. "I I ‘Iwi‘ ’,;Il'4 UNHIILII 2 " FOOD SECURITY AND COMPARATIVE ADVANTAGEf-{j-g‘:€13 f IN SENEGAL: ‘ AMICRO-MACRQ APPROACH I I 4' Dissertation for the Degree 0f- Ph D MICHIGAN STATE UNIVERSITY FREDERIC MARTIN 1988 PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due DATE DUE DATE DUE DATE DUE IMICH. STATE JNIV. .A00 0 i 2002 “3104200? JUN ‘3 5323““) 3 1/98 chlRC/DmDuest—p.“ MICH. STATE: UNIV. FOOD SECURITY AND COMPARATIVE ADVANTAGE IN SENEGAL: A MICRO-MACRO APPROACH By Frederic Martin A DISSERTATION Submitted to Michigan State University in partial fquiTTment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultura] Economics 1988 ABSTRACT FOOD SECURITY AND COMPARATIVE ADVANTAGE IN SENEGAL: A MICRO-MACRO APPROACH By Frederic Martin This research analyzes food security and comparative advantage in Senegal, from both a micro and macro perspective. The subject was chosen because of the continuing debate about whether poor countries with important food deficits such as Senegal should follow comparative advantage and import food financed by export earnings, or instead support local food production to increase self-sufficiency. The research objectives were: 1) to develop a conceptual framework for analyzing food security and comparative advantage at the farm and national levels; 2) to conduct an empirical analysis of: (a) Senegal’s current and projected comparative advantage; (b) the costs associated with higher levels of cereals self-sufficiency; and (c) the key variables which stimulate or constrain cereals production. The research methodology recognized the need for consistency between the farmers’ strategies and the government’s food strategy. First, the national food supply/demand situation was analyzed. _Ihen farm-level costs and returns were analyzed by preparing 181 crop budgets, used as the basis for 13 typical farm models. These LP models represented major production regions and incorporated micro food security constraints. Regional cereals supply curves were estimated assuming 20%-100% increases in financial cereals producer prices. These supply curves were incorporated into an agricultural sector model, which used separable programming techniques to identify the combination of local production and trade which minimized the cost of achieving given levels of cereals self-sufficiency. The results showed limited cereals supply response to price increases. Land for rice production is limited, and millet/sorghum remains less profitable than peanuts. Maize output increases substantially, which might create a surplus given the small demand for that cereal. Output increases more significantly when expansion of irrigated area is assumed, but irrigation development and higher producer prices are costly to government and consumers. Raising the rice price alone (e.g., by establishing a protected regional market) has negligible effects on production, but would impose substantial costs on consumers. Aligning producer prices with economic prices would produce benefits from Senegal’s current comparative advantage in peanut production, but may not be a viable long-run strategy. Copyright by FREDERIC MARTIN 1988 ACKNOWLEDGEMENTS To start with, the author wants to thank the Franco-American Commission for Educational Exchange for granting him a Fulbright scholarship, which helped finance the beginning of his Ph.D. program. He also thanks Michigan State University for providing further financing as graduate assistantships and, especially, for sending him overseas as research associate in a research project in Senegal financed by USAID. The years spent on the campus interacting with the faculty, the secretarial staff and fellow students were a great learning experience. The author wants to thank the ”Institut Sénégalais de Recherches Agricoles" for accepting him as part of its research staff. The author is greatly indebted to all the Senegalese and expatriate experts who contributed to this research. The author thanks his major professors, Carl Eicher and James Shaffer, for their very good advice. He also benefited from the technical expertise from Stephen Harsh at a crucial moment of the research. He feels a special debt to his thesis adviser, Eric Crawford, for his interest in the study and his constant availability to discuss research issues and to provide the best help possible to solve them. Finally, the author could not have completed this research without the support and the understanding of his wife, Lucie Cloutier. ii TABLE OF CONTENTS Page LIST OF TABLES .......................... vii LIST OF FIGURES ......................... xii LIST OF ABBREVIATIONS ...................... xiv Chapter 1 - INTRODUCTION ........................ 1 2 - CONCEPTUAL FRAMEWORK AND REVIEW OF LITERATURE ....... 5 2.1 - The concept of food security ............. 5 2.1.1 - The time horizon ................. 6 2.1.2 - The level of aggregation ............. 14 2.1.2.1 - Food security for the government ...... 15 2.1.2.2 - Food security for the farmer ........ 17 2.1.3 - The specification of the consumption level . . . . 20 2.2 - The concept of comparative advantage ., ........ 21 2.2.1 - The theory of comparative advantage ....... 22 2.2.1.1 - The role of production policies ....... 26 2.2.1.2 - The role of marketing policies ....... 29 2.2.1.3 - The role of consumption policies ...... 32 2.2.1.4 - The role of international trade policies . . 32 2.2.2 - The empirical analysis of comparative advantage . 34 2.3 - Other important aspects ............... 42 Chapter Page 2.4 - A theoretical model of food security and comparative advantage ..................... 43 2.4.1 - The farmer model ................. 43 2.4.1.1 - Graphical analysis ............. 43 2.4.1.2 - Mathematical analysis ............ 51 2.4.2 - The policy-maker model .............. 54 3 - ANALYSIS OF THE FOOD SITUATION IN SENEGAL ......... 58 3.1 - Food security .................... 58 3.1.1 - Food security for the Senegalese government . . . 58 3.1.1.1 - The level of cereals self-sufficiency . . . . 59 3.1.1.1.1 - Evolution of the national cereals balance sheet from 1974 to 1985 . . 60 3.1.1.1.2 - The regional cereals balance sheets for 1983-85 ............. 68 3.1.1.1.3 - Projected cereals balance sheets . . . . 81 3.1.1.2 - Other food security concerns for the the Senegalese government ........ 87 3.1.2 - Food security for the Senegalese farmer ..... 93 3.2 - Comparative advantage ................. 94 3.3 - Food habits ...................... 96 3.4 - The budget implications ................ 97 3.5 - The foreign exchange implications ........... 97 4 - ELABORATION OF CROP BUDGETS AND MARGIN ANALYSIS ...... 101 4.1 - Methodology used to elaborate the crop budgets . . . . 102 4.1.1 - Agricultural zones ................ 102 4.1.2 - Technical modules ................ 107 4.1.3 - Agricultural inputs ............... 109 iv Chapter Page 4.1.4 - Crop calendar and labor requirements ....... 112 4.1.5 - Yields and states of nature ........... 112 4.1.6 - Input and output prices ............. 118 4.2 - Crop margin analysis ................. 119 4.2.1 - Margin analysis for the Groundnut Basin ..... 121 4.2.2 - Margin analysis for Eastern Senegal and Casamance ................. 129 4.2.3 - Margin analysis for the Senegal River Basin . . . 140 4.2.4 - Margin analysis at the national level ...... 150 4.2.5 - Summary of findings and policy implications . . . 153 5 - METHODOLOGY FOR THE EMPIRICAL MODELLING EXERCISE ...... 155 5.1 - The stages of the empirical modelling exercise . . . . 156 5.1.1 - The construction of a set of farm models ..... 156 5.1.2 - The aggregation of agricultural supply at the zone level and at the national level ..... 159 5.1.3 - The derivation of agricultural products supply curves .................... 160 5.1.4 - The construction of an agricultural sector model . 161 5.1.5 - Policy analysis and experimental design ..... 163 5.2 - The structure of the farm models ........... 165 5.2.1 - Activities .................... 165 5.2.2 - Objective function coefficients ......... 166 5.2.3 - Constraints ................... 169 5.3 - The structure :f the agricultural sector model . . . . 181 5.3.1 - Activities .................... 181 5.3.2 - Objective function coefficients ......... 182 5.3.3 - Constraints ................... 184 Chapter Page 6 - ANALYSIS OF THE RESULTS OF THE EMPIRICAL MODELLING EXERCISE ........................ 193 6.1 - Farm level analysis .................. 193 6.1.1 - Cereals self-sufficiency at the farm level . . . . 193 6.1.2 - Profitability at the farm level ......... 197 6.1.3 - Most binding constraints ............. 200 6.2 - Analysis of a general cereals prices increase policy . 203 6.2.1 - Implications for the cereals self-sufficiency rate ..................... 204 6.2.2 - Implications for the production of major crops . . 209 6.2.3 - Implications for the level of technology and the quantity of inputs used .......... 213 6.2.4 - Cost implications ................ 216 6.2.5 - Implications for the agricultural trade balance . 218 6.3 - Analysis of other price policies ........... 221 6.3.1 - Analysis of a policy raising only rice and wheat prices ................. 222 6.3.2 - Analysis of an economic prices policy ...... 224 6.4 - Identification of key policy variables affecting the food system performance ............ 231 7 - SUMMARY OF FINDINGS AND POLICY IMPLICATIONS ........ 239 APPENDIX A - Other data tables ..................... 249 B - Crop budget sample for millet/sorghum in the Center of the GrOundnut Basin ...................... 280 BIBLIOGRAPHY ........................... 286 vi Table 10. 11. 12. 13. LIST OF TABLES Page Major characteristics of the micro and macro levels of analysis in this research applied to Senegal ...... 19 Crops and technical modules in the different zones of Senegal ......................... 110 Definition of the categories of rainfall quantity and distribution in Senegal ................. 114 Definition of the states of nature in Senegal ....... 115 Probability of occurrence of the states of nature by zone of rainfed cultivation in Senegal ............ 116 Assumptions about crop yields for each state of nature in Senegal ....................... 117 Margin rankings in the Center of the Groundnut Basin (zone 1) ........................ 122 Margin rankings in the North of the Groundnut Basin (zone 2) ........................ 123 Margin rankings in the Southwest of the Groundnut Basin (zone 10) ........................ 124 Margin rankings in the Southeast of the Groundnut Basin (zone 11) ..................... 125 Margin rankings in the Center of Eastern Senegal (zone 6) ........................ 130 Margin rankings in Upper Casamance (zone 7) with low rainfall ........................ 131 Margin rankings in Upper Casamance (zone 7) with high rainfall ........................ 132 vii Table 14. 15. 16. 17. 18. 19. I 20. 21. 22. Page Margin rankings in Middle Casamance (zone 8) with low rainfall ........................ 133 Margin rankings in Middle Casamance (zone 8) with average and high rainfall ................ 134 Margin rankings in Lower Casamance (zone 9) with low rainfall ........................ 135 Margin rankings in Lower Casamance (zone 9) with average and high rainfall .................... 136 Margin rankings in the Delta and the beginning of the Middle Valley of the Senegal River (zone 3) ....... 141 Margin rankings in the Middle Valley of the Senegal River (zone 4) ..................... 142 Margin rankings in the Upper Valley of the Senegal River and in the North of Eastern Senegal (zone 5) ...... 143 Excerpts of the LP tableau for the farm model in zone 1 concerning minimum cereals self-sufficiency constraints . 172 Excerpts of the LP tableau for the farm model in zone 1 concerning minimum income constraints .......... 177 23. Excerpts of the LP tableau for the agricultural sector 24. 25. 26. 27. 28. model concerning separable programming constraints . . . . 185 Farm size and cereals self-sufficiency rates for the typical farms of Senegal in average and worst rainfall years ..................... 195 Ranking of the zones in Senegal according to the net margin per hectare for their typical farm ........ 198 Most binding constraints in the typical farm models in all the zones of Senegal ................ 201 Levels of cereals self-sufficiency in Senegal according to several levels of increases in the prices of all cereals and assumptions about land use ......... 205 Estimation of the potential for an increase in the cultivated area by zone of rainfed cultivation in Senegal ......................... 208 viii Table Page 29. Production of the major crops in Senegal according to several levels of cereals price increases ........ 210 30. Production of the major crops in Senegal with a 40% increase in the level of cereals prices and several assumptions about land use ............... 214 31. Estimation of the cost implications of alternative levels of cereals self-sufficiency in Senegal ...... 217 32. Agricultural trade balance for Senegal according to several levels of cereals price increases ........ 219 33. Comparison of financial and economic prices for major agricultural products in several zones of Senegal . . . . 225 34. Comparison of net margins per hectare by zone in Senegal using financial and economic prices ....... 227 35. Comparison of the production of major crops in Senegal using financial and economic prices ........... 228 36. Data used for the experimental design procedure ...... 233 37. Variance decomposition table for the cost of meeting the population’s cereals needs minus the value of exports in Senegal ................ -. . . 236 38. Variance decomposition table for the agricultural trade balance deficit in Senegal ............... 237 Tables in appendix A: 39. Evolution of the national supply of cereals for Senegal from 1974 to 1985 .................... 249 40. Evolution of the national supply of cereals for Senegal in percentage by origin from 1974 to 1985 ........ 250 41. Evolution of the national supply of cereals for Senegal in percentage by product from 1974 to 1985 ....... 251 42. Evolution of rainfall and gross cereals production in Senegal frcn 1974 to 1985 ................ 252 .43, Evolution of cereals supply per capita by product for Senegal from 1974 to 1985 ................ 253 ix Table 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 543. 59L Regional distribution of the cereals supply for Senegal in 1983-85 ...................... Regional distribution of the cereals supply for Senegal by origin in 1983-85 ................. Regional distribution of cereals supply for Senegal in percentage by product in 1983-85 ........... Estimation of the consumption of cereals per capita for Senegal in urban and rural areas by region in 1983-85 . Regional cereals balance sheets for Senegal estimated by product in 1983-85 ................. Projection of the cereals balance sheet for Senegal by FAO ........................ Projection of the grain balance sheet for Senegal in the Abt Associates report ................. Projection of the cereals balance sheet for Senegal by the 'Schéma National d’Aménagement du Territoire” . . . Evolution of the world prices of selected agricultural products traded by Senegal from 1970 to 1984 ..... Evolution of yields for major crops in Senegal from 1970 to 1985 ..................... Percentage of self-sufficient farms in cereals in Senegal according to several surveys ......... Trade balance for agricultural products projected from 1985 to 2000 for Senegal by the Abt Associates report . Trade balance of agricultural products projected from 1982 to 1992 for Senegal by the Senegalese government . Estimation of economic prices for fertilizers in Senegal in 1986-87 ...................... Estimation of import parity prices for cereals to be bought by farmers in Senegal in 1986-87 ........ Estimation of import parity prices for millet/sorghum to be sold by farmers in Senegal in 1986-87 ...... Page 254 255 256 . 257 258 259 262 . 265 268 269 271 . 272 . 273 . 275 . 276 Table Page 60. Estimation of import parity prices for maize and rice to be sold by farmers in Senegal in 1986-87 ....... 277 61. Estimation of export parity prices for groundnut shells to be sold by farmers in Senegal in 1986-87 ....... 278 62. Estimation of export parity prices for cotton to be sold by farmers in Senegal in 1986-87 ............ 279 xi Figure 1. Utility function for the profit maximizing farmer ..... 2. Optimum production, consumption and trade for the profit maximizing farmer ................... 3. Utility function and level of production of the food product for the farmer concerned with food security 4. Utility function for the farmer concerned with food security ........................ 5. Optimum production, consumption and trade for the farmer concerned with food security .............. 6. Optimum production, consumption and trade for the country whose government is concerned with income maximization . . 7. Optimum production, consumption and trade for the country whose government is concerned with food security . . . . 8. Evolution of the national supply of cereals in Senegal from 1974 to 1985 ................... 9. Evolution of cereals national production and rainfall in Senegal from 1974 to 1985 ............... 10. Evolution of cereals supply per capita in Senegal from 1974 to 1985 ...................... 11. Evolution of national production of cereals by product in Senegal from 1974 to 1985 .............. 12. Evolution of commercial imports of cereals by product in Senegal from 1974 to 1985 .............. 13. Evolution of food aid in cereals by product in Senegal LIST OF FIGURES from 1974 to 1985 ................... xii Page 45 48 49 50 55 56 64 66 67 Figure 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. Map of administrative regions in Senegal before 1984 . . . . Map of cereal deficits by region in Senegal in 1983-85 . . . Disaggregation of the regional supply of cereals by product in Senegal in 1983-85 ............. Map of millet/sorghum deficits and surpluses by region in Senegal in 1983-85 ................. Map of maize deficits and surpluses by region in Senegal in 1983-85 ...................... Map of rice deficits by region in Senegal in 1983-85 . . . . Map of wheat deficits by region in Senegal in 1983-85 Projection of grain needs and domestic production in Senegal from 1985 to 2000 ............... Population growth under alternative fertility hypotheses in Senegal from 1980 to 2030 ............. Evolution of constant world prices for cereals from 1970 to 1984 (basis - 1980) ................ Evolution of constant world prices for major export crops of Senegal from 1970 to 1984 (basis - 1980) ..... Evolution of cereals yields in Senegal from 1970 to 1985 . Evolution of yields for other agricultural products in Senegal from 1970 to 1985 ............... Projection of the balance of trade for agricultural products in Senegal from 1985 to 2000 ......... Map of all zones considered in Senegal for this research . xiii Page 70 71 75 77 78 79 80 84 86 89 9O .' 91 92 99 . 102 BAME CSS CFAF CILSS FAO IMF ISRA LP OECD SAED SODEFITEX USAID HARDA LIST OF ABBREVIATIONS Bureau d’Analyses Macro-Economiques (part of ISRA). Cereals self-sufficiency. CFA Franc (currency in Senegal). Comité Inter-Etats de Lutte contre la Sécheresse au Sahel. Food and Agriculture Organization of the United Nations. International Monetary Fund. Institut Sénégalais de Recherches Agricoles. Linear programming. Organization for Economic Cooperation and Development. Société d’Aménagement et d’Exploitation des Terres du Delta du Fleuve Sénégal et des Vallées du Fleuve Sénégal et de la Falémé. Société pour le Développement des Fibres Textiles. United States Agency for International Development. West African Rice Development Association. xiv CHAPTER 1 INTRODUCTION The research presented in this thesis is an analysis of food security and comparative advantage in Senegal using both a micro perspective and a macro perspective. This topic was chosen for two reasons. First, from a disciplinary point of view, efforts have been expended in the last fifteen years to incorporate uncertainty in the analysis of comparative advantage. Since much of this work has been Itheoretical in nature, there is a widespread need for more empirical research. This research tries to make a contribution in this area with its attempt to add a food security perspective both at micro and macro levels to the standard comparative advantage analysis. Second, from a policy-making point of view, there has been and still is much debate about the desirable agricultural development strategy for poor countries with important food deficits. Senegal is a good example of such a country with the government and donors relying more on ideological grounds and short term financial and political concerns than on scientific analysis to decide on the desirable strategy that Senegal should follow. This research provides policy-makers in Senegal with a better estimation of the long run costs associated with different levels of self-sufficiency and hence hopes to contribute to a better policy dialogue. This research had the following objectives: V/- To provide a conceptual framework for analyzing the interactions between food security and comparative advantage at the micro level and at the macro level; - To elaborate a methodology for empirical analysis of the same topic; - To illustrate the potential of this methodology by applying it to the case of Senegal, making it possible to: i) determine Senegal’s comparative advantage in agricultural products; ii) estimate the costs associated with different levels of cereals self-sufficiency; iii) provide some insights about the relative importance of key agricultural variables on several food system performance indicators; - To draw some implications for food strategy and food policies in Senegal. The thesis includes seven chapters, including this brief introductory chapter. Chapter 2 presents the conceptual framework developed, the review of literature and a theoretical model of food security and comparative advantage with a micro-macro approach. Chapter 3 analyzes the food situation in Senegal mainly from a macro perspective. The evolution of cereal balances over the period 1974 to 1985 is described first, followed by an analysis of several projections of these balances made until the end of the century. Chapter 4 switches to a micro perspective with the presentation of the crop budgets elaborated for Senegal and an analysis of the financial margins for the major crops. This provides insights on the profitability of different crops from the producer’s perspective under current prices. Chapter 5 presents the methodology used in the empirical modelling exercise. This exercise starts with a micro perspective, but ends with a macro perspective. To incorporate food security considerations at the micro level and look at their interactions with the profit objective, a number of typical farm models were built. These models were run with several output price vectors to derive cereal supply curves. These supply curves served as an interface between the micro farm models and the macro agricultural sector model. The latter model considered the trade-offs between the benefits of comparative advantage and food security. Chapter 6 presents the results of applying this methodology to the case of Senegal. First, some interesting microeconomic results coming from the farm models are presented on the possible cereals self-sufficiency rates, the financial profitability of agriculture in various zones, and the most binding constraints. Second, the implications of three price policies on several performance indicators are analyzed: a general cereals prices increase policy, an increase in rice and wheat prices, and an economic price policy. Finally, some key policy variables affecting food system performance are identified. The concluding chapter 7 includes a summary of findings and an analysis of several implications for food strategy and food policies in Senegal. CHAPTER 2 CONCEPTUAL FRAMEWORK AND REVIEW OF THE LITERATURE This research is concerned with two key concepts: food security and comparative advantage. 2.1-WM Food security is a vague concept which can be defined in several ways. Reutlinger and Knapp (1980) defined it as ”the assurance of a minimally adequate level of food consumption” (p. 1). Alternatively, they defined food insecurity as ”the probability of per capita food consumption falling below a specified level" (p. 2). Another definition given by one of the authors is that food security is "the access to a sufficient and continuous food source at all times" (Reutlinger 1984 p. 2). Valdes (1981) defined food security as "the ability of food deficit countries, or regions or households within these countries, to rneet target consumption levels on a year-to-year basis" (p. 2). Staatz (1984) defined it as "the ability of a country or a region to assure, Ion a.continuous, long—term basis that its total population has access to a timely and reliable supply of food adequate to meet minimum nutritional needs" (p. 2). From these definitions, it is clear that food security involves three‘key dimensions: the time horizon, the level of aggregation and the specification of the consumption level. 