and is worked jointly. 1Among other reasons, the NMAPC administration emphasizes the pre- sent range production relations, for not extending credit to the farmers to cover critical operations (e.g. weeding and cotton picking). 77 .Acaxaea_ use: eases" use. sausages "oueaom N.m m _.— _ c.N N c.m— m— c.cm Ne m.eN mN mac—uaum __< c.o o o.c o c.o c m.m — N.NN m— m.m — m=.—_—a o.o_ — o.c c o.o o o.c¢ v c.cm m o.oN N atoms; o.o o o.o o N.m m m.o N o.oN a c.em N. .pmauox c.o o o.o c c.mN N c.mN N m.Nn m m.N— _ .co_ap :4. o 9c o age a N; _ m0m 2. as o 35:“ N.NN N p.—p — N.NN N N.NN N N.NN N o.c c enema 34¢ o.o c o.o o c.o c c.cN — c.oc n o.cN — oumaaaem 2: o.o c o.c o c.o o o.c o c.o a c.oo_ — ohmmea< a wemasam a mcoELam a mummguu u weaseou u mcmauom a mcoscuu $0 .02 MO .02 $0 .02 b0 .02 5° .02 $0 .02 cavaflum ~c.cm-—.cm. ~o.cm-_.on ~c.oN-_.m—v Ao.m_-—.opu Ac.c—-—.mv Ao.m-c. umpupau ace; we copusnweumwo mem N_.m m—no» 78 The main component of the NMAPC modernization program is the mechanical cultivation services it offers to participants. Theoretical claims for the yield-related advantages of mechanical land preparation, together with the current problematic situation of these services in the NMAPC schemes, are discussed in more detail in Chapter VI. Table 3.18 shows the average per feddan yield levels of sorghum and cotton for NMAPC participants. At present, NMAPC participants' yield levels are not much different from traditional farmers (see Table 3.18). Other factors blamed for low yields under NMAPC are discussed later in Chapter VI. Agricultural Services In addition to mechanized cultivation, NMAPC offers to its partici- pants chemically treated seeds, cotton pest control, and credit services. These services are briefly described below. Mechanized Cultivation NMAPC offers to its participants as a central agricultural service, mechanized land preparation. The operation is performed by a disc/seeder ’(a wide-level disc harrow) mounted on a tractor. It involves essentially shallow discing of the soil, with the aim of preparing the seed bed and controling weed growth. In the case of sorghum, planting is also done mechanically, usually simultaneous with the second discing. Two discings are stipulated; however, at present due to lateness of operation, gaso- line shortages and mechanical break-downs, very few schemes are disced twice. 79 Table 3.18 Average Yields Per Feddanfor'NMAHZParticipants: 1979/80 Season NMAPC Sorghum Cotton Station (90 kg Sack§)i (Small Kantar = 100 lbs.) Abassya 4.50 2.00 Um Brembita 1.71 0.54 Abu Gebeha 6.24 2.26 Kulogi 4.47 2.87 Talodi 2.37 2.38 Kadugli 1.58 1.89 Lagawa 3.52 2.61 Dilling 4.53 2.54 Source: Computed from the survey data (1979/80). 80 The present practice of NMAPC is to prepare the land for cotton first (to ensure it is planted), before preparation of land for sorghum. The NMAPC charges a flat rate of Ls 1.00 per feddan for one discing and Ls 1.60 for two. It is relevant here to mention that under the present system, there are many organizational problems, resulting in late and low quality performance of these operations. This has created an ele- ment of disatisfaction among farmers with the NMAPC system. Provision of Seeds The NMAPC provides chemically treated cotton seeds, free of charge, to both traditional and NMAPC farmers. In recent years, the NMAPC have been encouraging participants to grow improved sorghum varieties, by providing the short maturing varieties (Umbenien and Gradam—el-Hamam) to the farmers at cost. NMAPC has no seed multiplication facility, and varietal degeneration has already been observed for both cotton and sorghum [38]. Cotton Pests Control The cotton crop is attacked by two main pests: flea beetle in the early stages of growth and cotton bollworms around late September and October [6]. Thymul 35%, Using ULV sprayers, was found to give good re- sults in control of both pests. The NMAPC recommends two sprayings for the cotton crop. However, at present only around 50 percent of the crop is treated, and this with one spraying at the rate of one liter per fed- dan. The NMAPC provides the chemicals and Sprayers, and requires the farmers to do the job themselves, under the supervision of the crop pro- tection department. Timing and management of the operation was reported as being unsatisfactory [38]. 81 Another minor cotton pest, is the cotton redstainer bug. The NMAPC controls this pest by spraying trees and shrubs near the schemes, through an intensive campaign in the dry period. Costs of this campaign are assessed to all farmers collectively and are debited to the "joint account“ 'described earlier. Tenant's Credit At present, participants are offered credit facilities only for me- chanical land preparation costs, usually subtracted from the farmer's share of his cotton return. If the farmer is not growing cotton, or if his share will not cover these costs, he is expected to pay in cash at harvest time. The payment record is unsatisfactory, and farmers debits are transferred from one season to the other.1 Farmers hope that NMAPC will begin extending credit to cover other agricultural operations, namely weeding and harvest. Under its constitu- tion (as in other agricultural public schemes) NMAPC has the potential of doing this. However, implementation is hindered by the unclear produc- tion relationship (especially the issue of land ownership), and by the poor repayment performance of the land preparation services. 1The NMAPC administration mention the mounting farmers debits, as the reason for shutting down almost all schemes in the 1974/75 season. CHAPTER IV REPRESENTATIVE PRODUCTION MODELS AND THEIR LINEAR PROGRAMMING STRUCTURE The previous chapter presented a general description of smallholder agriculture and its environment in the Nuba Mountains. The material in this chapter is divided in three parts. The first part is a discussion of smallholder farming systems in the area, and the representative pro- duction models used to analyze these farm systems. The second part pre- sents the structure of the formal linear programming (LP) models (objec- tive function, activity and constraint sets), and the procedures used to estimate model coefficients. The last part describes the changes made in the basic LP models to allow the conduct of experiments which address the major research objectives. Smallholder Farming Systems and Representative Production Models An important purpose of this study is to identify and examine in a quantitative sense the smallholder's production structure and constraints. This includes investigating the effects of certain policies and/or tech- nologies on the existing smallholder system. The program of NMAPC repre- sents an important departure from traditional farming, hence it is treated as a separate production system. Therefore, traditional farming and NMAPC modernization schemes are the two smallholder farming systems in the re- gion which are modeled. Similarities of important production characteristics within each of the two systems render a representative farm model approach well suited 82 83 for the intended analysis. Many of the significant conceptual and metho- dological problems of the representative model approach center around the criteria for classifying groups of farms on the one hand, and on the choice and simulation of a model(s) to be representative of each group on the other hand, Collinson [17]. However, in this case, as in many traditional smallholder settings, it is relatively simple to identify and simulate representative conditions.‘ As Collinson has noted: Most of the obstacles to using representative farm techniques derive from the problems of selecting criteria for grouping the farm population. These problems are created by the proliferation of market opportunities and technical possibilities in ad- vanced agriculture which distort the pattern that would result from natural advantages of climate and soil [17, p. 103]. However, despite general similarities characterizing smallholder production within each of the two farming systems, there remain some important differences in production structure and constraints that should be incorporated in the representative production models. Such differences are primarily in the level of resources of the individual smallholders. Differences in area cultivated in turn reflect the variability of labor resources among smallholders. With land being relatively abundant in ' the region, labor is the principal scarce factor of production for the- smallholders. In addition to influencing the cultivation size and enterprise c0mbinations at the farm level, the availability of labor also interacts with other factors to determine the labor utilization profile in smallholder production. The latter has important implica- tions for the design of the representative model. 84 Classification by cultivation size is useful since area cultivated is correlated with other farm resource use levels. For example the sim- ple correlation coefficient (r) between the cultivation size (GRLND) and total labor (TL) use at the farm level was found to be 0.73 (i.e., rGRLND, TL = 0.73)]. Cultivation size is also highly correlated with total operating capital (00) use (rGRLND, 00 = 0.78)2. This indirectly reflects the influence of labor, essentially because hired labor (HL) is the principal item to which 0C is devoted. The correlation coefficient of the latter two variables was found to be very high (r00, HL = 0.96). For these reasons, the cultivation size criterion was used to select three categories of representative resource conditions for the analysis of the two smallholder production systems in the region, i.e., the tradi- tional and NMAPC systems. The choice of three categories (to represent small, medium, and large resource endowments) was based on the following considerations: 1. Three categories of farm level resources are deemed sufficient for representing the important differences relating to scale and cultivation size within each of the two smallholder systems. It was considered that using fewer than three levels might not capture the important differences across the sample populations. 1In fact an even higher correlation could probably have been obtain- ed if TL (the sum of family, hired, and nafir labor use) was disaggre- gated by certain periods (namely the labor intensive periods). Use has been made of this observation in formulating the family labor resource constraints of the production models (see next part of this chapter). 2As explained later, total cash expenditure (in agricultural opera- tions) is used as a proxy for DC. 0n the other hand, more than three categories would not add enough to the analysis to justify the cost and effort in-- volved. 2. As discussed below, differences in cultivated area, reflected in these farm categories, are also reasonably reflective of the other important differences in household characteristics (demo- graphic composition, consumption, off-farm work opportunities, etc.) which also must be controlled for in the production models. 3. Although the NMAPC smallholder system is planned and operated with a standard size and crop composition, the use of the three resource categories in the NMAPC smallholder production model allows an analysis of the merits of current and proposed NMAPC farm plans and policies. The production resources and characteristics of the three farm size categories are shown in Table 4.1. As already discussed in Chapter III, dura dominates the cropping pattern in the area. Cotton has almost disappeared from the traditional smallholder area. Family labor use as a percent of total labor decreases as the culti- vation size increases. It is high in category 1, less so in category II, and least important in category III. Hired labor use is the reverse of this case; it is high in the large category III, less so in the medium category II, and very low in the smallest category I. Nafir labor is a different case in both absolute and relative terms, its use being the highest for medium sized farms in category II. The large size of the labor expense share in operating capital ex- penditure is clearly shown across all the three categories. Other 86 Table 4.1 Production Resources and Characteristics: Distribution by Farm Size Category, for Farms Surveyed Cat or I Cat or II Cat or III Characteristic (0.10-2.D 733.) (5.1-5.9 733.) (10.6-35.; 733.) No. of Holdiggs 51 43 19 Percent of Total 45.1 38.1 16.8 Resource Usea (and Use (feddan) 3.10 5.59 14.37 ""' (1.4) (1.3) (3.1) cottonb 0 0.11 0.28 (0.4) (0.9) Oura 1.93 4.15 _ 9.28 (1 4) (1.2) (2.5) Groundnuts O 0.89 1.05 (1.3) (2.4) Sesame 1.18 1.4 3.76 (0.4) (1.4) (2 5) ngor Usa,(man-oaysl 133.75 155.40 250.57 '—‘—* (53.7) (59.2) (131.4) Family Labor 112.25 95.40 133.00 (77.7) (37.2) (77.2) Hired Labor 3.00 30.20 106.83 (3.0) (28.7) (150.3) Nafir Labor 18.50 38.8 ' 10.83 (23.9) (40.1) (25.0) Labor inputs/feddan 43.15 25.10 17.44 0 eratin Ca ital LS 7.85 40.99 92.95 (5.9) (31.5) (109.5) Labor Expensec 5.57 34.03 79.40 (5.5) (25.4) (103.2) Seeds 0 1.13 5.96 (2.8) (8.7) Chemicals 0 0.16 0.34 (0.3) (0.3) Sacks 1.28 5.57 7.25 (1.5) (7.03) (11.3) OC/feddan 2.53' 5.22 5.47 ( ) - Standard Deviations Source: Computed from survey data (1979/80). aEstimates from 42 cases with complete labor records. bFigures are missing or small since very few (only 6 farmers) of the traditional sub-sample grew cotton. cIncludes transport of crops to home or market. 87 purchased inputs vary directly with farm size; being highest in the largest category III, and almost zero in the smallest category I. Selected household characteristics corresponding to the three cate- gories of production are given in Table 4.2. Both average family size and number of active workers are directly correlated with farm size. Average family size varies from 5.25 (in category I) to 7.08 members (in category III). This variation in family size between the three farm size categories is directly correlated with household consumption levels. Dura consumption, expenditure on food, and other annual expen- ditures are all directly related to the family size. The differences in dura consumption levels (in 90 kg sacks) betWeen the three farm size categories are incorporated in the minimum dura consumption constraints of the LP models. However, the absolute amounts of the dura consumption of the three categories are adjusted by judgement to be 8, 10, and 12 sacks, respectively.1 Crop sales by the household are also directly related to the farm size category. This is not only because of the obvious increase in cul- tivation size, but more importantly because of the increase in dura pro- duction (see Table 4.2), and therefore the increase in potentially marketable portions (i.e., after allowance for household consumption needs) of dura in the relatively larger farms (category II and III). Livestock holdings show the same direct relation across categories. Investment in livestock, as previously indicated, is one of the 1Given the variances of the dura consumption estimates (see standard deviations, Table 4.2), these levels are still within the observed range. Additional discussion and justification of these levels is given in the next part of this chapter. 88 Table 4.2 Selected Household Characteristics: Distribution by Farm Size Category, for Farms Surveyed Cats?? I Q5533; 51 Cat II mm Average Family Size ’ 5.25 ' 5.93 7.08 ‘ (3m mJ) (an Active Population‘ 2.50 2.93 3.33 (0.6) (0.9) (1.7) Consular Eguivalentsb 4.05 4.41 . 5.13 (2.4) (1.6) (2.0) M Cons “on 7 05 8 76 11 57 m" ““9 3“") (2.1) (430) (512) c 260.0 274.04 282.16 6 27.75 95.53 113.42 are Production (90-kg sacks) 9.25 18.60 37.67 (4.3) (11.8) (29.6) QM Sela (Ls) 15.00 68.80 144.69 (30.0) (89.6) (113.1) Livgmk Holding; (110.) Cattle 1.27 5.71 5.53 (3.5) (9.3) (12.8) Sheep 1.30 0 4.86 (6.3) (14.9) Goats 4.61 8.58 7.95 (8.3) (8.1) (7.5 matey! 0.09 0.38 0.82 (0.3) (0.6) (0.8) Am—gw w Off-rm ac tion“ :0." :1; :9; ".1: :9: '1_' 1. Wage Labor in MFC 26 2 17 3 22 2 2. Wage Labor in mac 23 5 18 2 21 3 3. liege Labor in Traditional 25 3 18 2 20 4 4. Trade 24 4 19 1 18 6 6. Cutting Hood/Hay Z7 1 17 3 23 1 6. Bringing Water 25 1 l9 1 23 1 7. @vernment do 23 5 17 3 23 1 8. Miscellaneous 14 14 10 lo 19 S 9. Livestock Husbandry 28 0 20 0 24 0 10. Livestock Trade 28 0 20 0 24 0 ( ) - Standard Deviations Source: Counted froe survey data (1979/80). ADefined as sales and feneles in the age bracket 15-65 years. bA weighting procedure suggested by F40 was used. Heights used were as follows: males and tales (0-4 years) - .2; males and females (5-9 years) - 0.5: males (10-14 rs) - .75; females (10.14 years) - .7: sales and tales (15 years and above) - 1.0 [61. p. 138)?I channel expenditure on: meat. vegetables. oil, and tea. dAnnual expenditure on: clothes. health, children education. household physical mintenance. and cemnial. "1" - if involved in occupation; '0' - if not. This is based on responses, not respondents. 1'Include occupations such as honey collections. house building, charcoal burning, shoueker. labor in the market. well digging. etc. 89 important measures taken by the smallholders to counteract risks and uncertainties associated with crop production. Off-farm employment opportunities have been discussed in Chapter III. The distribution of off-farm activities by category, as seen from Table 4.2, shows some contrasting features. Very few farmers in cate- gory III engage in the common pattern of off-farm work, namely off-season work in hay or wood cutting, charcoal, rope-making, hut building, and other'miscellaneous jobs. A majority of individuals in this category are traders and shopkeepers. By contrast, the individuals in category 1, beyond their participa- tion in the common pattern of off-farm work, are relatively more engaged in work as agricultural laborers outside their own plots. Nineteen per- cent of this work is done during the growing season. The next part of this chapter will present the formal LP model. The features and characteristics of the three production size categories will be detailed. This will be done in the context of the two production models (traditional and NMAPC) of the Nuba Mountains area. Structure of the LP Models Introduction In this section the details of the production models are discussed in terms of the three major parts of the LP model: the objective func- tion, the activity set, and the constraints set. First, a general out- line of the basic production model and its variations is given. Figure 4.1 is a schematic representation of the basic production model. The agricultural operations of the four crops are represented as different activities. Two sets of the same activities are included to 90 4223 a: 3...— :ie. 23.33.39. .3: ..2.: —33 .o 225— .33.: Fate: a. upmom we. we cowumecmmmcamm u_umsmsum a ... oe:m.a .33: .83- e5 ... 32.8! e... .e .25. . ... .23 ......e .... ...... .83 3 =3. .9. «53.33“. a u ......uzseou 9.33.8. - o .638 a .3. a... .33. 3:03 a .8 "9.3 . 8 23.8 . 5. some, 1151135 1135» ...... aoww coconbc - .-H .- Bane Imam. .' “I w no» .3: 8 5:23.; a a .-.-.- -.-.--.-.- .- .- ...1.- .- .. ..- "......llulcza. :1)... - I 1115M ILSAM° ...m 3 .... . 3 ...: 3.3 a 3 we ... . o . 1 a. z .8 an .5... .- . ...... a... ...... ..2.. .‘i 5:3. ....e .. ... ...... .5... .. u ...... .... 9 .vi 9 ... ...... .133 u u . l ...e .33 u 4 . . ...... ...... .e - .3~ 83.3.... t 88“ 5.8.8. o w. 55. .833 a o 3. 8. 2 8..-... . 7 r 8. 2339.08 . .- a. S. . 3...? 1 an. .. ... S. 33 ”32932.3 9 o a a p. w. w. e . 5:93 5.83. s . a m m a m a W 3 . ".... 2.3.... 8...... 1 01 d represent 'early' and 'late' cultivation. The main difference between these two agricultural operations sets is their labor time distribution. Table 4.3 gives the timing of these crop operations by model periods. Also, in the basic LP model, labor in both the activity and constraint sets is disaggregated by the three types of labor [i.e., family, hired, and nafir (exchange labor)]and by the twelve monthly model periods. This basic model is modified in a number of experiments to permit analysis of the following: 1. Traditional Smallholders a) Analysis of the three production size categories. b) Crop yield sensitivity analysis. c) Planting time experiments. d) Credit experiments. e) Cotton price variation experiment. 2. NMAPC Participants a) 'Status-quo' model analysis. b) 'Full-phase' model analysis. c) Credit experiments. d) Cotton price variation experiment. The Objective Function The identification and specification of small farmers' production objective(s) represent a special problem which has been the subject of much theoretical and empirical study. Two main objectives have received much emphasis in modeling small farmer decision making. These two ob- jectives are: S12 ...eoa.aoaxo 53.. we..ea.a as. so. .... e. ...am. .me.u:o—n no us.a a. aepucouuo can. can xpcouu .aaaa zoseam co\mmm— seem vouuaeumeou .uuezom :o..oeenuca :o—uoeoaogm :o.uacoaoem can. use. ago. so: N— eopuoeanocm :o.uoeanoca eo_aocaaocm =o_uacoqmem can. can. need was. .sa< __ eo.ue~qoea nee. .co: c— .nom m me_:moc=h m:_xu—g .eaa m uevzmoegh m=.=moegh ae—uaau me.moegp m:.xu.m .uoo aN oe...:u m=_mocsh a=._.=a m:...=u .>oz am m:.:moggp me...=u ae__.=a ..oo .8 ~_ m=.voa: __ me_uooz __ me—uuoz .aom av . o=.uooz __ me.eooz _ me.ooaz __ a=.eooz __ u=.uoo: ._ a=.eooz . me.eooz .. m=.oooz .a=< an _ ae.oooz 5:.ueopm _ mepuoo: m:.u:o_¢ _ ae—uooz m=.ueo—m _ m:.uoo: o:..=o—a _ m=.uoo: ._:a «N a:..=o_a me..eo_a a:..=e.a ae.oee.a .eaa . was. a-eou mung x—cou can. a—cou can. x—cau sweet .82 «somom was: cesocu esgmcom eouuou wee—aoeoac pocaupzu.emmlu.mom vo—Loa —muo: no.2oa .oooz nee aosu Na ago..eaoao eo..e>.e.=u .o eo..:a_a.m_a m.a o_eee l. The maximization of some formulation of profits, output, or income.1 2. The satisfaction of basic family consumption requirements. The first objective (profit-maximization) was derived from the theory of the firm. Objections to the use of the profit-maximizing goal in analyzing traditional smallholder production question the validity of the theory's assumptions when applied to the case of the small farmer ”firm." Production in the latter case, it has been argued, differs in the nature of its resources, organization, objectives, and the place it holds within the complex of the traditional household.2 Production function studies have generally assumed some maximization of profits or net revenue [19, 36]. In these studies the test for pro- fit-maximizing behavior involves analyzing the resource allocation of the smallholder. A necessary and sufficient condition of profit-maximization requires equality of the marginal value product (MVP) and marginal factor cost (MFC), both for each input and across the inputs set . Although these restrictive conditions were not found to apply en- tirely in all of these studies, still, many of the inputs were found to be allocated efficiently by farmers [36]. Also, incorporation of the second objective (consumption/security) results in different optimal con- ditions. As Massel and Johnson note: 1Utility maximization has been shown to be the general case (Dillon and Anderson [19]) of which profit maximization is one of the special cases. 2For examples of the earliest literature discussing the issue, see Chaynov [l4], and Roden in Clifton [15]. 