2.1.1 - Ihe time horizon Food security can be seen as a short-run or transitory problem whereby some groups of pe0ple cannot satisfy their nutritional requirements on a temporary basis. It can also be looked at as a long-run or chronic problem when some people cannot meet their food needs on a permanent basis (Reutlinger 1984 p. 2) The number and variety of policy instruments that the government can use to reduce food insecurity depend on the time horizon considered. In the short-run, the policy options are limited since many variables such as technology, inputs availability or food needs cannot be changed. Short-term food security policy is therefore mainly concerned with the provision of adequate storage, commercial imports and food aid. In the long run, the range of available policy options expands. For example, the government might want to increase the average leVel of local food production by starting irrigation projects, or limit population growth by encouraging birth control programs. These policies aim at solving a chronic food security problem. Most of the research conducted so far on food security has looked at.short-term food security. There is an extensive literature on «thimal storage policies and more recently on optimal trade policies (see Valdes (1981), Chisolm et al. (1982), Bigman (1985) for a general discussion and McIntire (1981) for a study of the Sahelian case; see Newbery et al. (1981) and Schmitz (1984) on price stabilization stocks). The general conclusion to be drawn from these studies is that keeping national food security stocks is rather expensive compared to trade. However, encouraging stocks at the producer level can contribute significantly to rural food security. Using food stocks to stabilize prices does not seem desirable since there are in theory alternative less expensive policy options such as insurance schemes and trading on futures markets. Price stabilization through the use of food stocks seems hardly feasible, in particular on international markets. There has also been some research on the impact of alternative pricing policies on different performance indicators including production and consumption of important crops, foreign exchange and government budget. Two models have been built to analyze this issue in the case of Senegal: the SONED/SEMA model (SONED/SEMA 1979, 1980, 1981) and Braverman’s model (Braverman et al. 1983). Even though these studies were not made in a food security perspective, they offer insights on the impact of pricing policies on short-run food security and on the suitability of different methodologies used for agricultural ' sector modelling. The SONED/SEMA model was designed as a decision tool to help the government in setting the level of official producer prices for each agricultural campaign and in estimating the consequences of such prices on the public budget. The model contained two separate components: a set of producer models and a public budget model. Senegal was dividedfiinflIszon s and a linear programming model of production wasTBUilt for each zone. The producer had to decide on the area planted in the major crops cultivated according to several possible technical modules. Technical modules were defined as alternative input/output combinations based on different levels of input use for a given crop. Given a set of input and output prices offered by the state, the producer was assumed to maximize profits under a set of resource constraints. The public budget model was an algebraic model which calculated the net effect on the public budget of choosing a particular price vector. Optimal food imports were calculated by subtracting the sum of the regional optimal food outputs from estimated food consumption levels. Unit subsidies or taxes were calculated at each stage of the marketing of each product by calculating the difference between the real cost of production and the set of prices used to run the production models in each zone. Total subsidies and taxes were then calculated by multiplying the unit subsidy by the relevant optimal quantity produced or traded. Using the model involved picking a set of producer prices, running the regional linear programming models using these prices to obtain an optimal pattern of production from the farmer’s point of view, estimating' the impact on the public budget and then starting again with a (jifferent set of prices until one was satisfied with the pattern of production and trade as well as with the effect on the public budget. The SONED/SEMA model presented a number of interesting features: - It tried to address an important issue in the eyes of the government, namely at which level should official prices be fixed, by looking at the probable implications of different sets of prices on production, trade and public finance. - It was disaggregated at the regional level and considered explicitly the producer’s interest. However, this model had a number of limitations: - The allocation of production between official marketing channels, the parallel market and home consumption seems to be exogenous, which restricts the value of the model; - It was concerned exclusively with financial prices and did not consider economic prices that reflected the estimated opportunity costs of a resource or of a product; - Farmers are not simple profit maximizers as was assumed by the model. Their other important objective is food security; - The partition made of Senegal into 15 zones as well as the contents of some technical modules in the producer’s models have been criticized by a number of ISRA researchers as being an inaccurate representation of Senegalese agriculture in the mid-eighties; - To date, the model has not been operationalized and used by the Senegalese government. The other model dealing indirectly with short-term food security issues has been developed by a team of economists from the World Bank headed by Braverman (Braverman et al. 1983). The objective of the model was to assess the effects of agricultural pricing policies in 10 Senegal on the government budget, foreign exchange, real income disaggregated by region, and the production and consumption of important crops. The model was based on the multimarket analysis framework, which simulates the behavior of several key markets and their interactions. This approach has been advocated by Braverman and his colleagues as a good compromise between overly simplistic partial equilibrium models and very complex general equilibrium models. The model for Senegal divided the country in four regions and included the markets for major food and cash crops. Each region was endowed with a fixed supply of land and labor while other agricultural inputs had an elastic supply. Domestic crop production was modelled using translog restricted profit functions as in Lau (1976). The demand side was modelled using an Almost Ideal Demand System as in Deaton et al. (1980). Braverman et al. analyzed the effects of several changes in official prices of major agricultural inputs and outputs on several policy objectives. In particular, they looked at the trade-offs 7 between several policy objectives, e.g., between increased government net revenues and lower food consumption by the poor as a result of an increase in the consumer price of rice. They also incorporated uncertainty in yields and producer prices and showed the well-known ‘trade-off between the level and the variability of producer’s income. Their approach presented a number of advantages: - A good representation of key market interactions; - A quantification of the trade-offs between different policy 11 objectives; - A model based on microeconomic foundations with profit-maximizing producers and utility-maximizing consumers; - The attempt to incorporate uncertainty. This model had also a number of limitations: - A number of its results and hence policy conclusions could have been at least roughly estimated; - The policy contribution of the model was primarily its quantification of important trade-offs between several government objectives. However, the lack of adequate data in Senegal, in particular on the demand side, raises questions about the validity of the parameters used in the rather sophisticated demand and supply functions. Thus, the precise quantification of trade-offs among policy objectives must be interpreted cautiously; - The level of regional and technological aggregation was too high, which led to questionable conclusions. For example, the model predicted that an increase in the producer price of rice would result in a decrease in the production of cotton. This conclusion came partly from the lumping together of Casamance and Eastern Senegal.1 1In fact, the only regions where rice and cotton are both cultivated are the Upper and the Middle Casamance. Cotton is not grown in Lower Casamance and rice is not grown in Eastern Senegal, except as a Inorginal crop. Even in the zones where both crops are grown, the possibilities of substitution between these two crops are very limited. Cotton is grown on the plateau as a rainfed crop while rice is mainly grown on lowland. Cotton cannot be grown on lowland. There 15 some rainfed rice grown on the plateau, but it is usually grown on land recently cleared from trees and bushes. Rainfed rice is a risky crop and the producer is not likely to expand much its production. The ' IUHUJIOg together of all types of land resulted in making the 12 Another assumption, i.e., that the supply of agricultural inputs such as seeds, fertilizers, agricultural equipment is elastic, might be questioned, given the breakdown of the public input procurement system in 1980. Very little has been said about the role of long-run structural changes in supply and demand in increasing food security. Labonne et al. (1978) looked at a small part of this issue for the FAO when they analyzed the input requirements of meeting the Senegalese government’s objectives for rainfed agricultural production in 1980-81 and 1984-85. They built an agricultural sector linear programming model for Senegal including five regions and several technical modules for each major crop. The objective function maximized was the gross revenue over all modules, crops and regions. Resource constraints were defined at the regional and national levels. Production objectives were considered as constraints at the regional level. They found that the government’s objectives were too optimistic, since the model had no feasible solution. This model was useful for estimating the feasibility of certain production objectives and for identifying the binding constraints to production growth. This model also had some limitations: - The level of aggregation was too high. All rainfed cereals were lumped together. There were only five regions which, furthermore, were administrative regions, i.e., heterogeneous zones from a physical, human and agricultural points of view; substitutability between the two crops appear higher than it really is. r‘IIlIIIIl‘ I. (it! It‘ll 5“ l3 - The optimal national pattern of production identified by the model was not necessarily consistent with the optima of the producers in each region; - Uncertainty was not taken into account. I From a more general perspective, this model would not be very useful in analyzing the broader issue of the impact of structural changes in supply and demand on food security in the long run since it does not include irrigated crop activities and import/export activities. Sarris (1983) is apparently the first to have tackled this issue, in his research on Egypt. He started from a general welfare optimization problem under conditions of international price and domestic production uncertainty. To reflect a food security perspective, he set the consumption bundle to be obtained either through domestic production or trade at its current level. Assuming the country’s welfare function was an exponential utility function of the net foreign exchange gains, he derived an objective function in terms of the expected value and the variance of this variable. This objective function was a quadratic function which was rnaximized under the standard resource constraints after dividing Egypt 'in three producing regions. Sarris concluded that the current pattern (If production was suboptimal, and that the current area controls had a high efficiency cost. He showed that the area of different crops should be increased depending on the level of the national aversion to foreign exchange risk. Crops that were socially profitable at low levels of risk 14 aversion (mainly cash crops) became quite unattractive at high levels of risk aversion and gave place to subsistence crops. He also concluded that food self-sufficiency was impossible, although domestic reallocation of resources could yield substantial improvements in food security. Some aspects of Sarris' methodology can be criticized: - Some of his assumptions seem arbitrary, for example his use of net foreign exchange gains to approximate social welfare; - Using actual levels of consumption as nutritional norms for food security seems too rigid and too optimistic; - His refusal to submit his model to an empirical verification through a simulation of the current pattern of production and trade because his model was normative seems questionable and raises doubts about the validity of his conclusion that the current pattern was inefficient; - Even if the model provided the optimal pattern of production and trade from a national point of view, it remains quite uncertain whether such a pattern was also optimal for the producers. In spite of these criticisms, Sarris’ approach to the food security issue through changes in the pattern of production and trade seems a promising one. Our research uses the same approach and focuses on identifying the variables that most affect the cost of reaching food security. 2.1.2 - eve The analysis can be conducted at various levels of aggregation. One'can look at food security at the international, national, regional, 15 village, household and eventually individual levels. Food security for a given group does not necessarily imply that food security is obtained for all subgroups within that group. This raises the issue of income distribution among subgroups, e.g., between urban and rural groups. It implies that a food security analysis should be as disaggregated as possible. Most of the research so far, in particular the optimal storage literature already mentioned, has been conducted at the international and national levels. An important literature connected to the food security issue at the international level is the literature on uncertainty in international trade, although its theoretical orientation makes it less relevant for our policy-oriented research (see Helpman and Razin (1978) and Pomeroy (1979) for a review of this literature). Our research is concerned with food security at two levels: the government level and the farmer level. 2.1.2.1 - Eeed_§eenrity for the government From a macro perspective, at least four aspects of food security are likely to be important to a government in a typical developing country: i) The government does not want to depend on other countries to feed its population. This political objective of national independence favors a high level of food self-sufficiency. No country in the world is totally food self-sufficient, but many try to keep a high level of food self-sufficiency. This objective is a valid one. At the same time, it is useful to m) 16 evaluate the economic costs associated with different levels of food self-sufficiency. The government in a typical developing country must take into account the importance of agriculture as a source of employment and of income. Agricultural activities play a critical role in food security through the production of food crops which are mainly consumed on the farm and through cash crops that generate rural income used partly to buy extra food. Even if agriculture is not very competitive on the world market, the government may foster its growth to assure the food security of a large part of the population. The government may implement a regional land management policy. Even if cultivation is not economically efficient in a region, it may feel that investment in agriculture in that region is justified by the social and political necessity of assuring food security in that part of the country, as well as contributing to national food security. The government may want to minimize fluctuations in the cost of the food bill. These fluctuations are caused by price variations and quantity variations. World prices of agricultural commodities fluctuate widely over time. Domestic prices of nontraded or little-traded agricultural products fluctuate also within a given year as a result of supply variations. Regarding quantities, the main source of uncertainty comes from domestic crop yields that fluctuate from year to year mainly as a result of variations in rainfall quantity and distribution. Lerll’III IN,[ ii!‘ I I' I 17 In reducing the impact of this uncertainty, the government faces a dilemma in terms of food policy: variations in world prices and in yields of export crops argue for a greater level of food self-sufficiency, yet variations in cereals yields argue for a lower level of self-sufficiency. The relative importance of the‘price variations and the yield variations will vary by country and by crop and must be empirically estimated. 2.1.2.2 - Feed seeurity fer the farmer From the farmer’s perspective, food security is likely to involve questions analogous to those at the macro level concerning the appropriate degree of self-sufficiency in production. 1') iii) The farmer will try to produce enough cereals to meet a large share of his family needs. He will therefore plant each year a sizeable part of his land with food crops. _ The farmer also wants a minimum income to be able to finance purchases of complementary food and basic consumption goods. He will therefore also allocate a minimum share of his land to cash crops. This second objective may explain that a poor farmer with limited land might still grow some cash crops even though this means reduced acreage for food crops. The farmer may also want to minimize the fluctuations in the quantity of food produced and in his income. The implications of an income stabilization objective for the desired food self-sufficiency level are not clear. An often-heard argument is that the farmer would favor cash crops if their prices were effectively guaranteed by the state, as is often the government’s 18 desire in developing countries. The farmer would not cultivate cereals for sale because their price on the market fluctuates too much. This argument is debatable. In fact, in an uncertain climatic environment in which yields fluctuate from year to year, maintaining a fixed price directly transforms yield variations into income variations. In contrast, fluctuating prices contribute to stabilizing income by counterbalancing yield fluctuations. Hence, given our current knowledge about farmer behavior, it is not clear whether the farmer will favor industrial crops or cereal crops to minimize his income variations. Our research is disaggregated at the regional level. ~Thirteen 'ID zones in_§enegal were identified according to physical, human_end } aerjeelterel criteria. Of these thirteen zones, eleven are I agricultural, one is for livestock grazing and one is urban. In each agricultural zone, a typical farm is modelled, including household food security objectives. Then an agricultural sector model isflcenstructed, . -. -flH'-‘ *WH.’ ‘- “9—,... incorporating the results of the farm models, and reflecting national food security considerations. ”I’There are several important advantages to this multilevel approach: - .It reconciles the micro and the macro levels of analysis. It is based on the recognition of two key economic agents with different perspectives: the farmer and the government. The key characteristics oi= each level of analysis in our research are presented in Table 1; L/ 19 TABLE 1 MAJOR CHARACTERISTICS OF THE MICRO AND MACRO LEVELS OF ANALYSIS IN THIS RESEARCH APPLIED TO SENEGAL Level of Economic Objective Exogenous Endogenous Uncertain analysis agent function variables variables variables Micro Farmer Maximize .Producer .Hectarage .Domestic profits prices for each crop prices of under food .Technologies .Food purchases millet, security available .Food crop and maize, constraints .Input avai- cash crop cowpeas and resource lability sales .Farm constraints .Farm yields .Choice of yields technology Macro Policy- Minimize .World prices .Producer .World maker cost of .Short-run prices prices satisfying regional .Input imports .Regional population supply curves.Food imports yields food needs .Level of food minus the self-suffi- value of ciency desired exports NB: The proposed classification of variables as exogenous, endogenous or uncertain is only valid for the modelling exercise conducted as part of this research. 20 - The approach adopted provides good micro-foundations and a national optimum that is consistent with the producers’ optima; - Coefficients in the farm models are based on the technical expertise of many ISRA researchers, thus increasing the quality of the data; - The model can be validated at the regional level by comparing the regional model’s optimum pattern of production based on past prices with the observed pattern of production; - The implications of a given policy can be examined for each region. 2.1.3 - W In the food security literature, the consumption level used as an acceptable nutrition standard has either been the observed average consumption level or a norm defined in terms of nutritional elements. A widely used norm in the literature is the FAO norm defined in terms of calories per capita. This individual norm varies depending on the age, sex and physical activity of the person (FAO 1973 and WHO 1985). FAO calculates an average individual norm for each country by weighting the individual norms by the relative importance of each category of population disaggregated by age, sex and physical activity. This national norm is 2380 calories for Senegal (FAO 1977). There has been some discussion in the literature about the adequacy of defining a nutritional norm exclusively in terms of caloric needs, i.e., disregarding protein and vitamin needs. The consensus now is that a cereal diet which is adequate in terms of calories will probably also be adequate in general in terms of proteins. However, tiris diet might imply deficiencies in terms of vitamin A, iodine and iron (Sukhatme 1970 and Reutlinger et al. 1976). 21 Since we are interested in the most important components of an adequate diet for the Senegalese population, the FAO caloric norm for Senegal is adopted. Moreover, the consumption of each product is constrained to remain in an acceptable bracket to reflect food habits. A programming approach seems appropriate for our research, both at micro and macro levels, since we are analyzing how the farmer and the government can allocate scarce resources in the most cost-effective way to achieve food security goals, in particular this nutritional norm. This research focuses on the supply side of the food security issue, namely on the provision of a nUtritionally adequate quantity of food at the minimum cost. It does not examine whether the real income of urban consumers is sufficient for them to afford a nutritionally adequate quantity of food. However, our research does consider the demandfisideflof the food security issue at the farm level by setting appropriate constraints in the farm models: i) Farm households must satisfy their total food needs either by producing their food or by purchasing it; . ii) Farm households must satisfy a given and significant percentage of their food needs through production of food crops. 2.2 - [he eeneeet of eomeeretive adventage The second key concept of this research is comparative advantage. Food security is not necessarily synonymous with a high level of food self-sufficiency. Food security can be achieved through a combination of local production and trade. Pearson et al. (1986) distinguish three alternative strategies to satisfy food security: self-sufficiency, self-reliance and import dependency. 22 Akself-sufficiency strategy favors a level of domestic food production far above the comparative advantage level. The self-reliance strategy favors a mix of production and trade that will ‘\ “E ,‘W-Hu fl correspond roughly to the comparative advantage of the country. The importdependency.strategy favors the import of food at concessional terms to benefit urban consumers. This research attempts to identify the self-reliance strategy for Senegal, namely the pattern of production and trade that minimizes the cost of meeting the population’s food needs minus the value of exports. It also tries to estimate the opportunity cost of a self-sufficiency strategy, which would involve greater emphasis on local cereals production. 2.2.1 - Ine theorv of coneenetjye_eeyenteee The concept of comparative advantage is quite old since it dates back to Ricardo (1821) and Torrens (1815 and 1958). Focusing on technology, Ricardian theory explains patterns of specialization and trade by means of relative productivity differentials among countries (Schydlowsky 1984). A distinction is made between absolute advantage and comparative advantage. I If Senegal can produce good X at a lower cost than the rest of the world, it has an absolute advantage in the production of X. In contrast, comparative advantage involves comparing the opportunity cost of resources used in the production of several goods. Even if Senegal has an absolute disadvantage in producing X and Y, this country will be better off producing the good in which it is the more efficient, i.e., the good with the lower opportunity cost for the domestic resources 23 used, and trading that for the other good. The modern Heckscher-Ohlin theory explains international trade by means of differences in-relative factor abundance. It assumes identical linear homogeneous technologies in all countries. There are two versions of the Heckscher-Ohlin theory. The "weak" version states that a country will export the good whose production is relatively intensive in the relatively abundant factor of production of the country according to the value definition of abundance. Relative factor abundance is defined then in terms of pretrade relative factor prices. If the wage/rental ratio in autarky is higher in the foreign country than in the home country, the foreign country is said to be relatively capital abundant (Ohlin 1933). According to the "strong” version of the Heckscher-Ohlin theory, if both countries have identical technologies and identical homothetic preferences, a country will export the good whose production is relatively intensive in the factor of production in which that country is relatively abundant under the quantity definition of relative abundance. Relative factor abundance is then defined in terms of relative factor endowments. If the endowed capital/labor ratio is higher in the home country than in the foreign country, then the home country is capital abundant (Jones 1956-57). Both the Ricardian and the Heckscher-Ohlin theories of comparative advantage are macro-oriented, deterministic and static. The research on comparative advantage so far has been macro-oriented without explicit consideration of the individual strategies of the microeconomic agents. The main originality of this research is to 24 consider both a micro and a macro perspective and to make sure that the two are consistent. Farmers are the ones who make the essential decision of allocating land to different crops. This is a major determinant of the production pattern. The state can affect this decision indirectly by modifying the environment of the producer, particularly in Senegal by setting a number of input and product prices. This research considers the objective functions and the constraints of typical farmers and of the Senegalese government. The consistency between the two levels is achieved by the derivation of supply curves for agricultural products. There have been several efforts to incorporate uncertainty into the theory of international trade (see Helpman and Razin (1978) and Pomeroy (1979) for a review of this literature). The most relevant contribution in this domain for our research is Jabara’s thesis (1979) on the comparative advantage of Senegal under international price uncertainty. This study is reviewed later in this chapter. Uncertainty pervades the food system and more generally the whole economy in a developing country like Senegal. Therefore, it seems essential to incorporate it in an analysis of comparative advantage. The food security perspective adopted by our research reflects this concern. Few studies appear to have been done on comparative advantage in a dynamic context ItliiIPPPV'tEDPID...be..$.193.!‘_-ebnu1.tbe_nse.of-.thgWOY‘d ,anamice Comparative advantage is dynamic in the sense that it changes over time. The methodology used to study dynamic comparative advantage \v 25 can be comparative statics, i.e., comparing a situation at two points of time. Or the methodology can be dynamic; i.e., looking at the process of change. Most studies fall in the first category, namely a comparative statics study of dynamic comparative advantage. Several have looked ex post at the changing nature of comparative advantage in Japan and the newly industrialized countries, in particular the "gang of 4": South Korea, Taiwan, Singapore and Hong-Kong (Balassa 1977a, 1977b, Heller 1976, Whee Rhee et al. 1984). The only research that presents a dynamic theory of comparative advantage appears to be a study by Klein (1973). Klein incorporates a learning factor of production in the traditional production function. This learning factor consists of scientists that learn about and improve the product’s production process. Some countries such as the United States have a "learning" advantage while other countries have a "static” advantage, i.e., a comparative advantage with regard to the other factors of the production function. The total comparative advantage in the production of a particular product shifts from the first category of countries to the other category when the production process for this product is completely known. This research is a comparative statics study of dynamic comparative advantage. However, it differs from existing studies by its food security perspective and its ex ante approach to dynamic comparative advantage. It does not try to explain past changes in comparative advantage, but attempts to analyze how Senegal’s 26 comparative advantage situation is likely to change as a result of changes in key exogenous and endogenous variables. Comparative advantage will change over time because of changes in exogenous variables which the Senegalese government does not control. The most important exogenous variables are the world prices of .__.-________.___-_ _fi. agricultural inputs and products traded by Senegal as well as the share .LL.“_MH_HM_ ,”,-_mm of groundnuts in the fats and oils world market. However, a basic hypothesis of this research is that comparative advantage is not something given to the government, but rather a situation that can be modified by appropriate policies. Our analysis tries to identify the key variables that determine Senegal’s pattern of production and trade and the cost of reaching food security. This will give insights on the kind of policies that would result in the biggest reductions of the cost of meeting food needs. There are four major policy areas where changes could result in a lower cost of food security: production, marketing, consumption and international trade. The major policies in each area are briefly reviewed hereafter. 2.2.1.1 - r l i There are two possible mechanisms through which production policies can reduce the cost of food security. One is an increase in the productivity of Senegalese agriculture which reduces unit costs.of production, and the other is a reduction in the uncertainty faced by I-O“ the producers which makes specialization possible. The first mechanism, a productivity increase, implies the adoption x/ of more intensive agricultural techniques. This in turn requires a V 27 certain number of conditions to be satisfied including: - The availability of the right quantity of the right inputs at the K nightgtime‘inithe right place. This concerns seeds, fertilizers and agricultural equipment, credit and irrigated land with single or double crop seasons. In part, this involves increasing the availability of inputs, e.g., through state efforts to promote a thriving private input supply system that can satisfy farmer’s needs. The state can also provide inputs with public good characteristics such as certified seeds and irrigated land through public investment. Improved vertical coordination in the input marketing system is also necessary. This is discussed in the marketing policies section; The provision of adequate EXE§Q§122.