94 But if the farmer emphasizes security rather than profit maximization, the standards of efficiency are different. One cannot gauge efficiency by examining economic performance in a single year only; rather it is necessary to have time series data to permit analysis of the farmer's success over time in achiev- ing self-sufficiency....or else to have information on variability of crop yields and prices [36, p. 29]. A further criticism of the profit maximizing objective, in the con- text of smallholder production, has been offered by Lipton [34]. In the case of farmers producing in a risky production environment, he states: Owing to rainfall variability, there is no unique marginal physical product (MPP) associated with any factor, but only a probability distribution of MPP. By acting as if he used the calculus of expected values, an optimizing peasant can nevertheless find a long-run, profit maximizing algorithm analogous to marginal value product equalization. However, in the nonequatorial tropics, rainfall variance is much higher than in most temperate agricultures, so that for rainfall and hence MPP's - expected value is a much poorer predictor of actual value. In particular, the smaller is mean rainfall, the greater is the coefficient of variability [34, p. 330]. Under such a condition, Lipton proposes instead a "survival algorithm“ that constitutes an "explanation of rational, security-centered peasant conduct." Many empirical tests and studies by Norman [48], Wolgin [60], and others conducted in less 'risky' environments seem to support the , relevance of both profit maximization and security objectives. In this view, the smallholder is considered as an "efficient, risk-averse" producer. . In the Nuba Mountains area, both objectives are relevant, hence they are incorporated in this study. In addition to the safety-first 95 feature described below, the smallholder in the area practices other se- curity strategies which are not represented explicitly in the model. For example, production on the "Jubraka" (home plot) is mainly oriented towards the household consumption/security objective. Livestock owner- ship and investments also constitute an important strategy followed by smallholders in ensuring against risks of their crop production activities. The LP models used in this study maximize net revenue from crops, but ensure the production of sorghum to at least satisfy household con- sumption needs. This, in other words, is net revenue maximization sub- ject to a safety-first constraint. Other risk-elements are also con- sidered in special experiments analyzing crop yield sensitivity and planting-time strategies. Mathematically, the statement of the objective function and the LP problem is as follows: max R = n s.t. Z a. x. < b. and x. :_0 where: R = Net revenue, or total returns to fixed inputs (i.e., to family labor and land). C. = Net revenue/feddan (i.e., Gross Revenue-Variable Costs) of the ith activity. ai. = Per feddan input-output coefficient of the jth resource 3 used or contributed by the ith activity. 0. = Vector of resource availability. x. = Level of activity. 96 The following two sections give a description of the structure and estimation of the activities and constraints used in the model. The Activity Set The activity set includes the following activities: 1. Crop Production Activities Labor Hiring Activities Dura Consumption and Buying Activities Selling Activities 01-pr Transfer Activities 1. Crop Production Activities The crop production activities are the core of the model. As such, their specification determines to a large extent how well the model re- presents the two smallholder farming systems considered in this study. Collinson [17] discusses three methods of choosing a representative model: a) The choice of selected farm(s) which are representative. b) The use of averages (or any other measure of central tendancy) in constructing the representative model. c) Synthetic (component by component) construction and estimation of the representative model. In practice, the use of the selection method is greatly limited by the need to specify and weight the criteria to be used in arriving at a representative farm(s) (Collinson [17]).1 The use of the other two 1Not only are these difficult to arrive at, but the method requires the tabulation and ranking of all sample farms (by these criteria) in order that a representative one(s) can be chosen. 97 methods is much more common. Generally, depending on the sample in question, the use of averages in synthesizing a model unit from sur- vey data brings with it the problem of aggregation bias. An important aspect of this bias is reflected in the constructed average profile mis- representing the observed labor use profile. This problem would be more pronounced and serious if the interfarm differences in agricultural opera- tions timing (agricultural calendar) are large. In such a situation, peak periods in one farm are offset (through the averaging process) by slack periods in another so that the whole labor profile is flattened. The third method of constructing a representative model, which is essen- tially a modification of the averaging method, is particularly relevant and appropriate in dealing with this problem. This was the method used in this study. For the two farming systems in the area, three components of the production activities interact to determine the shape and hence represen- tativeness of the resulting labor profile.1 a) Agricultural operations. The NMAPC smallholder models differ from the traditional model by including only dura and cotton in the rotation. Two additional activities representing mechanical ploughing are also added to the NMAPC models. b) Calendar of agricultural activities. The timing of the above operations is an important element of the labor profile. This component of the production activities was estimated 1From the supply side, the distribution of family labor (by period for the three categories) is obviously crucial, and is discussed under the constraint set. 98 1 Two timing from the survey data using modal times of these operations. schemes, "early" and "late" (based on the time of planting), are repre- sented in the traditional models. The second scheme ("late") was added after initial tests and runs with the basic model and was found to be useful in improving the representativeness of the model in general and that of the cropping pattern and labor profile in particular. In contrast, the NMAPC smallholder model includes only one calendar of operations. The NMAPC of mechanical ploughing activities, which are done very early in the season, results in a generally similar time pat- tern of activities for all NMAPC participants. In the design of the LP model, the operations of any one calendar are forced together in sequential order (see balance constraints dis- cussed later), but the program is allowed to choose either "early" or "late" operations for crops entering the optimal solution. - c) The per feddan labor coefficients of the agricultural activities. This third component of the production activities affects the labor profile in a more visible and straightforward manner. Averages from a selected group of fields (discussed below) were used to estimate these coefficients. Other sources were used as the basis for adjusting the estimated coefficients. Ensuring a representative cropping pattern did much to achieve representative labor profiles. 1The mode and the second most frequent value were used. 99 Table 4.4 gives a summary budget for crops grown in the two systems. The labor cost component actually depends on the type of labor allocated to the production of any one crop. The three types of labor available to the smallholder (family, nafir, and hired labor) are represented in the model. Their levels and distribution by model period (for the three categories) are discussed later under the constraint set. In Table 4.4 it was assumed that 25 percent of total labor is hired, in order to facilitate the comparison of total costs and returns both between crops and for the two systems. The disaggregation of total labor by different activities across the model periods is shown in tables 4.5, 4.6, 4.7 and 4.8. These tables show the production activities of the traditional farm model for cotton, dura, groundnuts, and sesame, respectively. Differences per- taining to the NMAPC model are also shown, and discussed in detail later. In what follows we discuss the procedures and assumptions adopted in the estimation of coefficients for these production activities. In general, estimation of coefficients involves not only statistical. rigor but also appropriate subjective judgements. In the case of this study, FAO survey data were available to compare, cross check, and adjust the coefficients estimated from this researcher's 1979/80 survey data, using 71 selected fields with complete labor use records.1 Estimation of these coefficients was based on statistical averages. The input- output coefficients are calculated on per feddan basis which is the 1Each of the 71 fields has at least a complete labor record for one crop, but not all fields have complete labor records for all four crops considered. 100 Table 4.4 Per Feddan Crop Budgets for Traditional and NMAPC Systems Traditional NMAPC Items Ground- ., Cotton. Dura nuts Sesame Cotton Dura Variable Costsg(Ls) Tractor Ploughinga - - - - 1.6 1.6 Seeds/Chem. 0.0 0.50 5.17 0.57 0.0 0.50 Total Laborb(m.d.) 43.54 28.91 35.28 29.75 29.70 20.22 Hired Labor (LS) 7.24 4.81 5.87 _ 4.95 4.95 3.36 Sacks/Transport 1.96 3.96 6.63 2.57 1.96 3.96 Total V.C. 9.19 9.27 17.67 8.09 8.50 9.42 Returns (LS) Yield (Sacks)c 3.0 4.78 5.15 2.79 2.23 3.58 Price/Sack 3.80 6.60 4.70 9.20 ‘3.80 6.60 Gross Returnsd 11.40 31.55 28.95 25.67 8.47 23.63 Gross Margins 2.21 22.28 11.28 17.58 - .03 14.21 Source: Computed from (1979/80) survey data. aA fixed charge of Ls 1.50 per feddan. One Ls = $2.0. bAssuming 25 percent of the required labor is supplied by hired labor at an average cost of LS 0.665 per man day. cSack weights of different crops Cotton = one Kantar (45.0 kg) Dura = 90 kg Groundnuts = 45 kg Sesame = 75 kg dActual gross margin would vary according to the percentage use of hired labor (see b), but the figures are indicative of the relative pro- fitability of the different crops. 101 .23.... E ..8: ...8: 2358...... 3: .3: .38 .o ..2.. .....a: ... 3333 .8 :26— 3.33..- .33‘ 2‘! .83.. 3.3 ...8-.938. 38:88 "3.53 .2. a a.” a 8.3 n .— a... o 3 u... ... e 3 w o 6.2 . .... e 3 u. ...: u 3 um ...: . a. n ...... 8. o 3 u: .58 e 3 . : ...8 a 3 We. ..8 o 3 Ho ....em a 3 no .21 a 3 w . ...... 2..: .. no .38 ...: 8 ...... u ... 3.-.. 8 ......u . .... 8 ...... u . ...... 8 ...... n ~ .... .... «N ......u . ...... N... ...... u N. .8. «N a..." : .2... 2. ON ....I n 2 Joe .3... 2.. 2. 8 ......u a ..8 and 2.... 8 ....a... a .21 2.. ...... 8 .9... . . ...... 3.. 8 ...-n o .63-... . . . a T ..e a a a a a ... 4.4. a d... a u a d a a ..- .3. q o .: :23 .. .2. : aw. ...! .5. ... 3.. a. 3... at : =2. .5... a: g .. =3 : aw. ...! ...... ... 3... a. 3... t. . B (B 6:8 33 . .. we...>..o< eo..o=eo.a eo..ou ... o.ee. 102 55"“ .23....» .3 .32 a..-» 2.3.8.939. 3: .3: .33 5.. 225. 73...... .2 C333 ..8 :28— 3.530... ...-...... 8‘! 8...... ..8 35:88. 932‘... "3.33. T — as..- 2.7 "a; m :53 33 VIVIVIVUVIVIVIVIViVIVIVIvlvlvlvlvlvlvtvlvalVIAI a 0.0: 00.” OO.M r .... .... a... .... 2.. . N... o... 6 aaxaszzuaaaz§°°°°°°°°°° . . . .....- ad ”a. a o a m . u - .1- a ‘4‘. a a .. ...: 1-- d-.- a «a = 52. _ 5»: Z a! . a! 5.: ... 3.: ._ 3.: z = 32. _ 52. ._ ..8. _ cl. 3‘ a: 33; a. at a .3 3.. i. ||§I ||I 9. p..- mm...>..u< co..u=uo.. o.=a ... m.am. 103 .32.: .... 3.... ...... 83.32.... .... 3.... .33 ... ..2.. 23...... ... 9.383 ..8 :25— 3.533.- ...8... 8:! ..8... :8 333...... 3.3!... 8958 59- .1 3.0 2.... mn.~ :6 2.0. 3.0 n ". 801 N :8 .36 21-8 eeoeooooeoo -----nnuuauunuuunu h‘I'ICddddddddddddd vlvlvlevhlhrlvlvlvlvlvlvlvlvlvbvlvlvlvlvcvtvlvlnlv 2.... no.3 Q6 S§§53§£i§§1333§3§§£55 and ..2. .33-: UUUUU Ibfl-hblhb :6 an... g «a..- o .. 5.... . a... % Jo Z6- 2 a! _ a! pat LE Ilifl ur a..- J ... 3.. .. 3... 2 LB mm... = as: a a a 2d- ur a..- _ pm»: : nu: _afl. E a: at u— at - a ...: . Hr . 3 L mm...>..u< =o..u=uo.. .==u==O.w ... m.nm. 104 42.3.3 ...: .35: 3.1: 2.3.33.3... 3: ..2.-— .33 so :0»: pas—£4 .2 5333 .5.— So‘. 3.530..- ..33 9...... 3...... ...8 3.2.938. 98.5.89 .333 . 2d- o..~- a w ..8... .3. 3.3 3 n =8 e... o 3 w ....-8 o 3 n ....2 o .. w ...... a 3 u do. a 3 n ...... «.9 3d a a. u J8 .. «. ... .3: ...: ~ g 3 u ..8 e 3 w ...8 a 3 u .21 3. o 3 w ...... S. 3.... 3 w ...... 8.. 8 .....- n 3.-.: 3.. 8...... w ......2 8...... u ...... 8 ...... n ...o. #6.... u ...... 3.... 3.~ - 3. w 68 3.. a... - .... n .6: 8 3. w ..8 8.. 8.. on S. N ...am 2.... $6. a .... u .3. 8..” a: on 8.. w ...... 8 .... v ...8-.. b p .b 3 v we: «6 3...- o o e 3.- o: d..- a S.N. a a a 1.... a: a..- a . a... .. 52. .52. .. a... ...! .3... ... I... .. I... s! .. pm»: .52. .. 3.. ...! 5... ... .8... c. 3.. E a“. a a d... Ba Ba BEE... ...... ...... mm...>..u< co..u=vo.. mammmm m.. m.... 105 unit of the production activities. In tables 4.5, 4.6, 4.7 and 4.8 coefficients are assigned a positive or negative sign to indicate the use of or addition to a particular resource inventory, respectively. a) Objective function coefficients: All coefficients of the crop production activities except for plant- ing (PLNT) and second harvest (HVST II) operations carry a zero value in the objective function. This is because the only cash cost component of these operations is that of hired and nafir labor, which is accounted for by the hired labor (HL) and nafir labor (NL) activities in the model. The objective function coefficients for planting represent the average cost (in Ls) of seeds and chemical dressing per feddan. As for second harvest (HVST II) the estimate is for the cost of sacks and transport from the field to the home (in case of dura) or market (in case of other crops).1 b) Labor coefficients: Coefficients are average total man-days per feddan for the particu- lar operation. A man-day is equivalent to seven hours of work provided by adult males, but females and children also participate. In computing man-days, labor for weeding provided by children (male and female, 0-15 years of age) is discounted by a factor of 0.5, based on the assumed lower productivity of this group in weeding. The labor input contributed by all three types of labor: family labor (FL), hired labor (HL), and nafir labor (NL),was totalled in deriving the average per feddan coeffi— cients. 1In the case of cotton this would be the nearest collection center where local cotton markets are established. 105 Table 4.9 compares the per feddan labor coefficients estimated from available sources. As one might expect there is no exact equality between the three sets of available estimates, but overall they compare reasonably. c) Operating Capital Coefficients Operating capital (0C) is used mainly for hired labor expenses; these are accounted for directly through the hired and nafir labor (HL, NL) activities of the model which draw directly from the available OC resource stream. Other cash requirements are those of planting and harvest opera- tions. Planting (PLNT) includes seeds/chemical dressing. As for har- vest (HVST II), the coefficient represents the costs of sacks and trans- port. 2. Labor Hiring Activities Twelve activities (one per month) for both hired labor (HL) and nafir labor (NL) are incorporated in the model. Table 4.lO shows the labor activities portion of the LP tableau. The unit of the activity is man-day (equivalent to seven hours a day). The availability and use of these resources is governed in the model partly by the availability of operating capital (DC).1 No limits were set on the use of HL (other than that implied by the availability of OC); however, the NL activities are restricted to the upper limit specified by the correspond- ing constraint set- 1Including replenishment of 0C inventory internally in the model (through crop sales) or through borrowing activities (in some of the eXperiments). 107 .>m>.=m m..m=o.mwmmm mg. :. couaou so. co..~.aam.a u:m.u=m.m.:.:.:o.pacmaw.a can. mwu=.u:. a .msmmmm .mvcz “smegma mm ucm mgau Lave: coca acousma m. g..: gotta. a so. um:.nsou mcm mmuoe.umm as; mgsux.s a uozm .mo. ... 53 332% .3258. 3.55.. so... 3.2.33 525.. ...8 3 3.32.8 m .mnxm.m. .»o>.zm c<. .m.asmmnn=m .mco...umg.~ .om\m.m. .»m>.:m m..mgu.mmmmm .m.asmm-n:m .m:o.u.uwg.. "mmogzom .mmogm.o ..m ..m. o.o. ax.mum umm>gm= mm.m No.o em.¢ ... umm>cmsv me.. .N.o .e.o .. amm>cusv em.o. .m.m o.m e.¢. m.. m.m ..... mm.~. m..o. m.w. umm>go= mo.. mm.~ N... a... ... m=_uaazv mN.o. em... .... mm.e .. a=.uamzv o... mm.o e.o. m.m. N... ~.m mm.e. m..¢. om... m..~. m:.umm3 ¢.N m.¢ o.m m.~ mm.m e¢.m o..m .N.~ m:.u:m.m m.. m.m ~.~ m.. mm.. om.m o..m co.un.mam.a new. muzz «mammmm :ouuou mammmm muaz «can coupou wsammm musz acne coauou .w \mgao .o - .c mmhx mxw>cam ow>.:m m..m;u.mmmmm =o.umgmno "soc. .m..c.e :.v mmum5.umu 39.38 am...» :5... 3:23:80 .85... 22.3... .8. .0 533%.... m... «32. ..A.aa:m ... cogs. a..e~. a=.a:omo.na. ....V Lona. .auoa we m.o>u. .....:.a ... auoouauu so. m.u>o. oucaomoxn gauze-cu "ousaom 1CN3 ma..~ a. ...u can .3..- 8 ..n< ...: .30. .cae mm. .c. mm. mm. mm. .9. mm. mm. mm. .0. mm. pa. mm. ... mm. an. an. ... an. em. mm. Nu. .csa- u: .5e..= .goa — No.- In “9 mvnccmmvoccoaoooocococoo on -J p IDIOQ o I o .u .n cu p I F I O N '— I P I O N p I p I O N p I P I N N p I p I P I '— I NN NN 8 8 8 p I p I vlvlvlvlvlvlvlvlvlvlvlvlvlvlvlvlvlvlv1vlvlvlvlvIvlvlvlvlvlvlvlvlvlvlvllel — I p I O N llH 1|. U - nn.u mm.- mm.. mn...mn... mn.u nm.a mm.. mm.. mm.u mn.u nn.- .c.i no.. .0.. n... ma.» . um.o ans. men. can. Mme. .- O .1 m ZlTN ll1N 011M nmwmmwmmmmmmmm .... I. 9 S .7 8 z I. o o oo.- ... m... ....mm... Loan. ..mmzl annalwomwnl mm.._>_.u< Lona. o... m.nm. 109 a) Objective Function Coefficients Coefficients for HL are averages of monthly wage rates, estimated from this researcher's l979/80 survey data. For HL, both cash and in-kind expenses are incurred. For convenience here, in-kind amounts ' were translated into cash values in Ls. Also, activities which in reality are not paid per man-day, but per unit land (as sometimes for weeding), or per unit production (as in threshing and cotton picking for example), are also converted in the model to a per man day basis. For nafir labor (NL), the estimated average value of the in-kind costs per man-day (Ls 0.33) was applied in the objective function with no monthly variation in the rate. b) Input-Output Coefficients Both HL and ML activities directly supply the total labor (TL) re- source inventory.1 This is represented in Table 4.l0 byra series of ne- gative ones (-l) to indicate addition to the TL resource. In the 0C rows, the HL and NL coefficients are the corresponding wage rates that appear in the objective function. 3. Dura Buying and Consumption Activities A dura consumption activity' is included in the model. The con- sumption activity has a unit of one sack (90.0 kg), with a zero value in the objective function, implying no cost to the household from the activity. It has a coefficient of positive one (+l) with the dura con- sumption constraint, to indicate satisfaction of the constraint by one unit. 1Theinitial supply (right hand side) values for the total labor rows are estimated averages of family labor (FL). 110 A dura buying activity is also included in the model, to be activated only if farm production falls short of satisfying the required level of dura consumption. It has a coefficient of -l in the consumption con- straint row, indicating the replenishing of the dura inventory by one unit (90 kg sack). It has an objective function coefficient of negative Ls 11.0. This implies a price per sack of dura much higher than harvest prices which farmers receive for dura sold. The reason for this is that it represents the more likely situation of farmers buying dura well after the harvest period, when prices rise mach higher than their har- vest levels. 4. Selling Activities Four selling activities for cotton, dura, groundnuts,and sesame are represented in the model. These are shown in Tables 4.5, 4.6, 4.7 and 4.8,respectively. Selling is assumed to take place immediately following harvest. Prices used are the average prices received by the farmers in the- area . in the season l979/80. These are represented as positive co- efficients in the objective function. They also appear as negative co- efficients, replenishing the OC streams, corresponding to the selling months. 5. Transfer Activities These are activities which allow the passing of unused capital from one period to the next. The last activity carries over accumulated 111 capital to the next season. These activities have a zero value in the objective function and a series of +1 and -l coefficients in the corres- ponding 0C periods.1 The Constraint Set The constraint set of the model includes the following restrictions: 1. Farm Level Resources a) Land Restrictions b) Labor Restrictions c) Operating Capital Restrictions 2. Minimum Dura Consumption 3. Operations Balance Constraints 4. Non-Negativity Constraints l. Farm Level Resources The first part of this chapter has discussed the availability and distribution of farm resources for the three farm size categories. In what follows we describe the level of each resource made available in each model period by farm size category. a) Land Restrictions Chapter III described the land situation in the area. It was em- phasized that there is no current land shortage, and that limitations of land size are largely those brought about by the limitations of other production inputs (namely labor and operating capital). 1These activities are shown together with the capital borrowing activities discussed in the next part of this chapter (see Table 4.17). 112 Land referred to in this context consists of the field plots, locally known as Saraya. Such land is assumed to be homogenous. No distinction by soil type is employed in the model. The upper limit (in feddans) of each farm size category was used as the limit in the restriction. For the NMAPC models, the average of the existing farm size and the proposed tenancy size under full-phase development was used. Land restriction limits in the basic LP models are shown in Table 4.11. Table 4.ll Land Restrictions Limits in the Basic LP Models Model Restriction Limit (in feddans) Traditional: (i) Category I 5.0 (ii) Category II 9.9 (iii) Category III 25.9 NMAPC: (i) Status-Quo 6.0 (ii) Full-Phase l0.0 b) Labor Restrictions (l) Family Labor To derive the family labor supply for the three production categories, assumptions based on the observed ayerage labor use profile are employed. The FAO survey data, which has a larger sample size than this researcher's survey and a higher percentage of cases with complete labor records, was used. Table 4.12 gives the monthly distribution of family labor use by 113 Table 4.12 Monthly Distribution of Family Labor Use by Production Cate- gory (in man-days): Sample Averages category I Category II Category III Month (0.0-5.0 (5.1-9.9 (10.0-25.9 ' fed.) fed.) fed.) Jan. 16.5 14.2 18.2 Feb. 5.8 11.5 28.5 Mar. 4.4 8.1 14.8 Apr. 4.8 12.7 16.4 May 7.8 15.7 22.6 Jun. 13.7 23.7 34.3 Jul. 22.6 45.1 53.9 Aug. 24.8 42.1 62.7 Sep. 16.8 21.2 46.3 Oct. 10.9 19.8 24.9 Nov. 12.2 23.2 26.8 Dec. 12.1 '19.5 25.2 TOTAL 152.9 262.6 387.4 (125.2) (226.6) (365.8) ( ) = Standard Deviations Source: Computed, FAO survey (1978/79). 114 production category. Figure 4.2 illustrates the observed labor profile graphically. Three distinct labor periods can be observed. The first period ex- tends from last season's harvest operation until the beginning of the current year cropping season, which is from February through June. This period is characterized by the lowest labor use by the household.1 The second period is the planting and weeding period (from July through Sep- tember). This is the most labor intensive period. The third and last period is that of harvest and post-harvest activities (from November through January). This is again a labor intensive period, but less so than the second period. Based on the characteristics of the labor use profile described above, Table 4.13 shows the family labor supply levels in the LP models by category and model period. The figures are based on averages from the FAO survey data.2 (2) Hired Labor No restrictions on the amount of HL are employed in the LP models. HL use in the models is limited only by the availability of operating capital. 1October labor use also fits the pattern of this period. It repre- sents a slack between the weeding period and the harvest period. 2 . . In the first period (February through June) and during October, family labor supply is adjusted (upwards) to be 20 man-days per month for each of the three categories. 115 mm: Loam. a..sm. um>camno mmmgm>< .o co.umucmmmgamx .mu.:amcw m.¢ m.=m.. 3...... 2..; ..2.: vacuum 373.. an... .8: - .>oz - ..uuo - dam - .22 ...... l .5... as. 1 3:2. ...5. o I / / m o L n” m . m ~_. "I o. .. cm .. ¢~ L mm a .r mm mm .. ... ~m on .. .. on 3 . . cc 3 1. .I 3 me. -o. mm ... , .. .. wm cm 1 ... .2335 a I 1 om .. 1 cc 8 . . 2 .2838 u D on. . .. 3 . ... / we: , . .2338 n u an . um .. a x In N... .m52.=s==.. = a... .. . . N. 116 Table 4.13 Family Labor Supply (in Man-Days) in the LP Models: 13y Production Category and Model Period . Category I Category II Category III ”13‘?" (2.23-0 $.13: “0.2.35: 1 Jun. 20.0 20.0 20.0 2 Jul. 22.0 36.0 54.0 3 Aug. 22.0 36.0 54.0 4 Sep. 22.0 36.0 54.0 5 Oct. 20.0 20.0 20.0 6 Nov. 20.0 22.0 26.0 7 Dec. 20.0 22.0 26.0 8 Jan. 20.0 22.0 26.0 9 Feb. 20.0 20.0 20.0 10 Mar. 20.0 20.0 20.0 11 Apr. 20.0 20.0 20.0 12 May 20.0 20.0 20.0 (3) Nafir Labor Unlike hired.lab0r, nafir labor (NL) is governed by different in- stitutional arrangements which limit its availability. Usually this type of labor is available during periods of peak demand for planting, weed- ing, and harvesting. Limits employed in the models are derived from this researcher's survey data for the three farm size categories shown in Table 4.14. 117 Table 4.14 Nafir Labor Availability (in Man-Days) in the LP Models: By Production Category and Model Period . . Category I Category II Category III “133‘?“ 9.23: $.13: “0.2.35: 2 Jun. 3.33 5.58 2.70 3 Aug. 3.33 4.58 2.70 4 Sep. 3.33 3.53 0.00 6 Nov. 3.33 5.42 1.67 7 Dec. 3.33 9.21 8.20 8 Jan. 0.00 4.10 1.30 c) Operating Capital Restrictions In Chapter III, the sources and uses of 0C are discussed. Small- holders in the region need cash for both consumption and production uses. The sources are also varied, including savings, crop and livestock sales, off-farm earnings, and borrowing. Borrowing from traditional money lenders (the "sheil" system) is widespread [8]. In general, the estimation of operating capital availability is a difficult problem in a small farm setting. For this study, amounts spent on crop production estimated from this researcher's survey data were used as a proxy for farm 0C. The estimated figures were Ls 6.93, Ls 27.83, and Ls 152.27, for the three categories,respectively. A11 0C is made available in the first period (June), and unused amounts are trans- ferred as needed to the subsequent periods. The supply of DC in the models is augmented by crop sales later in the season, and by capital 118 borrowing activities in the credit experiments. An ending capital restric- tion is imposed in the models to ensure that an amount at least equal to the starting 0C is available for the next season. 2. Minimum Dura Consumption The importance of dura consumption for smallholders in the area has already been discussed in Chapter III. It was also pointed out that devoting a minimum feddanage of dura to at least satisfy household con- sumption needs is one of the risk strategies practiced by the smallholder. A study in Eastern Sudan EMT] estimated that 14 sacks of sorghum are con- sumed by a slightly larger than seven member household. Average dura consumption estimates from this researcher's survey (see Table 4.2) were 7.05, 8.76 and 11.57 sacks for the three farm size categories, respec- tively. These figures were adjusted based on average family sizes (5, 6 and 7 members) and it is assumed in the models that an amount equal to eight, ten, and twelve sacks (90 kg) must be devoted to consumption in categories I, II, and III, respectively.l 3. Agricultural Operations Balance Constraints These constraints are employed in the model to force the cultiva- tion for a particular crop timing to occur together in the proper sequence. A series of (-l) and (+1) coefficients are used to place land at the dis- posal of an agricultural operation, and to transfer it to the next opera- tion, respectively. (These levels are judged to be reasonably on the safe side of con- sumption, since>they don't include the "Jubraka" (household plot) production. The latter (not accounted for in the models), is devoted mainly to house- hold consumption early in the season (August-October). 119 4. Non-Negativity Constraints This set of LP constraints requires that no activity be operated at a negative level. Experiments and Changes Made in the Basic LP Model The basic model described above was modified in order to carry out the planting time and credit experiments, and to represent the NMAPC farming system. NMAPC Models Only cotton and dura crops are grown in the NMAPC schemes. Both the time distribution of the agricultural operations and the input-output coefficients are different in the NMAPC model. Chapter III compares the NMAPC and traditional smallholder farming systems. Further description of NMAPC farming system is given in Chapter VI. Changes made in the basic model in order to represent NMAPC farming conditions are as follows. a) Land Size and Crop Composition Both the tenancy land size and crop composition are determined by NMAPC. The initial "status-quo" model has a size of six feddans, divided equally between cotton and dura. The contemplated "full-phase“ model has a size of fifteen feddans, to be grown with cotton and dura (five feddans each), and five feddans to be left fallow in the rotation. b) Mechanized Cultivation Two mechanical plowing activities are added. On NMAPC farms, simul- taneous with the second plowing operation, dura planting is done mechani- cally through a mounted box (seeder) in the tractor. Therefore the dura planting coefficient is zero. The cost of these operations in practice is subtracted from the farmer's cotton returns at the end of the season.1 These costs are Ls 1.0 and Ls 0.6 for the first and second plowing, re- spectively. In the LP models these charges are costed to the objective function and to the DC of March. c) Agricultural Operations Calendar Only one calendar of operations is represented in the NMAPC models, since the delaying of the mechanical plowing (under the control of NMAPC) results in a late and similar activity calendar for all participants. In the NMAPC models,operations are assumed to take place as shown in Table 4.15. Table 4.15 Agricultural Operations Calendar in the NMAPC LP Models Activity Time For: Activity cotton Dura PLNT August August NEED I August August NEED II September September HVST I January December HVST IIa May January aFor cotton this refers to cotton stalk-uprooting and disposal ("AL-awdi"). 1If a farmer is not growing cotton, he is supposed to pay these costs in cash at the end of the season (around March). 121 d) Input-Output Coefficients The labor and yield coefficients for the NMAPC model are also dif- ferent from those of the traditional model. Based on estimates from the 1979/80 survey data, the weeding, harvest, and yield coefficients of the NMAPC model are given in Table 4.16. Table 4.16 Labor and Yield Coefficients in the NMAPC LP Models Item , Cotton Dura Labora NEED 1 4.45 4.87 NEED II 2.40 2.66 HVST I 13.30 4.29 HVST II 7.04 2.19 Yieldb 2.23 3.58 aAs before,labor units are 7 hour man-days. bCotton yield in small kantars (100 1b); dura yields in sacks (90 kg). Credit Experiments Capital borrowing and repayment activities are added to both tradi- tional and NMAPC models in experiments to evaluate the potential contri- bution of credit. As described in Chapter III, only non-formal credit sources are available to smallholders in the area at present. However, NMAPC contemplates extending credit to its tenants in the future. 122 Table 4.l7 shows the borrowing activities together with the original capital transfer activities. The unit of borrowing, transfer, and repay- ment activities is one Ls. Coefficients are a series of negative ones (-l) to indicate addition to the corresponding monthly 0C resources. A series of positive (l.07) coefficients ensures the repayment of princi- pal plus seven percent per year. Objective function coefficients of the borrowing activities (-.07) represent an annual interest rate of seven percent.1 The figure was chosen because it is the official interest rate charged by the Agricultural Bank of Sudan for short term loans. It will be discussed later in the context of interpretation of model results. Planting Time Experiments These experiments are designed to shed light on the effect of three different planting times. Each of the three time- of planting sequences has different yield coefficients. Therefore a third calendar of opera- tions was added, and coefficients of weeding, harvest and transport were adjusted proportionally.2 Resource availabilities are disaggregated in half monthly periods for selected months (July through December) for the purpose of these experiments. A summary of the changes in model design made for these experiments is given before the results analysis in the next chapter. The basis of the experiment's assumptions and the full account of experimental data used.are given in Appendix II. 1Actually the use of the same objective function coefficient (-.07) for all borrowing activities slightly over estimates the cost of borrow? ing for the later borrowing activities. However, given the small interest rate used, resulting discrepancies are insignificant. 2For dura only one timely weeding is required in these experiments (see discussion in Appendix II). 123 .2 9833-... ..3 223— 3.532.. $339.3 "3.50m T 3.. 3; 3.— 3.. 3.— 3._ 34 3.. 3.— 3.. 3.— 3.— a 3 w ~33. Eu _- 3.3 3 M Eu Ea _ T T o 2 w aux-8 _ T T c 3 w ...:E p T T c m, w ...o: _ T T o 3 .u .53 — T T o 3 .w ...2. _ T T o m. a doc _ T T o a. .. >oz — T T o 3 n .33 — T T c 3 u dam — T T o 3 .- .m:< _ 7 .- o as” 4% p T 8.3 u. . ...:Eo o c o o o c c o o c o a c 3.- 3.- 3.- 3.- 3.- 3.- 3.- 3.- 3.- 3.- 3.. 3.- a a an m n n 0. .... n m u n m m m n n n m w u .... n n n n a . . M up I. «In 5 m H 9 C. h K R / 2 H m 5 on I. 9 c. V E 7.. I. u .A U U I. 6 8 U on S 7 ..2. a 7. I. 0 l .. ..u 2.2 3 mu .6. up cmzfsmn m$fi>3u< Latest. 3338 use mcmzotom 24‘ ~33 CHAPTER V TRADITIONAL SMALLHOLDER PRODUCTIOH: A LINEAR PROGRAMMING ANALYSIS The previous chapter described the production models and their LP structure. This chapter utilizes the linear programming models to ana- lyze smallholder production.1 Information drawn from the LP solution includes the value of the objective function, the optimal enterprise combination, levels of resource use and their respective marginal pro- ductivities (MVP's), the nonoptimal activities with the cost of forcing each of them into the optimal solution,and the stability limits of the optimal plan. The analysis here focuses on the optimal cropping pattern and the associated resource use and productivities. Resulting changes in farm income are also discussed. The organization of the analysis and discussion in this chapter generally follows the order of objectives stated in the first chapter. In the first part the base models of the three traditional production categories are discussed, together with results from hypothesized changes in rainfall variability. The second and third parts investigate resource use and productivities under credit-aided resource expansion and planting 1The LP program used was the AG ECOH LINEAR PROGRAMMING PACKAGE Ver. 2.20 (Stephen B. Harsh and J. Roy Black, 1975)_28]. 124 125 time experiments. The fourth part discusses cotton production problems in traditional agriculture, and presents results of the cotton price variation experiment. Within the above analysis, relevant policy issues are discussed. A brief summary of the results is given in the last part of this chapter. Basic Solutions and Optimal Production Plans for the TraditionaI SmaTlhoner Categories The validity of optimal solutions derived from LP models is in general limited by the degree of accuracy and representativeness of the model's assumptions and coefficients. This, in addition to the nor- mative nature of the LP analysis in general, suggests a need for caution in using these optimal solutions for interpretation and inferences re- garding the farmer's behavior. The optimal solutions and production plans of the three traditional smallholder farm size categories are given in Table 5.1. Cropping Patterns Despite the tendancy of LP to generate highly specialized enterprise plans, the results obtained in this case are fairly comparable to the observed cropping patterns of smallholders in the region.1 This is due to the dominance of dura and dura-sesame combinations in the cropping pattern. The dura consumption constraint might be expected to explain the large share of dura feddanage in the plans, but in fact this is mostly due to the relative probability of the crop since in all three farm size 1Comparisons with the observed situations are made with Table 4.1 (Chapter IV) for cropping pattern, and with Table 3.12 (Chapter III) for the time distribution of the cropping pattern. 126 Table 5.1 Basic Optimal Production Plans for the Three Categories of Traditional Smallholders Category Category Category I II III (0.1-5.0 (5.l-9.9 (10.0- Item Unit ' fed.) fed.) 25.9 fedi), Cr0pping Pattern: Early Dura (0R1) Feddan 3.94 5.91 12.05 Early Sesame (5M1) Feddan -- 0.69 2.67 Late Dura (0R2) Feddan -- 2.32 6.29 Late Sesame (5M2) Feddan 0.49 0.25 2.42 Resource Use: Total Land Feddan 4.42 8.85 23.88 Family Labor Man-Day 110.03 178.93 275.81 Hired Labor Man-Day 11.51 56.50 403.88 Nafir Labor Man-Day 6.66 21.50 14.90 Total Labor Man-Day 128.20 256.92 694.59 Total Operating Capital Ls 28.61 83.36 374.46 Objective Function Value: Ls 55.19 124.55 269.75 Total Gross Margina Ls 107.99 190.55 348.95 Average Productivities:b Per Feddan Ls 7.35 6.60 6.20 Per Man-Day Ls 0.84 0.73 0.49 Source: Computed aIncludes the value of dura retained for household consumption. bSee the text for assumptions and procedures for calculating these averages. 127 categories dura feddanage is well above the levels required to satisfy household dura consumption. The relative share of sesame in the optimal cropping patterns for the three categories is less than in the observed situation by 27 percent for category I, 10 percent for category II, and only 5 percent for category III. However, for both observed and com- puted optimal cropping patterns of the three categories, the absolute size of sesame represents small feddanage (especially for category I and II). This situation, despite the seeming relative profitability of the crop,1 is mainly attributed to the labor intensive and time exacting de- mands incurred at sesame harvest. Groundnuts does not appear in the optimal solution for any of the categories. In the observed situation farmers may devote a small area to the crop. Relatively large groundnut areas are observed only in the northern regions (Northern Abassya and Dilling) where there'are suit- able loamy soils.' Cotton does not enter any of the three optimal plans. This agrees with the empirical observation that cotton is almost disappearing from traditional farms. As the issue is of great importance to the government, the fourth part of this chapter will analyze government cotton price policy in a parametric price programming experiment. An important feature that is displayed by the solutions is the time distribution of the cropping patterns. Both early and late schedules of cropping were observed in the 1979/80 survey, but the major portion of cropping (80 percent and 88 percent for dura and sesame, respectively) is 1For a picture of the relative profitability of the four crops con- sidered, see Table 4.4 I28 done in the early schedule. In the LP optimal plans, however, 100 per- cent,; 72 percent, and 66 percent of dura (for the three categories, respectively) is in the early schedule; and for category II and III, which grew sesame, 72 percent and 52 percent, respectively, is done in the early schedule. In practice, adoption of both early and late crop- ping schedules by the smallholders in the area results from the need to smooth out the stringent labor bottlenecks, especially those developing early in the season. As discussed earlier in Chapter III, this practice is aided by the use of different time maturing varieties. The issue of cropping schedule time is further analyzed in an experiment presented in the third part of this chapter. Resource Use Levels of resource use for the three production categories are shown in Table 5.1. Total land use is 4.42, 8.85 and 23.88 feddans for the three categories, respectively. These figures represent 88 percent, 89 percent, and 92 percent of the total allowable land in these models, re- spectively. Total labor used in each category is exactly proportional to cropped land sizes, amounting to 29 man-days per feddan. Family labor use in the three categories is 110, 179 and 276 man-days, respectively. Nafir labor use is highest in category II, but in all three categories its relative share in the total labor input is very small. Hired labor increases more than proportionally when moving from the smallest to the largest category. Its relative share in the total labor input is 9 per- cent for category I, 22 percent for category II, and 58 percent for cate- gory III. This is essentially because of differences in availabilities of operating Capital (0C) for the three categories. The amounts 129 available in the models were Ls 6.93, Ls 27.83, and Ls 152.7, respec- tively. However, the actual amounts used in the optimal plans (shown in Table 5.1) are higher: This situation is allowed for in the models through supply of additional 00 from cr0p sales later in the season, for example sales of sesame, which occur as early as November. Returns and Average Productivities Monetary returns from the optimal plans are Ls 55.18, Ls 124.55, and Ls 269.75. Total gross margin figures, which include the value of dura produced but retained for household consumption, are Ls 107.99, Ls 109.55, and Ls 348.95. Such low returns are typical of the smallholder farming in the area and they have a number of important implications which will be discussed later. These low returns are the product of low producti- vities of both land and labor, resulting from low crop yields. Average productivities for land and labor are given in Table 5.1. Drawing on Parhusip [53, pp. 88-89], the following procedures and assumptions were adopted in computing these averages: l. The market value of dura retained for household consumption (8, 10, and 12 sacks for the three categories, respectively) is added to the corresponding objective function value to obtain total gross margin (TGM), or total value product (TVP) less the variable costs (VC). This procedure reflects the total returns to both land and labor. 2. In computing the average returns to land and labor, an interest rate of 7 percent per year is charged on initial amounts of operating capital (00) in the models. Although the intent is to account for the opportunity cost of capital used in the 130 production process, the 7 percent rate used probably under- estimates the real cost or productivity of 00, especially when 1 However. no reliable esti- inflation is taken into account. mate of the real rate is available, and the use of any other figure is esentially an arbitrary one.2 This particular figure is chosen because it is the official rate charged by the Agri- cultural Bank of Sudan for short-term (up to 14 months) loans. 3. To compute the average returns to land, not only capital costs but also all labor costs must be accounted for. It is assumed that the prevailing wage rates of hired labor (HL) in the area reflect the opportunity cost of labor provided by the three types of 1abor(family, hired, and nafir)represented in the models. Hence, total labor used in the optimal plans (provided by the three types of labor) is costed at the prevailing wage rates of HL. This procedure amounts to costing family labor (FL), which 3 is notcosted in the LP model, at the corresponding HL wage rates, 1No official statistic for inflation in the Sudan is available; how- ever, it is believed to be running well above 50 percent for the past few years. 2An idea about the productivity of capital within the smallholder farming in the area. can be had from the model results (discussed later with the credit experiment) on shadow prices of capital in the different periods. 3 This is not because, as some have argued (see discussion of the issue in Eicher and Witt [20]), of zero marginal productivity, but be- cause family labor is viewed here to be rewarded by net returns from crop sales and by the dura consumption portion retained from production. 131 and adjusting the nafir labor (NL) fixed charge of Ls 0.33 to the corresponding HL wage rates.1 4. In computing average returns to labor, no cost is assumed for land and other fixed assets. This judgment derives from the abundance of land in the region at the present, and from the small level and value of fixed assets (mainly hand tools) uti- 1ized in smallholder farming in the area. 5. with the above four points in mind, the average productivities shown in Table 5.1 are calculated according to the following formula: Average Productivity per Feddan = [TGM - (FL costed at HL rates + NL cost adjusted to the HL rates) - (00 x .07)] e No. of feddans cropped Average Productivity per Man-Day = [TGM - (0C x .07)] + Total Labor (in man-days) Computed average productivities for both land and labor (see Table 0 5.1) are very low by comparison to returns under irrigated farming.“ This situation is a result of the low physical crop yields obtainable under the environment of the region, and the seasonal constraints which limit expansion of cropped area. This latter factor is of particular im- portance and will be discussed in the subsequent analysis. A number of 1As explained earlier (see part two in Chapter IV) nafir (exchange) labor (NL) is paid only in-kind (food and drinks), and the value of these expenses was estimated as Ls 0.33 per man-day; this rate. is then applied uniformly to objective function coefficients of NL activities. 