§§EXISPS to increase the producer’s knowledge about more intensive technical practices. This concerns primarily crops that hold the biggest growth potential in the future: corn, cotton and irrigated crops. To be effective, extension must use a bottom-up approach in which extension agents try to advise farmers at their request; The attractiveness of more intensive practices in the eyes of the producers. Our analysis evaluates attractiveness in both financial if§§9119VEII and economic (national) terms. ..... H One production policy the government can use to improve the attractiveness of the more intensive modules is to support agronomic research on higher yielding varieties and better cultivation practices. The government can also use marketing policies to reduce the cost of inputs and increase the price of products for the 28 producer. This issue is discussed in the marketing policies section. The second mechanism through which production policies can reduce the cost of food security is a reduction of the uncertainty faced by the producers. Reduced uncertainty makes it possible for producers to specialize more in the production of products where they are the most efficient. There are two major uncertain variables for the producer in Senegal: fann;level yields and output prjces;_n - Yield uncertainty can be reduced by state support for agronomic research on drought resistant varieties and for the development of irrigated agriculture in the Senegal River Basin; - Price uncertainty varies depending on the crop. For cash crops and for rice, the official price is generally enforced. The relevant prices for millet/sorghum, maize and cowpeas are the market prices. The uncertainty about the official price comes from the rather arbitrary price-fixing mechanism used and the announcing of these prices after the planting time. Government policies to reduce uncertainty about official prices could involve linking the official price level with the world price level and the domestic cost of production, and announcing the new official price level before the planting time. The uncertainty about the market price comes from the producer’s lack of knowledge at planting time about the price he will receive at harvest time for most food crops. This price depends principally on the levels of output of the crop and its close substitutes, which depend mainly on rainfall. Government policy to reduce this uncertainty could involve 29 announcing a floor price before planting time and stepping into the market as a last-resort buyer at harvest time to guarantee this price. However, there are serious questions about the cost and long run viability of such a price stabilization scheme in Senegal (see Martin (1986) and Ouedraogo and Ndoye (1986a and 1986b) for a review of this issue). 2.2.1.2 - WM There are two mechanisms through which marketing policies can reduce the cost of food security. One is to increase the productivity of the Senegalese agricultural.marketjnghsystem,mthusflreducing.unit costs of marketing. The other is to reduce traders’ uncertainty, which R -. >3...“ - ' In"... M- I'M-W vflimdflr‘l‘, In. ~ ”MJv-fl' would reduce the risk premium incorporated in the marketing margin and foster specialization in agricultural production and marketing. Greater productivity in the marketing system can be pursued by policies in three majorflareas: transportation, storage and processing. Transportation policies could involve: - Improvements in the road infrastructure. The road network of Senegal is good compared to that of neighboring countries. However, repairs of existing roads and construction of new roads in more isolated regions such as Eastern Senegal could reduce the time of transportation and increase the lifetime of trucks. The construction of a bridge over the Gambia river would reduce considerably the transportation costs from Casamance to the northern part of Senegal; .. Public support to a competitive private transportation industry. The private sector is best suited to minimize transportation costs. Its profit-seeking nature makes it very cost-conscious and results in 30 maximum loading of trucks and two-way hauls. The role of the state should be to encourage this sector through the suppression of pan-territorial pricing to reflect true transportation costs, and the promotion of interregional trade to favor market integration. These transportation policies might lead to reduced transportation costs. However, it is hard to predict a priori the consequences of a reduction in transportation costs on the comparative advantage of different regions. If we take for example the case of rice, locally produced rice would become more competitive in Dakar, but imported rice would also become more competitive in rice producing regions. What is clear, however, is that transportation costs could affect significantly the comparative advantage or disadvantage of different regions to produce rice (Stryker 1978). Storage policies could involve: - A partial transfer to the private sector of an extensive and largely underutilized public storage capacity; - A suppression of official price fixity over time to recognize the economic need for the trader to cover his storage costs. This fixed level could be replaced by a large price band; - An extension program to show farmers how to reduce storage losses at their level. Processing policies could involve: ‘-.A transfer of the rice milling industry to the private sector to make it more efficient (Morris 1986); - ‘The creation of a more competitive wheat and groundnut processing irwdustries. Right now, wheat processing is done by two private firms 31 that are given wheat grain quotas in fixed proportions by the state. Groundnut processing is in the hands of two public companies that do not compete with one another. The second mechanism through which marketing policies can reduce the cost of food security is a reduction of traders’ uncertainty. Two kinds of policies can play a useful role here: - Policies that clarify the rights and the responsibilities of the traders, i.e., who has the right to buy what, from whom, when, and where? The establishment of clear and durable trade legislation would reduce considerably the uncertainty of traders and thus the provisions they must make for possible fines, bribes or even seizure of their stocks; - Policies that enhance coordination in the agricultural system. The state has a major role to play in promoting vertical and horizontal coordination through the provision of facilitating services. Examples of such services are: - The collection and dissemination of regional and international market information (in particular on prices and on marketing opportunities); - The promotion of standard grades and measures; - The promotion of contracts between producers and traders or between traders; - The provision of export insurance. Whee-Rhee et al. (1984) showed that the public provision of facilitating services was a key factor in South Korea’s impressive growth record . 32 2.2.1.3 - The nele ef eeneenetien eoljejes There are two major mechanisms through which consumption policies can reduce the cost of food security. One is a change in food habits and/the other is a reduction in population_growth to curb the growth of food needs in the futbre. A change of food habits will result in a lower food bill if the populution reorients its consumption toward cheaper food products. Several policies can be used to modify food habits such as adueutisjng and nutritional education campaigns, and adequate processing and packaging of the advertised products. In any case, food habits are slow to changeand respond primarily to the relative price of each food product. The second mechanism through which consumption policies can reduce the cost of food security is a reduction in population growth. The major policy in that area is the promotion of birth central. This can only be a very long-run endeavour because of the sociological and psychological resistances to birth control in Senegal. Even if accepted and put into effect, birth control would have an impact an the level of food needs only in the very long run because of demographic momentum. _At an even longer time horizon, the reduction of the population will also reduce the size of the labor force. J 2.2.1.4 - WWW Senegal is a small country and a price-taker on all world markets where it is present. It is directly subject to world price fluctuations on the import side and on the export side. Furthermore, this country belongs to the French Franc zone which means it has no 33 control over the value of its currency, the CFA Franc being pegged to the French Franc. This increases even more the fluctuations of world prices expressed in domestic currency. There are nuufutures markets in the products traded by Senegal which could be used to protect this country from uncertainty in world prices (see Gordon and Rausser (1984), Lewis and Fry (1984), Thompson and Bond (1987) for an analysis of this trade policy). However, the Senegalese government can try to reduce world price uncertainty by negotiating long-term contracts with major suppliers and customers. The government also wants to protect the domestic market from too much price variation and has set up a parastatal in charge of absorbing the positive or negative differences between the world price and the domestic price. Too much isolation from world markets can lead to gross inefficiencies and a huge deficit for the parastatal, but too much variation in domestic prices is politically unacceptable and increases uncertainty for all economic agents. To conclude this presentation of policies that can potentially change Senegal’s comparative advantage and reduce the cost of food security, it seems clear that a number of these policies are related and must be consistent with one another to be effective. For example, it does not make sense to increase the productivity of a particular crop and thus probably its production if the marketing system is not ready to market the extra output and if there is no domestic or international demand for it. 34 2.2.2 - iri al f iv dv t A number of methods have been proposed to analyze comparative advantage. They fall into three broad groups: the Hecksher-Ohlin theory tests, the revealed comparative advantage methods and the comparative costs methods. The Hecksher-Ohlin theory has been tested empirically by three different kind of studies (Leamer 1984). The first set of studies consists of factor content studies which take measures of trade and factor intensities and infer from them the factor abundance. The best known study in that category is by Leontieff (1953) who found that American imports were more capital intensive than American exports. He interpreted this result to mean that the United States was scarce in capital compared with labor. This famous Leontieff’s paradox spawned an enormous literature. The second kind of studies consists of cross-commodity studies of trade and input intensities. These studies also measure trade and factor intensities to infer factor abundance in the same way as the first set of studies, but they use regression of net exports on factor input intensities (Baldwin 1971, Branson et al. 1977, Harkness 1978). The third kind of studies consists of cross-country studies of trade and resource endowments which use data on trade and endowments and infer trade intensities. They regress net exports across countries on measures of factor endowments (Leamer 1974, 1984, Chenery and Syrquin 1975). The second group of methods used to analyze comparative advantage consists of the revealed comparative advantage methods. They attempt 35 to determine comparative advantage on the basis of actual trade flows (Balassa 1977b, Vollrath 1985). A number of them use the export share measure of revealed comparative advantage that entails dividing the share of a country’s exports in world trade of a given commodity by the country’s share of the total world trade. A basic problem with these methods is that a number of trade barriers affect trade flows, in particular teniffs, quotas, licences anuuuubsidies. It has been suggested to work with export data, which have fewer distortions than import data (Hillman I980). The third group of methods consists of cempgrative costs methods. Two methods have been used. The first method is a partial equilibrium I approach and uses concepts of net social profitability, resource cost ratio and domestic resource cost. I I V The net social profitability of an activity is the difference 5 between the value of the outputs and the value of the inputs used. I,/’ Both values are calculated using shadow prices expressed in terms of border prices. The_uesource cost ratio of an activity is the ratio of the Value of inputs used over the value added by the activity. \Both values are calculated using shadow prices expressed in terms of border puices. The domestic cost ratio is the same as the resource cost ratio except that the value of inputs (the numerator) is calculated using shadow prices expressed in terms of local currency. An activity is considered efficient, or alternatively a country is considered to have a comparative advantage in an activity, if one of the three conditions listed below is satisfied: \- 36 - Its net social profitability is greater than or equal to zero; - Its resource cost ratio is smaller than or equal to one; - Its resource cost ratio is smaller than or equal to the shadow exchange rate of its currency. This methodology has been applied to the case of rice in Senegal by a Stanford/WARDA project. Tuluy (1978, 1979), Pearson et al. (1981) and Craven (1982) showed that rice production was inefficient in Senegal under the price and production conditions of the mid-seventies. However, some activities such as upland animal traction and improved manual swamp cultivation in the Casamance as well as manual irrigated production in the Senegal river basin would have entailed relatively small social losses. Moreover, some activities had a positive social profitability at the farm level. The authors pointed out one major reason for this inefficiency in rice production, namely the remoteness of the producing areas (Casamance and the Senegal River Basin) from the most important consuming area (Dakar) which adds high transportation costs to the local costs of production. Rice production might still be considered desirable if other objectives are considered such as an increased security in food production and a better regional distribution of income. The Stanford/WARDA study brought important empirical insights about the efficiency of rice production in the Senegal. However, the partial equilibrium methodology used does not provide the policy-maker with a global view of the interactions between major input and output markets. Also the study based its conclusions on data from the 37 mid-seventies. Rice production conditions are changing considerably in the Senegal River Basin with: - The restructuring of SAED (the parastatal involved in rice production and marketing in this region); - The completion of two major dams on the Senegal river which make generalized double cropping technically possible; - The increased experience of Senegalese farmers in rice production. Thus, the study’s conclusions need to be periodically reevaluated. The second comparative costs method is to build a model of the agricultural sector to find the optimal pattern of production and trade. Jabara (1979) built such a model for Senegal using linear programming. Her objective was to look at the impact of the introduction of uncertainty on comparative advantage. Her model incorporated production, processing, domestic and international trade and demand activities for six major products in six regions. Following Duloy et al. (1975), the model’s objective function was to maximize the sum of consumer and producer surpluses to obtain competitive equilibrium prices. Moreover, the objective function included a penalty for trade risk, which was modelled following a MOTAD (Minimization Of Total Absolute Deviation) specification (see Hazell (1971) and Hazell et al. (1974) for a review of this specification). Jabara found that under certainty conditions, Senegal had a comparative advantage in peanut production and a comparative disadvantage in cereals production. With uncertainty in international prices and in domestic export production, the production of cereals increased at the expense of peanuts. With risk in domestic production 38 of import substitutes as well as the above mentioned sources of uncertainty, the pattern of comparative advantage was not clear and depended on the relative weights associated with the various sources of uncertainty. Jabara’s methodology presented a certain number of advantages: It took into account the interdependence of major agricultural input and output markets; It took into account the demand side and calculated competitive equilibrium prices; It tried to incorporate the effect of uncertainty on comparative advantage. However, the model presented a certain number of limitations: Its results were not very conclusive. To conclude that uncertainty associated with international prices and domestic production of export crops favors domestic cereals production seems reasonable enough and might not require the elaboration of an agricultural sector model. Moreover, the design of the model seems to lead inevitably to the result obtained, given the penalty on uncertain trade activities in the objective function. The more interesting and realistic case is when all sources of uncertainty are acknowledged, including uncertainty in the domestic production of import substitutes. In this case, the model did not give a clear answer. Jabara considered production as an uncertain variable, which is true from the government’s perspective, but not completely true from the producer’s perspective. The level of production is a function of the 39 yield (an uncertain variable for the producer) and the area cultivated (a decision variable controlled by the farmer). One needs to distinguish between sources of uncertainty at the producer level and at the government level. Jabara also ignored the uncertainty for the government and for the producer about market prices; - Mo farm survey data were available to build the model. Input/output coefficients were taken from various secondary sources. Available data on the demand side is limited and not very reliable in Senegal. No sensitivity analysis was conducted on the demand parameters, costs of production or international prices. This raises questions about the empirical validity of the optimal pattern of production and trade derived from the model. Finally, since the analysis was based on data from the mid- seventies, it should be updated to take into account changes in the domestic and the international economies during the last ten years. Out of all the methodologies which can be used to measure comparative advantage, the agricultural sector modelling methodology was selected since: - It provides a global view of comparative advantage; - It is suitable for the simulation of different policies. There are some difficulties in estimating comparative advantage empirically using comparative costs methods. The theory assumes an accurate estimation of national costs of production and of the world price for the product under consideration. This is far from easy, however. Rainfed cultivation in many developing countries, especially in 40 Africa, does not use purchased inputs and agricultural equipment intensively. The main factor of production is labor. As a result, the guutfief production under rainfed conditions depends significantly onTTT the value given to labor. I Agricultural wage labor is not extensively used in many countries. This reduces the validity of the agricultural wage as the basis for valuing labor. Another method is to value labor according to the average net margin per man-day of on-farm labor, calculated from crop budgets.1 Still another method is to value labor at its shadow price estimated in a linear programming farm model. Whatever the method adopted, one must be cautious in the analysis of costs of production for rainfed agriculture. The cost of production in irrigated agriculture is also hard to "M estimate. Should the initial investment required to create the irrigated infrastructure as well as the cost of foreign technical assistance be included in the cost of production? A erjeri, it seems reasonable to decide that past investments are sunk costs and, therefore, should not be counted as costs of production. However, future planned investments and recurrent costs on existing irrigated perimeters should be considered as costs. Foreign technical assistance should not be counted if it is a gift to the country or if it is a very long-term loan with very little chance of being ever paid back. However, if the country is paying for some of the cost of living of the foreign experts such as housing, 1After deduction of all other costs than family labor. 41 these expenses should probably be considered as part of the cost of production. Whatever the method adopted, the inclusion or the exclusion of these costs will influence the total cost of production significantly, given their magnitude. Therefore, the cost of irrigated cultivation is also subject to alternative estimates. One must also be cautious when selecting world prices for the anelyeis. First, these prices fluctuate over time and it is not always easy to know which reference level to use. Second, the economic significance of world prices can sometimes be questioned, in particular for cereals. The United States and the European Community compete vigorously in the world cereals market, using direct and indirect subsidies to lure potential buyers. As a result, the export price falls well below the real cost of production in Europe or in the United States. This benefits food deficit countries in the short run since it lowers the opportunity cost of obtaining cereals. However, A comparative advantage analysis should take a longer-run view, asking_ uEEther current_world prices are likely to be maintained over a _ significant period. If this ishngt‘eeutain, then local production will be uure attractive relative to imports. In this regard, the world price of broken rice possesses some special characteristics. Only 4% of the world production of rice is traded in the world market.1 Therefore, the world price of rice is lSiamwalla and Haykin (1983), p.13. 42 set in a residual market that does not necessarily reflect the costs of production in Asian countries. This phenomena is all the more important since Senegal imports 100% broken rice, which is a by-product of paddy rice processing. The world market for broken rice is even narrower than the market for whole-grain rice. ‘23 -cher;_importan_t_asp_ects Three other aspects are important in this research: food habits, the budgetary implications and the foreign exchange implications of alternative food strategies. Food habits constrain food policy Options in the short to the medi um run. In the long run, one can expect food preferences to be open to change. The budgetary implications of alternative food strategies are hard to estimate in detail. However, it is important to try to come up with an estimate of the rough impact of different strategies on the national budget. Most developing countries experience high budget deficits and try to limit their expenses. Food strategies aiming at increasing the level of food self- suf"Triciency through increases of the producer prices of food crops imply corresponding increases in the consumer prices of food and/or SUbs 1dies from the state. It is interesting to estimate empirically the amount of subsidies corresponding to different levels of producer pricEs that would be required to avoid any increase in consumer prices. I\ food strategy oriented toward more food self-sufficiency would a 13° mean more investment. Estimating the amount of investment 43 required to reach a certain level of food self-sufficiency is also a parameter of interest for the decision-maker. Apart from the budget implications, the impact of different food strategies on foreign exchange needs to be considered. A strategy oriented toward more food self-sufficiency will probably result in reduced food imports. However, a higher rate of food self-sufficiency usually means less land allocated to export crops, a smaller production of export crops, and reduced export revenues. Also, higher food production might require the use of foreign exchange-intensive inputs such as chemicals, equipment, and fuel. The question of the net foreign exchange savings of a food self-sufficiency strategy is thus an empirical issue. 2.4 - : 1‘0 - 1 00-7 . fooq - _ ' ,1: orn 11: v' ,o a to- The theoretical model proposed consists of two models: a farmer model at the microeconomic level, and a policy-maker model at the macroeconomic level. Both models have a similar structure. 2.4-1- thfameLmodel 2.4.1.1- W The farmer can produce two crops: one CEEE,EIQP (X) and one food or- \“‘—‘"' crap (V). Leuus make the following definitions: F t Food needs of the farmer’s household (in kilos). N = Average level of food crop production desired by_the farmer which covers a high share of the household’s food needs. ’1 3 Minimum level of food crop production desired by the farmer which covers a minimum share of the household’s food needs. F’ i Ratio of the cash crop price over the food crop price 44 (PX/FY). Px and PY are farm gate prices. c : Consumption index. p : Production index. The farmer can adopt differeut strategies which correspond to different utility functions and to different shapes of indifference curves. Two strategies are considered here: an income maximization strategy and a food security strategy. - me i i i n s The income maximization strategy follows the standard neoclassical model of production and trade under certainty. The farmer’s utility function has two arguments: u - u(Xc,Yc). The utility function is unique and continuous in the space defined by the three dimensions X, Y and u (see Figure 1). The corresponding indifference curves in the [ilane (X,Y) are continuous and monotonic, with the first two derivatives being negative. If the farmer cares only about maximizing his income, his optimal strategy, given the production frontier ab and the price ratio P as drawn on Figure 2, is to specialize in the production of the cash crop X. The farmer produces X3 and Y1. Then he trades X3 - X2 to buy Y3 - Y1, so that he can consume X2 and Y3. The production pairrt (B) and the consumption point (C) can be separated thanks to trade (see Figure 2). Without the possibility of trade, the farmer is at A. He produces and Consumes X1 and Y2. Point C is on a higher indifference curve (“1) than point A (no) and so the farmer is better off when trade 18 Possible. 45 4G FIGURE 1 UTILITY FUNCTION FOR THE PROFIT MAXIMIZING FARMER 46 FIGURE 2 OPTIMUM PRODUCTION, CONSUMPTION AND TRADE FOR THE PROFIT MAXIMIZING FARMER 47 - Watson: The farmer’s utility function has three arguments: w - w(Xc,Yc,Yp). The utility function is different when Yp > N and when Yp s N, which creates a discontinuity at Yp - N. Indifference curves are also discontinuous at that level. For Yp > N, the utility function is as in the income maximization case: w - u(Xc,Yc). For Yp s N, the utility function is w . u(Xc,Yc) * H(Yp) where H(Yp) is a function as drawn in Figure 3. The slope of the indifference curves are much flatter when Yp s N, reflecting the very high value given to food production below N. The indifference curves are asymptotical to the M horizontal line, refTecting the fact that the farmer wants absolutely to produce a rninimum M of product Y (see Figures 4 and 5). The farmer can reach complete food self-sufficiency at E, which 'Iies on the "0 indifference curve (Figure 5). However, with the iruiifference curves as drawn in Figure 5, the highest indifference curve the farmer can reach, while staying on the production possibility curve, is "’1 at point G. It is worth pointing out that the Production solution is invariant to any reasonable range of price ratio P as long as Y SN. p Then the farmer sells X6 - X5 to buy Y4 - N. The farmer’s household consumes at H on indifference curve wz and its food needs are satisfied since Y4 > F. 48 H (Yp) FIGURE 3 UTIUTY FUNCTION AND LEVEL OF PRODUCTION OF THE FOOD PRODUCT FOR THE FARMER CONCERNED WITH FOOD SECURITY 49 HGURE4 UTHHYFUNCNONFORTHEFNHMERCONCBWNEHNWHFOODSECURWY 50 F V N \'\ M 0 D X X5 X6 b FIGURE5 OPTIMUM PRODUCTION, CONSUMPTION AND TRADE FOR THE FARMER CONCERNED WITH FOOD SECURITY 51 2.4-1-2 - WW Let us describe successively the income maximization strategy and the food security strategy. - m m' ' r The formulation presented for this strategy is similar to the one used by Jabara (1979 pp. 23-26) for a national model of production and trade. Let us make the following definitions: X : One cash crop; Y : One food crop; Yp - f (Xp) production possibility curve with f1 < 0 and f11 < O; c : Quantity consumed index; p : Quantity produced index; 5 : Quantity sold index; P : Ratio of prices PX / PY; Then Xc - X - Xs p Yc - Y + P Xs p u - u (XC,YC): Utility function of the farmer "del" means partial derivative. The unknowns are Xp and X5. The first-order conditions for utility maximization are: del u del u del u del Y + del X del Xc del Yc del X (1) " P P ' U1 + U2 f1 52 - 0 del u del u del Xc del u del Yc (2) - + del Xs del Xc del Xs del Yc del Xs - - "1 + u2 P - 0 From (I), we obtain: u _1.