2For example, Zaki [61] estimated the mean net returns to house- hold 1abor per feddan in the Rahad Irrigation Project (Eastern Sudan) from the rotation crops as follows: cotton = Ls 74.69; groundnuts = Ls 31.70; leguminous fodder = Ls 22.52. 132 important implications arise from the situation of low productivity and returns of smallholders' farming in the area. First, under such circumstances, alternative sources of income for the smallholder are a necessity to meet household consumption obligations.I One such important source of income to smallholders in the area is off- farm employment. Off-farm work, as described in Chapter III, is concen- trated in the off-season period and is markedly influenced by the general savanna climate of the area. An estimate of the share of off-farm earn- ings relative to the total household income is available from the FAO survey data. For the three categories the percentages are as follows: 2 The larger share 37 percent, 23 percent, and 33 percent, respectively. (in absolute value) for individuals in category III reflects their rela- tively better position in off-farm occupations (shopkeeping and trade), as pointed out earlier. Another source of income to smallholders is livestock sales. Smallholder investment in livestock constitutes one strategy against risks and low returns from cropping activities.3 This source of income is tapped especially in "bad" cropping years. One final alternative source that smallholders in the area often resort to, 1The average consumption expenditure levels for the three categories are given in Table 4.2 (Chapter IV). 21h absolute value the respective amounts for the three categories are Ls 46.82, Ls 43.8, and Ls 105.69. 3Almost all smallholders interviewed in the researcher's survey (1979/80) gave this as the primary reason for investing in and keeping ' animals (especially cattle and goats). 133 mostly during the growing season, is borrowing from local money lenders (under the "sheil" system) in the form of both cash and goods [8, 38]. Second, the present low productivity and returns of the smallholder farming clearly indicates the need for improved production technology. It is for this reason that the NMAPC modernization program has been in- troduced. The record of the NMAPC is discussed in the next chapter. However, in this chapter, two experiments (given in the second and third parts, respectively) are directed towards investigating possibilities of increasing smallholder productivity and returns. Seasonal Constraints and Marginal Productivities The previous section has emphasized the low productivity and returns of smallholder farming. It was pointed out that in general this situation is attributable to the low physical crop yields obtainable under the en- vironment of the region. It was also argued that seasonal labor con- straints limit the extent of what could be a somewhat compensating effect through either increasing the size of cropped area, or achieving per feddan increases in yield through better (labor intensive) management levels. Table 5.2 gives the marginal productivities of the resources used in the optimal production plans for the three categories. There are two noteworthy points that should be observed in the interpretation of these MVP's: 134 Table 5.2 Shadow Prices for the Limiting Resources: Basic Production Plans of Traditional Smallholders Shadow Prices (in Ls)i Category Category Category I II III (0.1-5.0 (5.1-9.9 (10.0- Resource Unit' fed.) ' fed.) '25.9 fed.) Family Labor Jan. Man-Day -- -- 0.457 Apr. Man-Day -- -- 0.560 May Man-Day -- -- 0.670 Jun. Man-Day -- 0.637 0.967 Jul. Man-Day 2.390 1.476 1.046 Aug. Man-Day 0:744 1.001 1.134 Oct. ManeDay -- -- 0.449 Nov. Man-Day 0.680 0.729 0.762 Dec. Man-Day -- 0.975 1.019 Nafir Labor Jan. Man-Day -- -- 0.127 Apr. Man-Day -- -- 0.230 May Man-Day -- -- 0.340 Jun. Man-Day -- -- 0.331 Jul. Man-Day 1.076 0.664 0.471 Aug. Man-Day -- 0.189 0.558 Nov. Man-Day .350 0.375 0.392 Dec. Man-Day -- 0.621 0.649 Operating Capital Jun. Ls 2.964 1.459 0.744 Jul. Ls 2.964 1.459 0.744 Aug. Ls 2.964 1.459 0.744 Sep. Ls 0.0001 0.071 0.744 Oct. Ls 0.0001 0.071 0.744 Nov. Ls 0.0001 0.071 0.120 Dec. Ls 0.0001 0.071 0.120 Jan. Ls 0.0001 0.0001 0.0001 Feb. Ls 0.0001 0.0001 0.0001 Mar. Ls 0.0001 0.0001 0.0001 May Ls 0.0001 0.0001 0.0001 Source: Computed 135 According to economic theory, MVP is defined as the total value product (TVP) obtained from the use of an additional (marginal) unit of the input in question, other inputs held constant.1 However, in LP with the linear and fixed proportion production function, inputs are not held constant but rather they are in- creased in a fixed proportional manner in order to obtain an additional unit of output.2 Nonetheless, within the context of LP, shadow prices (MVP's) are useful in making inferences about resource productivities. For the farm situation at hand, these MVP's indicate the increase in the objective function value (farm income) that would obtain if a particular resource is ex- panded by one unit. Hence, they reflect the maximum price a farmer would be willing to pay to augment a given resource. According to this evaluation, an MVP equal to zero means that the given resource is not constraining, while a positive MVP indicates scarcity of the resource. The higher the MVP, the 1Mathematically this is represented by differentiating the TVP func- tion, in order to obtain the partial derivative with respect to the input in question: TVP = f(x], x2 ... xnlxn+1 ... xm) MVP“. =§fl3= fxi; i =1, 2, n x‘ f = TVP function f . = partial derivative with respect to input X1 (xi); and i = 1, 2 ... n, is the set of variable inputs. 2 For the LP technology, which represents a fixed proportion produc- tion function, NVP's are defined only at corner points (proportionate in- puts combination) and zero elsewhere. 136 more scarce and hence the more valuable/productive the given resource is. This is the rule for augmenting a resource and expanding output in the LP context. The valuation of MVP in the LP context is unique within a specific range of resource levels in the model. Usually when a given resource is increased, another resource becomes limiting, and consequently the resalt- ing MVP's of all resources change. 2. The yields and returns used here correspond to rainfall condi- tions and levels for the 1979/80 season. Different conditions and rainfall levels would result in different yields and hence, different MVP's for the resources. Table 5.2 shows the MVP's of the resources used in the optimal pro- duction plans. The basic limiting resources in the smallholder produc- tion are the labor resources. 00 resources are also limiting, which leads to an indirect labor constraint since labor is the basic resource DC is expended on. Land is not limiting. Family labor is limiting mainly in two periods: June through August, and November through December. As one might expect from previous dis- cussions, these correspond to the most labor intensive periods in the grown season. The first period, early in the season, is when weeding takes place. The second period, late in the season, comprises harvest and crop disposal. Across all three categories, these two periods are the main constraining periods. For the largest category III, family labor is also constraining in a third period (April-May). which is the I37 preseason period of land preparation. This is because of the relatively large cropped area for this category. The MVP's corresponding to the two periods just discussed are posi- tive, and high in comparison with the respective monthly hired labor wage rates, in all three categories. This indicates that expansion of the labor resource during these periods would be profitable. In fact, in these basic optimal plans, nafir and hired labor use occured mainly during these two periods. For category I, HL man-days were divided approximately equally between July (first period) and November (second , period). In category II, HL man-days were allocated 61 percent in July and 39 percent in November and December. For category III, where HL came in months other than those of the two periods, 86 percent of HL still occurred in these two periods, with 50 percent in July and August and the remaining 36 percent in November and December. As noted, nafir labor provision is largely limited to the two peak periods. The MVP's were ex- pectedly high in July (first period) and November/December (second period). The OC MVP's are positive and extremely high for all three categories in the early months of the season (June through September), and then they drop abruptly to almost zero. The reason for this is the scarcity of cash required to pay for labor in the first period, whose MVP's indicate its profitability as discussed above. Low MVP's of DC in the second period, however, probably occur because no direct returns can be achieved by using 00 later in the model year. The other likely reason is that later in the season the OC inventory is replenished through sales of the sesame crop (which was in the solution of all categories) which occurs 138 as early as November. This effect has an important implication for credit needs and is discussed later in the credit-aided resource expan- sion experiment. The labor MVP's examined in this section suggest that technologies that are oriented towards breaking labor constraints in the two Peak periods might be the most rewarding for smallholders. The use of chemi- cal weeding and the use of stationary or combine grain harvesting would tend to relieve the first and second labor constraints. Needless to say, the development and adoption of such technologies depends on their technical and economic feasibility. At present the work by Hamdoun [26] on chemical herbicides indicates some potential if the current problems related to chemical use areovercomea.1 with family and nafir labor being fixed through the set of insti- tutional factors discussed before, expansion of the labor resource would depend on expanding the use of hired labor. The extent and benefits of such an approach are examined through a credit-aided resource expansion experiment in the next part of this chapter. Rainfall-Induced Yield Variability In this section we investigate the situation of smallholder's re- turns under conditions of "bad" rainfall years. In the area, extremely drastic rainfall_years are rare [38]. Consequently, incidences of 1At the research level the basic problem observed is that of plant toxicity, its long residual effects under rainfed conditions, and the human hazards it poses (for details see Hamdoun [26]). At the field level expected problems are those of management and application [25]. Water shortage at field sites in the area is an example of potential problems that need to beovercome to make such a technique practical. 139 complete crop failure due to rainfall are few, and probably obtain only coupled with poor or no crop management. Nonetheless, it is instructive to investigage less than favorable rainfall years as the major factor in depressed cropped yields. This is done here through simple sensitivity analysis around the yield coefficients in the models. Twenty-five per- cent and 50 percent reductions in yield levels are assumed for situation A, and situation 8, respectively.1 The optimal production plans for situation A and B are given in Tables 5.3 and 5.4, respectively. Situation A: ‘ a) Cropping Pattern Dura feddanage showed little change, with most being grown in the early schedule as before. The sesame share increased almost by 50 per- cent for category I, while slightly dropping for both category II and III. Interestingly.cotton entered the solution of category III; however, only 6 percent of the cropped land is devoted to cotton, grown in the late schedule. b) Resource Use Total cropped area went up slightly for all three categories, which was made possible through the reduction of required harvest labor of crops. In terms of labor composition, the share of NL remained approxi- mately the same, while that of HL dropped in all three categories. FL dropped slightly for category I and II, and increased slightly for cate- gory III. In all three categories 00 dropped reflecting reductions of HL use. 1In both cases, harvest and transport coefficients of the production activities in the models was adjusted downwards by the same proportions. 140 Table 5.3 Optimal Production Plans for the Three Categories of Tradi- tional Smallholders: Under Twenty-Five Percent Yield Reduction Category Category Category I II III (0.1-5.0 (5.1-9.9 (10.0- Item Unit fed.) fed.) '25.9 fed.) Cropping Pattern: Early Dura (0R1) Feddan 3.64 6.12 10.74 Early Sesame (5M1) Feddan -- 0.35 3.56 Late Dura (0R2) Feddan -- 2.33 8.37 Late Sesame (5M2) Feddan 0.93 0.17 0.42 Late Cotton (CNZ) Feddan -- -- 1.45 Resource Use: Total Land Feddan 4.57 8.96 24.54 Family Labor Man-Day 106.81 173.72 307.61 Hired Labor Man-Day 3.97 39.88 280.16 Nafir Labor Man-Day 9.41 21.50 13.60 Total Labor Man-Day 120.19 235.10 601.37 Total Operating Capital Ls 19.54 71.19 324.63 Returns: Objective Function Value Ls 26.91 72.80 138.34 Total Gross Margin3 Ls 79.61 138.80 ‘ 217.54 Source: Computed aIncludes the value of dura retained for household consumption. 141 Table 5.4 Optimal Production Plans for the Three Categories of Tradi- tional Smallholders: Under Fifty Percent Yield Reduction Category category Category I II III (0.1-5.0 (5.1-9.9 (10.0- Item Unit fed.) ‘ fed.)' 25.9 fed.) CroppingiPattern: Early Dura (0R1) Feddan 3.72 6.37 6.37 Early Sesame (5M1) Feddan -- -- -- Late Dura (0R2) Feddan -- 2.26 5.12 Late Sesame (5M2) Feddan 0.90 0.41 1.48 Resource Use: Total Land Feddan 4.62 9.30 13.19 Family Labor Man-Day 98.54 170.46 273.21 Hired Labor Man-Day 3.93 34.05 58.92 Nafir Labor Man-Day 6.66 8.37 11.03 Total Labor Man-Day 109.13 212.88 343.16 Total Operating Capital Ls 23.93 63.11 95.38 Returns: Objective Function Value Ls ~10.82 12.30 25.66 Total Gross Margina Ls 35.48 78.30 A104.86 Source: Computed aIncludes the value of dura retained for household consumption; in Category I the value of 0.98 dura sack bought is not included. 142 c) Monetary Returns Witha.25 percent reduction in.yield levels, the objective function values were 51 percent, 41 percent and 49 percent lower for the three categories respectively, indicating the sensitivity of returns to reduced rainfall and hence reduced crop yields. Situation 8: a) Cropping Pattern The share of dura in the crop mixture remained approximately the same for categories I and II, while dropping substantially for category III. The latter was caused mainly by the overall reduction in cropped land size for category III. The sesame area dropped in all three cate- gories, especially so in category III. b) Resource Use While total cropped land went slightly up for categories I and II, it dropped significantly (by 45 percent) for category III. For the first two categories the increase could be explained as one way of com- pensating for the severely depressed crop yields. However, with yields reduced by half, the profitability of HL in category III is substantially reduced. Since HL normally makes up 60 percent of the total labor force in this category, there results a large reduction in cropped area. FL and ML remained at approximately the same levels for the three categories. The HL share is reduced greatly in all three, being highest (85 percent) for category III. 0) Monetary Returns The same trend is observed as in situation A. With a 50 percent re- duction in yield levels, objective function values were reduced by 120 143 percent (with a negative objective function value),1 91 percent, and 90 percent for categories I, II and III, respectively. The results of modelling situations A and 8 reveal the sensitivity of smallholder returns to physical crop yields. A given drop in yields leads to a proportionally greater drop in returns. Credit and Land Expansion Experiment: Traditional Model The primary objective of this experiment is to examine the effects on crop mixture, returns and seasonal labor use of an expansion of land availability and provision of credit. The main component of this experi- ment is introduction of capital borrowing activities (with 7 percent annual interest rate)2. Land availability is also increased by 50 per- cent for each of the three production categories. The optimal plans re- sulting from this experiment are given in Table 5.5. CroppinggPattern The dura crop is still dominant, being grown in both early and late schedules. For category III the majority of dura is in the early sche- dule. By comparison to the basic model results, the area under sesame increases substantially for categories I and II to 2.67 feddans; while the category III sesame area dropped to almost half, from 5.09 to 2.67 1With the dura production in category I less than the specified con- sumption level (8 sacks), the difference (0.98 sack) had to be bought, resulting in the negative objective function value. glhe reason for using this particular rate, as discussed before, is that it represents the official rate charged by the Agricultural Bank of Sudan for short-term (up to 14 months) loans. 144 Table 5.5 Optimal Production Plans for the Credit and Land Expansion Experiment: Traditional Modela Category Category Category I II III (0.1-5.0 (5.1-9.9 (10.0- Item Uhit fed.) fgg.) 25.9 fgg.)' CroppingiPattern: Early Dura (0R1) Feddan 2.53 5.81 29.81' Early Sesame (SMI) Feddan 2.67 2.67 . 2.67 Late Dura (0R2) Feddan 2.30 6.37 6.37 Late Sesame (5M2) Feddan -- -- -- Resource Use: Total Land Feddan 7.50 14.85 38.85 Family Labor Man-Day 151.13 218.00 266.00 Hired Labor Man-Day 54.61 182.33 844.47 Nafir Labor Man-Day 13.32 31.20 14.90 Total Labor Man-Day 219.06 431.53 1125.37 Total Operating Capital: Ls 70.29 700.56 756.07 Initial Capital Ls 6.93 27.83 152.70 Crop Sales Ls 30.36 108.43 256.45 Borrowed Capital Ls 30.85 60.10 324.23 Interest Paid Ls 2.16 4.21 22.70 Returns: Objective Function Value Ls 97.81 186.22 374.67 TotalGross Marginb ‘ Ls 150.61 252.22 ‘453.86 ‘ Source: Computed aIntroduction of capital borrowing activities, with the availability of land increased by 50 percent in the three categories. b Includes the value of dura retained for household consumption. 145 feddans. All three categories therefore grow 2.67 feddans of sesame. The explanation for this as suggested by the changes in resource use levels (discussed below), is that the resulting increase in family labor use in categories I and II helped in raising their sesame feddanage, and that the decrease in family labor use which occurred in category III had the reverse effect. The particular figure of 2.67 suggests a limit for the optimum area of sesame grown by family labor under conditions of unrestricted availability of hired labor. Resource Use With the supply of land increased by 50 percent, the respective cropped areas for the three categories were 7.5, 14.85 and 38.85 feddans. This is made possible by more hiring of labor. The hired labor share in the total labor force increased substantially in all three categories, being 25 percent, 42 percent, and 75 percent, respectively. The increase in land and hired labor also made it possible to utilize more of the exist- ing family labor. For categories I and II, FL man-days in this plan com- pared to the original plan increased 37 percent and 21 percent, respective- ly. For category III, where there is substantial HL, FL man-days dropped by 4 percent. Borrowed capital amounts were Ls 30.85, Ls 60.10 and Ls 324.23 for the three categories, respectively. These amounts represent those needed just for production purposes, and hence could be regarded as the lower limits for any meaningful credit expansion, since smallholders also borrow for consumption purposes. Capital borrowing occurred early 146 in the season (June) to cover 0C needs of the first period (June-October). The second period 0C needs are mostly for harvest and transport expen- ditures, and these were met in the model through crop sales (especially of sesame which occurs as early as November). Smallholders in the area , generally borrow through the "sheil" system mainly in the first period when (as a HTS study notes) interest rates on borrowing are very high: Sheil credit, usually advanced in-kind, is characterized by very high interest rates which reach their maximum (up to 300 percent per annum) in mid-August and start to decline once the harvest season commences in mid to late September [38, p. 108]. Effects on Seasonal Constraints Table 5.6 gives the MVP's-of the resources used in the optimal plans in this experiment. With all of the land being exhausted, its MVP is Ls 11.35, Ls 7.38, and Ls 6.80 for categories I and II, and III, respec- tively. The differences by category are due to the fact that if an additional feddan is brought into production it can be worked with a higher share of family labor in category I, and to a lesser extent in category II, than in category III, where family labor resource is already fully utilized.