-,1 u2 From (2), we obtain: u .1...) u2 II>P=|-f'-1 The marginal revenue of a unit of X is equal to the marginal cost of this unit. This solution corresponds to the graphical optimum at B on Figure 2. -lhe_Eo_o_d_secur_i_t.L§_tLateez The difference with the previous case is the specification of the utility function: w (Xc,Yc,Yp) - u (Xc,Yc) when Yp > N - u (Xc,Yc) * H(Yp) when Yp s N where H(Yp) -- aYpz +pr+c $53 want H’(N) - 0 and H(M) - 0 (see Figure 3) .So - ZaN + b - 0 -> b - ZaN and c-aM2 - bMa-aM2 - 2aNM Then H(Yp) - - aYp= + 2aNYp +aMz - 2aNM 53 Without loss of generality, we can set a . 1. 1 + 2NY + M’ - ZNM Then H(Yp) - - Yp p When Yp > N, we have the same solution as in the income maximization strategy. When Yp s N, W is maximized with respect to Xp, XS and Y . P p' del w del u del u del Yp del H(Yp) del Y (1) - H(Yp) + H(Yp) + del Xp del Xc del Yc del Xp del Yp del X del H(Yp) f . u H(Y ) + u H(Y ) + u 1" z p delYpl = 0 del w del u del Xc del u del Yc (2) ' H(Yp) T H(Yp) del Xs del Xc del Xs del Yc del Xs del H(Y ) del Y + p p u del Yp del Xs del H(Y ) a “1 H(yp) + “2 P H(Yp) + -——--£-(-P) u del Yp . 0 del w (3) - u (-2Yp + 2H) del Yp = O From (3), we obtain: vp - N -> H(Yp) - o Raplacing in (1) and (2), we have del w del w “‘—-—-- 0 and d9] Xp del XS = 0 P U 54 The farmer produces N of Y and f'1(N) of X which corresponds to point G on Figure 5. 2.4.2 - model The eolicy-maker model is the farmer model transposed at the macro -___..—-— level. Therefore, we will only present the graphical model. .- Let us make the following definitions: -dn. ”~- _ .- X : One cash crop Y : One food crop AB : The production possibility curve of the country T : The food needs of the country Pw : Ratio of the cash crop price over the food crop price PXw / PYw° PXw and PYw are border prices. As in the farmer’s case, we will consider two strategies: an income maximization strategy and a food security strategy. - II . . . I' ! ! This is the standard neoclassical model of production and trade under certainty. Given the production frontier AB and the world price ratio Pw as drawn on Figure 6, the decision-maker will transmit Pw to the producers so that the country produces X3 and Y1. Then, X3 - X2 is exported to import Y3 - Y1. The country consumes at point E which puts it on indifference curve U1. This .solution is preferable to autarky (point C on indifference curve U0). - Wm If the decision-makers want to achieve a minimum rate of food sel f-sufficiency as part of a food security strategy, they would like the country to be at G on the production frontier (Figure 7). To reach 55 E Y3 I \\ A Y C U1 2 U0 Y D 1 PW 0 x B T X X1 X2 3 FIGURE 6 OPTIMUM PRODUCTION, CONSUMPTION AND TRADE FOR THE COUNTRY WHOSE GOVERNMENT IS CONCERNED WITH INCOME MAXIMIZATION 56 >X szo B FIGURE 7 OPTIMUM PRODUCTION, CONSUMPTION AND TRADE FOR THE COUNTRY WHOSE GOVERNMENT IS CONCERNED WITH FOOD SECURITY 57 that objective, they have to offer P to the producer rather than P”. P is smaller than Pw in absolute value, i.e., the relative price of the food product must be increased. The country then exports X5 - X5 to import Y4 - M. It consumes X5 and Y4 (point H). The decision-makers are sure there is enough food to meet the food needs of the population since Y4 > T. To review this chapter, a conceptual framework was proposed, based on a micro-macro approach to food security and comparative advantage. The relevant literature was reviewed at the same time. A theoretical model was presented to illustrate the relations between the key variables of the conceptual framework. The next chapters present the empirical part of the research. This starts with an analysis of the food situation in Senegal in Chapter 3. CHAPTER 3 ANALYSIS OF THE FOOD SITUATION IN SENEGAL The food situation in Senegal is analyzed using the five key concepts identified in the previous chapter: food security, comparative advantage, food habits, the budget implications and the foreign exchange implications. In this chapter, only secondary data are used. Their quality is sometimes less than desirable, in particular regarding consumption. Therefore, the general magnitudes of the data presented are more reliable than the specific numeric estimates. 3.1 - W Food security in Senegal is analyzed, first, from the policy- rnaker”s perspective and, secondly, from the farmer’s perspective. 3.1.1 - Eeee_§ecuritv for the Senegaleee gevernnent - As mentioned in the previous chapter, there are four aspects of food security likely to be important to the Senegalese government: the level of food self-sufficiency, the importance of the agricultural sector, the regional land management policy and the variations in the food bill. 58 59 3.1.1.1 - The level ef eereele eelf-eeffieieney The position of the Senegalese government has changed recently regarding this issue. Since well before independence, Senegal has followed a strategy based on specialization in the production and the export of groundnut products, which pay for the import of Asian broken rice to feed its population. This strategy was questioned for the first time in 1977 in the Food Investment Plan published by the Ministry of Rural Development. This plan called for a voluntary policy of substitution of local cereals for imported cereals, but the policy was never put into effect. More serious modification of the specialization strategy occurred in 1984 with the definition of a New Agricultural Policy, in 1985 with the publication of the Seventh Plan of Development, and in 1986 with the elaboration of a Cereals Plan (Ministere du Développement Rural 1986). In these documents, the government set a goal of 80% food self-sufficiency by the year 2000, implying a shift of priority from industrial crops to cereal crops. It is not really clear what is meant by food self-sufficiency. We will interpret it as meaning cereals self-sufficiency, given the importance of cereals in Senegalese consumption.l We will also use data for milled quantities of cereals rather than for unmilled 1According to surveys conducted by the "Organisme de Recherche sur I’Alimentation et la Nutrition en Afrique Noire" (ORANA), in 1978-81, annual per capita cereal consumption made up on average 57.8% of total calories consumed and 50.3% of total protein consumed in urban areas, and 60.3% of the calories and 49.4% of the protein consumed in rural areas (Secretariat Chargé de la Décentralisation 1984 p. 278). 60 quantities to have a consistent measurement across cereals and across sources of supply. In order to estimate the feasibility of the Senegalese government’s new objective, the evolution of the cereals balance sheet from 1974 to 1985 is analyzed, followed by a study of the regional cereals balance sheets for 1983-85. Finally, several projections to the year 2000 of the national cereals balance sheet are presented. 3.1.1.1.1 - volu ion h ' na r al alance eneet from 1974 to 1985 The national cereals balance sheets expressed in tons of milled product are presented first for all cereals combined and second for each cereal. The total cereals supply increases 29% from 845,170 tons in 1974-76 to 1,093,070 in 1983-85 (see Figure 8).1 National cereals production decreases 17% from 513,640 tons in 1974-76 to 428,100 tons in 1983-85. Moreover, the share of national production in the total supply of cereals, i.e., the self-sufficiency level, decreases from 61% in 1974-76 to 39% in 1983-85.2 The absolute value and the share of the national production fluctuate a lot from year to year in response to rainfall (see Figure 9).3 This reflects the predominance of rainfed agriculture in Senegal. Commercial imports of cereals increase 75% from 294,110 tons in 1See also Table 39 in Appendix A. ZSee Table 40 in Appendix A. 3See also Table 42 in Appendix A. 61 IIIIIIIIIIIIIE== \\\\\\\\\\\\\\\\\\\\\\\\\\\V Ilr""‘ L““ \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\V IIII"" L“‘ \\\\\\\\\\\\\\\\\\\\\\\\\\\\V ‘ \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\V IIII"" u“‘ \\\\\\\\\\\\\\\\\\\\\\\\\\\V --': sexexoxexexexexex FIII IIIIHII: ;\\\\\\\\\\\\\\\\\\\\\\\\\\ III-U .\\\\\\\\\\\\ IIIIIIIIIIIIIE=E .\\\\\\\\\\\\\\\\\\\\ 700000 600000 2__J_ru.lc 500000 0 200000 T Source: Table 39 FIGURE 8 EVOLUTION OF THE NATIONAL SUPPLY OF CEREALS IN SENEGAL FROM 1974 TO 1985 62 mam. O... :3 20m”. Aémzwm Z. d.Ea:m gmwo “.0 20.5403. 3 mung". me man... 6958 gw> mm. v9 mm. mm. Fm. cm. on. as. us. mm. mm. vs. I \ . VI .7... .I... u. «1.7% . o/rHaIM «Ha/- oral. All I . won -I. o ._.> + NOE no. wgwog 4.. Q/ 0 oIo/eIoIoiio /¢\o\ . u oo— cmw com omN m gmwofisgmmgx‘z ".0 853% am fine... “850w mass; mm memymzomm mu .2zzoéaxenquz _- mmmw Oh :3 20m... 4<0wzmm Z. Z. meOE mm.vm.mm.mm..m. m¢ om.mh. Qh.~n.mh.mn.vn. ooocom oooocm .ooooov ooooom cocooo i-Ozrn ::.—.J.J|u¢3 67 D MlLLET/SORGHUM Ina: ”WHEAT llllllIIIIIIIIIIIIllllllllll!!Lr\\\\ .' \\\\\\\\\\ .F‘L‘l‘ \\\\\\\\\\\ ‘3‘:- \\\\\\\\ t L‘L‘B \\\\\\\\\ “““\ 400000 o o o o o c c o c o o o o o o o o o o o o o o o c o o o In 0 in o In 0 10 co no N N - '- 2-4.111“: I—Ozcn '74 '75 '76 '77 '78 '79 '80 '81 '82 '83 '84 '85 Source: Table 39 FIGURE 12 EVOLUTION OF CWMERCIAL IMPORTS OF CEREALS BY PRODUCT IN SENEGAL FROM 1974 TO 1985 68 wheat (16%) and rice (14%). The respective shares of the different products fluctuate a lot from year to year (see Figure 13).1 3.1.1.1.2 - The regienel eeneele ealence eheets for 1.9.8.3;85 As for the national cereals balance sheets, the regional cereals balance sheets2 are presented first for all cereals combined and second for each cereal. The cereals balance sheet varies significantly from one region to the other3, even though all regions experience an overall deficit4 (see Figure 15).5 In general, the farther north in the country, the smaller the share of production and conversely the higher the shares of commercial 1See also Tables 39 and 41 in Appendix A. 2The regional cereals balance sheet is obtained by calculating the difference between the estimated cereals consumption and the cereals production in that region. Regional consumption is the sum of both urban and rural consumption. The latter is calculated by multiplying the estimated per capita consumption (see Table 47 in Appendix A) by the corresponding population. 3Senegal is disaggregated into eight regions according to A administrative boundaries: Cap Vert, Casamance, Diourbel, Eastern Senegal, Fleuve, Louga, Sine-Saloum and Thies (see Figure 14). Time series on production, trade and food aid are only available for those administrative regions. Administrative regions were redefined in 1984 as follows: Cap Vert, Ziguinchor and Kolda corresponding to the old Casamance, Diourbel, Tambacounda corresponding to Eastern Senegal, St. Louis corresponding to the Fleuve, Louga, Fatick and Kaolack corresponding to Sine-Saloum, and Thies. 4When net production represents less than 50% of net consumption in a given region, this region is called a large deficit region. When net production represents more than 50% and less than 100% of net consumption, it is called a small deficit region. 5See also Table 48 in Appendix A. 69 m8. O._. :3 20¢“. 4m 38 Z Q< 000“. “.0 zQSd>m 8 2.3 .858 8.x E H32 B 29.3 S 22105952.... I mpmm20#_ mm mm.¢m.nm.~m.pm.om.ah.mh.h~.mh.mh.v~. 5 \ \ .-. \\\ — \‘\ ‘ ““ \\\\\\\\ \\\\\\‘ \\\ \\\\\\\\\\\\ \ \ \ ‘ \ I? \ .- cocoa cooov ooooo oooom oooco— oooowp oooov— oooomw 1—020) 15.—.4.41u¢: vamp memmm .__._. A20 .20wa ES 71 3-89 2. géwzmw z. 20.0mm >m who-“mo 4m mammo “.0 >.En5w .220.me m1... ”.0 ZOF<0mm00m wwmfifimmam oz< who-"mo EDIOCOQEAS m0 n22 : mmDOE 3 223 6225 .832 31.5» .fiEm .332 :83 SoN :28. .38 :25 .332 :38 3.3 E III—II EII .2332 _o no.5 2: .o «coacaou .2 3 came co m 2053. cam (Mao..- 78 3-89 2.1—<0mzww Z. 20.0wm >m mmmqjmmam DZ< mtoimo wN_<_2 ”.0 .22 3 meGI we 2nd... “858 5.62 3.93 693 + + 8.8. 3.8 =2»... B .862 :38 693 _ I '— >9. .9558 _o 3%. 05 .0 2825.3 3. vs 093 cc N 2053. wow ++ 79 3-83 2. géwzmm Z. 20.0mm >m. atomic mo;— “.0 mg a. $50.". 3 page ”858 .228. a a8... 9.. a 88.5.8 .2 z 88 .5 m 22.8. 8m :28. 32.8 .35 l .852 :33 09...... fl >mx 80 3-89 2. 4<0w2wm Z. 20.0wm >m whoimo .2ng ”.0 as). am wIDOI me 033. 3225 .8292 _o 3%. as .o «Sagan 3. vs 33 co w 2053. 3m .352 :32. $54 m E 81 3.1.1.1.3 - Prejected eereel; belenee §heets Projections of the cereals balance sheet depend of course on the assumptions made. Projections of cereals production depend on the assumed growth of cultivated areas and yields. The predicted availability of credit, agricultural inputs (seeds, fertilizers), and agricultural equipment, and the predicted relative prices of inputs and outputs, will also influence the projections. Regarding irrigated cultivation, assumptions about the increase in irrigated area and the 1 play a crucial role. coefficient of cropping intensity On the demand side, assumptions about population growth and per capita cereals demand are critical. In general, the planned evolution of the institutional environment will have an important influence on the future cereals balance sheet, in particular the role of the state and of the private sector in the production and the marketing of agricultural inputs and outputs. The first approach to making projections is to start from estimated future cereal needs and examine what this implies for production taking as a constraint the government’s objective of 80% cereals self-sufficiency by the year 2000. This is the approach used by FAO (1985).2 The advantage of this method is to show clearly the 1The coefficient of cropping intensity measures the extent of double cropping. For example, a coefficient of 1.4 means that 40% of the irrigated zone grows two crops per year, and 60% one crop per year. 2FAO makes projections only until 1995 and therefore uses a cereals self-sufficiency objective of 75% by this date. See Table 49 in Appendix A. 82 infeasibility of the self-sufficiency objective, in light of the heroic assumptions required to achieve it. Apart from very optimistic assumptions1 about the intensification of rainfed agriculture, the most unrealistic assumptions are those made for the development of irrigated agriculture. The FAO estimates that the government’s objective can be reached with an annual increase of 3,800 hectares of new irrigated perimeters, a coefficient of cropping intensity of 1.8, and yields of 6 tons of paddy rice per hectare. At present, however, only about 2800 hectares of new irrigated perimeters are created every year. Moreover, a number of irrigated perimeters are not fully cultivated and must be rehabilitated. Therefore, the real annual increase in irrigated zones is around 2,500 hectares, i.e., below the 3,800 hectare target. The coefficient of cropping intensity on irrigated perimeters is presently a little above 1, i.e., there is very little double cropping. Perimeters that have two cropping seasons per year (for example, rice in the rainy season and tomato in the cold dry season) often do not represent true double cropping since each crop is grown on separate plots. The completion of the Manantali dam in 1987 will make the generalization of double cropping theoretically possible. It remains lRainfed millet/sorghum production must go up 29.5% on average, from 562,770 tons for the period 1974-85 to 729,000 tons in 1995. The production of rainfed maize must go up 292%, from 80,400 tons for the period 1983-85 to 315,000 tons in 1995. 83 to be seen if farmers are ready to make the investment in money and time to double crop if single cropping can satisfy most of their food and cash needs. The realism of a coefficient of cropping intensity of 1.8 thus appears questionable. Finally, the present average rice yield on irrigated perimeters is around 4.7 tons of paddy per hectare. The FAO assumption of an average yield of 6 tons seems a bit optimistic. In short, the assumptions necessary to meet the government’s cereals self-sufficiency objective seem unrealistic. Thus, the 80% self-sufficiency objective itself appears unattainable. Another approach to forecasting the cereals balance sheet is to project production using realistic assumptions and then compare it with projected cereals demand. This approach is adopted by Abt Associates (1984)1 and by the ”Secretariat d’Etat Chargé de la Décentralisation" (1984). Both studies come to the same conclusion: even with optimistic 2, the level of cereals and grain assumptions about production self-sufficiency will probably stay unchanged from 1983-85 to 1995 (see Figure 21 for the Abt Associates report projections). The only apparent hope for improving the level of food self- sufficiency in the very long run is to reduce population growth. The birth control issue is a complex and touchy one. However the following figures suggest the importance of this issue. A study by the research lThe Abt Associates study makes projections for the grain balance sheet, which includes cowpeas. See Table 50 in Appendix A. 2See Tables 50 and 51 for the detailed assumptions. 84 coow 0... momw 20m". ._<0w2ww 2. 20303093 omesOo oz< wowmz 225 ”.0 msgmama «.8 2.3 ”858 «N $50.". \hflullc c \4 \ com CO? 060 can coc— comp oovp coop comp ooom PIOZDUJ_._.m Z<20>> awn. ooom >m Z> mum 2320; En. 533:0 c 835.02: >559. Emmmmm -o. 29.54.10 m 0... 29.—Gama 1.. 2954.10 m 0... 20.5.0309”. no. «N Mme—n. .\\\\.\ H . \ H _\ .. QWOQAUJ 2-44—02 87 3.1.1.2 - chgr food security gonggrns for the Senegalese government - rt n f th a ri r l The government must take into account the importance of agriculture for Senegal as a source of employment and of income. In 1980, 81% of the active population was employed in agriculture (World Bank 1987). Even if Senegalese agriculture turns out to be not very competitive on the world market, the government may choose to foster its growth in the short run as a way to assure the food security of a large part of the population. In the long run, the government may prefer to favor other sectors such as fisheries or tourism as sources of employment if Senegal has a comparative advantage in those sectors. - r 'on l and man em The government may implement a regional land management policy. The importance of such a policy is obvious for the Senegal River Basin. The only hope for this region, which has very low and erratic rainfall, lies in the development of irrigated agriculture. Even if irrigated cultivation is not economically efficient given present conditions in the world cereals market, the government may feel that irrigation development is justified by the social and political necessity of assuring food security in that region, as well as contributing to national food security. However, alternative uses for the funds, which might produce more food per CFAF invested than irrigation projects, must be considered. 88 - he variations jg the food bill As mentioned in the previous chapter, the government may want to minimize fluctuations in the cost of the food bill. These fluctuations are caused by price variations and quantity variations. Which type of variation is more significant in the case of Senegal? If we use the coefficient of variation (CV)1 to compare the variability of world prices and of domestic yields from 1970 to 1984-85, we obtain the following relationships (see Figures 23 to 26)2: - The average CV of world prices denominated in constant CFA Francs (basis 1980)3 is slightly greater than the average CV for national yields4, respectively 0.29 and 0.23. - The CV’s for world prices vary more from product to product than the CV’s for national domestic yields. The range of coefficient values for world prices in constant CFA Francs goes from 0.2 for cotton to 0.4 for rice. The range of coefficient values for national domestic yields goes from 0.19 for cotton to 0.26 for groundnut and paddy rice. - The CV’s for world prices denominated in nominal CFA Francs are in general higher than the CV’s denominated in constant CFA Francs. - The CV’s of domestic yields by product are higher at the regional 1The coefficient of variation is defined as the standard deviation divided by the mean of a time series. 2See Tables 52 and 53 in Appendix A. 3The deflator used is the Senegalese consumer price index. 4Based on official yield statistics. 89 3.23 mm 88 5:: 32 0p 22 .205 95me mo“. 30?. Smog pzfimzoo no 202306 8 $56.“. mm mink ”mosom mm vohmhm. — 960356n.hh.oh.mu.vn.nh.wm. p Pom. c oooom .2nzwaom AH cocoa. #2: -x. 29mm: #513 -o. 5E0 hz c9233.8.3.2.2.3.3.2.2.“?For c ocooo ooocc. 43.2 525,. -u. 882 ooccom 29.50 -I. coocmm zoEmonzuo ezw mm 292. ”850m 2:10:03.ij -I. wN_<_2 -o. moi 59$ -0. mm mmawfi mm mm.vm.mm.~m.pm.om.am.m~.hn.oh.mh.¢h.Mh.~h.Fh.on. O can ccop comp OOON comm ILLIJC I< ._.4 _;._..N E... _.\.__.r._...../__.\ :f Ioov mun: I< ¥__JOUJ 93 northern regions (Louga, Diourbel) than in southern regions (Casamance and Eastern Senegal), where rainfall is less erratic. In summary, this analysis of year-to-year variability, which shows that world prices vary slightly more than national domestic yields, tends to argue slightly in favor of a high rather than a low level of food self-sufficiency in Senegal. 3.1.2 - curit f n farm r The farm household head is responsible for assuring that the food needs of his family are met. This obligation will dictate the farming strategy. The farmer will first plant short-cycle cereals in the home gardens to make sure there is something to eat during the hungry season. He will also try to produce enough cereals to meet a large share of his family needs.1 The cereals balance varies also from farm to farm in any given region. Even in high producing areas such as Sine-Saloum or Casamance, a number of farms do not succeed in achieving self-sufficiency in cereals.2 A study by Goetz (1987) found that the average net sales3 per farm from October 1986 to July 1987 were 154 kilos for millet, sorghum and maize and -128 kilos for rice in Southeastern Sine-Saloum, and ~91 kilos for millet, sorghum and maize and -86 kilos for rice in Middle Casamance. The average coarse grain production per farm in 1986 was respectively 2988 and 1323 kilos in the two regions. 1For example, see Kelly (1986). 2See Table 54 in Appendix A. 3Net sales are equal to gross sales minus purchases. 94 Most farms experiencing a cereals deficit are small. Apart from the constraint on land available for cultivation, the decisions of small farmers to grow industrial crops on part of their land and to sell part of the cereals they produce also contribute to their cereals deficit. This a priori surprising phenomenon can be explained by the necessity for all families, including the poorest ones, to have enough income to buy basic consumer goods such as sugar, tea, and cooking oil, and to satisfy their social and religious obligations. 3.2 - Comparative advantaga As mentioned in Chapter 1, a Stanford/WARDA projectl conducted a study of the comparative advantage of several Western African countries, including Senegal, in the midoseventies. They concluded that Senegal did not have a comparative advantage in rice production under the prevailing technological and price conditions at that time. This conclusion remains valid in the mid-eighties for the main rice consumer market: Dakar. Depending on the assumptions used, the cost of rice produced in the Senegal River Basin and processed and transported to Dakar is estimated to be in a range of 160-250 CFAF per kilo. The CIF price of imported Asian broken rice in Dakar varies between 50 and 100 CFAF per kilo. For groundnuts, however, local costs of production are much more competitive with world prices as will be shown in Chapters 3 and S. ‘ 1For a detailed analysis of this question, see Pearson, Humphreys and Stryker (1981), Jabara (1979), Craven (1982) and Tuluy (1979). These Studies all use data for the mid-1970’s. 95 It therefore makes sense to produce and export groundnuts and to import rice.1 The comparative advantage of Senegal is not fixed, but can be modified as a result of changes in prices and costs. Let us mention a few examples: - The extension of double cropping planned after the completion of the Manantali dam could reduce considerably the costs of production for irrigated cultivation, making locally produced rice more competitive with imported rice. - The decrease in the oil price in 1985-86 led to a decrease in the cost of imported fertilizers. For example, SAED2 sold urea at CFAF 118 per kilo in 1985 and at CFAF 80.5 per kilo in 1986.3 Such changes in fertilizer prices modify significantly the costs of production in Senegal. However they affect costs in other producing countries as well, hence the net impact on comparative advantage is not clear. - The institutional context is changing rapidly with the withdrawal of the state and the progressive transfer to the private sector of agricultural input and output marketing. These reforms will, among \ 1Jabara (1979) showed that under certain conditions local production of rice (hence a move toward self-sufficiency) was more economic when uncertainty was considered than when it was not. ?The SAED is a parastatal responsible for the development of Irrigated cultivation in the Senegal River Basin. 3The 1985 price excludes the CFAF 20 subsidy from USAID. The 1986 Dr‘ice is for purchase on credit, which is the most common mode of Payment. The cash price was CFAF 74.5 in 1986. 96 other things, modify the availability and the cost of inputs, and hence the costs of production in Senegal. , It is hard to evaluate the impact of these reforms ax_anta. This topic would require a separate study. The hope is that economic costs will decrease as a result of greater efficiency in the private sector. This assumes real competition among traders, rather than a situation of oligopoly or monopoly. 3.3 - food habits In Senegal, the food habits issue arises most importantly for rice. National production amounted to 61,500 tons of processed rice in 1983-85, i.e., 14% of the total rice supply (419,120 tons). Optimistic assumptions on the development of irrigated rice cultivation lead to projections for rice production of only 250,000 to 350,000 tons of processed rice in the year 2000. However, projecting current trends in rice consumption gives a demand of 700,000 to 900,000 tons. Quite obviously, the level of rice self-sufficiency is going to remain low (between 28% and 50%).1 The government is very hopeful about the policy of processing local cereals (millet/sorghum and maize) into easy-to-use products which can compete with broken rice. This policy is commendable, but it seems unlikely to have a major impact on food preferences by the year 2000. The preference for rice seems well-established in urban areas, in Particular in Dakar. Rice is presently consumed by the quasi-totality ‘ lSee projections in Tables 49, 50 and 51 in Appendix A. 97 of Dakar households every lunch, and one dinner out of two (Ross 1980). The national dish of Senegal, the 'Tiebou-Dienne", has rice as its main ingredient. Because it is hard to see how food preferences would turn rapidly against rice by the year 2000, there appears to be little hope for a significant reduction of demand for rice. 3.4 - im i i The budget implications of a self-sufficiency strategy are hard to estimate in detail. At this stage of the analysis, however, we can get an idea of the cost of developing the irrigated zone in order to meet the self-sufficiency objective. The 1985 FAO study estimates that 38,000 hectares must be irrigated by 1995 in order to meet the 75% cereals self-sufficiency objective. If we accept a cost of 1.5 million CFAF per new irrigated hectare, we obtain a required investment cost of 57 billion CFA Francs. This amount can be compared to the overall deficit in the national budget of - 55.47 billion CFAF in 1983 (IMF 1987). It is clear that the development of the irrigated zone required to meet the self-sufficiency objective will impose a major burden on the state budget, unless donors take over part of the investment costs. 3.5 - The foreign exchange implications The strategy followed until recently by Senegal, which was based on specializing in the production and exportation of groundnut products and the importation of broken rice, has counted on exports to generate enough foreign exchange to pay for imports. Unfortunately, the latest forecast of the world market prospects for groundnut products is not encouraging. Projections indicate that the size of the world market, 98 the world price, the market share of Senegal, and thus exports of groundnut exports will all stagnate.1 The two major purchasers of Senegalese groundnut products are the European livestock breeders who feed their animals with groundnut meal, and European consumers who use groundnut oil for cooking. The livestock breeders are tending to substitute soya meal for groundnut meal because the former gives leaner and thus higher value carcasses than the latter. Consumers are tending to substitute lighter oils such as sunflower or corn oils for groundnut oil. Projections for cotton exports do not indicate any significant growth either. The world market for cotton is expected to remain stable during the foreseeable future. Moreover, a rapid increase in cotton acreage seems unlikely; because cotton exhausts soils quickly, new crop rotations (cotton, cereals and legumes) would have to be introduced, the adoption of which would probably take time. Therefore, cotton exports are likely to remain around their present level.1 The balance of trade in agricultural products is projected to deteriorate (see Figure 271). Traditional exports will be less and less able to generate the foreign exchange required to buy the quantity of cereals necessary to meet food needs. However, the growth of fish exports and of tourism should help alleviate the foreign exchange constraint. From 1974-77 to 1982-84, the level of coverage of cereals imports by groundnut products exports went from 389% to 161%, the level of coverage of cereal imports by fish product exports from 73% to lSee Tables 55 and 56 in Appendix A. 99 ooom O... mmmp EOE qéwzwm Z. wPODn—Omn. 33530.6( m0”. was”: “.0 H255 ml... “.0 29.—.0me hm wing“. mm 2an ”350m 3890 no. 339:. .0. c < .—C ow CV ON— OZ OLL E * 1.25 Average E * 1.25 > et > E * 0.75 Good E * 0.75 > et where: qt : Quantity of rainfall for year t. Q : Average quantity of rainfall for the n observations. et : Rainfall deviation from the mean for year t. E : Average rainfall deviation for the n years. 115 TABLE 4 DEFINITION OF THE STATES OF NATURE IN SENEGAL Rainfall Rainfall quantity distribution Very low Low Average High Very high Bad State 1 State 2 State 3 State 4 State 5 Average State 6 State 7 State 8 State 9 State 10 Good State 11 State 12 State 13 State 14 State 15 116 TABLE 5 PROBABILITY OF OCCURRENCE OF THE STATES OF NATURE BY ZONE OF RAINFED CULTIVATION IN SENEGAL (in percentage) State of Zone nature 1 2 5 6 7 8 9 IO 11 1 O 11.1 5.6 O O 5.6 8.3 O 6 3 2 2.8 O 2.8 O 2.9 O O 3 O 3 2.8 O 2.8 O 5.9 2.8 2.8 3 O 4 2.8 O 0 2.8 2.9 0 2.8 3 6.3 5 11.1 8.3 11.1 11.1 14.7 8.3 11.1 12.1 12.5 6 25 13.9 8.3 13.9 0 5.6 5 6 15.2 12.5 7 5.6 2.8 8.3 16.7 26.5 13.9 11.1 6.1 6.3 8 13.9 5.6 13.9 16.7 11 8 16.7 19.4 21.2 15.6 9 5.6 8.3 5.6 13.9 5.9 19.4 11.1 3 9.4 10 5.6 13.9 5.6 O 0 0 2.8 9.1 3.1 11 O 5.6 O O 0 O O 0 O 12 5.6 16.7 13.9 2.8 11.8 11.1 5.6 15.2 3.1 13 8.3 8.3 13.9 13.9 17.6 16.7 16.7 9.1 21 9 14 11.1 0 8.3 8.3 O O 2.8 0 3.1 15 O 5.6 O 0 0 O O O O Note: The rainfall data used are for the following cities: Zone 1 : Diourbel Zone 6 : Tambacounda Zone 9 : Ziguinchor Zone 2 : Louga Zone 7 : Vélingara Zone 10: Nioro du Rip Zone 5 : Bakel Zone 8 : Kolda Zone 11: Koungheul Source: "Direction de la Météorologie Nationale" for monthly rainfall data from 1951 to 1986. 117 TABLE 6 ASSUMPTIONS ABOUT CROP YIELDS FOR EACH STATE OF NATURE IN SENEGAL Rainfall Rainfall quantity distribution Very low Low Average High Very high Millet/sorghum Bad MA*O.9 MA*1.1 M0 M0 MA Average MA*1.1 MO BO BO MO Good MO BO*O.9 BO*1.2 BO*1.2 BO Maize Bad MA*0.7 MA MO MA MA*O.7 Average MA*O.9 MO 80 MO MA Good (MA+MO)/2 (MO+BO)/2 BO*1.2 80 MO Rice Bad MA*O.6 MA*O.8 MA MO BO Average MA*O.8 MA MO 80 BO*1.1 Good MA MO BO BO*1.1 BO*1.2 Cowpeas Bad MA (MA+MO)/2 MO (MA+MO)/2 MA Average (MA+MO)/2 MO*1.1 BO MO*1.1 (MA+MO)/2 Good MO*1.1 BO BO*1.2 BO MO*1.1 Groundnuts Bad MA*O.9 MA*1.1 MO (MA+MO)/2 MA*1.1 Average MA*1.1 MO 80 (MO+BO)/2 MO Good MO BO*O.9 BO*1.2 BO (MO+BO)/2 Cotton Bad MA*1.1 MO*O.9 MO (MA+MO)/2 MA*1.1 Average MO*O.9 (MO+BO)/2 BO (MO+BO)/2 MO*O.9 Good MO*1.2 BO*1.1 BO*1.2 BO MO*1.2 NB: MA means estimated yield in a bad year. MO means estimated yield in an average year. 80 means estimated yield in a good year. 118 nature. Only the major factors were taken into account. In irrigated cultivation, the yield and the state of nature are not -p.._._.- correlated, except during the hot- off season. In the rainy season and M the cold off-season, the yield is relatively stable from year to year for a given technical module. However, a number of climatic events can negatively affect the yield during the hot off-season. For irrigated crops, a unique yield is assumed for the rainy season and the cold off-season, and three possible yields are assumed for the hot off-season corresponding to bad, average and good years. For flood recession crops, the yields are relatively stable from year to year; rainfall influences them only marginally. Flood recession craps are treated like irrigated crops during the cold off-season, i.e., a unique yield is specified. 4.1.6 - Ioout and outogt orjoeo Input and output prices used in the budgets are 1986-87 farm-gate prices in each zone. Input prices exclude temporary subsidies on input prices, but include subsidies that have existed for a long time. Valuing labor is difficult, since there is no well-organized agricultural labor market. Labor is valuedtat a cost similarflto the agricultural wage received by temporary agricultural workers who —""‘""--—.. perform certain operations, such as paddy rice threshing. Cash crops, i. e. , groundnut shells andkcotton, and,piCe in the Senegal River Basin, are sold by the farmer at the official producer price set each year by the government. The other rice producing area, Casamance, has a large deficit for that cereal and most of the production is home consumed. Therefore, 119 the opportunity cost of producing rice is the official consumer price 'Vof rice converted into paddy equivalent. Tomatoes grown in the Senegal River Basin can be sold at the official price to the companies that manufacture tomato paste. The farmer can also sometimes sell his production to private traders at a higher price. The official price was used in the budgets. The other agricultural products, i.e., millet/sorghum, maize, cowpeas, cowpea hay and groundnut hay are valued at their market price. This price depends mainly on the supply in any given year, which is in turn largely a function of the state of nature. Therefore, market prices were estimated for these products in each state of nature and each zone, based on data available from market price surveys. 4.2-Walla; \/ Ingorder to identify the most profitable crops and technical modules from the farmer’s perspective, crop margins are calculated on the basis of current input and output prices, i.e., in financial FQEEE; This analysis does not evaluate economic profitability based on the use of shadow prices intended to represent the opportunity cost of inputs and of outputs, nor does it not consider the food security objective of the farmer. Both considerations are analyzed in the modelling exercise presented in the next two chapters. -§5Four margins were calculated for each combination ofcgroo, technicalimodule, and_zone: - The gross margin with labor cost, i.e., gross revenue minus variable costs including labor cost; - The gross margin without labor cost, i.e., gross revenue minus 120 variable costs excluding labor cost; - The net margin with labor cost, i.e., gross revenue minus fixed costs minus variable costs including labor cost; - The net margin without labor cost, i.e., gross revenue minus fixed costs minus variable costs excluding labor cost. These margins were calculated on a per hectare basis as well as 90.. a_pe:fiman:daymbasis: The margins per hectare were ranked for thg_wor§t possible state of nature (state 1) and for the most frequgnt_§tates of nature in each zone. For rainfed cultivation zones, margins were first ranked by technical module to identify the most profitable crops for a given leggltof:intensification. Only modules 1, 2 and 3 were considered in that ranking since they correspond respectively to high, medium and low levels of intensification. A second ranking was made including all crops and all technical modules for a given zone to find the most profitable crops and modules overall in that zone. All five technical modules were included in that ranking. For_1rrigated zones, the ranking by module is not relevant since mostfiirrigated crop budgets were built for only one level of Ii intensification. This ranking was replaced by a ranking by the number of crops per year, i.e., single or double cropping. Apart from the ranking by zone, two rankings were made at the national level. First, margins in the different regions where a given crop is grown were ranked to determine where it made most sense to promote the crop. 121 Second, a ranking was made across all crops and all zones to identify the most profitable agricultural activities in Senegal. Both kinds of national rankings were made for state of nature 8, which corresponds to average rainfall quantity and distribution. Overall, 724 rankings were made._ A summary of the rankings and their policy implications is presented below.1 The rankings by zone are discussed first. The zones have been combined into three regions: the Groundnut Basin, Eastern Senegal and Casamance, and finally the Senegal River Basin. 4.2.1 - Margin analysis for tho Groooonut Bagio The Groundnut Basin combines four zones: the Center (zone 1), the North (zone 2), the Southwest (zone 10) and the Southeast (zone 11). The margin rankings for these zones are presented in Tables 7, 8, 9, and 10 respectively. Three important issues arise in that region: 1) Are cowpeas more profitable than groundnuts? 2) What is the potential for cereals production expansion in that region to contribute to the government’s objective of 80% food self-sufficiency in year 2000? \l 3) Is production intensification financially attractive? The first issue is important for the center and the north of the Groundnut Basin. Faced with declining yields in the North (zone 2), especially for groundnuts, the Senegalese government is promoting the substitution of cowpeas for groundnuts since cowpeas are more resistant 1For a complete presentation of the rankings, see Martin (1988). 122 TABLE 7 MARGIN RANKINGS IN THE CENTER OF THE GROUNDNUT BASIN (ZONE 1) 1 - Rankings for all levels of intensification One ranking was made for each of the three levels of intensification for states of nature 1, 5, 6, 8 and 14. To summarize those rankings, the most often encountered rank for a crop is given to that crop in the table below. GMWL means gross margin with labor cost. GMNL means gross margin with no labor cost. NMWL means net margin with labor cost. NMNL means net margin with no labor cost. Mil./sorg. means millet/sorghum. Rank GMWL GMNL NMWL NMNL 1 Groundnuts Cowpeas Groundnuts Groundnuts 1 Cowpeas 2 Mil./sorg. Groundnuts Mil./sorg. 3 Cowpeas Mil./sorg. Cowpeas Mil./sorg. 2 - Rankings for all technical modules One ranking was made for each state of nature 1, 5, 6, 8 and 14. To summarize those rankings, the average rank for a given crap was calculated and then a ranking of the average ranks was made, which appears in the table below. Rank GMWL GMNL NMWL NMNL 1 Groundnuts Cowpeas Groundnuts Cowpeas 2 Mil./sorg. Groundnuts Mil./sorg. Groundnuts 3 Cowpeas Mil./sorg. Cowpeas Mil./sorg. 123 TABLE 8 MARGIN RANKINGS IN THE NORTH OF THE GROUNDNUT BASIN (ZONE 2) 1 - Rankings for all levels of intensification One ranking was made for each of the three levels of intensification for states of nature 1, 6, 10 and 12. To summarize those rankings, the most often encountered rank for a crop is given to that crop in the table below. GMWL means gross margin with labor cost. GMNL means gross margin with no labor cost. NMWL means net margin with labor cost. NMNL means net margin with no labor cost. Mil./sorg. means millet/sorghum. Rank GMWL GMNL NMWL NMNL 1 Mil./sorg. Cowpeas Mil./sorg. Cowpeas 2 Cowpeas Mil./sorg. Cowpeas Mil./sorg. 2 - Rankings for all technical modules One ranking was made for each state of nature 1, 6, 10 and 12. To summarize those rankings, the average rank for a given crop was calculated over all technical modules and then a ranking of the average ranks was made. Since this table is the same as the above table, it is not repeated. 124 TABLE 9 MARGIN RANKINGS IN THE SOUTHWEST OF THE GROUNDNUT BASIN (ZONE 10) 1 - Rankings for all levels of intensification One ranking was made for each of the three levels of intensification for states of nature 1, 5, 6, 8 and 12. To summarize those rankings, the most often encountered rank for a crop is given to that crop in the table below. GMWL means gross margin with labor cost. GMNL means gross margin with no labor cost. NMWL means net margin with labor cost. NMNL means net margin with no labor cost. Mil./sorg. means millet/sorghum. Rank GMWL GMNL NMWL NMNL 1 Groundnuts Groundnuts Groundnuts Groundnuts 2 Mil./sorg. Mil./sorg. Mil./sorg. Mil./sorg. 2 - Rankings for all technical modules One ranking was made for each state of nature 1, 5, 6, 8 and 12. To summarize those rankings, the average rank for a given crop was calculated over all technical modules and then a ranking of the average ranks was made. Since this table is the same as the above table, it is not repeated. 125 TABLE 10 MARGIN RANKINGS IN THE SOUTHEAST OF THE GROUNDNUT BASIN (ZONE 11) 1 - Rankings for all levels of intensification One ranking was made for each of the three levels of intensification for states of nature 1, 5, 6, 8 and 13. To summarize those rankings, the most often encountered rank for a crop is given to that crop in the table below. GMWL means gross margin with labor cost. GMNL means gross margin with no labor cost. NMWL means net margin with labor cost. NMNL means net margin with no labor cost. Mil./sorg. means millet/sorghum. Rank GMWL GMNL NMWL NMNL 1 Groundnuts Groundnuts Groundnuts Groundnuts 2 Maize Maize Maize Maize 3 Mil./sorg. Mil./sorg. Mil./sorg. Mil./sorg. 2 - Rankings for all technical modules One ranking was made for each state of nature 1, 5, 6, 8 and 13. To summarize those rankings, the average rank for a given crop was calculated over all technical modules and then a ranking of the average ranks was made, which appears in the table below. Rank GMWL GMNL NMWL NMNL 1 Maize Maize Maize Maize 2 Groundnuts Groundnuts Groundnuts Groundnuts 3 Mil./sorg. Mil./sorg. Mil./sorg. Mil./sorg. 126 to drought than groundnuts. .While the decline of groundnut cultivation seems inevitable in zone 2, it is worth looking at the comparative profitability of cowpeas and groundnuts further south in the Center of the Groundnut Basin. One problem with cowpea cultivation is the significant labor requirements, in particular for manual harvest. For example, module 2 in an average year requires 95 man-days for cowpeas compared with 42 for millet/sorghum and 38 for groundnuts. As a result, cowpeas are last after groundnuts and millet/sorghum in the rankings by module according to the margins with labor cost, whatever the intensification level. However, in the rankings according to the margins without labor cost, cowpeas rank higher in general than millet/sorghum and sometimes higher than groundnuts. Cowpea cultivation implies slightly higher fixed costs than those for groundnuts because of the insecticide sprayings, which require a sprayer. Therefore, cowpeas rank better according to the gross margin than according to the net margin. Cowpeas can be considered as a potential alternative to groundnut if the producer can mobilize enough labor at harvest time and if the insecticides required for cowpeas cultivation are available at the right time. This conclusion depends on the assumption made for cowpea price. The government policy pushing producers to raise the area planted in cowpeas will probably result in an increased supply. The capacity of domestic or foreign demand to absorb this extra supply and the evolution of the cowpea price is hard to estimate precisely. Presumably, cowpea prices will fall from their past level. 127 In the past, prices of 250 or 300 CFA Francs per kilo could be found on local markets. The cowpea budget assumes an average price of 100 CFAF, a low of 60 CFAF and a high of 140 CFAF per kilo. These prices were based on the most recent price data, and on experts’ opinions. The second important issue in that zone concerns the development of cereals cultivation, which is a government priority. One important requirement for the expansion of cereal area cultivated is that cereals should be more profitable than other crops, in particular groundnuts. The rankings give ideas about the present relative profitability of the different cereals. The conclusions are different depending on the cereal. In all the zones covering the Groundnut Basin, millet/sorghum is almost always less profitable than groundnuts whatever the type of margin, gross or net, with or without labor cost and whatever the level of intensification. Thus, it is hard to see the financial interest to the farmer of increasing his millet/sorghum production for sale under present price conditions. This does not mean that millet/sorghum is not an interesting crop, in particular from a food security perspective at the farm level. In the rankings with all crops and all modules, module 4 (home garden) for millet/sorghum is often highly placed. This position can be explained by the limited use of inputs by this module and by the good manual care given, which results in high yields. One must remember that this module is not grown for sale, but to contribute to the farm household’s food security. In any case, areas reserved for these home gardens are limited to the village surroundings and could 128 not be expanded significantly. Maize is ranked much higher than millet/sorghum. In the Southeast of the Groundnut Basin, which is the only zone of the Groundnut Basin where it can be grown, maize is more profitable than groundnuts in a low rainfall year and comes just after groundnuts in an average or high rainfall year. This difference in ranking can be explained partly by the difference in the price setting mechanisms for the two crops. Groundnuts are sold at the official producer price which does not vary according to the yield. This means that yield variations are automatically translated into income variations. On the other hand, maize is sold at the market price which fluctuates inversely to supply. If we assume a positive correlation between the typical farm yield and the market supply, the market price of maize will fluctuate inversely to the yield variations, which results in a certain income stabilization. As a result of this difference in price setting mechanisms, the groundnut income falls more than the maize income in a low rainfall year. Conversely, in a high rainfall year, the groundnut income increases more than the maize income. The third issue concerns the financial attractiveness of intensified production. The highest intensification level (module 1) is financially attractive in a good rainfall year. However, in a low rainfall year, the best modules are module 3 for groundnuts and millet/sorghum and module 2 for cowpeas (module 3 does not exist for that crop). These results seem logical since the high input costs of 129 module 1 are worthwhile only in the case of good rainfall, which converts the high level of intensification into high yields. This is another illustration of the classical correlation between the level of profit and the level of risk. The level of intensification chosen by the farmer depends on his risk aversion. Given the important climatic vagaries, the persistence of the drought, and the subsistence orientation of many farmers in that region, especially in the northern part, the level of intensification chosen is likely to be average or even low. The late planting module is in general not very attractive financially, but one must realize that the alternative to that module is not another module, but not to plant at all. If there is a labor bottleneck during the normal planting period, the farmer may well decide to do some late planting with the idea that the resulting yield, however small, will at least cover the seeds used and bring some surplus. 4.2.2 - Margjn analyois for Eaatoco Sooooal ago Caoamaooo Eastern Senegal (zone 6) and Casamance (zones 7, 8 and 9) benefit from better climatic conditions than the Groundnut Basin. This means yields are higher and the range of crops that can be grown is wider in general. The margin rankings are presented in Tables 11 to 17. Three poliey issues arise in those regions: v/1) Are maize and cotton financially attractive? ,, 2) Are the different types of rice cultivation profitable? 3) Is production intensification attractive financially? Maize and cotton are two crops that could potentially be developed 130 TABLE 11 MARGIN RANKINGS IN THE CENTER OF EASTERN SENEGAL (ZONE 6) 1 - Rankings for all levels of intensification One ranking was made for each of the three levels of intensification in states of nature 1, 6, 7, 8, 9 and 13. To summarize those rankings, the most often encountered rank for a crop is given to that crop in the table below. GMWL means gross margin with labor cost. GMNL means gross margin with no labor cost. NMWL means net margin with labor cost. NMNL means net margin with no labor cost. Mil./sorg. means millet/sorghum. Rank GMWL GMNL NMWL NMNL 1 Groundnuts Groundnuts Groundnuts Groundnuts 1 Maize Maize 2 Maize Maize 3 Mil./sorg. Mil./sorg. Mil./sorg. Mil./sorg. 4 Cotton Cotton Cotton Cotton 2 - Rankings for all technical modules A ranking was made for each state of nature 1, 6, 7, 8, 9, and 13. To summarize those rankings, an average rank was calculated for each crop over all technical modules and then a ranking of the average ranks was made, which appears in the table below. Rank GMNL GMNL NMHL NMNL Maize Maize Maize Maize Groundnuts Groundnuts Groundnuts Groundnuts Mil./sorg. Mil./sorg. Mil./sorg. Mil./sorg. Cotton Cotton Cotton Cotton #wNO-i 131 TABLE 12 MARGIN RANKINGS IN UPPER CASAMANCE (ZONE 7) WITH LOW RAINFALL 1 - Rankings for all levels of intensification One ranking was made for each of the three levels of intensification in states of nature 1 and 7. To summarize those rankings, the most often encountered rank for a crop is given to that crop in the table below. GMWL means gross margin with labor cost. GMNL means gross margin with no labor cost. NMWL means net margin with labor cost. NMNL means net margin with no labor cost. Mil./sorg. means millet/sorghum. Low. rice means lowland rice. Rain. rice means rainfed rice. Rank GMWL GMNL NMWL NMNL 1 Groundnuts Groundnuts Groundnuts Groundnuts 2 Maize Maize Maize Maize 2 Mil./sorg. Mil./sorg. 3 Cotton Cotton Mil./sorg. 3 Cotton 4 Mil./sorg. Cotton 5 Low. rice Low. rice Low. rice Low. rice 6 Rain. rice Rain. rice Rain. rice Rain. rice 2 - Rankings for all technical modules A ranking was made for each state of nature 1 and 7. To summarize those rankings, an average rank was calculated for each crop over all technical modules and then a ranking of the average ranks was made, which appears in the table below. Rank GMWL GMNL NMWL NMNL 1 Groundnuts Groundnuts Groundnuts Groundnuts 2 Maize Maize Maize Maize 3 Mil./sorg. Mil./sorg. Mil./sorg. Mil./sorg. 4 Cotton Cotton Cotton Cotton 5 Low. rice Low. rice Low. rice Low. rice 6 Rain. rice Rain. rice Rain. rice Rain. rice 132 TABLE 13 MARGIN RANKINGS IN UPPER CASAMANCE (ZONE 7) WITH HIGH RAINFALL 1 - Rankings for all levels of intensification One ranking was made for each of the three levels of intensification in states of nature 5 and 13. To summarize those rankings, the most often encountered rank for a crop is given to that crop in the table below. GMWL means gross margin with labor cost. GMNL means gross margin with no labor cost. NMWL means net margin with labor cost. NMNL means net margin with no labor cost. Mil./sorg. means millet/sorghum. Low. rice means lowland rice. Rain. rice means rainfed rice. Rank GMWL GMNL NMWL NMNL 1 Rain. rice Low. rice Low. rice Low. rice 1 Low. rice Groundnuts 2 Groundnuts Groundnuts Rain. rice Rain, rice 2 Maize Groundnuts 3 Mil./sorg. Rain. rice Mil./sorg. 3 Maize 4 Cotton Cotton Cotton Maize 4 Mil./sorg. 5 Maize Cotton 6 Mil./sorg. 2 - Rankings for all technical modules A ranking was made for each state of nature 5 and 13. To summarize those rankings, an average rank was calculated for each crop over all technical modules and then a ranking of the average ranks was made, which appears in the table below. Rank GMWL GMNL NMWL NMNL 1 Rain. rice Low. rice Rain. rice Low. rice 2 Low. rice Rain. rice Low. rice Rain. rice 3 Groundnuts Groundnuts Groundnuts Groundnuts 4 Maize Maize Maize Maize 5 Mil./sorg. Cotton Mil./sorg. Mil./sorg. 5 Cotton Mil./sorg. Cotton Cotton 133 TABLE 14 MARGIN RANKINGS IN MIDDLE CASAMANCE (ZONE 8) WITH LOW RAINFALL 1 - Rankings for all levels of intensification One ranking was made for each of the three levels of intensification in states of nature 1 and 7. To summarize those rankings, the most often encountered rank for a crop is given to that crop in the table below. GMWL means gross margin with labor cost. GMNL means gross margin with no labor cost. NMWL means net margin with labor cost. NMNL means net margin with no labor cost. Mil./sorg. means millet/sorghum. Low. rice means lowland rice. Rain. rice means rainfed rice. Rank GMWL GMNL NMWL NMNL 1 Groundnuts Groundnuts Groundnuts Groundnuts 1 Maize 2 Maize Maize Maize 3 Mil./sorg. Mil./sorg. Mil./sorg. Mil./sorg. 4 Cotton Cotton Cotton Cotton 5 Low. rice Low. rice Low. rice Low. rice 5 Rain. rice 6 Rain. rice Rain. rice Rain. rice 2 - Rankings for all technical modules A ranking was made for each state of nature 1 and 7. To summarize those rankings, an average rank was calculated for each crop over all technical modules and then a ranking of the average ranks was made, which appears in the table below. Rank GMWL GMNL NMWL NMNL l Maize Groundnuts Groundnuts Groundnuts 2 Groundnuts Maize Maize Maize 3 Mil./sorg. Mil./sorg. Mil./sorg. Mil./sorg. 4 Cotton Cotton Cotton Cotton 5 Low. rice Low. rice Low. rice Low. rice 5 Rain. rice Rain. rice Rain. rice Rain. rice 134 TABLE 15 MARGIN RANKINGS IN MIDDLE CASAMANCE (ZONE 8) WITH AVERAGE AND HIGH RAINFALL 1 - Rankings for all levels of intensification One ranking was made for each of the three levels of intensification in states of nature 5, 8, 9 and 13. To summarize those rankings, the most often encountered rank for a crop is given to that crop in the table below. GMWL means gross margin with labor cost. GMNL means gross margin with no labor cost. NMWL means net margin with labor cost. NMNL means net margin with no labor cost. Mil./sorg. means millet/sorghum. Low. rice means lowland rice. Rain. rice means rainfed rice. Rank GMWL GMNL NMWL NMNL 1 Low. rice Low. rice Low. rice Low. rice 1 Rain. rice Rain. rice Rain. rice Rain. rice 2 Groundnuts Groundnuts Groundnuts Groundnuts 4 Maize Maize Maize Maize 4 Cotton 5 Mil./sorg. Mil./sorg. Mil./sorg. 6 Cotton Mil./sorg. Cotton Cotton 2 - Rankings for all technical modules A ranking was made for each state of nature 5, 8, 9 and 13. To summarize those rankings, an average rank was calculated for each crop over all technical modules and then a ranking of the average ranks was made, which appears in the table below. Rank GMWL GMNL NMWL NMNL 1 Trans. rice Trans. rice Trans. rice Trans. rice 2 Low. rice Low. rice Low. rice Low. rice 3 Maize Groundnuts-Maize Maize Groundnuts 4 Groundnuts Groundnuts Maize 5 Mil./sorg. Mil./sorg. Mil./sorg. Mil./sorg. 135 TABLE 16 1 - Rankings for all levels of intensification MARGIN RANKINGS IN LOWER CASAMANCE (ZONE 9) WITH LOW RAINFALL One ranking was made for each of the three levels of intensification in To summarize those rankings, the most often encountered rank for a crop is given to that crop in the table below. states of nature 1 and 7. GMWL means gross margin with labor cost. GMNL means gross margin with no labor cost. NMWL means net margin with labor cost. NMNL means net margin with no labor cost. Mil./sorg. means millet/sorghum. Low. rice means lowland rice. Trans. rice means transplanted rice. Rank GMWL GMNL NMWL NMNL 1 Trans. rice Trans. rice Trans. rice Trans. rice 1 Maize Maize 2 Groundnuts Groundnuts Groundnuts 2 Maize 3 Groundnuts Maize 4 Mil./sorg. Mil./sorg. Mil./sorg. 5 Low. rice Low. rice Low. rice Low. rice 2 - Rankings for all technical modules A ranking was made for each state of nature 1 and 7. To summarize those rankings, an average rank was calculated for each crop over all technical modules and then a ranking of the average ranks was made, which appears in the table below. Rank GMWL GMNL NMWL NMNL 1 Trans. rice Trans. rice Trans. rice Trans.rice 2 Maize Groundnuts Groundnuts-Maize Groundnuts 3 Groundnuts Maize Maize 4 Mil./sorg. Mil./sorg. Mil./sorg. Mil./sorg. 5 Low. rice Low. rice Low. rice Low. rice 136 TABLE 17 MARGIN RANKINGS IN LOWER CASAMANCE (ZONE 9) WITH AVERAGE AND HIGH RAINFALL 1 - Rankings for all levels of intensification One ranking was made for each of the three levels of intensification in states of nature 5, 8, 9 and 13. To summarize those rankings, the most often encountered rank for a crop is given to that crop in the table below. GMWL means gross margin with labor cost. GMNL means gross margin with no labor cost. NMWL means net margin with labor cost. NMNL means net margin with no labor cost. Mil./sorg. means millet/sorghum. Low. rice means lowland rice. Trans. rice means transplanted rice. Rank GMWL GMNL NMWL NMNL 1 Trans. rice Trans. rice Trans. rice 1 Low. rice Low. rice Low. rice 2 Trans. rice Low. rice 3 Groundnuts Groundnuts Groundnuts Groundnuts 3 Maize Maize 4 Maize Maize 4 Mil./sorg. Mil./sorg. Mil./sorg. Mil./sorg. 2 - Rankings for all