“ ' Reliance must therefore be mainly on the relatively costly hired labor. The two labor constraint periods observed in the basic plan are also noticeable in this plan: July-August (first period) and November- December (second period). The difference in this case, however, is that through hiring of labor the value of the MVP's is much lower than in the basic model. With the labor MVP‘s in this plan being equal to or less than the corresponding HL wage rates (except for July and August MVP‘s- Table 5.6 Shadow Prices for the Limiting Resources: 147 Expansion Experiment, Traditional Model Credit and Land ‘Shadow Prices (in L5) Category Category Category I II III (0.1-5.0 (5.1-9.9 (10.0- ResourCe Unit fed.) fgg.) 25.9 fed.) Land Feddan 11.35 7.38 6.80 Family Labor Jan. Man-Day -- 0.87 0.87 Apr. Man-Day -- 0.56 0.56 May Man-Day -- 0.05 0.11 Jun. Man-Day -- 0.56 0.55 Jul. Man-Day 0.64 0.64 0.64 Aug. Man-Day 0.69 0.70 0.69 Oct. Man-Day 0.01 0.13 0.40 Nov. Man-Day 0.61 0.68 0.73 Dec. Man-Day 0.91 0.91 0.97 Nafir Labor Jan. Man-Day -- 0.54 0.54 Apr. Man-Day -- 0.23 0.23 Jun. Man-Day -- 0.20 0.20 Jul. Man-Day 0.29 0.28 0.28 Aug. Man-Day 0.34 0.34 0.34 Oct. Man-Day -- -- 0.04 Nov. Man—Day 0.28 0.35 0.37 Dec. Man-Day 0.58 0.58 0.62 Operating Capital Jun. Ls 0.07 0.07 0.07 Jul. Ls 0.07 0.07 0.07 Aug. Ls 0.07 0.07 0.07 Sep. Ls 0.0001 0.07 0.07 Oct. Ls 0.0001 0.0001 0.07 Nov. Ls 0.0001 0.0001 0.07 Dec. Ls 0.0001 0.0001 0.0001 Jan. Ls 0.0001 0.0001 0.0001 Feb. Ls 0.0001 0.0001 0.0001 Mar. Ls 0.0001 0.0001 0.0001 Apr. Ls 0.0001 0.0001 0.0001 May Ls 0.0001 0.0001 0.0001 Source: Computed 148 which are marginally higher than their corresponding wage rates)-HL use and profitability to further augment labor resources in these periods is greatly diminished. Returns with Credit - In this credit experiment, the resulting returns (total gross mar- gins) were Ls 150.61, Ls 252.22, and Ls 453.86, for the three categories respectively. These levels represent substantial increases from the base plans of 77 percent, 50 percent, and 39 percent for categories I, II and III, respectively. Increases were highest in category I and II, where labor and capital are most limited in the basic model. It is noteworthy that in these two categories (I and II), the increase of land and hired ' labor availability made it possible to use more of the existing family labor resource (by 37 percent and 21 percent for the two categories, re- spectively), as compared to the base plan. Smallholders' Credit Situation and Possibilities The analysis in this part has shown that credit use can increase the low farming returns of smallholders in the area. Although the in- terest rate charged for borrowed capital in this experiment was 7 per- cent (to represent the official rate of the Agricultural Bank of Sudan for short-term loans), it is worth noting that the MVP's from the base optimal solution (Table 5.2) indicate the profitability of borrowed capital, within the smallholder system, at extremely high rates: 296 percent, 145 percent, and 74 percent for categories I, II and III, re- spectively. It is further evident that the profitability of borrowed capital is highest in category I and II, where 00 is most limited, and 149 hence credit use is most rewarding. Another important result revealed in the analysis is that credit is mainly needed in the first period (June-October), when smallholders in this period require not only produc- tion credit but also consumption credit to see them through harvest. The second period 0C requirements could be financed, as farmers normally do, through initial crop sales. In this context, the inclusion of sesame in the cropping plan (with its crop sales occurring as early as November) is particularly important and useful. Smallholders at present have no access to formal credit and many of them resort to traditional money lenders under the prevalent "sheil" system in the area [8, 38]. Although the establishment of the Agricul- tural Bank of Sudan (ABS) in 1957 was intended to provide formal credit to smallholders, and despite the fact that the by-laws of the ABS state that "preference for loans should be given to small and medium sized farmers and to cooperatives" [8, p. 107], the ABS historical record was disappointing to smallholders. Instead, the majority of ABS investments and activities were directed towards financing large private cotton schemes in the White Nile and Blue Nile provinces, and later towards private large-scale mechanized rainfed schemes.1 The need by the ABS for tangible security requirements was identified by Stickley and Abdallah [58] as the main reason limiting the participation of small farmers. l A concise review of ABS history and investment activities ' i in Ahmed [8, pp. 76-86]. 15 g ven 150 In the Nuba Mountains area, the Dilling Branch of the ABS was es- tablished in 1970. The activities of the Dilling Branch were concen- trated in offering short and medium term loans to private large scale farmers in the mechanized areas of Habela and its extensions.1 Before 1977 no attempts were made by the Dilling Branch to extend credit to smallholders in the area. Ahmed indicated that "the high cost incurred in providing finance to small undefined areas was the main reason that discouraged lending" [8, p. 113]. In 1977, however, the Dilling Branch of ABS introduced what was con- sidered by Ahmed, in his study of "Lender Behavior and the Recent Perfor- mance of Rural Financial Markets the Sudan," as "the first serious attempt in the Sudan to provide institutional credit to typical small farmers" [8, pp. 113-114]. In this attempt, credit was extended to traditional smallholders through their cooperative societies, where intensive credit supervision was used to replace collateral requirements.2 The first trial conducted in the 1977/78 season was limited to the finance (with a total amount of L5 14,838) of harvest operations for groundnuts I For example, all of the short-term dura production loans, which is the main line of activity for the Dilling Branch, are advanced to owners of large schemes (1000 to 1500 feddans) in Habela and other mechanized schemes in the area. For the 1979/80 season the total amount of dura cultivation loans advanced was Ls 608,366.85 (loans granted on the basis of Ls 2,350.0 per 1000 feddans) and the total for dura harvest loans was Ls 432,792.65 (loans granted on the basis of L5 1.85 per dura sack). 2A coordinated effort involving ABS officials, staff of Cooperative Department, staff of Crop Protection and Extension from Ministry of Agriculture, in coordination with village heads (Sheikhs) was used in superv1sion of this credit program. 151 and sesame crops of 658 smallholder members of two cooperatives in Um Ruaba area. The following 1978/79 season the scope of the trial was enlarged (with an amount of Ls 80,000) to increase participating small- holders to 972 cooperative members, and to finance both cultivation and harvest operations for the two crops.1 With the record of these two seasons showing 100 percent repayment of loans, the bank was encouraged to increase the scope of its smallholder credit for the 1979/80 season to include 1300 small farmers in 20 villages. In the same season the program was extended to include 650 smallholder cooperative members in 24 villages in El-Debaibat area, north of Dilling. The promising potential of cooperatives in securing smallholders' access to formal sources of credit is revealed in this recent ABS Dilling Branch experience, cited by Ahmed: Innovations intended to reduce lending and borrowing costs should promote the development of cooperative societies, who in turn, could extend services to a large number of farmers. The Um Ruaba pilot project introduced by the ABS revealed the possibility of ex- tending comparatively small amounts of credit (an average of $200) to a large number of farmers [8, p. 208 . 1‘Loans are granted on the basis of Ls 10.0 per Mukhamas (1.75 feddans) for cultivation of sesame, and Ls 18.0 per Mukhamas of ground- nuts. Harvest loans are granted on the basis of Ls 6.53 per Mukhamas of sesame, and Ls 19.49 per Mukhamas of groundnuts. 152 With the growing number of smallholder cooperatives in the area,1 pro- viding the framework for expanding the ABS smallholder credit program, the extent of such efforts could be substantially increased. Already, this experience of lending through cooperatives encouraged the World Bank to consider financing traditional agriculture in the Sudan through the ABS with a little less than Ls 7 million in a three-year program as part of an Agricultural Services Project [8]. Planting-Time Experiments and Model Results The previous part examined the effect on productivity and returns in the traditional farming model through provision of formal credit and a 50 percent increase in land availability. This section focuses on the impact on productivity of changes in time of planting. One of the widely claimed reasons for the low productivity of smallholders is their in- ability to take timely decisions and their poor crop management practices. The Experiment The planting-time experiment is based on the important husbandry practices of smallholders in the area. These include: choice of crop variety, planting time, and weeding rates/times. These factors have al- ready been discussed in Chapter III. Table 5.7 summarizes the implica- tions of these factors as deduced from experimental material and research. 1As of June 1980, the number of registered cooperatives operating in the four districts of South Kordofan Province was 170 cooperatives with total membership of slightly less than 30 thousand. The majority (80 percent) of these cooperatives at the present are "consumption" co- operatives, which are primarily involved in procurement and distribution of consumer goods. 153 The supporting research, experiments, and data are given in Appendix II. These comprise experiments carried out at the Kadugli Research Station lo- cated in the Nuba Mountains area, and the Kenana Research Station lo- cated in the central-eastern Savanna of the Sudan. It is important to comment briefly on two methodological points: 1. Being mindful of the limitations of the experiments discussed in this section and Appendix II, they still remain useful in highlighting the implications of timing sequences included in this experiment. 7 2. In particular there is the much more difficult question of relating yield levels obtained under experimental environments to actual farm conditions. Judgment and comparison with other sources [6, 38] are used in the choice of realistic yield levels that could be achieved by farmers under optimum conditions (suffix I yield levels in Table 5.7), while using research im- plications for the rate and timing of operations in addition to the yield discounting factors for delayed schedules (suffixes II and III yield levels in Table 5.7). In the LP model modified for this experiment (see discussion in the last part of Chapter IV), three timing sequences are incorporated. In addition, coefficients of weeding, harvest, and transport corresponding to yield levels in this experiment are adjusted accordingly. It should be noted that in this experiment (see Table 5.7), following the assump- tions developed in Appendix II, dura (short-maturing variety) requires 1534 ".mx ms .mc—ucu—a see» axon: oz» .eue.:aoe me—vuu: oco x—eo "ASmEo: .33 mum-ea» ac—eauaa-ueosm a cannon «.mx cm n neat “.33 mc u :ouuou .__ x-v:oaa< c. commaumpu ago-«assume we. .o.guua§ poacoa—eoaxo sou» county; .coueoaeum so mam—ea) paged» 2.3.33 aunts. 2233-2050 w ._ umm>eaz :. once 3:.8u—a couuou he ueaueoa mu an.) .¢u .mx mv u muacvcaoea ecu unsuppou mu m— mango acoeo~u_u as» so» ago—u>v=ao xuom - a .__ xvucoaa< c. cummaum_u mapsmog paucuspgoaxo nae» consume meouuou m=_u::oum—v uc—ma eu:.ouao use ___ we: __ max—553m so omega u mac—u—ccou canvugo Love: menses» x; apnea—cane m—u>o— vazsmmo as» one u x—uuam 5o nope—a gonna "aueaom a he.~ ~.o~_ _~ .>oz m. .>oz m. .eow s .a=< _ .asc .2. useoesoeu mm.m m.co~ s .>oz _ .>oz — .aom —~ ..33 m_ .—:e .~ uscvcaoeu we.“ ..omn .N .uoo m. .eoo m. .a=< N ..sa _ ._=e e~ “aeoeaota ee.m m.a- _~-m_ .>62 2 .>ez _ .eom m. .u=< _ .a=< ... oseuom 2.” :8 p .5. a .38 2 .2: _ .91 ...: ...... : 953m mn.m. n..m~ .N-m_ .uoo N .uoo m. ..se m. ..aa , . ..ee 0. «semen m: 2.8 2 .98 a .2... 2 .92 _ .22 E ...8 e... e.aom _ .ooo N .sez _ .a=< m. ..ae _. etso No.» n.6ce m. .>oz .N .uoo m. ..ae _ ..ae o. .238 m~.~ e.m~. . .eea _ .ooo _ .eom m. .o=< _ .m=< .u. eoueou am.m e._m~ m. .ooo m. .>oz m. .m=< _ .a=< m. ..ae __ eouuou o..o °.mem _ .oea _ .>ez on ..aa m. ..sa _ ..sa o_ eooueo ..oo. “.55. __ _ __ _ ooeoaaom mxuumv .mx. umu>eaz umo>eaz 3:.uua: 3:.3uo3 a:_u=a—m 3:.3—p 65.6.» eoeoommm co.uneoac amen mcpacmpm mo wave An maceu eo upow> umuuquu m.m opnmp .. ...--"" 155 only one weeding, two weeks from planting. Capital borrowing activi- ties (as before, with 7 percent annual interest rate) were later added to the LP model of this experiment to investigate the implica- tions of the planting-time sequences when a formal credit option is available. Results of LP Analysis Optimal Production Plans: Optimal production plans for the planting-time experiment are given in Table 5.8. Despite the high physical yields associated with early planting of crops, the results reveal that it is optimal for smallholders to grow crops in the late planted (second and third) schedules. The optimality of late planting arises from the stringent time requirements of early planting and its interaction with seasonal resource availabili- Planting crops at different times is one way smallholders can ties. Nonetheless, all three farm smooth out these seasonal bottlenecks. size categories made use of the high physical yields of the early planted crops, as much as their resources could allow; category III, which has the highest resource endowment, was able to grow 75 percent of its crops in the early schedule. The resource use figures for these plans, in comparison with the basic plans (see Table 5.1) show the size of cropped area slightly in- creasing for category I, while dropping 20 percent for category II and 27 percent for category III. This result is explained by the extent to which the three farm size categories made use of the relatively labor intensive, early planted schedule in this experiment; category II and III grew the majority of their crops (51 and 75 percent, respectively) in the 156 Table 5.8 Optimal Solution for the Traditional Model in the Planting lime Experiment [Category Category Category I II III (0.1-5.0 (5.1-9.9 (10.0- Item Unit fed.) fed.) 25.9 fed.) Cropping Pattern:a Cotton I Feddan 0.15 0.15 -- Dura I Feddan 1.35 2.91 10.30 Sesame I Feddan 0.06 0.58 2.63 Dura II Feddan 0.83 1.14 1.30 Cotton III Feddan 0.18 1.53 1.58 Sesame III Feddan 0.96 0.67 1.64 Resource Use: Total Land Feddan 4.91 7.07 17.46 Family Labor Man-Day 116.60 190.98 252.64 Hired Labor Man-Day 6.71 42.10 297.62 Nafir Labor Man-Day 3.34 6.95 4.20 Total Labor Man-Day 126.65 247.09 554.46 Total Operating Capital Ls 19.51 56.46 262.75 Returns: Objective Function Value Ls 48.91 101.01 303.74 Total Gross Marginb Ls 101.71 167.01 382.94 Source: Computed aSuffixes I, II and III refer to time of planting as defined in Table 5.7. bIncludes value of dura retained for household consumption. 157 early schedule, resulting in a decrease of their respective cropped areas, while in category I, which grew only 31 percent of its crops in the early schedule, this was compensated for by the increase of its cropped area. The total labor use was slightly reduced for category I and II (by 1 percent and 3 percent, respectively) but significantly for category II (by 20 percent). While reduction of cropped area (in cate- gory II and III) helps explain part of the observed reduction of labor use, the primary influence is the reduced labor requirements for weeding (It should be recalled that only one timely weeding operating is dura. As in the original plans, dura required in these planting sequences). dominates the cropping pattern. Reduced harvest labor requirements (for crops grown in the second and especially the third sequence) could be a factor too, but given their small share of total cropped area, this is of small significance. The returns picture (in comparison with the original plans) also captures the interaction of the new resource requirements and labor time distribution of these three planting time sequences, with the resource availabilities of the smallholders. For categories I and II, returns are down by 11 percent and 19 percent, respectively. be noted that the production opportunities (and corresponding yield co- The returns in However , it should efficients) are modeled differently in this experiment. category III, whose resource availabilities are more suited to take ad- vantage of early planting, have risen by 13 percent despite the sizeable (27 percent) reduction of its cropped area. 158 Effects on Seasonal Constraints: The limiting resources and their corresponding shadow prices for the planting time experiment are given in Table 5.9. Given the present resource availabilities of the smallholders, early planting places an The labor extra strain on the already observed labor constraint periods. MVP's in the second half of July and the first half of August (the first peak period) are highest. The second peak period is pushed back to in- clude the second half of October until the end of November, correspond- ing to harvest of early planted crops. The MVP's of labor in both periods are high (relative to prevailing hired labor wages), being highest for the most resource-scarce farm size category. In addition, the operating capital MVP's for almost all the cropping season (June through the first The importance of seasonal half of November) are also positive and high. constraints in this version of the model, and the rewards captured by relaxing them, are demonstrated by the next experiment. Credit with Planting-Time Experiment Results: To allow the augmenting of labor resources in the constraining per- iods through hired labor, credit is made available in the models through addition of borrowing activities. Optimal plans for this version of the model are given in Table 5.10. With seasonal labor constraints largelyovercome- through the use of hired labor, sizeable returns resulted in the optimal plans. The pro- fitability of dura , especially that of the early schedule (Dura 1), makes it dominate the cropping pattern. While this situation is a bit unrealistic in comparison to both actual practice of farmers and other model solutions discussed so far, it demonstrates the effect of breaking 159 Table 5.9 Shadow Prices for the Limiting Resources: The Planting Time Experiment Shadow Prices (in Ls) Category Category Category I II III a (0.1-5.0 (5.1-9.9 (10.0- Resource Unit fedL.) fed.) 25.2 fed.) Family Labor Apr. Man-Day -- -- 0.560 May. Man-Day -- 0.141 - 0.210 Jul .1 Man-Day -- -- 1.241 Jul .2 Man-Day 1.887 1.633 1.241 Aug.1 Man-Day l .654 l .769 1.344 Aug.2 Man-Day 0.764 0.412 0.526 0ct.1 Man-Day -- -- 1.179 Oct.2 Man-Day 1.792 1.550 1.179 Nov.l Man-Day 0.024 0.034 0.678 Nov.2 Man-Day -- 0.938 1 .059 Dec.l Man-Day 0.040 0.154 -- Dec.2 Man-Day -- -- -- Nafir Labor Apr. Man-Day -- -- 0.230 Jul .1 Man-Day -- -- 0.558 Jul .2 Man-Day 0.849 0.735 0.558 Aug.1 Man-Day -- 0.871 0.662 0ct.1 Man-Day -- -- 0.496 Oct.2 Man-Day 0.754 0.653 0.496 Nov.2 Man-Day -- 0.483 0.545 Operating Capital Jun. -Nov.1 Ls 2.145 1.722 1.068 Nov.2 Ls 0.542 0.379 0.558 Dec.1 Ls 0.542 0.379 0.558 Dec.2 Ls 0.542 0.379 0.558 Jan . Ls 0 . 0001 0 . 0001 0 . 0001 Feb . Ls 0 . 0001 0 . 0001 0 . 0001 Mar. Ls 0.0001 0.0001 0.0001 Apr. Ls 0.0001 0.0001 0.0001 May Ls 0.0001 0.0001 0.0001 Source: Computed aSuffixes 1 and 2 refer to first and second half of the month for July through December. fl 160 Table 5.10 Optimum Production Plans: The Planting Time Experiment with Credit Option Category [Category Category I II III (0.1-5.0 (5.1-9.9 (10.0- Item Unit fedg) fed.) 25.9 fed.) CroppinggPattern:a Dura I Feddan 3.85 8.01 23.72 Dura II Feddan 1.15 1.89 2.18 Resource Use: Total Land Feddan 5.00 9.90 25.90 Family Labor Man-Day 86.50 124.70 149.50 Hired Labor Man-Day 64.86 174.56 645.33 Nafir Labor Man-Day 3.34 7.68 2.80 Total Labor Man-Day 154.70 306.94 797.63 Total Operating Capital: Ls 71.79 172.22 564.60 Initial Capital Ls 6.93 27.83 152.70 Crop Sales Ls -- -- -- Borrowed Capital Ls 60.99 129.92 355.16 Interest Paid Ls 4.27 9.09 24.86 Returns: Objective Function Value Ls 78.55 189.83 464.22 Total Gross Marginb Ls 131.35 235.83. 543.42 Source: Computed aSuffix I, II and III refer to time of planting as defined in Table 5.7. b Includes value of dura retained for household consumption. 161 seasonal labor constraints. The resource use figures show the somewhat higherlabor levels employed in these plans. Not only are these levels high relative to total labor use in the three categories, but they re- quire the recruitment and utilization of this labor during the two short peak labor constraint periods. In reality, achieving the benefits from such a management scheme would be limited by the scarcity of credit. Cotton Price Variation Experiment It should be recalled that the cotton crop, in agreement with the empirical situation, did not enter in any of the smallholder categories original plans (see Table 5.1) and was absent also from all model ver- sions and solutions discussed so far.1 This situation, being of impor- tant policy concern as indicated earlier (see Chapter I), will be taken for detailed analysis in this part. First, a brief background note on cotton problems in the area is given. Next, the results of the cotton price programming experiment are discussed. Background on Cotton Problems in the Area Unlike all other crops grown in the area, the cotton crop is en- tirely a cash crop aimed for the export market. The establishment of the cotton growing industry in the Nuba Mountains area dates back to the early'l920's, as described by Tothill: 1With the exception of two solutions (see Table 5.3 and 5.8), where insignificant feddanage was devoted to the cotton crop. 162 In pursuance of the policy of "turning swords into plough shares"1 the government [British Colonial era] decided in 1928 to endeavor to introduce the growing of cotton as a cash crop. It was accordingly decided that observational plots of American type cotton should be grown by the government during the rains of 1924. The results were promising, and it was decided that the Department of Agriculture and Forests should go ahead and endeavor to establish a cotton growing in- dustry....about that time a further attempt was made to initiate a company which would finance and handle the growing cotton industry [59, p. 842]. Although the "endeavors" were presented as serving the interests of national and regional security, as well as those of the local popula- tion, by establishing cotton as a cash crop, it is clear that the colo- nial interests were the overriding reason behind the effort.2 Despite the rapid expansion of areas under the crop in subsequent years, and the institutionalization of the cotton industry since then, potential problems of cotton in the local setting were perceived early on. The view then, however, was overoptimistic with regard to cotton; given the cultivator's high cash returns expected from the new cotton crop, its expected wide adoption by cultivatorsin the area was expected to come at the expense of other crops, as Tothill explains: When propoganda for more cotton was being spread, care was taken to emphasize the principle that it should not be grown instead of dura, the main food crop, but in addition to the usual areas under food crops [59, p. 843]. 1Reference is made to the "unruly" Nuba people and the government need to secure peace in the area. 2Representatives of England-based industries (Lanchshire industries and spinners) were among the originators of the idea [59]. 163 Needless to say, the unfolding of events in the historical record of cotton in the area, especially in recent years, rendered the above concern irrelevant. Instead, the growing importance of other crops in the smallholder farming system, together with the high inflation rates of recent years, have led to the decline in cotton production [38, 39].1 The decline in cotton production reached serious proportions in the late 1970's, with the crop almost disappearing from traditional small- holder areas. In 1979, aministerial committee was formed to study and make recommendations regarding the decline of cotton production in the Nuba Mountains area. The committee report [39] singled out the un- rewarding farm-gate cotton price as the main cause of cotton decline in traditional agriculture. Another factor mentioned was the increasing cost of production for cotton (mostly labor costs). The corresponding recommendation was for a higher cotton price. Recommended prices for cotton delivered by farmers were: Ls 7.00, Ls 4.50 and Ls 3.50 per kantar (45 kg) of cotton for grade I, II and III, respectively.2 To raise the current level of prices (Ls 4.25, Ls 3.25, and Ls 2.75 for the three grades, respectively) to the recommended level, and since the NMAPC pric- ing system cannot put this into effect .given the current production levels, it was suggested that the resulting deficit from adopting the reconmended prices be absorbed by the government, until such time when it is no longer l . The HTS study [38] est1mated that for the period 1968/69 to 1978/79, the price of seed cotton has decreased in real terms (allowing for infla- tion) by some 60 percent, while the price of dura has risen by about 20 percent. Generally, over 90 percent of cotton produced falls in the first two grades. 