technical modules A ranking was made for each state of nature 5, 8, 9 and 13. To summarize those rankings, an average rank was calculated for each crop over all technical modules and then a ranking of the average ranks was made, which appears in the table below. Rank GMWL GMNL NMWL NMNL 1 Trans. rice Trans. rice Trans. rice Trans.rice 2 Maize Groundnuts Groundnuts-Maize Groundnuts 3 Groundnuts Maize Maize 4 Mil./sorg. Mil./sorg. Mil./sorg. Mil./sorg. 5 Low. rice Low. rice Low. rice Low. rice 137 in those regions much more than they are presently. Maize area expansion could contribute to reaching the government’s self -sufficiency objective. Cotton area expansion could lead to an increase in exports. In Eastern Senegal and in a low rainfall year, the rankings in order of decreasing profitability are: maize, groundnuts, millet/sorghum and cotton. In an average or high rainfall year, groundnuts supersede maize at the top. This difference in the ranking order can be explained by the price setting mechanisms for each crop already mentioned for the Groundnut Basin. Cotton is nearly always last. Net margins for cotton are negative in a bad rainfall year, unlike for other crops. These poor results for cotton can be explained partly by the large quantity of inputs required for that crop. Variable costs without labor for module 2 amount to 63,975 CFA Francs for cotton, 9,340 F for millet/sorghum, 18,740 E for maize and 25,600 F for groundnuts. Cotton also requires more labor than other crops, in particular during the harvest period. Module 2 requires 66 man-days per hectare for millet/sorghum, 49 for maize, 69 for groundnuts and 81 for cotton. Similar results can be found in Casamance, leaving aside the rice modules. In short, maize seems a promising crop while cotton is not attractive at present prices. The second issue concerns rice, which is grown mainly in Casamance. Rice is important because it is the major staple for the urban population. One component of the government’s food strategy involves substituting local rice for imported rice. Rice can be grown 138 in only two regions of Senegal: the Senegal River Basin and the Casamance. It is therefore important to assess the potential for rice expansion in Casamance. In that region, rice can be grown in several traditional ways following the toposequence: rainfed on the plateau, on lowland or transplanted. Rainfed rice occurs mainly in Upper and Middle Casamance (zones 7 and 8). Lowland rice can be found in Upper, Middle and Lower Casamance (zones 7, 8 and 9). Transplanted rice is found mainly in Lower Casamance (zone 9). In a low rainfall year, rice ranks last. Conversely, in a high rainfall year, rice is very well positioned. This difference in the rankings comes from the especially high positive correlation between rainfall quantity and rice yield. Lowland rice ranks better than rainfed rice according to the margins without labor cost, but sometimes worse according to the margins with labor cost. This results from the large quantity of labor required by lowland rice. For example, lowland rice module 2A (mechanized) requires 182 man-days per hectare, lowland rice module 2B0 (manual), 162 man-days and rainfed rice module 2, 114 man-days. In Lower Casamance, transplanted rice is clearly the most profitable crop. This can be explained by the limited amount of inputs used, except for labor, and by high and stable yields. Overall, rice seems a financially interesting crop in Casamance, although a risky one for lowland rice and especially for rainfed rice. One major constraint for rice expansion in Casamance is land availability. The areas where transplanted rice and lowland rice 139 cultivation is possible are stagnating at best, and regressing in many areas, because of the lower level of the water table and the associated increased salt content of water. Rainfed rice so far is mainly cultivated as 'pam pam" rice on recently cleared forest. The government’s policy of protecting the fragile environment by classifying forest areas limits the possible expansion of this type of rice cultivation. The third issue is the financial attractiveness of production intensification. This is especially relevant for Lower Casamance where use of animal traction and agricultural equipment is minimal, except in the northern part. Two modules 2 were distinguished for maize and lowland rice in Lower Casamance, and for lowland rice in Middle and Upper Casamance: one mechanized (2A) and one manual (280). Both modules obtain similar ranks, although the manual modules tend to be slightly more profitable. Mechanizing the cultivation of these crops does not seem very attractive financially at first glance. However, mechanization helps avoid labor bottlenecks at critical periods, for example during rice transplanting. For an average year, lowland rice mechanized module 2A requires 162 man-days compared to 182 man-days for module 2B. In general, the more intensive modules for each crop are ranked higher in terms of net margins, which should result in a positive attitude of farmers towards production intensification. Finally, as in the Groundnut Basin, modules 4 for cereals generally rank high for the reasons already mentioned. 140 4.2.3 - Margin analy§i§ for the Sooogal ijer Baoio The Senegal River Basin includes three zones: the large perimeters of the Delta and the beginning of the Middle Valley of the Senegal River (zone 3), the Middle Valley of the Senegal River (zone 4) and the Upper Valley of the Senegal River and the North of Eastern Senegal (zone 5). The main characteristic of this whole region is the development of irrigated agriculture. The margin rankings are preégnted in Tables 18 to 20. Four policy issues arise in this region: 1) Is rice the most profitable irrigated crop? 2) 1s double cropping more profitable than single cropping? 3) Is irrigated agriculture more profitable than rainfed agriculture and flood recession agriculture? 4) Is irrigated cultivation on large perimeters more profitable than on small perimeters? The first issue, i.e., the profitability of rice, is important because a major component of the Senegalese government’s food strategy is the development of irrigated agriculture in the Senegal River Basin, with rice as the major crop. It is therefore important to assess the profitability of rice in the eyes of the farmer. Rice is clearly the most profitable irrigated crop. In single cropping, the rankings give, in order of decreasing profitability, rice, tomato, sorghum and maize. In double cropping, the order is a double crop of rice, one crop of rice followed by another crop, and one crop other than rice followed by one crop other than rice. In the Middle and the Upper Valley, tomato margins are close to 141 TABLE 18 MARGIN RANKINGS IN THE DELTA AND THE BEGINNING OF THE MIDDLE VALLEY OF THE SENEGAL RIVER (ZONE 3) Two rankings were made: one ranking by number of crops (single or double) and one overall ranking. To summarize those rankings, the average rank of each crop or combination of crops was calculated and then a ranking of the average ranks was made, as is presented below. GMWL means gross margin with labor cost. GMNL means gross margin with no labor cost. NMWL means net margin with labor cost. NMNL means net margin with no labor cost. Rice+Rice means a double crop of rice. Rice+0ther means one crop of rice followed by one crop of either sorghum or maize. Sorghum+Any means one crop of sorghum followed by one crop of either sorghum or maize. Maize+Any means one crop of maize followed by one crop of either maize or sorghum. Rank GMWL GMNL NMWL NMNL 1 Rice+Rice Rice+Rice Rice+Rice Rice+Rice 2 Rice+0ther Rice+0ther Rice+0ther Rice+0ther 3 Rice Sorghum+Any Rice Rice 4 Sorghum+Any Rice Tomato Sorghum+Any 5 Tomato TomatooMaize+Any Sorghum+Any Tomato 6 Maize+Any Sorghum Sorghum 7 Sorghum Sorghum Maize+Any Maize+Any 8 Maize Maize Maize Maize 142 TABLE 19 MARGIN RANKINGS IN THE MIDDLE VALLEY OF THE SENEGAL RIVER (ZONE 4) Two rankings were made: one ranking by number of crops (single or double) and one overall ranking. To summarize those rankings, the average rank of each crop or combination of crops was calculated and then a ranking of the average ranks was made, as is presented below. GMWL means gross margin with labor cost. GMNL means gross margin with no labor cost. NMWL means net margin with labor cost. _NMNL means net margin with no labor cost. SorghumI means irrigated rice. SorghumO means flood recession rice. Rice+Rice means a double crop of rice. Rice+0ther means one crop of rice followed by one crop of either sorghum or maize. Sorghum+Any means one crop of sorghum followed by one crop of either sorghum or maize. Maize+Any means one crop of maize followed by one crop of either maize or sorghum. Rank GMWL GMNL NMWL NMNL 1 Rice+Rice Rice+Rice Rice-Rice+Rice Rice+Rice 2 Rice-Rice+0ther Rice+0ther Rice+0ther 3 Rice Tomato-Rice+0ther Rice 4 Tomato Tomato Tomato 5 SorghumI Sorghum+Any SorghumI Sorghum+Any -Sorghum+Any 6 Maize+Any Sorghum+Any Maize+Any 7 Maize-Maize+Any SorghumI SorghumO SorghumI 8 Maize Maize+Any Maize 9 SorghumO SorghumO Maize SorghumO 143 TABLE 20 MARGIN RANKINGS IN THE UPPER VALLEY OF THE SENEGAL RIVER AND IN THE NORTH OF EASTERN SENEGAL (ZONE 5) Two types of rankings were made: one ranking for rainfed crops for states of nature 1, 5, 8, 12 and 13 and one overall ranking with irrigated crops and rainfed crops for states of nature 1 and 13. To summarize those rankings, the average rank of each crop or combination of crops was calculated and then a ranking of the average ranks was made, as is presented below. GMWL means gross margin with labor cost. GMNL means gross margin with no labor cost. NMWL means net margin with labor cost. NMNL means net margin with no labor cost. Mil./sorg.//sorg. means rainfed millet/sorghum. SorghumI means irrigated rice. Rice+Rice means a double crop of rice. Rice+0ther means one crop of rice followed by one crop of either sorghum or maize. Sorghum+Any means one crop of sorghum followed by one crop of either sorghum or maize. Maize+Any means one crop of maize followed by one crop of either maize or sorghum. Rank GMWL GMNL NMWL NMNL 1 Rice+Rice Rice+Rice Rice Rice+Rice 2 Rice Rice+0ther Rice-Rice Rice+0ther 3 Rice+0ther Rice Rice+0ther Rice 4 Tomato Tomato Tomato Tomato 5 Groundnuts Sorghum+Any Groundnuts Sorghum+Any 6 SorghumI Maize+Any SorghumI Maize+Any 7 Sorghum+Any SorghumI Mil./sorg. Groundnuts 8 Mil./sorg. Groundnuts Maize+Any SorghumI -Maize+Any - 9 Maize Sorghum+Any Maize 10 Maize-Cowpeas Cowpeas Cowpeas Cowpeas 11 Mil./sorg. Maize Mil./sorg. 144 that of rice margins, especially when including labor cost. In fact, the labor requirements for rice are much higher than those for tomato (respectively 233 and 149 man-days per hectare). Two reasons for this clear advantage of rice compared to competing cereals are 1) the higher yield of rice (4.5 tons per hectare for paddy rice, 3 for sorghum, and 1.7 for maize) 2) and the higher price of rice (CFAF 85 per kilo of paddy rice, 70 for sorghum, and 80 for maize). The tomato yield is much higher than the rice yield (15 tons per hectare of tomato), but the tomato price is much lower than the paddy rice price (CFAF 25 per kilo of tomato). If the farmer wants to diversify, his best alternatives in decreasing order of profitability are tomato, sorghum and maize. However, tomato and maize expansion are limited by a soil constraint; the soils of the Delta are too heavy and too salty for tomato and maize cultivation. The only crops there possible are rice and sorghum. Tomato expansion is also constrained by the absence of an organized marketing system in the Upper Middle and Upper Valley of the Senegal River. The only organized marketing system for tomato exists in the lower Middle Valley. The second issue in the Senegal River Basin concerns the profitability of double cropping versus single cropping. At present, technical reasons limit the area available for double cropping, but the completion of the Manantali dam in 1988 should make it possible to control the flow of water of the Senegal river all year round and, therefore, to generalize double cropping. This is an important issue because the extension of double cropping 145 could reduce production costs, dividing fixed costs roughly by a factor of two. Domestic irrigated rice would then become more competitive with imported rice. The rankings indicate that double cropping seems generally worth the extra effort in the case of rice. If the farmer goes from a single crop of rice to two crops of rice, his margin increases from 278,392 CFA Francs per hectare to 701,317 F, i.e., an increase of 422,925 F. Since double cropping requires 118 man-days of extra work per hectare, each extra day brings 3,584 F to the farmer. This figure can be compared to the agricultural labor wage of 500 to 700 CFA Francs per day. Therefore, it seems that rice double cropping is financially attractive. ~_However, a single crop of rice can bring more money than most double crops other than rice. Considering the extra work required for double cropping, the margin per day of labor is much higher for a single crop of rice than for non-rice double crops. For example, the net margin without labor cost for module 2 of irrigated rice planted early during the rainy season is 266,525 F per hectare for 103 days of work, i.e., 2,588 F per day. The same margin for the combination sorghum-tomato, which is the best combination without rice, is 200,516 F per hectare for 199 days of work, i.e., 1,008 F per day. The margin for the combination maize-maize, which is the worst combination without rice, is 50,082 F per hectare for 204 days of work, i.e., 264 F per day. An important issue is the opportunity cost of time. The opportunity cost used above is the agricultural wage labor. The 146 opportunity cost could instead be based on the returns to off-farm employment in trade or transport activities, but no data were available on this. It could also be based on the value of leisure, which is virtually impossible to measure empirically. Because of the difficulties of estimating the opportunity cost of labor, which is an important cost element, our conclusion that rice double cropping is more financially attractive than rice single cropping must be treated with caution. Other factors are important in evaluating the value of double cropping compared to single cropping. In particular, one problem with double cropping is that it requires the farmer to stick to a precise crop calendar, to prevent the two crop calendars from overlapping. Therefore, double cropping is much more rigid than single cr0pping. Also, problems of overlapping calendars mean that all combinations of crops are not possible. Often, one crop has to be planted late to allow enough time to finish the growing cycle of the previous crop and to harvest. From the rankings, it is clear that late planting means reduced yields, and so, reduced margins. The margin for double cropping is less than the sum of the two single crops. The third issue concerns the relative profitability of irrigated agriculture compared to rainfed cultivation and flood-recession cultivation. These last two types of cultivation were the only ones before the introduction of irrigated agriculture on a large scale in the mid-sixties. Both have seen their importance decrease as a result of the drought. Rainfed cultivation on dierj land is now mainly limited to 147 the Upper Senegal Valley and the North of Eastern Senegal (zone 5). Flood recession cultivation on ooalo land is mainly limited to the Middle Valley of the Senegal River (zone 4). Although flood recession cultivation is supposed to disappear as a result of the control of the floods with the dams of Diama and Manantali, this will probably not happen before 1995. Until then, flood recession cultivation remains an alternative in the years when the water level is adequate. Two rankings have been made for zone 5 to compare irrigated crops with rainfed crops in two states of nature: the worst possible state (state 1) and the best possible state in that zone (state 13). If the worst state of nature occurs for rainfed crops, irrigated crops rank better than rainfed crops according to the margins without labor cost. They also rank better according to the margins with labor cost, except for irrigated crop combinations including sorghum and maize. The latter are dominated by a number of rainfed crop modules, in particular groundnut modules. In the second ranking, with the best probable state of nature for rainfed crops, irrigated crops remain in general more profitable than rainfed crops, but the difference between the two is smaller, in particular for the ranking according to the margins with labor cost. Rainfed crops require less work than irrigated crops. Module 2 requires 47 man-days per hectare for groundnuts, 42 for millet/sorghum, and 95 for cowpeas, compared to 233 for rice in the rainy season, 165 for maize in the rainy season, 179 for sorghum in the rainy season and 129 for tomato. Overall, irrigated crops seem more profitable than 148 rainfed crops in zone 5. Let us now compare irrigated cultivation and flood recession cultivation in the Middle Valley of the Senegal River (zone 4). Sorghum, which is the only flood recession crop, is the least profitable of all irrigated crops in the rankings according to the margins without labor cost. In the rankings according to the margins with labor cost, it ranks ahead of only single or double cropping combinations with maize. In fact, flood recession sorghum is not grown for sale, but as a home consumption crop requiring minimal inputs, which therefore leads to low yields. To sum up, irrigated cultivation appears more profitable than rainfed or flood recession cultivation. Another advantage of irrigation is its independence from weather vagaries, except during the hot off-season. This contrasts sharply with rainfed and flood recession cultivation. The area available for flood recession cultivation varies considerably from year to year depending on the importance of rainfall in Guinea, and hence the magnitude of the flood downstream in Senegal. The area suitable for rainfed cultivation varies as well depending on the importance of rainfall in the Upper Senegal River Valley. However, irrigated cultivation also presents some drawbacks from the farmer’s perspective. First, it is limited by the amount of irrigated land available. Second, its profitability depends on the availability of the right inputs at the right time. It is much more dependent on the input marketing system than the other traditional types of cultivation. Third, at a more general level, it depends on 149 the policies of the parastatal in charge of irrigation development in the Senegal River Basin: the SAED. The fourth issue concerns the relative profitability of large versus small perimeters. Large perimeters are found in the Delta and the lower Middle Valley (zone 3). Small perimeters are found in the Middle and the Upper Valley (zones 4 and 5). There are some small perimeters in zone 3, but they follow the same cultivation practices as large perimeters in zone 3 and can therefore be grouped with them. Elsewhere, large perimeters and small perimeters use different techniques for rice cultivation. On large perimeters, land preparation is done mechanically by SAED and rice is seeded directly. On small perimeters, land preparation is done manually and rice is transplanted. In terms of gross margins, the profitability of irrigated crops on large perimeters is greater than on small perimeters. In terms of net margins, irrigated crops on large perimeters are less profitable than on small perimeters, except for rice. The profitability of irrigated craps on small perimeters is also relatively higher than on large perimeters when labor cost is excluded. These results can be explained by the difference in factor intensities on both types of perimeters. Large perimeters are more capital intensive, which increases the share of fixed costs in total cost and consequently reduces net margins. Small perimeters are labor intensive, which reduces margins with labor cost. Overall, it is hard to reach a definitive conclusion on the relative profitability of small versus large perimeters. If the net margin without labor cost was the sole criterion, then small perimeters 150 would be more profitable than large perimeters. Other factors must be considered in choosing between large and small perimeters. From past experience, it seems clear that farmers are far more enthusiastic about small perimeters than about large perimeters. Small perimeters are by definition easier to control and are usually settled by a more ethnically homogeneous population than large perimeters. However, small perimeters have generally been set up where manual land preparation was possible. Future perimeters will probably require some mechanical land preparation. Given the disadvantages of large perimeters and the limits of small perimeters, SAED is now favoring the concept of intermediate perimeters; large perimeters are being disaggregated into smaller units for the management of certain operations. 4.2.4 - Margin analyoj; at tho hatjohal Iovol Two issues arise at the national level: 1) In which region does it make sense-to promote a given crop? 2) What are the most profitable agricultural activities in Senegal? To give insights on the first issue, a ranking of the net margins underha;erage rainfall conditions (state of nature 8) with and without labor was made across all regions where a given crop is grown. In the ranking for millet/sorghum, irrigated sorghum in the whole Senegal River Basin comes first according to the net margin without labor cost. In the ranking by net margin with labor cost, irrigated sorghum in the Delta and the beginning of the Middle Valley (zone 3) is first, but irrigated sorghum in the Middle Valley and the Upper Valley (zones 151 4 and 5) rank at the 20th and 22nd positions. This results from the high labor cost on the small perimeters in zones 4 and 5. The second and third ranks for millet/sorghum go respectively to rainfed millet/sorghum in the South (zones 10 and 11) and in the Center (zone 1) of the Groundnut Basin. The last place is taken by rainfed millet/sorghum in the North of the Groundnut Basin (zone 2) if labor cost is excluded and to flood recession sorghum in the Middle Valley of the Senegal River (zone 4) if labor cost is included. In the ranking for maize, Lower and Middle Casamance occupy the first place. Irrigated maize ranks last. In the ranking for rice, irrigated rice in the Senegal River Basin is more profitable (financially) than traditionally grown rice in Casamance. In the ranking by net margin with labor cost, irrigated rice in the Delta and the beginning of the Middle Valley (zone 3) is more profitable than rice in the Middle and Upper Valley (zones 4 and 5). In the ranking by net margin without labor cost, irrigated rice in zones 4 and 5 is more profitable than in zone 3. This difference comes from the high labor use on small perimeters in zones 3 and 5. Transplanted rice in Lower Casamance (zone 9) is the most profitable of the traditional types of rice in Casamance, followed by lowland rice and rainfed rice. In the ranking for groundnuts, the south of the Groundnut Basin (zones 10 and 11) is the most profitable. The high intensity module (module 1) in the Casamance (zones 7, 8 and 9) is also highly ranked according to net margin without labor cost. At the bottom of the ranking are the low intensity modules in Casamance and in the Center of 152 the Groundnut Basin (zone 1). It seems clear that the center of gravity of groundnut production is moving from the Center to the South of the Groundnut Basin. In the ranking for cotton, Upper Casamance (zone 7) is the most profitable region, ahead of Eastern Senegal (zone 6) and Middle Casamance (zone 8). All net margins with labor cost are negative, except for zone 7. This means that cotton is not a profitable crap in zones 6 and 8 at present prices. Even in zone 7, cotton is outranked by most other crops. Cotton is a marginal crop which occupies 7% to 8% of the cultivated areas in zones 6, 7 and 8. This percentage is currently decreasing as cotton area declines. Cotton is not grown for its profitability, but as a way to get access to agricultural inputs and credit supplied by SODEFITEX, the parastatal initially created to foster cotton cultivation. In the ranking for cowpeas, the Center of the Groundnut Basin (zone 1) is more profitable than the North of the Groundnut Basin (zone 2), where the climatic conditions are less favorable. Whether cowpeas should be developed in zone 1 and not in zone 2 depends on the alternatives open to farmers. In zone 1, groundnuts are in general better ranked than cowpeas while in zone 2, there is no real alternative to cowpeas. In the rankings for tomato, small perimeters in the Middle and Upper Valley of the Senegal River (zones 4 and 5) are more profitable than large perimeters in the Delta and the beginning of the Middle Valley (zone 3). This result must be interpreted cautiously since the same price was assumed to be paid for all tomato producers in zones 3, 153 4 and 5. Right now, most of the tomato is produced close to the tomato paste plants located at the border of zone 3 and zone 4. However, tomatoes that could be produced in the future in the Upper Middle and Upper Valley would probably be paid a lower price given the distance from the plants and other markets. To identify the most profitable crops in Senegal, a ranking by net margin with and without labor was made for all crops in all zones. The overall ranking by crop gives: 1) irrigated rice 2) irrigated tomato 3) groundnuts 4) maize 5) millet/sorghum 6) cowpeas 7) cotton. To identify the most profitable regions in Senegal, a composite ranking of crops by zone was made. The results were not significant, indicating that there is no best agricultural region in this country across all crops. 4.2.5 - ngmarv of findingo_aho_oolioy_imolioat1oos The present price policy does not especially favor rainfed cereals. Millet/sorghum is clearly not an interesting cash crap, and is grown mainly for home-consumption. Maize is much more profitable and can be competitive with groundnuts in some cases, although overall, groundnuts remain the most profitable rainfed crop. This does not seem consistent with the Senegalese government’s objective of significantly 154 increasing the level of cereals self-sufficiency. The present price policy strongly favors irrigated rice in the Senegal River Valley. Irrigated rice is the most profitable irrigated crop and the most profitable crop overall in Senegal. This reflects the very high priority given by the Senegalese government to increasing domestic rice production and to developing the Senegal River Basin. Cotton poses no challenge to groundnuts as the dominant export crop. Cotton is less profitable than most crops and, quite often, not profitable at all. The current price structure does not seem consistent with the efforts of the parastatal SODEFITEX to promote cotton cultivation. The present price structure generally makes intensification of production financially attractive in the southern part of the Groundnut Basin (zones 10 and 11), Eastern Senegal (zone 6) and the Upper and Middle Casamance (zones 7 and 8). However, intensification does not seem very profitable in the rainfed regions located more to the north. This pattern makes sense since intensification should be favored primarily in the zones with the best agricultural potential. To review this chapter, a number of detailed crop budgets were constructed to describe agricultural production in Senegal. These crop budgets vary by crop, by zone and by technical module. Uncertainty on yields and prices was incorporated. Margins were calculated and analyzed to evaluate the profitability of alternative crops and technical modules in all agricultural zones of Senegal. The crop budgets were also used as data base for a modelling exercise described in the next chapter. CHAPTER 5 METHODOLOGY FOR THE EMPIRICAL MODELLING EXERCISE The crop budgets provide a good idea of the financial attractiveness of the major crops for the Senegalese farmer. However, the profjt motive is only one side of the coin. The other important component of the farmer’s utility function is the food security objective. 4 ‘ iflwfl‘This objective pushes him to grow food crops for home consumption and to select crops for sale in order to guarantee a minimum income i,/ whatever the state of nature. Both actions may run counter to the profit maximization objective. Therefore, the farmer often has to make trade-offs among conflicting objectives. The farmer also operates under resource constraints. Land, labor and capital must be allocated to competing production activities. Some resource constraints are binding and force a change of production strategy. 