164 necessary to have government support, and the system is able to pay such prices by itself.1 Beyond the main recomendation for higher cotton prices, the com- mittee report also gave some recomendations to improve cotton production within the NMAPC modernization schemes. The latter, mostly organizational and administrative in nature, are discussed in the next chapter. Below, we focus on the effect of changing cotton prices and discuss the results of the experiment for the traditional smallholder models. Experiment Results With the above background, LP analysis is used to examine the effect of varying cotton prices on the cotton feddanage within the smallholder The procedure used is parametric programing, sometimes referred models. Such a procedure is to as variable price programing or price ranging. generally used to estimate the normative supply function of a comodity [32]; the optimal output of the comodity is derived by varying its price, within an appropriate range, while other prices are held constant Before proceeding to present results and analysis of this experiment, it is important to note the following two points regarding this experiment: The supply response resulting from variable price programing l. is generally limited by its static and normative nature (in the sense of being what farmers ought to be producing).2 1The present NMAPC cotton pricing system, which is crucially depen- dent on total cotton production in the area, is described and discussed in detail in the next chapter. It was estimated that the present system can afford to offer the recomended prices, without government support, when production of cotton in the Nuba Mountains area reaches at least the 400,000 kantars level [39]. 2And is also believed to result in an upward bias of some degree [32] 165 2. Supply responses to price changes could theoretically be par- titioned into yield and area responses. The first may stem from an intensified or new input use, and the second may stem from substitution between crops, expansion of cropped areas, or changes in cropping intensity [33]. The yield response, within the context of cotton problems in traditional agricul- ture, is of relatively less importance at the present.1 While it is conceivable that intensified and more skillful labor use might result in higher cotton yields, at the present the real issue and the policy concern is that of crop substitution and massive reductions in cotton area. In this situation, it is more realistic to examine the feddanage response to cotton price changes. Cotton area responses for the traditional farm models are graphed in Figure 5.1. The results and conclusions drawn from this experiment can be summarized as follows: 1. There are differences in the feddanage response to higher price between the three farm size categories. Such differences are a result of differences in resource availabilities between the three categories (especially the labor resources, and the pro- portions of hired labor employed). Particular differences are: 1However, yield response is of utmost importance within the context of NMAPC smallholder schemes. With the participants cotton areas deter- mined by the NMAPC, price policy incentives aimed at increasing cotton output would have to induce per unit land yield increases. This point is further discussed, in the next chapter, within the context of NMAPC farm model '5 cotton price variation experiment. 166 Aneuvvou. 2:1 .338 0.. L m. I-n Q n I-O m .686: stem _eeo_eeeeee 6 O n on 5 once ..:.: . — n. 0>< 9:05.550 . ....-.........................S...........-..._.._..i::..nfl-.. ...nL G n 3.: €3-92. ___ meooooou ‘2 v . 0.0 v a.~ 1 0.0— fi 3... .eaa:~x\uav no.5; couuou ”mmmcogu move; :ouuou o» mmcoamom mae< :oupou p.m we:m_m . 9.: m.n o.n m.~ e.~ m.— o.— m.o o aneavoou. :.v m.n o.n m.~ c.~ m.— c.- m.o a we 1 i r h > s h F 1 k h > L r » oMc¢onnuuwu noe< couaou a 0.. . c.— - o.~ a.~ v at" F can «0.5; oaoco>< «coeeau -:.-::.::.:.::.:.:..:--.:...:.:..::-:. "Hm 09.5; oaoeo>< aeoeeau :;..;..:.:.:.:.:...:.:...:.u.:unuxx " own ......12 a o c a.» 2..... . 3.... v 0.0 . 0.0 . e.~ . c.~ he; as —ss§— .ea .. ....... -— 40.0 ...-—:.. f 0.9 6.2 . 92 ...: fl ...: .L..:.x\¢d. ...ae.x\ud. ..90. o.W-—.mv 09-5; ecuoou 09.»; couuou __ nsoooaau ...... 2-..... — usoooa-u 167 a) At slightly higher than the current price and around the price of Ls 4.0 per kantar, cotton enters the solutidn ' of all three categories. At this initial level cotton fed- danage is very small; 0.43 and 0.53 feddans for categories I and II. respectively, and only around one feddan in cate- gory III (5 percent of its total cropped area). This share for category III also remains stable over a relatively large price range (Ls 3.86 to Ls 8.43). b) To increase cotton feddanage to one feddan, an average price of Ls 7.00 is needed in categories I and II. c) To increase cotton feddanage beyond one feddan, higher prices are required. At around Ls l0.00 per kantar, cotton feddanage is raised to two feddans in categories I and III, and to three feddans in category II. 2. Prices higher than Ls l0.0 are not only unlikely at present but also the response they generate becomes smaller. 3. From the above, it can be concluded that although a small in- crease in price causes the cotton crop to enter solution, sub- stantially higher average prices (Ls 7.0 and Ls l0.0) are needed to induce a one and two feddan area response, respectively. JIt should be recalled that prices simulated in this experiment cor- respond to average cotton prices per kantar, and therefore they might have to be further adjusted to reflect differences between cotton grades (which means slightly higher prices for grade I, and propor- tionally lower prices for grade II and III). .Possibilities of offer- ing cotton prices such as those derived in this experiment are l68 discussed in connection with the NMAPC cotton pricing system given in the next chapter. Sumnary In this chapter, the analysis of results from experiments with the traditional linear programming model are presented. In addition to base plans for the three farm size categories, the following experi- ments were conducted: hypothesized rainfall-induced yield variability experiment, credit and land expansion experiment, planting-time experi- ment, credit with planting-time experiment, and cotton price variation experiment. Results of the base models show the cropping pattern of the small- holders to be dominated by the dura crop, with a small percentage of sesame. Groundnuts and cotton did not enter any of the solu- tions for the three categories. Eighty-eight percent, 89 percent, and 92 percent of the total available land in the three farm size models, respectively, was used in these basic plans. Resulting returns confirm the typical low productivity and returns of smallholders farming in the Total gross margin figures for the three respective farm size It was shown that area. categories were Ls l07.99, Ls 190.55, and Ls 348.95. these low returns are the product of low productivities for both land and labor, resulting from low crop yields. Computed average productivi- ties for land (per feddan) were Ls 6.20 to Ls 7.35, and for labor (per man-day) were Ls 0.49 to Ls .84, always highest in the smallest farm size category. Analysis of the marginal productivities has shown that the basic limiting resources in the smallholder production are the labor re- sources. Nhile operating capital is also limiting, this indicates an 169 indirect labor constraint, since labor is the basic resource 00 is ex- Land is not limiting. Family labor is limiting mainly in pended on. the first (June through August), is when weeding takes two periods: place, the second (November through December), corresponds to harvest and crop disposal. The family labor MVP's corresponding to these two periods are positive, and high in comparison with the respective monthly hired labor wage rates, indicating that expansion of the labor resource during these periods would be profitable. Results of the sensitivity analysis used to investigate hypothe- sized rainfall-induced yield variability, indicate the sensitivity of returns to reduced rainfall and hence reduced crop yields. |dith 25 per- cent reduction in yield levels (situation A), resulting objective func- tion values were 5l percent, 4l percent, and 49 percent lower, for the three categories, respectively. In the case of the 50 percent yield levels reduction (situation 8), also a proportionally greater drop in returns resulted; by l20 percent, 91 percent and 90 percent for the three categories, respectively. The credit and land expansion experiment results have shown that credit use can contribute significantly to increase the low farming re- turns of smallholders in the area. In comparison with the base plans, resulting increases in returns were in the order of 77 percent, 50 per- cent, and 39 percent for categories I, II, and III, respectively. Al- though the interest rate charged for borrowed capital in this experiment was 7 percent (representing the official rate of the Agricultural Bank of Sudan for short-term loans), the MVP's from the base optimal plans indicate the profitability of borrowed capital, within the smallholder 170 296 percent, 145 percent, and 74 per- system, at extremely high rates: It is further cent for the three farm size categories, respectively. evident that the profitability of borrowed capital is highest in cate- gories I and II, where 00 is most limited, and‘hence credit use is most For the latter two categories, in this experiment, the in- rewarding. crease of land and hired labor availability made it possible to use more of the existing family labor resource (by 37 percent in category I, and 2l percent in category II) as compared to the base plans. Another im- portant result revealed in the analysis is that credit is mainly needed in the first period (June-October), when smallholders in this period re- quire not only production credit but also consumption credit to see them 00 requirements of the second period (November-December) through harvest. In are financed, as farmers normally do, through initial crop sales. this context, the inclusion of sesame in the cropping pattern (with its crop sales occurring as early as November) is particularly important and useful. Results of the planting-time experiment reveal that despite the high physical yields associated with early planting it is still optimal for smallholders to grow late planted crops. The optimality of such a practice arises from the stringent time requirements of early planting schedule, given the already limiting seasonal availabilities of labor. Simply stated, late planting is one way of smoothing out these bottlenecks Nevertheless, all three farm size categories made use of the high physical yields of the early planted crops, as much as their resources could allow; category III, which has the highest resource endowment, was able to grow 75 percent of its crops in the early schedule. The returns picture (in 171 comparison with the base plans) also captures the interaction of the new resource requirements and labor time distribution of the three planting time sequences, with the resource availabilities of the smallholders. Fbr categories I and II returns are down by ll percent and l9 percent, respectively. However, it should be noted that the production opportuni- ties (and corresponding yield coefficients) are modeled differently in this experiment. The returns in category III, whose resource availabili- ties are more suited to take advantage of early planting, has risen by 13 percent despite the sizeable (27 percent) reduction of its cropped area. The credit with planting-time experiment results demonstrate the importance and rewards captured by relaxing seasonal labor constraints revealed in the previous experiments. Through credit, hired labor is used to augment labor resources in the constraining period. In compari- son to base plans, sizeable increases in returns resulted; 22 per- cent for category I, 24 percent for category 11, and 56 percent for category III. The profitability of dura, especially that of the early schedule (Dura I), makes it dominate the cropping pattern in this experiment. Results of the cotton price variation experiment reveal that there are differences in the feddanage response to higher price between the 'three farnisize categories. Such differences are a result of their 395 ac: CSEZT lllll 5825 179 corporation's activities. Particular difficulties have arisen in the ef- ficiency and coordination of mechanization services. The field (scheme) level managerial functions are under the supervision of a NMAPC inspector 1 The function and authority of aided by the scheme's farmer committee. the committee, and coordination with the NMAPC officers, has been unclear and in many cases problematic [46]. (ii) The Modernization Program The central activity of the corporation is its modernization program. Started in the season 1970/71, the program aimed at achieving the follow- ing objectives [31]: 1. Introduction of machines (tractorization) to raise traditional farmers'productivity and annual income. 2. Consolidation of traditional farms into collective farms as a nucleus for agricultural cooperatives. 3. Provision of social services. 4. Introduction of improved agricultural services, techniques, and rotation to raise the productivity of land and crops. 5. Emphasis on improving production and productivity of cotton, so that the NMAPC can stand in a better financial position. Not preceeded by a study, the arrival into the region of some sixty tractors/discs signalled the beginning of the modernization program. The actual opening of schemes was largely on an ad-hoc basis and mostly left 1This is composed of 10 members elected by farmers under the super- vision of Nuba Mountains Farmers Union. They serve their function on a voluntary basis. These functions are not well defined; at the moment they include helping distribute rotation plots to farmers, supervision of mechan- ization services, and in general acting as farmers' representatives in supervision of day-to-day field work. 180 to Uwediscretionof the field stations.In 1970/71, a total of 23 schemes were opened in the eight stations of NMAPC. The general procedure for opening schemes consisted of a group of farmers, after consulting with NMAPC field stations officers, clearing an area (between 1,000 to 2,000 feddans). The scheme was then officially registered and the land di- vided (as tenancies) between the participating farmers. This procedure of unplanned openings of schemes with associated land ownership problems has persisted with the NMAPC system until now. Future land surveys and land expansions (to complete contemplated rotations) are thus made more difficult [46]. The initial target of this modernization plan was to have 300 trac- tors/discs, and a total area of 300,000 feddans by 1974/75. Actual ex- pansion of schemes and areas of cotton and dura are shown in Table 6.1. While the areas of crops fluctuate from one year to the other, the in- tended balance of the two crops (approximately 50 percent each) was main- tained over the seasons. As can be seen from the table, this plan suffer- ed a great set-back in 1974/75, when all schemes were shut down except for three schemes in the Kadugli area. The reason given by NMAPC ad- ministration was the inability and refusal of farmers to pay their loans for mechanized cultivation which were mounting from previous seasons. While the program continued, a number of schemes were being shut down each year in the different field stations, primarily because farmers were strongly reluctant to grow cotton. There were other problems as well, including tribal/village conflicts over land ownership and distribution, managerial and administrative conflicts, and shortage of drinking water at schemes sites [46]. 181 .mueogum coupo=__aem c. sea: saga o» mucuuacu an wou¢>oca a=_oa m. saga as» .cm>m:og .=o_aaum um¢xz ca ac: a. .oxa< .comuom m_=u e. axon uagm «so: moeogum coauan.=cuuas p_a ..pmaua» e. mueogum omega cow uamuxo .uxou «so a. co:.a_axo mn< . own. awn mvm . I see . e—NN _- cm—p mom as. mwm uupoaeocom 5: oom_ . coc- o n . cm—— . —smm v~c_ ¢m_— «.m— ascm emu upmunn< . cm—m ova muem . . Nc—p cam mpmm nmmp came msco mmvm came ozpoaom an< mmpm mmmm Nam. nmmw u . mowm nmmp c~m~ umm_ ocom .mn- mam mm~ ozcmud cmc m—mc oos— «.mu . . cmN— emom mum emu «mm. _mm come mmmm .uo_uh om. omom mmmm emw . - .<.z .<.z . m-mN can. coo cmm u m:_—__a cm— pome n—m omom u . umo Nmo— pmvm mmnm mw—N Nmsn mam NNN .mo—ux cccm . oswn— smnm m_mm «we Ncom who. mcmv omcn mwmm .mcm muom camp nem— .ma—ox mesa eouuou «can coueou scan eouuou scam eouuou «Lao .nmmwuou nemmlilamuuou acaa .mowuou -\m~m— oNNWNm— um~\¢~m— eNNmso— m~\-m— -\_~m— —~\o~m_ m~\cmmp o» PN\o~mp .mmEmsum :owuchcgmcoz as“ cw cmpcmpm Amcmtcmm :Pv mgao vcm :ouuou we mmmg< nom<22 p.© mFQMH 182 Despite the problematic record, the modernization program was given special emphasis in the current Six-Year Plan (1977/78-1982/83). The government plan for revitalization of NMAPC was expected to benefit from the findings of the recently started foreign research projects in the area.1 By 1979/78, a total of 62 schemes were operating in the eight stations of NMAPC, commanding a total area of 72,868 feddans. Mechanized Cultivation in the NMAPC The mechanized cultivation service, as pointed out earlier, is con- sidered the corner stone of the NMAPC modernization program. Its pre- sumed capability to "raise traditional farmer productivity," as expressed in the modernization objectives, closely resembles what Binswanger [13] has called the "net contribution view." In general, proponents of this view believe that agriculture is characterized by a power constraint, and that tractor machinery is suited to breaking this power constraint in agriculture, such that a net contribution to production (which otherwise is not possible) is generated [13]. However, the origin of this view, within the Sudanese experience, traces back to the early 1950's when mechanization under rainfed conditions was first introduced in the Gedarif region of eastern Sudan? To provide a mechanized cultivation service, NMAPC relies on differ- ent types of machinery (tractors, discs, and Seeders) that are distributed 1Some of the relevant features of these projects to NMAPC moderniza- tion program are briefly discussed in Appendix IV. 2For an example of early advocates in this regard, see Shazli [52]. 183 throughout its field stations.1 The mechanical cultivation service con- sists primarily of shallow ploughing of the soil by the wide level disc. Two discings are stipulated; however, at present due to lateness of oper- ation, gasoline shortages, and mechanical breakdowns, very few schemes are disced twice. In the case of lands designated for dura, the sowing operation is also performed mechanically, through a mounted boxer/seeder on top of the disc, simultaneously with the second discing operation. Ideally, the cultivation operation should achieve a number of objectives including: preventing of soil erosion, conservation of soil moisture, control of newly germinating weeds, eradication of tufted and perennial grasses, and the creation of an adequate tilth for promotion of seed germination [45]. For farmers in the area, weed control is the single most important objective of the operation. At the present, the field performance of these operations is widely recognized as unsatisfactory. This judgment is based on both the timing and quality of the mechanized cultivation service. The farmers inter- viewed in the researcher's survey have singled out the service as the most unsatisfactory and problematic. One of the few available records of; field performance is that of the Lagawa station for the season 1979/80.2 It is believed to be typical of field performance in other NMAPC stations. Table 6.2 shows the total ploughed area by each of the thirteen tractors 1By 1979/80, NMAPC had a stock of 140 tractors, composed of five dif- ferent tractor makes; and 139 discer/seederS‘ of three different makes. 2Machinery records (use, performance, maintenance, fuel consumption, etc.) are lacking or incomplete. Efforts to organize and keep such re- cords have recently (1979/80) been initiated by the Agricultural Engineer- ing Section, NMAPC. 184 .Amv cowamgmm momma: "Amp .m .N .m .Nv mcmmo snow “Amp .PF .op .m .m .e ._V egom ”mzoppoe mm umuzawcumpu use Amp cmzocgp .v nexus couumgp .mu_zu umgmzo_q new. mmczpucun .mmmcaco new .mxmmamz .ocmmemw .xac< .mzmuxzz .PnonEo: "om\mnmp common as» :p mcpumcmao mew: mesmnum xpmm .u¢o eocm umuuzcumeou "mucsom map, mme omm swap ommp pom ompp omm woe mpm wmm mew _Nm Pouch mm m mN\.m=<-op\.m=< oe_ em we we Nem me. am m_\.m=<-m\.m=< ¢e_ ON so, m_ cm we, mm ~\.a=<-P\.ms< vow om_ mom «on mow mmm mo_ map app mFN com _m\»_=a--\zpaa mm mm omw _o_ mop ow me m_. mpp mpp pup _~\xpaclm_\xpza «mm oo_ mmm cow om_ we mNP Rpm mm mm mum op\»—=alm\»P=a mm com om— mme mm oep mm, mm Fae mmm mmm emu “\xpzaamm\m::a mp Np pp op m m N o m e m N — gfiwcacuom :PV mcouumc» an umsmzopalvcms uo_caa mom\a~mp cemaom .mueoeom eoeeawweeoeoz age co ao2< vacancy; Payee uua¢Pmo on muomcucou oop.o ¢._ o.oom.op Ammuwcco upmpcv cone; ech: - ...N o.om_.¢m_ ccaom m=_s=m eoooou toe mowea_am oo_.o m.o o.mnm.m mmmcmaxm mcpumos ooa.o m.e o.muo.mm coeeou co ecoameae» caoepcam aw; cwv mmmcwmwu ~w4cpwl mwmmm cepcmx pouch mmmcoaxm Em cm; eo “moo conm umuumgxm peace a“ mow\mnmp commmm one cow covuuzuoca cmuumaxu no women mwmcmaxm acaouu< u=_oa uuacmmmm peoemp m;u_coc mmom ;u_gz mgmzm m.u:w=mu as» co am mcpewmemc as» mcwuspu:_ no: ..m.wv am“ so commmm H~N_ .a .mmu we: ”wotzom me.. ee.~ m_.p ewe=a¥\ms .eeossaa mmooxm. _ao_eoeoose F.moe o.-e e.~_m Aooo.msv ”League a“ .eeossaa mmooxm. _ao_oocoa;e ep.m e_.m em.~ cae=a¥\ms ao_ea .o>< eeaepsmom e.mm~ m.m- N.NNm Aooo.msv maoeeaa o8 u_aa s._e=eo< oesoa< me._ oe.o oe._ caeeax\ms ao_2a .o>< oeao_=mom m.omm N.mm c.mmm Macao.msv atmgm .measeaa m.me¢ m.mN_ e.e_e Aooo.msv e_coea mmoeu P._me.P ..omo.P ..mem._ Aooo.msv ecsooo< eepoe to memou e.oem.2 . P.m_m._ «.mmN.N Aooo.msv ozcosam _aeoe mee._¢~ mms.ee~ _eo.m_m meeeeax =2 eoceoseoea _aeoe m~\m~m_ mm\eem_ N~\e~m_ Ease me\m~m_ co Ne\eemp .moosea .meoEtaa ace .ocsooo< “ghee .o==a>o¢ coeeou _aooe nua gopqmgu c? cm>_m «gm mmmugo>o «was» mcpuu_=o_mu cw vow: magnumuogq ago mcopuassmmu maps .cowuasamcou u_osmm=og go» umcwoums scan co m=_m> an» mmczpucga cmgagaoo "mugzom mm.o Pm.o -.o m4 amo-=mz cm; mm.mn Nm.mu om.m- m4 :muumm can ammpup>wuu=uoga mmmcm>< hm.mm mm.o¢ ¢N.em m4 mcvmcmz mmocw Pmuoh No.¢m- mo.mp- cm.wp- m4 mzpm> cappuczm m>_uumnno megapmm mp.p m~.~ m~.N m4 c_mg ummgmucfi _ mm.op mm.m_ mm._m m4 _mupaau uazoagom . _m.~ -.P~ m4 mmpmm aocu % 8.8 8.x 86 3 .338 5:5 1 3.2 3.3 24.6 3 _338 2.52on :38 mm.oep om.oe_ m~.ce~ saa-=az gonad Pabop om._ op.e mm.m amoucmz Loam; c_$mz m~.oN m¢.pm mo.nm amo-=oz gonad was“: e~.m__ mm.m._ mm.oo_ xoo-=mz L59”.... »_Psmm oo.o oo.m oo.o amoncmz can; _mpo» mm: muszommm oo.m oo.m oo.m conned ecu: oo.m oo.m oo.m cannon couuou zcmuumm napaaocu A.uac m.mN-o.on .