0” A farm model that tries to include these key variables and their interactions can provide a number of insights about the farmer’s b// opportunity set and his probable behavior when faced with different 155 156 price polities. Therefore, our concern for understanding the farmers’ micro strategies and their implications for macro food policy led us to build a number of farm models and link them into a larger agricultural sector model. The stages of the modelling exercise are presented first, followed by a detailed description of the typical farm model and of the agricultural sector model. 5.1 - a f m el in e r 5.1.1 - e o t c ' o t l A typical farm was modelled in each agricultural zone previously identified, exoopt inatwo zones where two submodels were SODEEEHELEQ~LQ—- account for the diversityciouthose zoneszl In total, 13 farm models were constructed. Linear programming techniques were used for these models. The objective function specified for the producer was to maximize profits ~// under a number of food security and resource constraints. Only major «/” crops were considered as production activities. Livestock activities were not included for several reasons: - Policy choices regarding theflivestock sector are less important than policy choices regarding the crop sector in Senegal. The most lIn zone 3, one model was built for the Delta of the Senegal River where the only crops possible are rice and sorghum. Another model was built for the Lower Middle Valley where tomato and maize cultivation are also possible. In the Upper Valley of the Senegal River and the North of Eastern Senegal (zone 5), one model was built for the areas close to the Senegal and Faleme rivers where both irrigated cultivation and rainfed cultivation are possible. Another model was built for the areas away from those rivers where only rainfed crops are possible. 157 important agricultural policy issue in this country concerns the optimal mix of food crops and cash crops; - The government’s policies have a much more important impact on crop production than on the livestock sector in Senegal. The government has a limited influence on two major components of the livestock sector: the subsistence activity at the farm level where goats and sheep are raised for the milk and meat needs of the farm household; and a commercial activity which is integrated to a large extent in international trade flows of cattle and camels within West Africa; {- The livestock sector is less well-known than the crop sector. ). Livestock data is limited and often not very reliable. ,. Nonetheless, the existence of a livestock sector was considered indirectly by this research in several ways: - In the farm models, a value is given to groundnut hay and cowpea hay. Hay, especially groundnut hay, is an important livestock feed in Senegal; - In the farm models, land available for crop activities is equal to total cultivable land minus estimated cultivable land used for) livestock activities; - In the agricultural sector model, food needs to be satisfied are equal to total food needs minus food needs satisfied through consumption of animal products and food crops other than cereals. 'The farm models were calibrated by running them with the finnncial V/// éEiEEi_:g:Jl:E:}$ and butputs for the mid-1980’s and then comparing the models’ results with the data available on 1) the regional pattern of cultivated acreage by crop and by technical module and 2) farm income 158 for those years. Data on the percentage of cultivated area allocated to each crop came from the Ministry of Rural Development and from farm surveys conducted by ISRA and WARDA researchers. The figures from the Ministry of Rural Development came from yearly area and yield surveys conducted by the regional services of the Ministry and by regional development agencies. The percentage of the cultivated area allocated to each technical module for a given crop, and farm income were estimated for each zone on the basis of farm surveys conducted by ISRA and the expertise of several ISRA researchers. No major differences between the observed pattern of production and the farm models results occurred. When some minor discrepancies existed, changes in the food security constraints, which are described later in this chapter, and in the structure of cereals demand by the farm household made a fine-tuning of the farm models possible. The only exception was for cotton acreage. As mentioned in the margin analysis section, cotton is the least financially attractive crop. As a consequence, the initial solution of the farm models, in which cotton was a possible production activity, did not include any cotton. In practice, cotton represents on average 7 to 8% of the cultivated area in Eastern Senegal and in Casamance. It is mainly grown to get access to credit and inputs, in particular fertilizers, from SODEFITEX, a regional development agency which was initially created to favor cotton cultivation. To reflect this situation, a minimum constraint on the cultivated area allocated to cotton was added R'— / in the madels run with financial prices. This constraint was of course eliminated in the models with economic prices. ’4- nclufihduhh— -_., ”_H .1 159 5.1.2 - a n a i 1 n v l ! ! I! !i 1 1 1 -- The number of farms in each zone was estimated in several ways: 1) By looking at the sparse data available on farm numbers; u——..—._._ 2) By dividing the actual total cultivated area of the zone (calculated using data from the Ministry of Rural Development) by the size of the typical farm modelled for that zone (calculated from farm surveys conducted by ISRA and WARDA)); 3) By dividing the actual zone production (calculated using data from the Ministry of Rural Development) by the optimal output from the typical farm modelled for each crop. This gave one farm number estimate by crop considered. The number of farms that seemed the most consistent was chosen for each zone. A final consistency check was done at a later stage by comparing the national production obtained by summing up optimal production from all the farm models over all zones for each crop and national production figures for this crop calculated by the Ministry of Rural Development. The production of each crop at the zone level was calculated by / \/ multiplying the optimal farmoutpnt levels by the number;0f farms,jn W the zone. National production was calculated_by adding production from .. ntfl,..fl _ , , RH -_..... F— ' “w" all zones. Net production at the zone and at the national levels was then calculated by multiplying gross production by a coefficient which incorporated the effects of storage and processing losses and I reconstitution of seed stocks. 160 5.1.3 - v i lt r l rv The farm models were run with several different price vectors. Each price vector is a particular set of input and output prices which is consistent across all zones. The price of a given input or output is not assumed to be the same in each region, but rather varies to \ ' reflect differences in transportation and marketing costs. ixiThe basic price vector used was the financial price vector composed of 1986-87 prices for inputs and outputs at the farm gate. Five higher price vectors were derived from this basic price vector by increasing all cereals prices by a given percentage: 20}, 40%, 60%, 80%, and 100%. The supply responses nf the typical farm were then multiplied by the estimated number of farms in each zone to provide the Z°flEr§EEBIY curves, which added together gave the national supply curve. The zone suppTy in tons corresponding to any price vector was converted into calories to be able to compare it to the food needs expressed also in calories. Thus, six points on the supply curve of each cereal were obtained. It is worth pointing out that these are not Etandard supply curves since all cereals prices were varied together and not separately. Also FEEFE,§"PP]¥"CUFVES are normative, i.e., derived by simulating the \/// imnact of different output price policies on the production of the typical farms modelled. They do not come from the statistical analysis of time series data. Nonetheless, the derivation of these supply curves provided M \ insights into the likely response of cereals supply to price policies 161 which are plausible for Senegal. They also served as the interface between the micro farm models and the macro agricultural sector model. §!—¢U$x.o.¢g.U.—Ieg. d.awld!— .8 diafld:838usd {cakudieadu— 3“.ng.— §:8.!.§§83.§§§d=§ ’Egg .8..EE6'H.§E§ggeggafifldlflagafiflgbaafipg I— :!§:"3§'EEgaabga.§!.-§Ufl§§§§ d3! EHS'UEEnodn §.§E3.SH§B§¢‘fl.d—:lgaagago. §§g=§iwfl... cg .23 2.20 28” Home 333. 2m. 3... a. a. g . g: 38.— :«i 3:. a“: m a. E“ .83 80.3 .g. . g... 0.3.. .2.“ . . g. .3 239 . . 3%. 2.8. 3.30 m 2.8. .80 .380 M g. 0.3.. a mg 308 a m 23.. 22. g m .fl... :0... 998 “as in :33 “can: a... aim-n “nu-non. a '3. . .333 as; :2 a a“ a! a a. 22. as.“ 0.32 039 E a. gem 8.8 g. g a 33m 33 :3» .33 935 g“ :3. I! 2n“ 23¢ 33:“ 0.90. «.2. a a Seam g 03 g ~33 :33.“ 0'... =- £- 83. 3.39“; an: 0‘ a at,“ 8.”. Ban 5 name E3...“ 33. B3 0 c g. .382 pa ~03 a." §§.m§o§~§ péinuoa 3n; 3.... Bangaggfi 93 Bang—egg gnaw-Sumo Big .1 final". 38.!” I: "Ems-nag." .03. gmgo 02.2% 33 3mg 3.3 g a. gnu-mg 3.3 so“ u! a mi. :3. 8.8. . 3.3 3.3.; ~93 g 90‘ I... . g I: . I... 3:33 ‘3 . 88 53.833 3 go gaugfiog auguaag 88 azwzs 2.8 §.§malug g. gems-H.230 83.88.“,338‘3 a 853 “a. m a g ”i. u so a Ma. M ‘03 an“: “a. m :3 053 H .I. m as ~I~n Ma. m a 338 mm... M '0. 29% mu... m B.‘ 3.03 m E. M EU 0.2.. “In. M g a We. . .33 283 M93. as... M fig .13- .! H..- Ii 33.1..- § U.- Hi 3.15 a! 3.- Hi 5...:— U3 3.3.3.!!!abpsgilflhaa 335...... 250 a... mg. «g 8.222.225.08E2EmQ—32a3 Quito-a. .28: 8.38.823!!!— ssggggaagxggg.fii .82 . 8 n9 8. . g. .82 . 8. a 80 . a. . . . . . .82 . 8. a n... . «a. . . . . . .82 . E. 8. 8o . n8. . . . . . .82 . 8 8.. n9. . a. . . . . . .82 . H. u: a... . .8. . . . . . .82 . 2. 8o 8. . a. . . . . . .82 . 8 no 8.... . 2h. . . . . . .82 . 8. 2n n... . 2.2 . . . . . .82 . 8 no. Er. . Rm. . . . . . .8. . 8 mm. 8n . 2a. . . . . . .82 . 8 an... 8. . n3. . . . . . .82 . 8 :3 8w . £2 . 9.39.0 . a... E. 3.; . . . .85— . a d.§ 5...! . g . n8. 2. .3. II“. 2.28 8 gig”. z. gagbug .3835! “35.5% 9 u!— 251 .dp... EH! H..- 3.‘ 5...... .13— EU! H... .86.. a... no.” u... a... mag. 36. 3... g... 3.3 mag. 3d. 3.3 n... n... #93. 8d. 3... £38 .g. m m m m m m 10.8. um. mm...” m... 3.: 10.8. no... a... u... Bfih macs. 3... no... 3.. 3... mad. ma... 3... 3.2. mg. . M348. mm... 3...” no... 3... “3.8. 3.3 so... sad. 3... “no.8. 3.. 3.3 a... no... Macao. nus. 8... 3.... mos. m “3.8. a... E... 3.. a... mag. 3.8 3.3 nu... 3.8 M360. 3... 3... u... 3.. W362 2... 3.. 2.: “as. m “3.8. u... no...” 3.. 3.: “3.8. a..." In... u... 3d. mics. 3.5. «fig 3.. 3.. Tbs. no... a... 3...... mm... m “3.8. um... um...” u..~ 3... 10.8. «0.... 3... In...“ go..." “3.8. 3... 3.... a... a... M369. .5... I; 3.3 “.8. m “3.8. 3.. 3.2.. up... in... macs... no... 3... n... 313 “3.8. In... 3.: as... u... “no.8. us... «r... 3... “.8. m “3.8. 3... 3...” a... no... “3.... u... «n... us... 3.2 “3.8. 3... 3.2 3... E6 M348. mm... 3.. 3.... “E. m “no.3. 3.2 3.3 mm... 3... m3.3. a... a... a... 3&5 “3.8. nu... no... 3.... 3... M36... 3... #6 5.3 M2... m “3.8. #6 no...” n... mm... “8.8. an... nu.~ an... up... Tog. 3... no... a... 3... Mag. 3... n... 3.... ME. m “8.8. and an...” n... B... “3.8. no... mud a...” Hot: “8.8. .5... nus a... 3.3 “3.8. “Md. 5.. no... “.3. m “3.8. a... R... u... 8... “3.8. 3...... 3.. no... no.3 “no.8. u...n fl... a... I... “8.8. 3... a... 36. “.3. m “no.8. 3.. no... {.0 3.3 “3.8. 3.. u... 3.9 3.: “8.8. 3.8 no... u... 3.... “8.8. 3.. a... no.3 “1... m . . . . . 3.! 33m in... new! H... 3.5 madam #3— 8: 3:! Ed... . m . .85. . 2:. 1.3!... . .83: i 5...! «3.22... gu—gkagfigagbpgg:§gbl=§ .o 3— 252 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 m¢3h430.20¢.4 NO ZO.hUN¢.G .mucaom . . n.m @Oh . Dam. . . . M.M .ND . tom. . . . .m¢ Nhh . Mom. . . . Mon ”mm . flaw. . . . mn¢ 000 . .00. . . . mm¢ N00 . 000. . . . 000 Ohm . mhm. . . . m.¢ fine . 05m. . . . whm ONO . bhm. . . . m¢w 00h . mhm. . . . MOW Ohm . nbm. . . . won 000 . €50. . . . II II . . .2.¢z m0 .mZOh . . m.t2 2.. GZCmZOIh. . . xmaz. zanhuzooza . . JJGKZ.¢¢ mJszmU . m¢¢w> . 4G20.h¢2 @0030 . . mom. Ch #50. tozm Jdmmzwm Z. ZO..UDQD¢Q mJGNZWU QGOKG 02¢ JJQEZ.¢¢ l0 20.h340>w N0 quch 253 .9... In... I: E 43.23:! a 23!. 1...: I... H53...»- .... U 3.5.. 43...; 5...! E 33.3.!3 I 3.5!.- ”g . ad... .6. a... 0... pd. . g . m . . . 0.0.. n... ma ed. 0.8 . a. . m . . . ..s.. o... ”.3 pd. . ads . .8. . . . . . 0.2.. .d. 9.2. a... h... . a. . . m . . oi... ad. 669 .4 n6. . a. . . . . . md... .d. 0.... ed. .fl. . .8. . m . . . "do. n... a... .6. ~65 . a. . . . . . ~68 o... , 0.... 0.. “.3. . E... . . . . . 0...... o... 0.“ .6 w... . .3. . m . . . .63 0... .6. ed 0.3 . Ra. . . . . . . m... ..... ed. .... o.‘ . 2h. . . m . . rt... .45.. .dn 0.. 06.. . Eh. . m . . . 9...». ad. «.3 ad. a... . in. . m m . . u . . «S g . . . .1 hi 3.. u...‘ bud—8 . g . 3:8 .8 Rug 3...... B 8...! n8. 9. 03. ll... gggfich.gux>§aughs.5§u .33. 254 33.53333.filliu33.—§.IE§§.§.S.§U:3.2.3.. as g 3 2. 3. 3 3:83. 33! w.— 3 3 w.— .. 3:23. 323...... 3..“ .E. 3.83. 532...... 3 3.3. “a 3.3.5.... 3..“ U.— 3. a! 2.9.9.... u 6.3. 3.9.3 EH3 h. 3.5.2.... 333 U.— 3. 38.53.633.39: 59... gbgggiuauagn .2. 8“. h. 323...... 33! u: 3“. a a g. 3.2 H... h. 333...... 3.3 U.— 3. g c 3...; 3:3 3. 3.36.3... U 3.5.. u .g . . . . . . . ~38. o..o. .3...” . 8...... . mm... ~.o.. . . 88 . 033 .8. .8... . . . 8...o . . 8.... . 8...». m in... no... .38 .36. .3...m ..... hm... . o 03.. m .00.. a... 3.8 .So «.8. m .33. . 83. 8.3m 3...... U...“ m .3... .8... ..... 2.... 0o... H o5 n... . on” 8.. m 8o... «0.. .3. ..o. .36. m 8.3 8.. 83. 8.... m g 3......“ m 0.3. 0.0.. ~....0o 88. .9... m 0.... n8. . Non. o~o.. m 29. o... 3... .on ono. m 8... 8.3 8o. 8.. m Us...“ m .3... 0... a. . .30. u .9... a... . . .3: H Eo. 8.. a. . . m 8:. . . go. m a...“ M So... .oo. 5o... . :9. m .8 ...n . . .3. u no... 8.. 3o... . . m 85 . . 8.... m an...“ m o3... 0... 2h... noon 09.0 u no... . 0.3. 0a a m 0.3 00.. 00.... so: .08. m 83 8a 88. go m gm “ 8m... «0.9 an... . so u a... . 8.. . 8o m .8... go 9...... . o m . . . . u in: 3” m m m m u m 838.. gain. 33.-.33 3.333 3.3. . M .5.— EH... H... m...‘ 5.... M 3... EU! H... u...‘ 5...... m 3».- ..U.. H... w...! 5...... . .5... H... .23. 3...... . 33! .85— .E. 3.8! . :93. 3...... 3 .3... 0.07.8. .. g 3. >3 an. U.— h. 333...... 33! 3.3. 255 33. .g . .5 .8. .r. d. n. .0. n. .3 . 8...: . . . 388. n... H... 3... . g 0.... . . . 388. a... .. .. 3.3 . g 8...... . . . 39.8. 3... n... 3... . 25...... . . . :88. 3... a... an... . an... . . . .86.. u. d. u. d. .. .... . g... . . . .86.. 3... B... nu... . g. . . . 30.8. 3.. 3.... 8.. m =3 3.... . . . u . . . . . . E. 8:93.. . . . d... 2. 8. dnlg 5..“ . 2..“ . . . . n8.-.8. .. 33.8 >. g 8. >§ a U..— h. 8.5.3... .3838 93— 256 “no.8— nnd— nndn no.0 no.9. “no.8. nude no.0 no.0 3.3 mno.8_ nud- nsdo no.0 no.0 “no.8— n0.0 no.0 no.8— m “no.8. nus no.2 nné nouns Mno.8_ nude no.0 no.0 no.3 mn0.8_ n04 2.0... n56 no.2... mn0.8_ no.0 no.0 Wotan m “no.8— n~..~ n10. n13 3.3 “no.8. 2&0 no.0 no.3 no.2 “no.3_ nod no.2 no.2 no.3 “no.8- nod no.3 nmd m gin—gm “no.8. nod no.0... nm.~_ no.6— mn0.8. no.0 no.0 no.8 n5.“ Mag. no.5 no.9 nod 2.0— mnooo. no.0. nod n..~ m “no.8. no.2 no.5 no.0 no.3 Mn0.8_ no.“ no.0 no.0 3.3 mn0.8_ no.2 nmdo no.0 no.0 “no.8— 36 no.0 no.8— m “no.8— n0.~_ no.3 no.0 n13 “no.8. nu.” no.0 no.0 3.3 “no.8. no.0. no... no.0 no.0 “no.8. no.0 no.0 #202 H “no.8. nu; no.3 no.2 no.3 “no.8. no.0 no.3 no.2 nu."- .Wno.8_ nod. no.” no.2 no.3 m.no.8_ 8.0m nn.n~ no...“ N “no.8. 2.3 nsdo no.0 n~.0 “no.8. no.0 no.8 no.0 no.0 “no.8. no.3 3.2. no.0 no.0 “no.0 no.0 no.0 no.0 m m m m m m “.50— EH... 8: 3:! bud—x m .32 EU! 8:. NE! a: a .83— pcws US $5. 5.1:: m d5... H:- H—i Fwd—cm . .50— “ 0:. g“. m 28.!- .58: m 3:? i :3! m 08.08. 8— ; >0 a; :— g E..— >§ g .5 3:33—20 53* 3m...— v! ill-‘1‘! t A 257 gaging—a .anug.§§g.g=§u .ggggaaomdu=gfioo=amn .3 .za :5: : .8: 6%... 5:... 6.... .§.§: EBB—Eu 23... 2.3.: we >- 6.5:: I. 638 8.5.... an! m.— ggghfiEiuéEEB-a: 5 Es. m m m a m m m n nu. mu. m 0 mm m 0 o m 3 00 m “- mu m 8:5 m m m m m m m n so. or.— m _ o— m 2 u. m cu 8 m .2 no m ‘55 EB“ m m m m . m m m 0: 3. m . 2 a ou n. m rd so u 8. «o m .8!“ £92 m m m m m m m u n2 no. u = no u a o m a on m B 24. m Home: m m m m m m m a no. mm. m n nu m o 0 m on no m fl_ 9. m 5.: m m m m m u m u nu. mu— m n no u 0 0 . 5 00 m a. mu m 35.2 m m m m m m m a $— 3. m . a. m an on m 2. 3- m _r. = m gum y m m m m m ... m n: 00. m n- 3 m o o m 0: 0: a u. u. m g a: m m m m m m m m m m m a m m ”gingimgimgimg 3!: m u n n u 0 983! m m d5— ” EU! a H-i m N:- mgnfl—E m 3:53 «W. g and—x h. g! Eggs...—gasp—gEQSBE—ngBBE—zl—Bw no N!— 258 gig I. .aE-Iusfla.fi=o .5..§§§§3.8!E~. a... gagging-.3303! .g ”is. a.gggaag's.sgfiagapgpi.§§gaggg.ag3:! m $- 00.- n- .n- an- m 3 5- 53? E.“- g... n 5.8. a .38. E. .03.. m 880. 0 a: 8.0. 33“ is.“ M .0- 00.- ST 3? m? m own...- Rno... 8.8- 2....- §- M an. ~03. ’3 2... 30 m 8.0. 0 0 0 8.0. M 3....“ m z..- 00.- 00.- no... .- m 33- ..07 :3.- .flh- .8...- “ mi :0. :3. .u... .3... M 00.“. o o 8.0. g.“ g Ham“ M 0n- 00.- ~0- 8. "a m 88.- .0~.- .Ho- .8. 3.... m a .8. .~.~. a. 8.0. M 3.. 0 83 go. 283 M g Iflgm “ .0- 00.- n- .7 .0- M .ao- flow... 030.- n.on- .83- ” .23 a 0.0.0 “.7. .30... m 830 0 8.3 8.. g m gdw “ ~0- 00.- 87 87 .7 M 0.80- 3... .0007 a? no.3- " 0.! g .000. ham 3 H 3... o 0 0 .0... M g.“ m on- 00.- 00.- 8.- .7 M 3.....- muon- $0:- E? a H 2 upon p.02 88 033 m 8.3 0 o o 9.3 m 35.-M m N. 8.- no- ~.- 8 m 3...- SR- .3..- .l? a. H a... a g .33 Zn! m 3 0 an an a. m gm m 00.- 02- 3? 8... 3? H ~033- 803- 3.- o...- n...- H «03.0 on.“ .3. 0.3 0...... m 0 0 o o o m =3 3“ . . . . . . . .U. .08.. 8.0 3.! 33...} i an! H..- B—i spud—um ‘3— EU: 3.. 8:! saw g .3 3.. 3:! nsa-w “3:608 don...“ U n . 3 gm 5‘ g m 3:3 i 5.: m In! a g . 8.0“ 259 on... E a. g 8.0 0.. a. .3... .u. a... .. =5; 38.. 9.... «g ism. #1.... 8... u...‘ h. g— 808 086.. .22 0.... .n. .2.E i=0. 8. Hi 3 I... 88. a. :02 0...— .u. a. E l .03. I... U.— x. g...— l...¢d§ 2. as. 3.80.. . . . . . . . .8. . .3 . .n. 2. . .3 . no. . a. . .3. . . . . . . . . . . . . a. . 0.0 . .0. 2. . an. . no. . 0.8 . no. . .n.~ 2.. .038: . . . . . . . . . . . . fig . . . . fig . Bin—us . .383 . .d§.§.s.w~.i.—§. U=.§=..§..§.. .!.§..B.§..a.§. g. .3... a... 3... «.89.. 33.8 3... 35.8.... n. .8. $53.58!.Eagig £§Z=§aoa§u=sT 4:3 .2 an. a 8.... ma 0.. .32. a 8.... . £3. 3 8.... n. .8... a 8.... 2.. £53.63... A. 8.... E .82 a. was. as... a. an. a aim. u: 8. - «:3 5.. Sin. .0 m3... .8 a. .5: .a 8.5. 3. .5... n. 8.... .. .8: .0 8.... 2. (5.8.5:... 8 8.... a .82 ... an... 3.8.. a. an. a. 2.5.. u: an. - 8:36. 1.... u.— a. 3.." a. .3... 3... a. :3 a... 5:: 5.... u: .. a... as... 2. 3.... amps. .30.... 3:56.. 52.. u: a. - 80. 0.. Sna.E§EaUE§U.—Bl=§ .d. a. 9u3§§§§§g33.5§ 0. an 260 3.3.8.26... .2 . :8...- ..~.. I a! . :8...- al: I. an! End 30.!- .263 5.3.82.3... 5a 3. gil!!i.h§:.hlfl§:§!fl§ i.bq§..~db§.§5§!3§ du!§l§flflllllil.taz..Ignaz-Illa... .Ifillllun.anagfigprBIIQ-ighfifillla. Eflglliohaz..bgfiuaigda ghuiicgfl.ilaualal=.g 2.5. . . .2 . 33!... 3.5 . . .3 . 8:15: . . . . . 3.3.3.6. . I: .55. .n . . .35.. n... 65....-. Z. .38.. 3.. . . 3.8. . .a. .382 3.8 . . aa- . . . . . . .gga... . . . . . . .58....35.-. .3 .88 3.. . . slu- . .3. .88 .n . . 8.....- . . . . . . 5:38.! -. I .5 I . . §.§-. .8 .9. .8 . . sail-38.5..-. .a .8: .8 . . 3:33..-. 2 .8. .n . . 8:33..-. . . . . . . 2.5 . . 3. . 3.3.1.2. . . . . . 35.5.6. . 3.8 . 3.... .n . . .85.. 59.3.3.1.-. . . . . 52335.5. 1:2: . .u . a... .3. . . .35- uiafii.-. 3 . . . . gust-.53.-. :8 . . . . 3.3.9.3-. . . . . . . .3. . . .5. . nag—3.... . a! . Isl. . g. a... . . .23.... .83.... al.-5. . . a. . .: . . 8a.... . . a: . I. . n: .IEII. a . 2.53.8... 3.. . «I . may... . §.§..§U.—Iigg§a2gggagueSign. Hagan—H.333: .fiit-fic'gan... 3...!EBE s3.flilzfla.n.o:gegfiflgnba=g .agp'ugspieglflg§§:!gu==§.fifig galauzgfigglfiilqgaa gum—i=2 3.5!. I! H..- b... Ole=fiagg flihudllw... to. as... 1.525.253.2522. gun. .‘B! a! g. o had.» 48.: ad URI. Carla! 353 5.!!- dug -. 5.55.85.32.35 23.3..3' gill—.8393... g-S‘UB'!!EUB!!I. 693%. .33; .U.-§&3.§=ddu=!u gum...‘U—Ic.‘l§.§2§83§.fi!2§§g inn—5.36:5.pig'g‘. 5.338ng gnu—8. gaggfiu £3.38... 3a.=egi.ou Husky-Sn ii:i§I§—agfllaug «do: 261 WOON “CON {Olmomu em .nm .mm .m_ .o. .n .aa _ x~azmaac ._m.a mo mecp o x.azmaa¢ .nom.. ecu .mmucnom .ngamuu 4¢:_z¢ com m~n¢z no mzo» coco” mmoadozu .m. .maaamam ¢ muchcn m>~b.moa c .p—u.uma c an uozc4¢m m>~h¢muz c .azctua nzc zanhuaaocn zmm3hmm muzmmuumua mzp mu uuchcm mt» .q. “mmpoz xmn m om¢u m mme. m .om. m xnn u «an: m mg» m con. m 4¢h0p m Ill m . . m a all m m Ill“ m m . m m m m . m xmmm m 0mm m snn ”.m. n_. m x.m m h: m n.» m .m. an m m~.cz m . . . . . . . . . xo m «mm: . o . «mm m :0 m mn_s m a m mm. m pcmzz m m . . m m m u m m m . . m m m m m tazomom m xms m mom: . mum m «me u xmo . mu: m nun m mew m \hm44_t m . . . . . . . . . x¢m m mew: . ace m moo m xn. m mun: 0 mm m ncc m no.2 m m . m u m m m m m uuuuuuuuuuuuu ”unusuanuu.uu . m muuuansuunmnnuuuuuuuumunuuuuuuumauuuuunnum >uzm~uuuunmm .u. m m “ >ozm.u_mu:mm .u. m m m m mem no 4m>w4m mochcm .zo_huanomam azcxmn ”qum mo 4u>u4m uuchcm .zo~»u:ao¢am azctma m m . m4¢m¢wo m . .mZOh OZCNDOIhv 0h *lemm— 10mm Admmzwm can hmmlm mUZGJ¢m deuzmu th to ZOuhumfioca ".mt UJQGh 262 EA w!— 23: ESE 5 .g ”868— mood "on." ”26 “88 m m m m m m m m 8.3: m 86 ”8.... m 3.” ~82 m m m m m m m m omdmn. m 34. m tutu m “.0 a 8a— m m m m m m m ”8.34... ”and mu—fi m 3.0 ”9‘ u m m m m m m u u a a - . m g m3_§ma—gma—§u m m 038 m .85— m d.- ” i a g m g i. x— g n.2- ..a a g g s— a. 0.3..“ “a 52!“. a! .EJZEi-I-usfid—E 0:: g 3... g .. .83»; i. H: E #6 ”a a! B—pgsganodBw—isiu SUN—=3! 3!: g 3 Us. .38”. 33:: .= .3:ch Some 93 §u§8u9_§§gfl=&3_§ Team!— nag—«_aigfidlflmE—EQ‘53U633-E on m!— 263 «I :3 5A BI. 38: 32.83 i .3 3:55. 3.!» iicfifi m . . m m m m m a a . mg a m a o:— m 3‘ a: m In, is: m . a m u m m an an up u cu g a m a. a 3 u N. a o. m gm m m a m m m m a. a: m .2. on: m .3 a. a a no: u g ‘5: m m m m m m . 5 o.- n 03 w: m ~= no u no 2. m U: (E . m . m m m m m . «N .3» 8. n mm— a- n- u 8 cm.— a m N: So— n... m 3: E5—‘— . m . m m m m m . m m m m a: g m .2 Io a m: a. an «2 '0 n m: a. an m '5; . m . m m m m m . m m m m H:- o m . u m m m m . nu. on: 8— m «2 23 n. m n. a 8. m on rub an m H: n m . m a . m m m . no. 3n '- m a 2“ a a a: no. as. m a: o: g. m .35: a m . m m m m m . s . > . c . a . b . C m . . a . c u g . p . c a . u a . . z . g. m ‘a . ’q . . g3! 5.23.!»inuga..§5a. .3 .882!_§‘g_§§.flbm¢§ «fling- .3533 a 389.. a: .35: 83... .3 .8829: n! «35! 8...; fig. .8: 65:: 8.- 623:3: .55! u. - .3 .8888: I! is! 8333“! .5558. 3.3.: .62! a: 93:3 585925- ..§aa&§q§.g§f .3 §2§.§g§flb§ 33.-z- sign—gatlufliic- :3 .§e§.§uifi§€_alfll gnu—.3235... 2:233—3 .§a3-3_: Ensign-S gag.§mglfla—§'igbl=§ mg m..— 264 n...» H!— 33: as: a .3 ”8.5 m 32 m a m 93. m as“ u m m u m a m ”3.9 m a m g a mu: m 82 m m m m m a . ”3.3 n «mu m a n «ma. u 85 m m m m m m m ”3.9 m as m 8. u can. m a. m m a m m m m . hum—“=5 m . . m a . . “aha” pun—QME—gm i m a m 5.3.3-2.: §E§_§m§§§§33253_§ nanug 265 an... .52.; 38: 3—5—5 5 a g 35.3 psag .3 m m m m m m m m 923 m 0.82 m n.“- m .68 m «.32 a g m gm m m m m m m m m «.33 m .a m 9.9 m 0.02 a mi“ m a m a. . m m m m a m m m g m . m m m .85— m Pad... 0 3;. m 5*! m H: m 8.5! m . u g. m g» .85 flan-E an. 8..— i . 8:; m .53... s: :6. m 3.5. a: a... .5: E 5 .8: s: I. 62.8 a azazgu8§ 9.3:: u: on 3.. a was: 3.3 .2 a «E a «may. 88 a. 3:825 .528 _E 393 a 8.. 5 .3 £5... a 8.. 1: as... a 8.. as. .32. a 8.. t: .8: a 8.. 3. .88 a. 3&8 a 2...! - .528 5.. 333 a 8.. 38 .3 §\§_. a 8.. «.8 2.5. a 8.. E. .5: a 8.. 3. .8: a 8.. t8 .8: z. 238 a“. 223 - .29Enbui§§-§§:2§_§u=&§§s 2.5-8:88:38: 525238.383 sag—55:38.13: gag—imaflgggba_§ _ .3 a— ..uu—E—imh a g..— 2253 U: >- gaaflg 38 U: “BE—g 3 an 266 .3... :8: 533—5 5 H g Ban.- »Sag .g m En— m m m n2. M £5— . . . . . . m mm. m g m 3. u so u H—im m m m m m m m 9 m m m a mum-:- a m m m m m m . m on. m 3...- m 8. m m gnu a m m . . . . a m nmm m g m 3 m n my; swam n m . . . m m m m .33 m 3 g 8 m g 5.5 n m m m m 5:: m . on. m 8. Hz. m .35.: a.» a as an! 3.5:! .53. a: .. 3318353333.; NEE- 5.5:. 3.... a up! :5! 3:89! gFfiax n: - .5325 $8.5. z— 3;. .35! g x— a! .2— g 5 3:95.: go—§_E§§E_§§gbs_§ a...» an 267 «um... 33: BEE—.3533 5 8 g 53.. 2:23 .g gca—ggb—I3333—hca—ggfic gig—aggm—cggggz. 5:! ”3.3 m 82- m 2.3 m :8 34.... m a? m 3h m 32 m 15— m m m m m m a m m m m “3.8 m 2. m nu. m an. an”: m a? n so u 3 m $2! m m m m m m m m m a u ”S... m at... m e m 2m “3.: m 27 m o m o: m EH: m m m m m m m m m m m man... u -~u m 93— a n3. ”3&3 a 3 m 80 m 3n n 33.3.: m m m m m m m m m m m “and. m in mono m 32 :52 m 37 “8. m an m H:- m m m m m m a m m m m ”Eu—Em 2. m m “So—Em 2. m m m m m 3 gm a 33:85.: i m a 2.: 3 i=3“ i m m u u u a m u g m i- m an: Dan-E !KV107 INMIIMIDI! 1’3 1’0 'm-m “(IE I ' o o o I 0 7H! 32 I975 1976 1977 1970 1979 1900 191 192 MININGDENNIGBUHEIBHIDLMMIMWMMIVONI” ! mm ! 1970 1971 1972 1973 1970 116! 0.30! 0.37! 0.00! 277 257! 310! 02612! 003 293 66‘1 was 7116 91‘) mm 90607 16099 11233 ! 01103 ! 2973 ! 330 030 272 30 036 99111 77373 9153120“ 76112 15053 72039 "M3! 5023323 Immune: 1m ! mo “11 rm 110078 m 07702 166010 23797 107772 smmmem 35733 37000 man ! 11.0.” ! 100 129 107 00! 0.30! 10033! 0.01! ”manna 10“ 03299 01N16637 79” «a: 0363 31203 30520 37003 no: 00900 01577 03127 031M100720!11!9! 0.27! 0 I33! !II1ll WM! 152” 17170 17600 31220 033” 31“ 31707 2533 20%| 300“ 36503 076M 516“ 5917 66061 ! 35906! 70 100 I” 109 133 103 120 160 I73 I75 160 157 153 ! inn (.2 !1|.S-M!5$ 27! 0.26! ! I07 ! 0.33 ! 26! 0.26! 350%! 0.31 1 7339! 0.22! 030! 0.23! 350! 003! 100 I36 136! IN! 116 125 131 90 I“ 129 I26 109 129 I21 ! 1m ! 1HMWOM!10W 15029 10112 20739 29161 23907 23103 21706 21199 2N2: 271! 30301 35006 0%73 52077 ! 27165! 11160! 0.01! ”manna 137697 330 33001 «no 533113213 30037 5010 R97 5021 271” 2065 20190 330130092133”! 101 ”12120112 93 121 “31101506“! 16120 16136 10112 21050 31012 23“ 5760 233“ R03 207“ 6373 532 30” 51016 590! ! 20760! 12205! 0.03! !I”50Flln am 00069 3300106397 9157 517 3769 2'16 27239 3151 26379 3307 M0 3730 331913595! !1L0.m152 :usmea 0 M1116 101 268 90760! 0.36! ! 1” 60 mo 116597 175131 101670 103913 m 1519 13070 16021 170170 13307 was 1W1: 9529 “I ”10 1132976 ! 1116.“!22025125039173903302337631.3305623303309 em 131 am n ma: 63300 6527 am 07193 um 92662 101097 13032 103“ 119919 1025“ 1690i 123709 132%! 152% 1115606 ! 09 ! 0.27 ! M 211 201 239 I09 1% 150! 102! summit-Inseam 27106 11700 51310 01130 2950 0&0 1320 “33 00903 $051 65“ 61007 70676 6%06100337! 10051! 0.31! N 122 266 170 I00 176 210 Hummus 17023 60370 7205313070 76662 0153 077” “15 at 0‘1 “1 613‘ 09113 5300 0051716319011“! 0.2! 10.0.“!102 25! 0.31 ! 6502! 0.02! 76 ! 0.27 ! m ! 0.01 ! ””029 m 56! 1‘012109102 006” ”700 10113 79720 7030 71297 m 7150 “10151612939! 10501! 0.20! o 3B 275 307 360! 277! 55 711 1017 ! 0061079 .9 0591003 21.0mm2215525301491“mommummnmmnmrummmlmwsums: 71703! 0.20! 0 “III 1711 Pm11fi251 122070 10727 I21N3 25509 1660! 165109 “116 20399 10993 101207 203“ mm 270053 000029 !fl3§0 ! 10.0.”!379001035061077770 2mm MI. 171 3325330330 339 DI “IIIHFM! 30360 37672 36162 0997 01603 61“ 71910 0&16 66“ 7203 71297 91107 00501170031570.11170097! ! 11.6.“ ! I30 13 100 220 171 "10101 IIEIII. “10. 071E! MOM 3-‘51 II III, CIF W 10118. 161 .1010: ”£010. 201G U 71‘ m m. 101 261911101 canal. CIF um. 131160. mummul. mums. 10111.0. HIE02MHIWMS. “I 111 IMMIIGPIIGCIFMMIIIVMW, MMGIKIIMII'IMIIIIE. mmmrnuam lemm 12111.0. '81. man. HI am 11” I IN). - 1" use) Mi. 9.5577 H 110' 11371 SMIISIIIES FMIM MEDIUM “)1. P. F052 m IIE man: 0011! III! Imuloamummasuua. um.mn.muummm. m:4.qunewunmuunmumumum-umcun-mnwm-mmlmmmss. '10I1H11I51 MIWII‘HNWEIMFIIMMEUW 1m1m191mulil1l‘! amnmemmxme 1979 11.653 1011.3 PEI 1‘01“) 1975 1976 1977 MWIOIGIIQNFIIMCIPSIIMFIIIMIUIIS 1971 1972 I973 I970 269 «38168189. 81181318226 666666 666666 aeanaa =23 assasg sags 16$ 20! 2” 339 910 7“ 700 03 1310 1130 95 256 055 000 10” 790 017 033 1213 713 150 010 a % 33$! §§§ g :2 a 2:3 I“ 207 1150 1160 012 709 031 706 § 1110 791 9” 300 I3 692 7” 623 $0 0” I5 0! 711 201 I25 30 29 632 600 9” 062 1105 me 3 730 919 uoou 3 figs: 532 - - — 0-0 -0 -. -l -1 -n-c -I-o -1 -0 -0 0 m1 no mu 1200' 1002. 1527 933 600 222 792 0” I000 750 0” 113 1160 033 1001 903 379 1% 1009 1009 391 997 1N3 mm 53 1M!!!” 0 I I O O O Im'm'smm'm IF! I I a O 0 1976 1977 1970 1979 I“ III we 1‘3 1m 1973 970 1971 1972 1973 1 .1970 !EV107101!10I1011191! 113 000 732 317 001 030 231 507 700 0 0 1225 103 1230 2170 2210 m a: so» 310 631 270 0” 701 707 1667 1010 060 1029 600 700 063 610 000 667 $7 an. 670 730 $7! mé 101' 019 710 1931 180 11$ 977 I! 000 329 1027 730 700 020 0.32! 0.00! 230! 0.07! 07! 3311 273! 73! 722 607 330 010 0“ as . 693 9” as 1201 ! 900 022 a an 61 I297 63 302 99 676 763 0“ 211 310 033 31 m 213 611 973 977 707 200 =52 Q o -0 -0 -0 -0 m: "min: E ummm 271 . 33: m 69 m .d a 38.. m m . m m 38: m Rm 3% 83.: :- m a! E 36.. m . a: gflam . a: was: . 38: m a. gs" m a» >- 5—5 m 3 S: 81.8. . «>3 m . m m . 38: m o. m gig—mm 33 m 5:81:53 m . . . . R. m Ids-U32 E— m 83: m m . 3 m m . 3;: m on as.” 2.35- m g m . g . m m m. in b m g a g H: m g . g m 33H m a g» m a»; .398 a 9.3-d“! g 3 3H— :- gu 3.9—5a “3 ”ES! or. 3— 272 ad an 33: ESE pa .5 m ~30: ~67. 0.5.. ~67 m g m m m m m ~62 .60. 0.3 n.‘ m $5.. .m m m m m «.3 3 ads 3 m 93.5 .m m u a m w... ad ad ~d m g .m m m m m ed. _.