~.umc m.m-..mv ~.uac o.m-P.oml “we: say“ HHa Accompmu HH agommgmu H accumumu Page: Etna mpcaawuppcaa uaoz mo.o ~o.o so.o m4 poo mo.o mo.o so.o m4 gum no.o so.o mo.o m4 ma< mc.o No.o mo.o m4 _aa No.8 eo.o No.8 m5 gee - _eepaeu me_peeeeo - - em.o aua-:ez ma< om.o om.o em.o zeo-:ez an: mm.o mm.o mm.o xeo-ccz sea gone; LPeez - - o~.o xeoucez ma< Nu.o Nu.o no.o xeaucoz an: mm.o mm.o mm.o xeo-=ez can Loam; >F_sem o~.op m~.cp m¢.m_ cacao; mean o_.~_- mp.~.- mm.o_- canned :ouuou m=o_au_gumom coco A.nm» m.m~-o.opv .~.umw m.¢-p.mvl A.um$ o.m-P.ov u_== wuezommm Hug xeommueu HH xeommueu H agommueu AMA epv ee_ee geeeem . _muoz Seem u¢a :.t:.::.:. «coccau ..eo. e.m~-e.o_. ___ us¢uu-u 9.0 c.~ a.a 0.0 6.9— . a... v 6.N- Asqucqxxna~ o.n m.~ o.~ .aenveou. 005‘ COquU ou¢ua ounce)- acoucau v...-.o . - a - ...e. ...-..m. __ Asooouau 5.0 o.n 0.0 6.0 6.:— o.—. a.~— asuucqxxadv 09—Lt COUaOU on” a.“ °c~ m.-. o._ are .ae.ee... - i 005‘ §d8 59.5: canto». . .: acoccau ..sn. 8.“-..8. _ agency-u 0.0 o.“ a.~_ .c-usauxnd. Ou.b& BOaaOU 207 Analysis of the Future Full Scale NMAPC Model The analysis of the NMAPC model is extended in this section to in- vestigate the future full scale model. First, a brief discussion of the features and model specification of the future full scale NMAPC model is presented. Next, the LP results are discussed in terms of the optimal production plans and net returns, seasonal constraints, and a proposed alternative to the future NMAPC model. The Future Full Scale NMAPC Model The NMAPC is expected to benefit from the research now underway in the area.1 In the analysis that follows, however, no changes in crop yields are assumed. Changes in the NMAPC are limited to the following features planned for future implementation: 1. Tenancy size is expanded to the 15 feddans, to be grown in a rotation of cotton, dura, and fallow with five feddans each (i.e., 10 feddans cultivated annually, equally divided between cotton and dura). 2. A second experiment with the model allows all crops to be culti- vated, not just cotton and dura. Guided by the cotton price analysis experiments conducted earlier, the cotton price was raised to an average of Ls 10.0 per kantar for this experiment. 1A brief description and discussion of some of the on-going foreign aid research projects is given in Appendix IV. 208 3. In the future, the NMAPC (as the case in most public irrigated schemes) is expected to offer credit to its participants to meet operating capital requirements. The present NMAPC credit system covers only the mechanized cultivation operations. Re- call that the present NMAPC model required credit to get feas- ible solutions. In the full scale model as before a 7 percent interest rate is specified. The introduction of one-third (5 feddans) fallow in the future NMAPC rotation is aimed at better maintenance of soil conditions and productivity over time. Since the use of chemical fertilizers is not practiced (or contemplated), fallowing is one way of maintaining soil fertility. Although the fixed tenancy size of 15 feddans is in line with policy decisions in all other public schemes, Zaki [61] criticized the fixed tenancy size standard. In discussing the problem in the context of the Rahad irrigated project (eastern Sudan), he identified three vari- ables affecting the tenancy size: 1. Tenant household size and composition (as both a source of labor and as a consumption group). 2. Income from the tenancy. 3. The impact of crop mix and rotation on labor requirements and returns to labor. In his analysis of alternative tenancy sizes based on different schemes, and in recognition of households heterogeneity (size and composition), Zaki concluded that "a fixed tenancy size for each tenant is unjustifi- able, whether in terms of the tenant's household needs or labor potential supply" [61, p. 164]. 209 In the Nuba Mountains area, the question of desirable cultivation size and crop composition depends on two basic considerations: 1. Productive capacity (mainly labor force) of the household. 2. Income and dura consumption requirements of the household. Both features were incorporated in the representative LP models. thimum Production Plans and Net Returns The optimum solution of the future full scale NMAPC model for the production categories is shown in Table 6.8. A total of 234 man-days of labor were required for the 10 feddan (equally divided) cotton and dura cropping pattern. Inclusion of the labor intensive cotton crop (5 feddans) in the rotation requires hiring of labor by all three categories. The percentage of hired labor to total labor input rises from 30 percent in category III to a high 47 percent in Category I. Labor is hired exclu- sively during the months of August, January and May. The latter two months correspond to cotton harvest and pest harvest (stalk uprooting) operations, respectively. Category I needed the most operating capital, Ls 66.79, of which Ls 38.29 had to be borrowed. Farmer- returns under the full scale NMAPC model show substantial improvement in comparison with the current NMAPC model(given in Table 6.6). As can be seen from Table 6.8, total gross margins (including the value' of dura retained for household consumption) are Ls 93.77, Ls 101.10, and Ls 124.90 for the three categories, respectively. These returns levels, as compared to corresponding levels of the current NMAPC model, represent an increase of 174 percent for category I, 118 percent for category II, and 72 percent for category III. These increases in re- turns are the result of expansion of cultivation size to the 10 feddan 210 .covga33mcou uponmmso; com no:_upmg asst mo ozpm> on» mmvzpu:_m umuznaou "muczom om.¢- o~.pop -.mm m4 m=_mcm: mmogo —muoh o~.m¢ op.mm no.o¢ m4 mapo> :o_uu:=m m>Puumnno mceauom . o~.o wo.~ no open ummeoucm - ne._ a~.mm m4 _ae_aau cassetem . mo.om mm.mp m4 mmpmm coco Ne.mm mm.~N m¢.a as Paaweau _eeaeeH mm.mm ~¢.mm om.me m4 _eu_amu mcfiaesoao _muop m~.em~ mm.¢mm o~.emm ancient cone; peach om._ on.~ mm.e maniac: gone; sweez om.mo om.em m~.mop anoint: L33 umswz mo.mmp em.pe_ om.o~p menace: gone; x—psmm co.o_ oo.op oo.op :muumm use; —muoh mm: ousaommm oo.m oo.m oo.m annum; mesa oo.m oo.m oo.m sauce; couuou stopped mcwaaogu .umw m.m~-o.o_ .cmw m.m-_.m .umw o.m-_.o a _HH agommumu v AH“ xsommamu V NH Accumuem M ups: Emu~ _aeaz e_aem p_=a deeded geezz eat mee_a eeaeeeeees _aeceao m.a apaae 211 specified for the three production categories in the full scale NMAPC model. Given the different resource availabilities (and dura consumption requirements) of the three production categories, the NMAPC standard of a fixed tenancy size for all participants is clearly unjustifiable. In particular, farmers in categories II and III would benefit from a more flexible tenancy policy that recognizes their relatively higher levels of both labor resource endowments and dura consumption requirements. Another important policy matter, related to the tenancy size impact on participants' returns, is the required crop mix. The NMAPC rule for minimum 50 percent cotton in the two-crop mix was shown in the previous section to have a direct influence on returns. This was explained in the light of the present low cotton yields achieved by farmers and the low cotton prices offeredaby NMAPC. The relative profitability of dura and cotton, and particularly the losses associated with cotton, are also maintained in the full scale NMAPC model. However, in the case of the full scale model, the expansion of the cultivation size helps to offset some of the losses associated with the cotton component. An additional feddan of cotton in the optimal plans of the full scale NMAPC model is associated with losses of between Ls 1.87 to Ls 5.23 (depending on the production category), while an additional feddan of dura earns net re- 1 turns of between Ls 10.29 and L5 12.50. The importance of the crop 1This is based on the MVP's of the specified cropping pattern re- strictions. See Table 6.9. 212 mix, and in particular the size of the cotton component within a more flexible cropping pattern, is investigated as an alternative to the fu- ture full scale NMAPC model by a following experiment. Seasonal Constrains Under the Full Scale NMAPC Model Table 6.9 gives the shadow prices for the limiting resources under the optimal plan of the full scale NMAPC model. Despite the reduced labor requirements of the NMAPC model (see last section in Chapter IV), the full scale NMAPC model displays basically the same two labor constraint periods of the traditional model. In comparison with the latter model results, the labor peaks are somwehat delayed, starting in August (and September for category I) for the first period, and January (or December for category I) for the second period. The month of May, as was the case in the current NMAPC model, is also labor constraining for all three production categories. This month would not be constraining if the cotton stalk uprooting ("Al-wadi") operation could be spread over the period March through May instead of 'delaying it to be finishedby the 'NMAPC deadline of May 3lst. The display of the two labor constraint periods, more or less simi- lar to the traditional model, in the NMAPC's full scale model has two important implications. First, unlike the current NMAPC model situation, where NMAPC participants who cultivate both private and NMAPC plots would be able to distribute the work and spread labor peaks between the two types of plots, in the full scale situation this would not be possible. Given the priority attached -by farmers to their private plots, cultivating both plot types under the NMAPC full scale situatiOn would likely result in further reduced management levels (especially for 213 umuzasou "musaom D r _ooo.o Pooo.o pooo.o m4 aazicen _ooo.o no.o mo.o m4 conicaa . _eupgeu mappesmao . . m~.o xeoicez new mm.o em.o em.o xmaica: m=< em.o em.o wm.o >45.52.. an: . cm.o mm.o mm.o manned: and . cone; spwez . . mm.o xmaicoz own i . mo.o amount: amm mo.o o~.o o~.o amoicoz m=< mo.o No.o mo.o xmoiccz at: no.o mm.o mm.o Assign: new . Lone; »__smm om.mp m¢.mp m~.op canned mean mm.pi mm.m- mm.e- annum“ couuou m=o_uowcummm coco A.umm m.m~-o.opm A~.umw m.mip.mv A.umw o.m-p.ov “we: mossommm HHH asoamumu Ha accumuou H acommumu ~m4 emu mocks; sommmMT _eeoz uaezz epeem Ppea we“ ea maeesemem me_e.e_4 use see mae_ea zoeaem 8.8 a_aae 214 the cotton crop) applied to NMAPC plots. Second, and more importantly, the display of the two labor constraints periods in the full scale NMAPC model indicates that the NMAPC modernization program has not significant- ly relaxed the main constraints of the smallholder production, as revealed by the analysis of the traditional model given in Chapter V. Although the introduced tractor ploughing, which farmers in the area view as the equivalent of "mechanical weeding," helps in reducing weeds and hence weeding labor requirements, this is not sufficient to remove labor bottle- necks associated with weeding operations. The suggestion given pre- viously within the traditional model analysis, for the need of consider- ing technologies which are particularly oriented towards breaking the labor constraint, is also relevant here. In particular, the work by Hamdoun [26] in chemical weed control has potential for relieving weeding labor constraints. Under the Nuba Mountain conditions some research in this area is already underway.1 An Alternative to the Full Scale Model The preceeding analysis suggests the need for changing the standard fixed tenancy size policy. Tenancy size should recognize the potential labor supply of the household (especially at the two constraint periods) and the household's dura consumption needs. 1The chemical weeding research is conducted presently in the area by a chemical company (Ciba Gigey) under the supervision of Kadugli Re- search Station. 215 Another important issue is the required crop pattern. At present the government policy of promoting cotton is at odds with smallholder interests. To shed more light on this issue, an experiment was conducted under the following assumptions: 1. To focus on the issue of crop mix, the 10 feddan size was kept the same, while all crops (cotton, dura, groundnuts, and sesame) were admitted in the determination of the optimal cropping p1ans, with no restrictions an individual crop areas. 2. As before, no changes in productivity levels are assumed. 3. As before, borrowing is allowed to meet the farmers' operating capital requirements. 4. Guided by the two cotton price programming experiments, the price of cotton in this experiment is assumed to be an average of Ls 10.0 per kantar. Results of this alternative full scale NMAPC model are shown in Table 6.10. Compared to the results of the planned full scale NMAPC mo- del, this alternative full scale model with unrestricted cropping pattern has several advantages from the farmer's point of view. Relative to the planned NMAPC model, the alternative model generates much higher returns. Percentage increases in returns for categories I, II, and III are 70 per- cent, 76 percent, and 52 percent, respectively. The higher returns are directly related to improved cr0p composition. The increased cotton price has little effect (compared to freely admitting all crops) as the areas of cotton in the solutions were all relatively small (under and around one feddan). 216 .co'uasamcou twosomao: toe umcpmpmg mean we mzpe> can mmuapuca a .gmucex can o.o_ mg on o» cmsammm copped we woven as“ use .mcoquFLummc nogu poauw>vucp oc cur: umzoppo maoco ppm .mceuumw op m~Pm vamp _muopm umuaqsou "ouczom HHH accompmo HH acommumo H xgommumu m~.mm. mo.wm_ em.mm— m4 acvmsmz mmosc Peach mo.o_p mo.mpp em.mop m4 ozpe> cowuoczu m>Puomnno mcezamm . i mm.— m4 ave; pmmcmucu - - mm.m~ m4 _ae_aeu edzaesom . mm.mm me._m m4 mmpmm aoeu mm.mm mm.nm mm.o m4 _ma_aeu Pepapcm Ne.mm mu.oa mm.mm as _ee_aeu me_eateao Pause oo.~mm ~m.me~ m¢.mm~ maniac: Loam; Pouch . co.m om.m maniac: cone; tweez. _~.~ mm.o~ om.~¢ monies: gone; ums_= mm.em~ oe.mpm oo.mmp xmoicez cone; »__Emm oo.o_ oo.o_ oo.o— cacao; use; peach mm: mosaomma eo.¢ m~.¢ mm.¢ :euomm «annum om.¢ m~.e mm.e cacao; wean No.p mm.o mo.o cucumm :opuou :Lmuamg mcwnaoso a cm» a mm o op“ A ewe m m P nu, a any 0 m p cm “we: smu— mxwz aosu umuuwgumogca cup: _muoz w—oum ppzm um037» eeeneea n.Neo «3:08:00 pug-chm; «wig; 44.0850: «230» m><3¢<¢ O38 acmZI .m .a .mmeu New ”muszom o.“ Una OOH - \ / ' oo'o|.|o'olln’. Y .s ..2., z. \ 0"" 0' D 0‘. ‘53‘ Al ’0' ol 0' 0' o I. o \.\ 1.1. .I ..... .... a z z . . . . . OI .\ I.’. m I/‘Illll I .\ I. . . r23. .\ /.l .\.s I. .\ .l .\ .\ .\ .s \ .\. .38. \ 3‘. .\ .\ M m3 m.~ + mm me. me_ am. __~ __~ m o.o + mm mam om. mmm mmm Fem m e.m + um N.N._ e.mm a.ee a.~m_ e.me_ a\eemaam mum: mcvzom mmuao mcwzom Aae\mmv e_a_> Sees Ae;\m¥v ape.» ee_3 weedseseaxu aspen e.zem eeeeau ..HH e_aae 240 Table 11.2 Dura Sowing Date Experiment, Kadugli and Habila (1979/80) Grain Yield (kg/feddan) Kadugl ia _ Habilab Top ' Topc Sowing Dates Grain Yield Rank Grain Yield Rank Mid—June 1237 1 Mid-July 1005 2 Early June 1003 2 Early July 628 3 Mid-August ' 497 4 560 3 Mid-July 383 5 685 1 Early September ' 460 6 278 4 Mid-September 409 7 173 5 Early August 374 8 685 2 Source: Kadugli Research Station [6] aKadugli: Early sowing was highly significant. Reduction in yield from early July were: 39 percent for mid-July, and 40 percent for early August. bHabila: Trend was the same as in Kadugli but less pronounced be- cause earlier dates were not tested. Reduction in yield from mid-Jul y were: 0 percent for early August, 18 percent for mid-August and 60 per- cent for early September. CT.U.B. (early-maturing) consistently outyielded DABAR (medium- maturing). 241 .mcopuumsmucv may go moms ummm on» mcw>_o>:_ mocwcmmwwc pcmuw$_:mpm oz .mmwumwsm> ozp mzu sow mamsm>m an we xpm>wuumamms .w\m new “\wP toe acmuemn mm ace acousma up ,mcpm_» cw :o_uo=umc cow; Asmuumn mcwma .m.=.»v xuowge> new moon mcpzom sow u—m?» :Pesm c? mmococmemwu aceupevcmwm nozozm mpmxpmcm _muwumwunumm .an copumum sugmmmmm Ppmzuex ”ougaom um.“ mm mm cop mm, mop Nap omm owe mom com: .I mew mm _e mm. mpm m_m amp new mom .e.g.s m mm + mm o_N Pm me. me_ _P_ we, eon Rom com .e.a.e a mum Pm app mm sup ¢m_ mNP omm mme .$.a.s e new: ¢N\m\mp ¢N\m\m ¢~\m\m~ ¢~\~\m ¢~\m\mp e~\m\m ¢~\~\m_ ¢m\~\m em comm mute o<¥ seapea> .m.:.h >«.52; mcpzom Aeeeeae\axv epa_> e_aeo m.Am~\¢nmpv “pmznmx .uzwswgmaxu mama comm new mama mcp30m muse m.HH mpnmh 242 3. Sesame The sesame crop has received less husbandry research at Kadugli Research Station. Results from a 1974/75 sowing date and seed rate experiment are given in Table 11.4. Due to these results the Kadugli Research Station report [6] indicated that because of its short-growing season (90-110 days), sesame is less susceptible to delays in the planting time. For sesame, the following assumptions are adopted: 1. Farmers would grow short-maturing variety Herahiri. 2. Optimum planting date is mid-June to early July. 3. Planting in July 15th and August lst reduces optimum yield levels by 5 percent and 10 percent, respectively. 4. Groundnuts The effect of sowing date on groundnut yields is shown in the results of an experiment carried out at Kenana Research Station given in Table 11.5. The Kenana Research report [7] concludes: The outstanding feature of the results, which was also reported in the first year of this experiment (1963/64), was the extremely sharp drop in top yield by more than 50%, as a result of mere 14 days delay in sowing from July 6th. Delayed sowing depressed pod yield to an even greater extent than top yield. For groundnuts, the following aSsumptions were adopted: 1. Farmers would grow either the local variety or "Barberton." 2. Optimum planting date is July 6th. 3. Planting in July 15th and August lst reduces optimum yields by 33 percent and 66 percent, respectively. 243 .xum_sm> mcpgaumsisawume mcu ummequzo .xum_sm> mapsapmsiucocm .Pspsmsm: .mmwum_tm> ozy «so see >Fw>wuomgm um... Jameson .3 EB ucmutoq m me: SE use in so» «232 so: 33» 5 53263. $326 @533. new 8.8. ummm .mmpumpge> ms» cmmzumo upowz ummm cw mmocmsmwwpu ucmuwmwcuvm zpsmw: vmzogm mpmemcm pmu_pmwumumm Hog copuoum gusaommm v_m:uex "moeaom «p + mm em_ me. am. eeF ae_ me_ mm. eeaz mm am we Ne me me as .t.a.s m e_.H Pm Fe Fe em ma 8.. m__ .e.a.e e um . em_ mNF em_ _m_ o_~ m_N m_~ .e.a.e m mmm New mam . . new emu e_m mmm .c.a.e N eta: me\~\_m e~\~\m ee\e\e_ ee\~\.~ e~\e\m ee\a\e_ apex eeem xompm xwmm: _muod xpowge> Pcpgec : Numwsm> Aeaeeae\mxv e_a_> e Ama\e~m_v ¢_meeex .eeeepedaxm seem eaam eea spec me.zam esemem e.- e_aee 244 Agonppwusmm u m .muma mcwzom n my HRH =o_»aom nosemmmm ececmx "musaom a. ...,. we I. m—m o .cmmz Amp + mmv x mmm mmm pmm mm_ NPN mam _me a a .cemz E m .5 EM as :m m we mmm emm cpm neF em. com “me mm < oem 5mm emm om_ epm emm eme ems m can moe opm mpm mmm mmm mmm umcoeoo —mm ewe mam omm mow mmm m_m couponsem new: 2N zo zm, 2w...M s PPPpsum zo cam: oN. u xumvge> xeae mm xpza m zuo_ee> flaw memo mcwzom Aeaeeae\exv e_e,> eoa ,=.e¢ smog: mmeumwse> uzcccsoeo snot co sm~,_vusmm maocmmoeuez new mac: acrzom to aumeem as» m.HH m_neh 245 Time and Rates of Weeding To achieve optimum crop yields not only the planting time.but also other subsequent weeding operations must be performed in time. The op- timum number of weedings and weeding times for the four crops considered is briefly described below. 1. Cotton Optimal weeding times and rates for weeding cotton are shown by the experiments carried out in Kenana Research Station, given in Table II.6. The Kenana Research work [7] concluded that two weeding opera- tions to be speced 15 and 30 days from planting, are required for cotton. This recommendation is adopted as an assumption for Chapter V's planting time experiment. 2. Dura The research work in Kadugli Research Station [6] has shown that: l. Dura needs one weeding as early as possible (15 days) during crop establishment. 2. Progressive decrease in yield is expected with delayed weedings. The further the delay, the greater the decrease. 3. Delayed weeding decrease yield irrespective of the number of weedings. _These results have been substantiated by work from Kenana Research Station as shown in Table 11.7. We adopt the assumption that dura needs only one weeding, two weeks from planting. 3. Sesame and Groundnuts No experimental work was available on the number of weedings and times for these two crops. We adopt the assumption that two weedings are required for sesame and groundnuts, at 15 and 30 days from planting. 246 mom“ .Amnapv sneeze: ”mugzom e-m~m.m m-ewm.u m-aem + mFe.N .H Nmm.m .H NNm.e .H .m.m momm mmmu mew. mp.” ew.em e_.m~ mm.mm Na.ak eat: mumm eemm .eem Name me-em ~o-ae mm-_o_ me-aa me .oe .me .on .m— amom Namm NMNN mmoe _o-em ao-mm mm-em op-.e me .oe .me .om wmpm ammm spam m_mm ~_-mm um-ae Ne-ea mo-om co .me .0m .m— Nmmu memm meNN mewm oe-mm me-om ..-Nm om-me co .me .om _mom omen em_m muem cm-am em-_m ea-ma em-Pm me .om .m. Nemw mamm Nee_ eemm --mm No-ae me-em a_-ae me .cm mmam __m~ oemp emem __-Nm m~-me c~-mo_ oe-e~ on .m— eom_ mew. awe, menu mm-mm we-m~ ea-aa _N-om om ewe. eee_ omm meNN eo,mm mm-he hm-ea m~-e~ me New e-_ mmm emo_ ae-Nm m~-e~ m~-mc_ am-aa o eaaz eemri meme lama. eeaz ouch. meme mam, Aae_sem ease usage Aaewmxv e_e_> eoeeao eeam Aezxooov mcoeuapaaom acme; nauseous» m:_ummz e.e_> eoepeu eeam eea ea_ea_=aea sea—a eepeeu ea mme.eeaz to tease: eee new» to “pace” e.HH e_aae 247 xumpsm> zmze mama .Amum_v enouae: ”moszom Amgxmxm upmw> :ngw ammxooov copumpzqmm ace—m e _mF.H e NNN + m mop.“ 0mm mi.+ gem up.“ mam _~.H .u.m Rem. _emm cam eop_ me ~e_ _m cm, mo mm, me am_ eaaz em_~ mowe‘ Nam _~N_ om me, _N am. No me. me NNN me .oe .me .om .m_ emoN emem mam mm~_ _e mm. mm aw. go Now as so, me .oe .me .om e__~ me.e coo, map. a“ ee_ em om_ e. me. me em. so .me .om .m. mam. swam emm mom. ee me_ me m__ mm mm. me em. on .me .om Noam nape e_m nee. am an. ap mm. mm em. .8 m_~ me .on .m. mmom moan _em a_N_ _m mm. me e_. we Ne_ e_ om, me .om m__N Paee _Ne Nae. __ ma_ em em. No mm. no em_ on .m. away mmam mam mme we we. mm mm, em _e_ _e mm. on mmm_ Pemm Nee. meo_ m. mm, on no. em mm. mm emu m_ em._ a_e~ 0mm Now an mm. o_ No, we so. No ea. c eaez (aoea_ meme mam. eat: cheer weep, moor Aaeezam ease asaov acmsumwse unecomz upmw> :wecu use co_umpaaoa peep; Ezsmsom co mmcpumoz mo smaszz can ms_p eo womeem n._m open» 248 Other Agricultural Operations Harvest operations are to be carried out promptly according to the crops maturity periods. Delayed harvest, especially for sesame, results in substantial losses, as the crop is dehiscent (opening of pods) and quickly shatters the seeds upon maturation. APPENDIX III ARTICLE 10 OF THE BASIC NMAPC CHARTER: THE JOINT ACCOUNT SYSTEM APPENDIX III ARTICLE 10 OF THE BASIC NMAPC CHARTER: THE JOINT ACCOUNT SYSTEM] The Joint Account The corporation (NMAPC) should establish a special account (joint account) to be credited by: 1. The total returns from cotton produced in the Nuba Mountains area. This includes returns of cotton fibers, seeds, skarto, etc. The total returns from any other crop that the operating cor- poration (NMAPC) may decide to include in the joint account system. The value of any sold machinery, equipment, or materials which have been bought from joint account funds. The value of fines.or compensations,collected for any damages inflicted on the cotton crop or any other crop included in the joint account system. The corporation (NMAPC) joint account is to be debited by: 1Translated from the Arabic text given in E1 hag [22],“NMAPC and the Development of Traditional Agriculture: A background review cotton growing and Agricultural Development." Presented in Agricultural Moder- nization Seminar, Kadugli, 1979. 249 250 l. The total costs incurred in sunning and sterilization of cotton seeds, spraying of the cotton crop by insecticides, or cotton protection from pests by any other means.1 2. The total costs of cotton packing materials (sacks, strings, etc.). 3. The total costs incurred in establishing cotton local markets, collection, storage and transport. 4. The total costs incurred in cotton ginning, packing and trans- port. 5. The total costs incurred in storing cotton seeds for any period, as might be necessary. 6. The total costs of any other expenditure incurred in cotton collection, processing.and disposal.2 7. The total costs of opening and maintaining cotton roads.3 Division of Cotton Net Returns Cotton net returns (after deduction of the joint-account costs) are to be divided according to the following sharing formula:4 1Mainly includes the costs of the cotton stainer-bug campaign under- taken by the NMAPC. 2 . ‘. . Includes such as labor wages, insurance and sales commiSSTons. 3Cotton roads are the roads opened in the different localities in the Nuba Mountains area which are intended to aid in cotton collection and transport to gins. 4In case of Abu Habil irrigated scheme, cotton net returns are to be shared 50 percent for NMAPC and 50 percent for the scheme farmers. 251 NMAPC 15.5% Tenant Reserve Fund 2.0% Farmers 78.5% Social Services Fund 2.0% People's Local Government 3.0% Payment of the Farmers' Share The farmers should be paid as soon as they deliver cotton to the collection centers. Initial payment to farmers should be according to the rates (prices) recommended by the NMAPC administration committee and approved by the board of directors of the Public Corporation for Agri- cultural Production (PCAP). The resulting surplus (deficit) arising from the initial payment to farmers and the final actual farmers' share (78.5 percent of cotton net returns) should be deposited (subtracted) to the Tenant Reserve Fund. Division of the NMAPC Share From the 15.5 percent share of NMAPC, the NMAPC should deduct all total expenditure (use funds and special allocations funds) which have been approved in the fiscal year (in which cotton was produced) budget. Any remaining surplus should be divided as follows: PCAP 85% NMAPC Reserve Fund 15% 252 Agricultural Season Account At the end of any season, the corporation (NMAPC) should establish a separate account for the season. The corporation should close the farmers' account and pay their final share for the season not later than June 30th of the next year.1 Farmers' Debits to NMAPC If the farmers' share in any one season is not sufficient to cover their debits to NMAPC, the NMAPC may transfer and subtract the remaining debits from the farmers' share for the next,or any other, season. 1This is in case more than 80 percent of the season's cotton crop has been sold. .mmm_ abuenmg< new .mwg< mnemucsoz can: cw ocepumo :o_uu=vog¢ coawoo we mmmzmu asp mcpumms» use mcwzuzum sow mwuuesEoo we» wo usoawm .mmmu mmoczommm peszuez can uoom ,mcza—zupcm< mo aspmwcpz ”ooezom 253 omm.~ Nom.m oom.u mawsm>< mom.m eme.e omo.e mc.em~.m_o,F omm.m me.m~m.mpp.m mm.me~.ne~ m~\mmm_ «mm.~ mom.m mpm.m mm.mmm.mmm —~N.h om.mpp.nmx,— lum.ome.~em m~\-m_ ou~.e sep.m mm~.m .~.mm~.mmo._ ~mw.m ae.o-.w~e.~ mm.emm.o_m m~\mmm_ Nam.m ump.e mpm.~ ee.mow.m~o.. omo.~ No.m~_.mom.~ mm.om_.mpm m~\m~m_ uno._ epm.p mee.m _mm.oo~.mo~ mcu.e mm.mmm,emm en.m~e.__~ m~\e~mp ~m4M Away Tdmav ”may Amdv Away Amseucaxv common Lone; smn_m page; saw couuou Acouuou momesosza smucax Leucmg Leucex unzooo< soap; soawuv :ouuou so; can. so; ucwoe saucex: mcsaumm Pouch weezm mcszumm uczouu< _muop so; cougou mamasmm umzl acemn magnumm Ame\memp - me\e~mpv maeee.eeeaxm eeseeae pe_ee eee meespem .memaedtae eeeeeu _apee "maezz ..H~_ a_aee APPENDIX IV FOREIGN-AID RESEARCH PROJECTS IN THE NUBA MOUNTAINS: OBJECTIVES AND RELATION TO SMALLHOLDER AND NMAPC AGRICULTURE APPENDIX IV FOREIGN-AID RESEARCH PROJECTS IN THE NUBA MOUNTAINS: OBJECTIVES AND RELATION TO SMALLHOLDER AND NMAPC AGRICULTURE In the 1970's a number of foreign-aid research projects have been initiated in the Nuba Mountains area. Among these, the three major projects targeted at the smallholder farming and/or NMAPC schemes are: the European Development Fund (EDF) Project, the West German Technical Aid (GTZ) Project, and the British Overseas Development Administration (ODA) Project. The following is a brief description of their objectives and contemplated activities. EDF: Nuba Mountains Rural Development Project The project was prepared and is being sponsored and implemented by the French Technical Consultants (SATEC). This project is conceived as an agricultural extension and rural development program designed to im- prove agricultural production with special emphasis on the introduction of animal-drawn farming equipment. As held by SATEC, the introduction of animal-drawn equipment and techniques has the following main advan- tages: (1) reduces the demand for, and increases the returns to labor (2) creates more rapid agricultural operations; (3) enables farmers to plant on or Close to the optimum planting dates; (4) saves the farmer money because animal-drawn equipment is cheap compared with tractors; and (5) gives the farmer full control over the timing of farming operations. 254 255 After the testing anddemonstration phases of the project implemen- tation, the animal-drawn farming techniques are to be introduced into both NMAPC and traditional agriculture. The proposed format is as follows: 1. Five development schemes are to be established in connection with the NMAPC schemes (one at Kadugli and four in other field stations). An initial cultivation size of 16 feddans per farmer is proposed. As farmers get more experience and skill with animal-drawn techniques, the cultivation size is to be gradually Rincreased up to 26 feddans per farmer. 2. For the traditional cultivation areas, the aim is to establish two development units in areas of light soils. An 18 feddans cultivation size per farmer is proposed to be managed in a three-course rotation (dura, sesame, and groundnuts). At present, testing of equipment and techniques is the main activity being undertaken. Preliminary test reports have identified a number of technical and equipment problems. The 1979 testing program report [42] concluded with respect to land preparation that: It will require a few years of trials on the testing farm to work out a cost/benefit comparison between ploughing costs and yield-cum-income increases. For the immediate future, thus, ploughing could only be recommended to rich farmers who can bear the risk of unconfirmed test [42, p. iii]. The 1979 weeding test results indicated technical problems especially on the clay and "gardened" soils. Weeding in light soils, however, was re- ported to pose no problems. 256 GT2: Pilot Project and Nuba Mountains Region Masterplan'for Rural Development The GTZ project was prepared and is being implemented by the West German engineering consultants Agrar and Hydrotechnik Gmbh (AHT). The project proposes a comprehensive rural development program. A pilot pro- ject for developing smallholder agriculture is to be linked and supported by a masterplan aiming at improvements of key sub-sectors (animal husband- ry, agro-industries, infrastructure, etc.) in the region. The objectives and assumptions of the pilot project, as identified by the GTZ main re- port [43] are: The proposals, designed to overcome the present deficiencies in the agricultural systems in the Masterplan Area, have to aim at an increase of agricultural production in general as well as at a higher income per farming family. The increase of production and income can be achieved through: - increasing the size of farm holdings. - increasing the yields per feddan through using improved seeds and crop maintenance. - increasing both the above through the in- troduction of capital-intensive methods (mechanized soil preparation and sowing), [43, p. 65]. Proposed specific measures and detailed strategy of the GTZ project include: 1. Expansion of rainfed mechanized crop production on individual smallholder farms, with some of these farms also including an integrated livestock component. 2. Improvement of crop yields through crop diversification, and the introduction of high yielding, drought-resistant,varieties of crops. 3. Introduction of a legume into the crop rotation to serve both as a fodder crop and in maintenance of soil fertility. 10. 11. 12. 13. 257 Creation of an agricultural machinery pool. Establishment of range management practices. Development of agricultural research programs and on-farm trials to test proposed changes in the existing farming activi- ties and cultivation practices. Organization of seed multiplication, inspection, certification and distribution. Encouragement of farmers' agricultural cooperatives. Increased effectiveness of the present agricultural extension serviCes. Development of an efficient marketing system for all agricultural products, including adequate storage facilities. Improvement of transport and communication facilities. Improvement of the supply and quality of both surface and ground- water to secure all-year-round provision of drinking water. Establishment of better education and health facilities. At present, however, only limited machinery testing is being carried out under the project. The proposed institutional and organizational set- up is similar to that of NMAPC, a point considered by AHT to be useful when actual implementation of the project is undertaken: It would be conceivable to make the implementation of both the Masterplan and the Pilot Project the responsibility of the Nuba Mountains Agricultural Corporation,....Basically this program could be interpreted as an especially intensive NMAPC scheme, and it is probable that farmers also see it as such, [43, p. 99]. 258 ODA: Mechanization and Rural Planning Unit Projects Since its conception, the British ODA program has evolved into two separate projects:] an agricultural mechanization project for the NMAPC and a rural planting unit project. The Mechanization Project for NMAPC Originally, all components of the ODA's project were directed towards the NMAPC, which was then expected to assume the responsibility of an overall development authority for South Kordofan Province. Later, as NMAPC did not assume this function, ODA decided to focus its program for the NMAPC by strengthening NMAPC mechanization activities. Main com- ponents of this program are: 1. Training of tractor operators and mechanics. 2. Strengthening or workshop facilities. This includes provision of specialized equipment to the major workshop at Kadugli, and three mobile workshops to aid maintenance at field schemes. 3. Testing of alternative mechanized cultivation techniques. 4. Evaluating a selected range of appropriate new machines and implements. This component of the ODA project was delayed and is expected to start in the season 1980/81. Rural Planning Unit (RPU) Project As indicated above, this project was originally intended to be im- plemented within NMAPC, with the aim of aiding the Government of Sudan in 1A third component (an oil milling study) was part of the initial ODA plans but was subsequently abandoned [38]. 259 planning, monitoring, and promoting agricultural development in South Kordofan Province. At present, the project is operating as a separate entity and is being implemented by the British consultants Hunting Tech- nical Services (HTS). The aim of the project is to establish a Rural ' Planning Unit, and during the first three years the principal objectives of the RPU, as indicated in HTS [38] would be to: a) Train Sudanese counterpart staff to fill the specialist posts in the unit and leave behind an effective rural planning capability. b) Collect and collate all available agricultural, human and physical resource data and to establish an agricultural data bank for the area. Gaps in data essential for planning would be identified and either surveys mounted from within the RUP resources or recommendations made for specific studies, to make good these deficiencies. c) Prepare an agricultural development plan for South Kordofan [38, p. 2] This component of the ODA project is currently underway. Completion of its first phase has resulted in drafting an indicative development plan [38] for South Kordofan Province central districts. The second and third phases are expected to be completed by 1983. BIBLIOGRAPHY N BIBLIOGRAPHY Adam, F. H. and Khidir. "Development of Small Scale Agriculture," Economic and Social Research Council Preparatory Conference, I.L.O. Comprehensive Employment Strategy Mission-Sudan Khartoum February 1975. . Adams, 0. W. and Graham, 0. H. "A Critique of Traditional Agricultural Credit Projects and Policies,“ Economics and Sociology»Occasion- al Paper #621, Agricultural Finance Program, Department of Ag- ricultural Economics and Rural Sociology, 0.S.U., June 1980. 3. Adams, M. E. and Howel, J. "Developing the Traditional Sector in 10. the Sudan," Economic Development and Cultural Change, Volume 27, No. 3, April 1979.‘ Affan,Kh 0. "Output, Employment, and Income Distribution in Mech- anized Farms," A Report on the Findings of a Socio-Economic Survey in Habila, Southern Kordofan. Economic and Social Re- segrch Council, NCR,_Research Report No. 2 Khartoum,7March 9 8. Agricultural Bank of Sudan, "Dilling Branch RecordsJ'Season 1979/80. Agricultural Research Corporation, Kadu 1i Research Station Annual Reports, Sudan, 1968/69, 69/70, 74,75, 78779, 79786. Agricultural Research Corporation, Kennana Research Station Annual Reports, Sudan, 1964/65, 65/66, 69/70, 70/71. Ahmed, A. H. "Lender Behavior and the Recent Performance of Rural Financial Markets in the Sudan,“ Unpublished Ph.D. Dissertation, 0.S.U., 1980. Ahmed, 8. "Farm Management and Agricultural Development: A Case Study of the Pakistan Punjub," Unpublished Ph.D. Dissertation, M.S.U., 1972. 'Ali, 5. M. "Nuba Mountains Cotton Industry: Before and After the Introduction of Modernization," (in Arabic), presented in Agricultural Modernization Seminar, Kadugli, Sudan, 1979. 260 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 261 Att-Konddu, Y. K. “Economic Optima in Resource Allocation for Smallholders Subsistence Farming in Ghana," Unpublished Ph.D. Dissertation, M.S.U., 1974. Berkoff, D. J. W. and Adams, M.E. "Which Way Sudan Agriculture?" Economic Development and Cultural Change, Vol. 28, No. 1, October 1979. Binswanger, H. P. "The Economics of Tractors in the Indian Sub- Continent: An Analytical Review," International Crop Research Institute for Semi-Arid Tropics, India, 1978. Chayanov, A. V. "The Theory of Peasant Economy," Edited by Thorner, D. and Kerblay, B. and Smith, R. E. F. Published for the American Economic Association by Irwin, Inc., Illinois, 1966. Clifton, R. W. (ed.) "Subsistence Agriculture and Economic Develop- ment," Adine Publishing Co., Chicago, 1969. Collinson, M. P. "Micro-Level Accomplishment and Challenges for the Less Developed World," 50th Anniversary, IAAE, 17th Conference, Banff, Canada, September 3:12,Tl979. Collinson, M. P. "Farm Management in Peasant Agriculture: .A Hand- book for Rural Development Planning in Africa," New York, Prager, 1972. Crawford, E. W. "A Programming-Simulation Study of Constraints Af- fecting the Long Run Income-Earning Ability of Traditional Drayland Farming Systems in Northern Nigeria," Unpublished Ph.D. Dissertation, Cornell University, 1980. Dillon, J. and~Anderson, J. "Allocative Efficiency in Traditional Agriculture and Risk," American Journal of Agricultural Econo- mics, Vol. 53, 1979. Eicher,.C. and Witt, L. "Agriculture in Economic Development," McGraw- Hill Inc., 1964. El Hadari, A. E. "Socio-Economic Aspects of Farming in the Nuba Moun- tains--Western Sudan," Department of Rural Economy, Faculty of Agriculture, University of Khartoum, Research Bulletin No. 21, January 1972. El hag, M. S. "NMAPC and the Development of Traditional Agriculture: A Background Review of Cotton Growing and Agricultural Develop- ment," (in Arabic), presented in Agricultural Modernization Seminar, Kadugli, Sudan, 1979 Freidrich, K. H. "Farm Management Data Collection and Analysis Sys- tem," Working Document, FAO, Rome, September 1977. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 262 Gemmill, G. and Eicher, C. K. "A Framework for Research on the Economics of Farm Mechanization in Developing Countries," African Rural Employment Paper No. 6 , M.S.U., East Lansing, 1973. Gotsch, C. H. "Technical Change and the Distribution of Income in Rural Areas," American Journal of Agricultural Economics, May 1972. Hamdoun, A. H. "Chemical Weed Control in Sorghum in the Central ‘ Rainlands of the Sudan," Agricultural Research Corporation 11th Research Colloquium Cereal Crops in Shdan,19779i Harsh, s. 3., Connor, L.TI;and Schwab, G. 0. "Managing the Farm Business," Prentice-Hall, 1981. Harsh, s B and Black, J. R. "Agricultural Econanics Linear Pro- gram Package Version 2, " A. E. Staff Paper No. 75- 10 Depart- ment of Agricultural Economics, M. S. U, April, 1975. Hyden, Goran. "Beyond Ujamaa in Tanzania: Underdevelopment and an Uncaptured Peasantry," London, Heimemann Educational Books Ltd., 1980. ILO/UNDP Employment Mission, "Growth, Employment and Equity: A Comprehensive Strategy for the Sudan," Geneva: International Labor Office, 1976. Khogali, M. "The Nuba Mountains Corporation: Between a Subsidized and a Productive Corporation," (in Arabic), preSented in Agri- cultural Modernization Seminar, Kadugli, Sudan, 1979. Kreniz, R. 0., Dawman, R.Vh and Heady, E. O. "Normative Supply Functions by Linear Programming Procedure," Agricultural Econo- mic Research, 14, No. l, 1962. Kriseney, B."Agricultura1 Technology and Rural Development: A Case Study of Mazabaka District, Zambia," Unpublished Ph. D. Disser- tation, Stanford Univ. , 1977.. Lipton, Michael. "The Theory of the Optimizing Peasant," Journal of Development Studies, 4(3), London, 1968. Lynam, J. K. "Analysis of Population Growth, Technical Change and Risk in Peasant Semi-Arid Systems: A Case Study of Machakos Bistrict§7genya," Unpublished Ph.D. Dissertation, Stanford n1v., . . Massel, B. F. and Johnson, R.W.M. "Economics of Smallholder Farming in Rhodesia-A Cross-Section Analysis of Two Areas," Food Research Institute Studies in Agricultural Economics,‘Tr§de and Develgpe ment, Supplement to Vol. VIII, 1968. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 263 Meigis, P. "World Distribution of Arid and Semi-Arid Homoclimates," Review of Research on Arid Zonepflydrology, UNESCO, Paris, 1953. Ministry of Agriculture, Food and Natural Resources. "South Kordofan, Central Districts Indicative Development Plan," prepared by Hunting Technical Services Limited, England, under assignment by the Ministry of Overseas Development (U.Ka) Sudan, April 1980. Ministry of Agriculture, Food and Natural Resources. "A Report of the Committee for Studying and Treating the Causes of Cotton Production Decline in Nuba Mountains Area,"(in Arabic) 1979. Ministry of Agriculture, Food and Natural Resources. "Medium Scale -and Village Level Crop Mechanization, A Proposal," Mechanized Farming Corporation, IBRD, Sudan, 1978. Ministry of Agriculture, Food and Natural Resources. "Nuba Moun- tains Region, Development Potential Survey, Annex 1,5," pre- pared by AGRAR-UND HYDROTECHNIK. ESSN. FRG , Sudan, 1977. Ministry of National Planning. "Nuba Mountains Rural Development Project, Testing Proaramme of Animal-Drawn Equipment," 1979 Campaign Report,. prepared by SATEC 110 rue de l'Universite, (Paris, France) Sudan, May 1980. Ministry of National Planntpg. "Nuba Mountains Region Masterplan for Rural Development, Main Report, prepared by AGRAR-UND HYDROTECHNIK, ESSEN, FRG, Sudan, 1980. Ministry of National Planning, "The Six-Year Plan for Social and Economic Development, 1977/78-1982-83," Sudan, Vol. II, 1977. Mohamed, A. E. H. "Tillage System Selection: A Case Study Drawn from a Dry Farming Region of the Sudan," Unpublished M.Sc. Thesis, Cranfield Institute of Technology. National College of Agricultural Engineering, U.K., September 1977. N.M.A.P.C. “Discussions and Proceedings," of the Working Agricul- tural Conference, Kadugli, Sudan, June 1980. Norman, D. W. "The Farming System Approach: Relevancy for the Small Farmer," Rural Development Paper No. 5, M.S.U., 1979. Norman, D. W. "Economic Analysis of Agricultural Production and Labor Utilization Among the Hausa in the North of Nigeria," African Rural Employment Paper No. 4, M.S.U., East Lansing, 1973. Norman, D. W. "Methodology and Problems of Farm Management Investi- gations: Experiences from Northern Nigeria," African Rural Employment Paper No. 8, M.S.U., East Lansing, 1973. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 264 Norman, H. N. gt_ol, "Statistical Package for the Social Sciences, Second ed." McGraw-Hill, Inc., 1975. Osman, A. A. "Planning, Agricultural Policies and Policy Implemen- tation in the Republic of the Sudan," Unpublished Ph.D. Disser- tation, M.S.U., 1968. Osman, M. S. and Magid, A. A. "Problems of Mechanization in De- veloping Countries with Special Reference to the Sudan," a paper presented to UNESCO Post Graduate Course on the Teaching of Tropical Agronomy, University of Khartoum, Sudan, 1972. Parhusip, U. "Strategies for Increasing Food Production and Income on Small Upland Farms in Lampung, Indonesia--A Linear Program- ming Analysis," Unpublished Ph.D. Dissertation, M.S.U., 1980. Saeed, M. H. "Mechanization in Habila," Economic and Social Research Council, NCR, Bulletin No. 45, Khartoum, October 1976. Simpson, M. C. and Khalifa, A. H. "A Study of Agricultural Develop- ment in the Central Rainlands of the Sudan, An Interim Report," Economic and Social Research Council, NCR, Occasional Paper No. 5, Khartoum, 1976. Signh, I. "A Note on the Economics of Agricultural Mechanization," Studies in Employment and Rural Development No. 33, IBRD, 1976. Spencer, 0. S. C. "A Micro-Level Farm Management and Production Economics Research Among Traditional African Farmers: Lessons from Sierra Leone," African Rural Employment Paper No. 3, De- partment of Agricultural Economits, M.SIU., East Lansing, 1973. Stickley, T. and Abdalla, M. H. "A Country Study Paper on the Dif- ferent Sources of Credit for Small Farmers in Sudan," Faculty of Agricultural Sciences, A. U. B., Beirut, Lebanon, December 1972. Tothill, T. 0. (ed.) "Agriculture in the Sudan," London Oxford Uni- versity Press, 1948. Woglin, J. “Resource Allocation and Risk: A Case Study on Small- holder Agriculture in Kenya," American Journal of Agricultural Economics, Vol. 57, 1975. Zaki, E. A. "An Ongoing Evaluation of the Planning, Implementation, and Tenancy (Farm) Size of the Rahad Irrigation Project of the Sudan," Unpublished Ph.D. Dissertation, M.S.U., 1979. HICHIGQN S TATE UNIV. LIBRARIES 111111111 1111111111lllllllllllllllllllll111 312 3104769124 | 9