= 2 od m new: .m m m a m «.9 0.9 T3 udn m H:- .m m m m m m E_ m h n m m m m m 3 3 3 on u dz: .m m m m m a o a a m 3:8 .m m m m m a a a 3 m a; :3 .m m m m m m 3335 m u n m m g 82 82 a. a as m 38. .3 hub is .3 33:2 page—aigsgausaa— a.gggifla—al—mgg— Bu!— 273 .2...— 3’: g .- 35 .g m «.8 ~..~ 18 0.8 . g m m m m m ~.~n m.” u.“ n‘ m .85— .. M ad ad nd ed M has .M M ndu 78 ad n.‘ M H:- .m m M s... M m m m m a...» «.3 n.“ r23 . dS— .. m ad ad a nd m 3:8 .m m oz: nfio as. on m g a; .M m m SIR-u m m 82 . an. a. a. m as m as. .3 *8 is nag—.3... iw‘mggggmuSuflm— S. 822 ~8— gagging—lagg— am!— 274 3523833.; m u m m u «.3 «.3 ~.«« «.3 u «.n m = m m «.3. «.3 «.3 m «.« m 2 m m «.«r. 13 «.8 «.2. «.3 . 1s m « m m «.3 «.= «.8 «.3 «.3 «.3 m «d m « m m «.3 «.«n «.3 ..«« «.«n «.«« m «d m a m m «.3 «.3 78 «.8 «.3 . —.« m « m m «.3 «.3 «s . «.« m g‘ u m m «.«u «.9 «.9 «.3 a «.3 m E—._ n m m 13 «.3 «.«« m «.2 m o m m «.9 «.3 «.«u . . «.« m « m m «.3 «.8 _ «.« m « m m «.«u a.“ «.«« m «d m . w m m m m m m m .g 3. H3! 9.5” m m m m u m m m m « «.« «.« «.« «d «.« «.« u . I—g 9:93»: m m a m m «..« «.«o «.9 «.«c _d« «.8 «.3 m m *8 U: 3 33833-3: m m ma l—g E 33.: m m u m u «.«« «.m« o... «..« «.«o «.3 «.«« m m :3 g 3.!— A! m m m m m 3 «.3 «.8 «.«n «.3 «.3 «.«o m m .33. g .35 m m m m m .3 C! no...« «73$ «*3 g.« «.7: m m m n u «a m m m ”I «mu—«=35 m 333;: m ~«n!. :- g I 3%.—Emu E a! g 5 5:25 daze-.5385!!!— 53:355.: lug—5.83333gaggizahauntggiueglsfig .3: «K in «b. RH!— Z75 m . m m m «as n.‘ a.“ a... .do no." and u = m a 063 .do 0‘ 0.9 a.” . ad a nd m o. m “a...“ 0‘ 0.. 0.“ ado - .d . -.— m o m a a! ado. n.8- 0.‘ a.“ o .J a nd m C m mug .40. ~40- ...“ a.“ and mod n a m m 0‘ “.3. .1 Ada o.“ a 0.6 a 1. m a a . 0J3 ”no. a! 0‘ a.“ u n.- ~ 0.: m n m a cg 18— n...- ofin a.“ a ad a...- m o m a a! 0.3 n.‘ a... 0.: a n6 u 0.. u n m m pd: —d n.' 0.: a.” and mud n u u o .63 v.3 u: add 0.: m .6 a nd m a m m a a m .! .- U-: 832 m a m a u no! ‘3 NJ. 0.3 u.‘ u m maglfll'gm m u m o u m m u m . m pd. u m m pg gag“ m a m m m m m m m i I 535;: m 0.0 0.0 0.0 O.“ a m m 8 '1’! .gam m m a u m m 0.: a...“ ads 0.." a.” u o m *3 U.- & 335:3” m a m mg 5:: E mg!“ m a m m m m «in s.“ 0.; 0d. a u m i an! an: a i a i o a! m g m m m i a. a. g “a i aging—Cg: m ggflaflflllugpcguluguu- e—i'gugflzihguzu bauagc-adaggna ’2 ad“ .a¢&_§ dS—u ii 5 an a.» 5:33.. 8 hub 23 - .moésofldsudflaahgsgB—g: .uudnl-iI-d-lbi! d: .3 2 3' ll. 3.9-E l—a u a; anagua ~T’.l§!g.:l.§!ath-lug 33.—‘33 276 as U.— 3 2:58 .3 . m . . m . a.“ m -.~ m ad . 2 m . _fln . .6 m _.. . 2 m . «.8 . 04 m n6 . a m . ad . 0.0 m .d . a m . a.” . s... m .6 . a m . 0.3 . 0.. m 06 . o m . n._~ . ad a 0d . n m . ufi- . o... m o..— . o m m a.“ . 0.0 m ad . n m m ~.~n . 0d . 0d . M. m m . u! :- U—g Hg” . . can . - U3 :- ung 3g" . . I m . m o._ . u g 2. pg 9.35;“ . . '3 N 2 g 3 3 5.3—38: . m 2 in: 3 8:3. .9; . use. 2 3.3583: on 9.6:. 5.8.9 m . m m m m m m g H: 3 3:83.33” . m . m m o d B 3..— 5—3 cm; 8:2: 9 I: U...— 53” o 0" .- m 1.: m a m is u 3:! m 3:5;— Suga—iugllmpgl=—c§3gmun .2 g. 3.1;!— Eggflaggga—uigufig—E g—x-Kgucb. 31!.8—ga—guB-dmfle— 8.8583333; >2! g— 3.82.233 on:— 277 g H..— 3 8:33 .3 m u . m m m new m .d m nd m : m m ".8 9.8 ad a .d m 2 m a m m 3... 98 18 ¢ .4 . d.- “ . m m .3.” 3... a... and u 9.. m ~ m m «.3 m ad 0 .d m o m u 1.» 93 ad. m u.- “ o... u d m 0 1” .3 ad. a fin a a... m . m . “.3 add 18 m 9.. m ed m n m m . m m 52 a. 8.: 2.3-Em m a m m m m m m m m m z m m . w... and... a 3:. 62 m m m m 58 9.33.: m m m m m m 3. in m m . m. 2: a. 8...: m m m m m m a m m .38 E 3:533: m 9.. 3 m m . a: 9.322 Has—2 . o g o o m QR is m m m Eb u: a 3:553” m m m 3: 3.38.8 SE. 8...: . . . . . a 3.... e..~ ” m a It. 8.! a”: 5g .35 m a! m g m m u a: U: 3.! m 1.; . 3:2;— m m .8: 6:... a gaahbnflaihaagflfiglfiaua gaggd—sH—E-.§ Saga-luck: souaqfigfilgugaue 8: s H_! 35 3:; >2; E. B 32:33 3 u..— 278 ggfiggafiia . . m m w . 83 . .4 m 3 m = m u 93 u a . 3 w 2 m u «.2 . o . . u a m m 3. . 3 a 3 m . m m 1.. . 3 a 3 m . m m 3» fi 3 . a... m . m m 3. . a m o... m a m . a... . 3 . .4 m . m m . . m is a a... 8982 . a . m m m m . m m m ad. m m u .3. a. 8.8” m m . m . m a . . m .3- a 3... as . m w m in SIN-g m m m m m m .2 3. u... 3. 3: . a . 2 I: 3 3...: . . . . . m 3 .a m “ u .3 2 is m m a a .5.- ..E 6.3.8.38: m m m m on 2!... .35 . . . m n m .d 3.. m m m it «a a 3:55; m a m .8. 5:2 E 8...: m . m m ... a n .4: u . a ll 3.! a! m 33 h d! n! .8 I! . . m u 3! an .3- 3. 3 . i . in m m .85— 5: gggiulluuaflbgs-KmnB—gazc—iu a.gnznnH—g!.fl§u_gu gigaflsagggflogan swan—liaiacgizslal- s!:!a:3.:...n..g.l.§- 32.-888.3. ~?'_u—§Si‘al2 ails—‘Ciglgg—B T! 279 g!‘8—§.g m u m m . m nfi. m 9. a pd. . a m m ~53 m a.» m «J. m n m m in»: m .6 . ad. m o m m m m m 53 :— H3... 2552 m m a m m m . u “ . m m m m m m «a: a: m m m 3.8“! .339: m m m m 23 9:5: 8:2: m m m w m m ad m a m 3:3 23 B as! . m m m .8 58 0:83.: m m u m m m .E m m m m: wa~ :— 8...: m m m m m m ed m m m .53 2 En m m m m g E 3:53;: m m m . on 0.39! .83....” m m m m m m .63 a m m *8 U3 3 3:83.332: . m m ”8... 3.; E 3.!“ . u c a o . .38 a m m I; 8.! 8.: . m m m m n :58 SEE . 3.! m E m m .2! g. m a; a 33¢; a m 8:8 i.§as..§&bna§a=¢§_ua~2§.- is 333.55 9...... - 32:228.... g Jawsfimacafiaa- ggub§.§&b3 a§.=!3_§§:g.gal§es Ed: I! gm 5 unsu8_x— Swain-£53382 g8 E 3.: >23 gm & 335:3 «can APPENDIX B Crop budget sample for mi11et/sorghum in the Center of the Groundnut Basin Z80 3!! 183 ’ .I-0-0--“.0.0--0-0--.0--.-0.0.0-0-..“--0.0.0“-0-0~-.0“--.0.l.0.0.0-0-0~“~“a i! 2 a 8 a ail . a -0 .- -0 c. -0 .0 .0 .0 -n -0 .- -0 .t .- 00 .0 .0 .0 .0 .0 .C -0 .0 -0 -a .- .0 .g -u .0 -0 .0 .o .. -- .0 -3 .Q on .0 -u .0 .c .1 .- .u g 3 - I 0 -l -0 .0 .0 .0 - .0 .0 -0 -0 -I -0 -0 .0 “ .. .0 .0 .0 .0 .0 -0 .0 -0 .0 u .0 - .0 .0 .0 - .0 - .0 .C .0 .0 -0 .0 - -0 -0 -I .0 .0 GI.0-..I.l.0-.I-0.0.0“...¢.0*.0.0.0-0.0.0-l-l.0-’.0.0.0.0.0-I.0.C.O.l.0.0.0.....0-‘.lm“.c 1-0....0o...-m-.-’o-o.¢.o-o.fl.o.fl.¢.n-o.n.o.fl“OOOQOOOCOIOOOOCIQOOCOOOIOOOOOOOO 0 ‘-O.-.---0-0.0.0-0.I’0‘.-0~“~.l.0...‘.0-0-0...0.0-l-l.0-l.0...0.0.0.0.0.0...0.0.C.l-0“..~ O . . u u a . . . . u . u a . ~ . u . . ~ . . u . o . . a . . . u . u . m a . . a a . . m . o .0... “B; a app-i! 3 .:—§ I... app-‘0“ a mp=§m C I I . I!m“.0.0..-“m”-...0-.0’.0.l.0...0.0.0.0.0-..“--.0.0.0-.0.0.0.0.0.0~.‘.0.C““' i“..-n.-.o.o-o-o.o.¢.¢ 33139 g 5353; alll' !. 3 :5 ga Oi. a" a as 00-0-0m--.0-0-..‘...0.0.0“..--~-’0’..-“-“-.0--.0-.0-...0”.0..—~“-0“”“ I.““~~“.0-“~"“---~.0...0“-.0-l-0.0“-0.0-.0-“.0-.0.C-“.‘-I-O.C-O-I.0.Q.. . :!§3.adi .‘Z . gas-:02“ gag-’2 gas'.§m . g..- g .03: . in: .05— .8: . i an: a l .8: it In. a: a E .2: a! :3. u 83 a: g . lilo—lug! .05 3!” . afloaos a l I!“ .aofl.l.§ lu- g .3: .2.805 a ' .3: . l.- .. s 3.! ‘8: u a a a: a g a: a ‘ :9. a g . i a: CA}: ‘ .3: . a .8: u and is u a! a: a I: .8: a. =_: a in m m 281 . . . . . . . . no . 0H . an u no . 0O . aadflu> 02—0206008100\HD. 000) 0000 .ufl... . v0 . 00 . ac . O0 . on . .OJU-> 02-OZOQINBSOOxNG. 000) 00¢¢U>¢ .ON-u. . .2.u2~ . hm . no. . an. . NH. . .040~> GZ~OZOQOW¢¢OU\-O. ¢¢u> O¢0 .O.-v. . . . . . . . . . . . . ZO-bUSOOCQ to bmou b-ZS. . . . m . m . OOh . 0&0 . 050 . nnb . non . .H.Omm\m... #02 .0.... . can . HMO . 60$ . 0N0 . n60 . .H.OO~\O-.. 00°80 .h.... . . . . . . . . . . . . >¢03003 Bun ¢¢0> 6000 0 Zn. . . . m . . . 000 . 050 . n05 . ~80 . ans . .N.OON\0O.. hm! .0.... . 000 . nan . 000 . ..h . 005 . .N.OON\0O¢. 00020 .n.... . . . . . . . . . . . . >¢OX¢OI Run 000) 00¢¢u>¢ 26 2.. . . . . . . . 00.! m 000 . 0am . no. . OOH . ...°O~\CO.. but .0—... . 0N! . hnn . .Nfl . HON . HON . .—.OON\NO-. 00080 .n.... . . . . . . . . . . . . >¢ax¢OJ Cut SCH) OGO C 2—. . . . . . . . 0006“ . MONO“ . 00cm” . 00N0n . “N006 . .Ofloflq. hOOU 000¢J bZOIh—S but .Nauv. . 0000 . 003.0 . 0006- . 0000- . 0000- . anflonq. #000 COICJ Ih-S hm: .uuuo. . 000"“ . 08000 . Oflnnn . Ohnlfl . Och-Q . .00!"-. #000 8000. hadzpua 00000 .0-... . OM00 . Oflncfl . QMON. . Ounn- . noun. . .n0ln.. b000 8006. In.) 00°10 .00... . . . . . . . . . . . . wc¢huu1 80¢ ¢¢u> 0000 C Zn. . . . . . m . 0000‘ . “Osnc . 60cc“ . 0030“ . ””0"” . .cfllflu. hoou 80004 h301h~3 ~02 .00... . 000. . 0050- . 0°00 . 00h” . 0000‘ . aflfilflnv hOOO 00004 Ih-3 #01 .90... . OflOh. . Ofln0o . Oflnhfl . 0500“ . 00h0fl . .O0II-o bmou 80034 hDOZh-Z 00080 .00... . ONO! . Omnmm . ONn.- . 0500 . “than . .N0!N—. #000 8°0¢J Ira) 00080 .00... . . . . . . m . . . . . mxdbuux cut :¢u> 00¢¢m>¢ 2C 2.. . . . . . u . 000”! . Mama. . 0600 . onN0 . 0N.h . .00!.-. #000 BOQQJ hDOI—ul but .06... . 0000.! . 000.! . 000“! . 0050! . 0hH°-! . anal-o. bflflu 80004 :b-I but .60... . O00! . OUONN . ONOH- . Oh”. . 0000a . .006-uv hmou COO¢J PDOZb-S 00020 .NO-o. . 0000! . 0N0. . 0M0 . 6000! . numb! . .n0lau. baou COOCJ Ih—S 00080 .uO... . . . . . . m . m . . . wcdbumz CUQ 0C0) QC. 8 2n. . . m . . m . . m . . m . . . . . . .n HJSOOI .0 “439°! .n 04:00! . 0 04:00! . a 04:00! . .m¢u0. 02-008! 0 to N Mada . hufloal 60:0 u.q 20.020) I:IO¢00\b044-8 .0000 .. uzau. Z—mdm hDZOZDOCQ 02h k0 KubZUU .mZO~ 282 '||'I"'|'|'l-I'I.|IIII'IIIIII'II'|'llIluIl'll—I'I'l'I'-'l'l"-‘|"'-II-'"n|.III‘Ilolll"|'|l"l'lll'll|"l' ‘I|"'l|'nlu ' 0 .-. N .- O. u “10 -0 -0 -0 -0 -0 .0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 000!202. 80020400 000 80004 ... 20.080) 0020800\h044.2 .0080 .. UZON. 2.000 #:2023080 02h l0 8uhzmu .UZON 283 . c . On an m NM Ge MD m @N m Nn m mm m Ne m 0? m mn m n! m mm m MN m 0” m m m m m w m m m m m m u m m m on m mm w nm m mm m QM m Mn m MN m on m 0m m MN n W m Mn m m m On m Mn m M— m MN m @n m ON m MN m m m on m Mn a “n m an m ¢ m m m h m 0 m m m c u h n O m 0 u h m M m n m 0 m n u n m n u n m m m N m n m m m m m m m u m m m m m m m m m m m m m N m 0 m 0 m 0 m 0 w m m O m D m 0 m O m m m m m m m m m m m m m m m m m m m m m m m m m m llllllllllllllllllllllllllllll I' m . m m m m . m m m fin WAJOOZ me 34300! ”N WJQOOZ m N WJDOOZ u n NJDOOZ . W m: ¢ NOGQ -0 -0 -0 -0 ¢¢m> 0000 nfl.omm. ¢¢u> MGGQNDG nm.ommv ¢¢w> OGQ nn.O@Nv. mm: COQCJ J¢b0h z¢m> 0000 G Zn 0Q JGbOb an.owmv z¢m> w0¢¢m>¢ ZG Zn “a JChOh nm.ommv 11m) 060 ¢ Zn nu Adhah an.0mmv ¢¢u> 6006 I GZnZOZZnZIGZnImwKZP nfl.mnm. ¢¢m> w3¢zw>¢ I OZnZOZZnZIOZnImszh ZN.Mwmn ¢¢m> adm I GZnIOZZnZIOZnImmzlh an.MnNn. ¢¢m> DOUG ¢ Zn @Znhmm>¢¢1 nn.nnmv ¢¢u> UG¢¢u>C ZC Zn @Znhmw>¢¢I nm.nnmn ¢¢w> Odfl ¢ Zn @Znhmm>¢¢z nn.nnflv m m nZnCC datum: m hmznu Nib ¢WPl¢ ON Oh m OZWmZ. flu ocnzuim u m 0 #Q JGbOh nonmn @Znnme nn¢flv . nZnGC Jfiummz m hmmnk th BNFEC 0 Oh h GZMNI. (Q Qanzwfl. m m . nm>¢OlZ¢tn m CGDZwJCU um: COQ¢J m . Illlllll'llll-"' 9&3232 lSZU n.n fonm¢m> ZZIGGOQ\buJJnI .0020 .— MZON. anIm bDZQZDOKG WIh to ¢thmU .mZO~ 284 .80) g .6.006.. . 0.6 . 6.6 . 0.0 . I6 . I n . .Iofl .66 .6.0 ...O .0 O n at g: .~.an.. .a.. .36 .n.0 .n.0 .. O . g a. 2.86.. . . . . . . . . . . . . . HO... {:3 18:0». . . . . . . . . fl . O . N . 6 . 6 . CCU) g 8 Zn It .136. .6.A.d6.. . Q . O . Q . I _ 6 . ICU» :2 .8. Zn an. 1:.— .6 .03.. . n .0 .6 ..I .C . ggCantbflh .n.006.. . I . O . O . 6 . 6 . 30> a Z Zn chip hi; .6.nn6.. . l . O . I . I . I .3 g: ‘ Zn £820.!- hli .6406: . u . O . fl . C . Q . 30> a C Zn .pih p.85! 3 Jan: . . . . . . . . . . . . . :3: an... . . . . . . . pant Z... (up! 0“ On I 0.3!. “6 8.8V... . . . . . . . . n . O . O . O . O . on dhflh .36.. . n . O . O . O . O . gnaw! :06... . . . . . . . . . . . . . nZn‘ dung . . . . . . . bOZnt u.» 80h... 0 Oh h 5!. On. 8.8V... . . . . . . . . O . n . n . n . n . 66 1:... 2.06.. . O . . . a . n . n . gnaw! 366.. . . . . . . . . . . . . . nZn: it! . . . . . . . p.87. Zh tuba 0 Ch 0 I16“... 6.. 8:5... . . . . . . . . 0.0 . n n . n . n . n . NA .13.: .066.. . O . n . n . n . n . snag .63.. . O . a O . O . O . O . .39.: g ‘9anth i .63.. . 0.0 . O . O . O . O . 2:53.48 2N6... . . . . . . . . . . . . . «Zn: flan-Hg . . . . . . . but—t U: flint 0 On 6 0.3!. N... 8.8:. . . . . . . . . O . 0.0 . O . O . O . nn. dun-h .03.. . O . a... . O . O . O . '32: 2 snug-cg: ”I; 3.6.. . . . . . . . . . . . . . nZn‘ .53 . . . . . . . want '3. (Ub& 6 O.— n guy. .6 sncua. . . . . . . . . O . 6.0 . 6.0 . 0.0 . 0.0 . a Zhflh .On6.. . O . 0.0 . 0.0 . I.O . 0.0 . 9338.46 306.. . . . . . . . . . . . . . nZn' g . . . . . . . pant u: ”at: 6.3! I. On. anew... . . . . . . . . . . . . . . . . . . . . .O>SO.!:’.. . .nflg.OU§.6U§.O§. n3. gigging! H . . . . . . “Bag h! alt... non snag §\bfljnZ .gu .n '5». Zn! pg. 2 a IUnZUU .gn 285 NOTES ON MILLET/SORGHUM CULTIVATION IN THE CENTER OF THE GROUNDNUT BASIN (ZONE 1) Module 1: high level of intensification Mechanized planting. High use of fertilizers. 2 mechanized weedings with spring-tooth harrow set on a ”sine" hoe or on a I'western" hoe. 1 thinning very well done. Module 2: average level of intensification Mechanized planting. Average use of fertilizers. 2 mechanized weedings with spring-tooth harrow set on a "sine" hoe or on a "western” hoe. l thinning well done. Module 3: low level of intensification Mechanized planting. No use of fertilizers. 2 mechanized weedings with spring-tooth harrow set on a "sine" hoe or on a "western” hoe. 1 quick thinning. Module 4: home gardens ("champs de case") Mechanized planting. Manure spreading. 2 mechanized weedings with spring-tooth harrow set on a "sine” hoe or on a ''western" hoe. 1 thinning well done. Module 5: late cultivation Mechanized planting. This planting often follows a first planting done according to the optimal crop calendar, but which failed because of a lack of germination or an insect problem. The farmer has to plant again at a later time. The seed cost and the time required for planting are very low and are counted only once. No use of fertilizers. l mechanized weeding with spring-tooth harrow set on a ”sine" hoe or on a "western” hoe. 1 quick thinning. BIBLIOGRAPHY BIBLIOGRAPHY ABT Associates. Senegal Agriculturg] Policy Analysis. Dakar: USAID, 1984. Ba, F.N. "Analyse de la situation céréaliere dans l’Unité Expérimentale de Thyssé Kaymor Sonkorong (Sine-Saloum)." Master’s thesis, Montpellier: Institut Agronomique Méditerranéen, 1982. Balassa, B. A "Stages" Approach to Cpmparative Advantage. Staff Working Paper 256, Washington, D.C.: World Bank, 1977a. -------- . " ’Revealed Comparative Advantage’ Revisited: An Analysis of Relative Shares of the Industrial Countries, 1953-1971," lug Manchester School of Econpmig ang §ogial Stgdigs 45-4 (December 1977b). Baldwin, R. "Determinants of the Commodity Structure of US Trade," American Economic Review 61-1 (March 1971): 126-46. Benoit-Cattin, M. s échan e ér l n'v e rodu t urs au Sénégal. Département des Systemes Agraires, Montpellier: Groupement d’Etudes et de Recherches pour le Développement de l’Agriculture Tropicale, 1984. Bigman, D. Food Policies and Food Secpritv Under Instpbilitv. Modelling and Analysis. Lexington, Mass.: Lexington Books, 1985. Branson, W. and N. Monoyios. "Factor Inputs in the US Trade," Journal of International Econpmics 7 (May 1977): 111-31. Braverman, A., J. Hammer and J. Levinsohn. Agrigpltural Priging Policies in Senegal; Their Implications fpr governmept Budget, Foreign Exchange and Regional Income Distrjputipn. Washington, D.C.: World Bank, 1983. Casey, H. "Dynamic Programming in Agricultural Production," in Planning Agriculture in Low Income Countries: A Symposium, Department of Agricultural Economics and Management, University of Reading, England, Department Study No. 14, 101-15, 1974. 286 287 Chenery, H. and M. Syrquin. Patterns of Development 1950-1970. London: Oxford University Press, 1975. Chisolm, A. and R. Tyers. Food Security. Policy and Perspectives from Asia and the Peeific Rim. Lexington, Mass.: Lexington Books, 1982. CILSS/Club du Sahel. Merketinq. Price and Storage of Foodgrains in the Sahel. Ann Arbor, Mi.: Center for Research on Economic Development, 1977. CILSS/Club du Sahel. Les politigues ceréelieree dans les.pey§ du Sahel, Actes du Colloque de Mindelo du Ier au 6 décembre 1986, Ouagadougou and Paris, 1987. Craven, K. "Peanuts and Rice: Some Obstacles to Senegal’s Drive for Food Self-Sufficiency." Ph.D. diss., Fletcher School, Tufts University, 1982. Crawford, E. A Simulation Study of Constreints on Tragjtjpnel Farming Systems in Northern Niqeria. Department of Agricultural Economics, Michigan State University, MSU International Development Paper No. 2, 1982. Deaton, A. and J. Muellbauer. "An Almost Ideal Demand System," American Eepnomie Review 70-3 (June 1980): 312-26. Duloy, J. and R. Norton. "Prices and Incomes in Linear Programming Models,“ AmerieeneJournel of Agricgltural Eeonpmiee (November 1975): 591-600. FAO. Energy and Protein Requiremente. Report of a Joint FAD/WHO Ad Hoc Committee, FAO Food and Nutrition Series No. 7, Rome, 1973. -------- . The Fourth FAO Worlg Food Survey. FAO Food and Nutrition Series No. 10, Rome, 1977. -------- . Mission d’étude sur la sécurité alimenteire- senegal- Diagnostic et éléments de politigue- Rappprt de la mission. Rome, 1980. -------- . Politigues et programmee eéréeliers - Une anelyse du secteur eéréalier et un programme d’eetipns pop: spn gevelgppement - Sénégal. Rome, 1985. Gabas, J.-J., J. Giri and G. Mettetal. A Preteeted Regionel Cereals Market: An Initial Exploretion of a New Idea. Paris: OECD/CILSS/Club du Sahel, November, 1987. 288 Goetz, S. Qbservatipns on RuralsSelf-Sufficiency, Transections . Behaviour and Prospects for Expanding Cereals Production in Sgutheastern Senegal. Institut Sénégalais de Recherches Agricoles/ Michigan State University Food Security Project, Draft Working Paper, November, 1987. -------- . Maize Production in Sogth-Eestern Senegal. A Briefing Note on Constraints and Prospects. Institut Sénégalais de Recherche Agricole / Michigan State University Food Security Project, East Lansing, January 1988. Gordon, K. and G. Rausser. Qountry Hedging fpr Reel Ineome Stabilization: A Cese Study of South Kprea ang Egypt. Department of Agricultural and Resource Economics, University of California, Berkeley, and California Agricultural Experiment Station, Working Paper No. 306, March, 1984. Gouvernement du Sénégal. Progremme d’ejgstement éepnpmjgue et finencier e moyen et long terme. Dakar, 1984. Harkness, J. "Factor Abundance and Comparative Advantage,” American Econpmic Reyiew 68 (December 1978): 784-800. Hazell, P. "A Linear Alternative to Quadratic and Semivariance Programming for Farm Planning under Uncertainty," American Journal of Agricultural Eepnomies 53 (1971): 53-62. Hazell, P. and P. Scandizzo. "Competitive Demand Structures under Risk in Agricultural Linear Programming Models," American Jogrnal of Agricultural Eeonomics 56 (1974): 235-44. Hazell, P. and R. Norton. Mathemeticel Prpgremming for Eepnomic Anelysis in Agrieulture. New-York: Macmillan, 1986. Heller, P. "Factor Endowment Change and Comparative Advantage: the Case of Japan, 1956-1969," Review pf Econpmics end Statisties 58-3 (August 1976): 283-92. Helpman, E. and A. Razin. A Thepry pf Internetiopal Trade pnger Uneertainty. New-York: Academic Press, 1978. Hillman, A. "Observation on the Relation between ’Revealed Comparative Advantcge' and Comparative Advantage as Indicated by Pretrade Relative Prices," eltwirtsehaftliehes Arehiv 116-2 (1980): 315-21. IMF. Annuaire de Statistigues Financiéres Internationeles. Washington, D.C, 1985. 289 IMF. Stetistigues ananciéres Internetieneles. Washington, D.C., March, 1987. Jabara, C. “Agricultural Comparative Advantage under Uncertainty: the Case of Senegal.” Ph.D. diss., Purdue University, 1979. Jabara, C. and R. Thompson. An Agriegltural Seetor Moeel for Senegal. ‘Agricultural Experimental Station Bulletin 278, West Lafayette, Ind.: Purdue University, 1980. -------- . ”Agricultural Comparative Advantage under International Price Uncertainty: The Case.of Senegal," Amerieee Jegrnel ef Agrieulturel Eeepomies 62-2 (May 1980): 188-98. Jolly, C, M. Kamuanga, S. Sall and J. Posner. Situeti en Cer eeliere en u l e '. n .- :- n .' n- n: ét .- Terrain Document de travail, Centre de Recherches Agricoles de Djibelor, Djibelor: Institut Sénégalais de Recherche Agricole, 1984. , Jones, R. "Factor Proportions and the Heckscher-Ohlin Theorem," Review ef Eeonemie Stedies 24 (1956—57): 1-10. Keita, M. Note on the 1987 Foog Needs Assessment Methods. Internal Memorandum, Dakar: USAID, February 13, 1987. Kelly, V. Acggisitien and Use ef Agtjeultgrel Inpets in the Context of Senegal’s New Agriculture] Policy. The Implications ef Fermers’ Attitudes end Input Purchesing Behevior for the Design of Agricultural Policy and Researeh Programs. Bureau d’Analyses Macro-Economiques, Dakar: Institut Sénégalais de Recherche Agricole, 1986. Klein, R. "A Dynamic Theory of Comparative Advantage,” Ameriean Eeenomie Review 63-1 (March 1973): 173-84. Kramer, C. Foed Consumption, Feed Demane in Senegal: Summery of Literature end Suggestions fer Neegeg Aeseemeh. Document 1036 A, Dakar: United States Agency for International Development, 1985. Labonne, M. and B. Legagneux. Modele regionelise de simulation de l’agriculture pluviale Senegaleise. Série Etudes et Recherches 33, Montpellier: Institut National de la Recherche Agronomique, 1978. Lau, L. "A Characterization of the Normalized Restricted Profit Function,” Jogrne! ef Eeenemie Theery 12-1 (February 1976): 131-63. 290 Leamer, E. "The Commodity Composition of International Trade in Manufactures: An Empirical Analysis,” Qxfemg_fieememie_£epems 26-3 (1974): 350-74. -------- . r e f n erna i n om at v A v n . nd Evigenee. Cambridge, Mass.: MIT Press, 1984. Leontieff, W. "Domestic Production and Foreign Trade: The American Capital Position Re-examined," Preeeedjngs ef the Ameriean Philesoehieel Seeiety (September 1953): 332-49. Lewis, C. and J. Fry. mm e A licati n for Subsaharan Afri n Countries an Pr d W rld Benk Reseereh Pregram. AGREP Division Working Paper No.95, Washington, D. C: World Bank, November 15,1984. Low, A. "Linear Programming as a Tool in Innovation Selection for Farms in South East Ghana," in Pl nn 'n A r t in Countries: A Symposium Department of Agricultural Economics and Management, University of Reading, England, Development Study No. 14, 64-86, 1974. Martin, F. La Referme ee Ie Peljtjgue Cereeljece ae Senel. Le Sénégal. Elliot Berg Associates for CILSS/Club du Sahel, Paris: OECD, 1986. -------- . Budgets de eulture eu Senegel. Document de travail, Bureau d’Analyses Macro-Economiques, Dakar: Institut Sénégalais de Recherche Agricole, 1988, to be published. Martin, F. and A. Dieng. Analyse de la situetiep eljmenteime ae Senegal. Evoletiog de 1975 a 1985 et perspectives. Document de travail 86.3, Bureau d’Analyses Macro-Economiques, Dakar: Institut Sénégalais de Recherche Agricole, 1986a. -------- Le eemmerce exterieur de pr e dejts egnjceles_et_dLintranL§ agricoles du Sénégal de 1975 e 1234. Document de travail 86.1, Bureau d’ Analyses Macro- -Economiques, Dakar: Institut Sénégalais de Recherche Agricole, 1986b. McIntire, J. Food Security in the Sahel: Veriable Import Lev y, Grain Reserves end Eereige Exehepge Ass istepee, Research report 26, Washington, D. C: International Food Policy Research Institute, 1981. Ministere du Dévelopement Rural. Pl "nv i me i en ire. Dakar, 1977. 291 Ministere du Dévelopement Rural. Le Negyelle Eeljtigge Agrjeele. Dakar, 1984. ........ . fitpge du Seetegr Agrjeele - Elem Cereelier. Dakar, May, 1986. Ministere du Plan et de la Cooperation. Ero rojet ee Se ptieme Plan de WWW Dakar. 1985. Morris, M. "The Cereals Subsector in the Senegal River Valley. A Marketing Policy Analysis," Ph.D. diss., Michigan State University, 1986. Ndiame, F. "Par quelle céréale remplacer le riz? Quelques problemes du mais en Basse Casamance." Conference on the Dynamics of Cereals Consumption and Production in West Africa, organized by the International Food Policy Research Institute, Dakar, July 15-17, 1987. Newbery, D. and J. Stiglitz. The Theemy ef Commodity Price Stebilization. Oxford: Clarendon Press, 1981. Ohlin, B. Interregional and Internatjenel Trede. Cambridge, Mass.: Harvard University Press, 1933. Ouedraogo, I. and 0. Ndoye. ve Polit' e i e vu ar les predeetegrs et les commereants du Bassin Arachidjer en 1285-86. Note d’Information 86-6, Bureau d’Analyses Macro-Economiques, Dakar: Institut Sénégalais de Recherche Agricole, 1986. -------- Réfl xion ur la lit n ricole eans la perspeetjve de "L’Apres- Barrage" au Sénégel. Note presented at the Seminar on the New Agricultural Policy organized by the ”Centre de Recherches Economiques Appliquées" in Dakar, May 7-10, 1986, Bureau d’Analyses Macro-Economiques, Dakar: Institut Sénégalais de Recherche Agricole, 1986. Pearson, S, D. Stryker and C. Humphreys. Rj ce jg West Afrjce. Policy end Economics. Stanford, CA: Stanford University Press, 1981. Pearson, S. T. Josling and W. Falcon. F od e -Relian n Food Self-ngfi1ejeney: Evaleetjgg the Eeljsy thjegs. Food Research Institute, Stanford University, 1986. Pomeroy, J. "Uncertainty and International Trade,” in Interpetienal Eeonomic Policy. Theory ane Lvidenee. R. Dornbusch and J. Frankel (ed.), Chapter 4, Baltimore, MD: The Johns Hopkins University Press, 1979. 292 RAPID. né l: ffet de F ct m hi u s r le Developpement Soeiel et Eeenemigge. Resources for the Awareness of Population Impacts on Development, Washington, D.C.: The Futures Group, 1984. Ricardo, D. lhe Eripeiples of Eelitjeel Eeenemy epe lexetiep. London: J. Murray, 1821. Reutlinger, S. Eeee leseegrjty: Megmjtgee epd Remedies. Staff Working Paper 267, Washington, D. C: World Bank, 1977. -------- WWW Agriculture and Rural Development Department, Washington, D. C. World Bank, 1984. Reutlinger, S. and K. Knapp. in F i it untri 5 Staff Working Paper 393, Washington, D. C: World Bank, 1980. Reutlinger, S. and M. Selowsky. Melpgtrjtiem epe Eeverty: Megnitude and Peljey. Baltimore, MD: The Johns Hopkins University Press, 1976. Ross, C. A Villa vel S ud P d r r n Tr a tion in Rural Senegal. Discussion Paper 81, Ann Arbor, Mi.: Center for Research on Economic Development, 1979. Ross, C. "Grain Demand and Consumer Preferences in Senegal," Feed Beliey 5-4 (August 1980): 273-81. Sarris, A. F... : ri nu ‘- ° ' co- . . e-'- nder stk in Egypt Working Paper 249, California Agricultural Experiment Station, University of California, 1983. Schmitz, A. omm d' Pri a ' i h r t Applieetion. Staff Working Paper 668, Washington, D. C: World Bank, 1984. Schydlowsky, D. ”A Policymaker’s Guide to Comparative Advantage," World Development 12-4 (1984): 439-49. Secretariat d’Etat Chargé de la Décentralisation. Sepéme Nationel g’Aménagement eu Territeire. Dakar, 1984. Siamwalla A. and S. Haykin. The W rl Ri e a t° r nduct ene Perfermenee. Research Report 39, Washington, D.C.: International Food Policy Research Institute, 1983. SONED-SEMA. Modéli ation des ri a ric e e in remiere phase. Methedolegie et donnees ge pese. Dakar, 1979. 293 SONED-SEMA. Modélisatien des prix agricoles. Mise a jour eu rapport de fjn ee premiere phese et rappept de fje d’apelyse. Dakar, 1980. -------- d t on e rix ric l Ra in . m . Les tests en yr ei e greneeur. Dakar, 1981. Staatz, J. Eeee Seeupjty jn Afpiee; flee ee We Cheese eer Beseereh Is ssees? Food Security Project Working Paper 1, Michigan State University, 1984. Stryker, D. Feoe Secerjty, Self:seffici and 0 mi rowth in the Seheljae geuetpies ef West Afpjee. Washington, D. C: United States Agency for International Development, preliminary draft, 1978. Sukhatme, D. "Incidence of Protein Deficiency in Relation to Different Diets in India," British eeernal ef Nutritjen 24 (1970): 477-87. Thompson, S. and 6. Bond. "Offshore Commodity Hedging Under Floating Exchange Rates," m ric n rm 1 r c nomi 69-1 (February 1987): 46-55. Torrens, R. nEssey en the Eyte epn nel tern Trede. London: J. Hatchard, 1815. -------- . "The Economists Refuted", in Th r' ra ti al Qeeyetion of Sir Robert Peel’s Aet ef a§4§ Esplejeee end Defeeded. R 3 edition, 1958. . Torrens (ed.), London: Longmans, Tuluy, H. C m r ive R r C nce ti i n e Ri e Preeeetien. Food Research Institute, Stanford University, 1978. Tuluy, H. Comparative Costs and Incentives,in Senegalese Rice Predeetjon .Food Research Institute, Stanford University, 1979. United Nations Economic and Social Council. A gomperetive Study and Anelysis of Qn-Going Feee Security Pregpams in Africa, Economic Commission for Africa, Item 6, Provisional Agenda, 8th Meeting of the Technical Committee of the WHOLE, March 23, 1987. Valdes, A., ed. Feod Security for Developing tountries. Boulder, CO: Westview Press, 1981. Vollrath, T. namics f m ar tiv v d t R i ance to Free Trade. Report 214, Economic Research Service, Washington, D.C.: United States Department of Agriculture, 1985. 294 Whee Rhee, Y., B. Ross-Larson and G. Pursell. Keree’s Competitive Edge. Meintaining the Entry into World Markets. Baltimore, MD: The Johns Hopkins University Press, 1984. WHO. Energy ene Epeteie Reeeiremepts. Paper of a Joint FAO/WHO/UNU Expert Consultation, Technical Report Series No. 724, Geneva, 1985. World Bank. Werle Develepment Report 1272. Washington, D.C.: Oxford University Press, 1979. -------- . Werld Development Report 1922. Washington, D.C.: Oxford University Press, 1982. -------- . Werld Development Report 1924. Washington, D.C.: Oxford University Press, 1984. -------- . Helf—Yeer Revision of Commeeity Priee Forecasts and Qeerterly Reviee of Commoejty Mertets for December 1986. Memorandum, Washington, D.C., February 10, 1987a. -------- . World Development Repert 1287. Washington, D.C.: Oxford University Press, 1987b. -------- . Demmedity Trade end Price Trenes. Baltimore: Johns Hopkins University Press, 1987c. Yung, J.M. tesFiliere Mil dees le Dessin Areehieier au Sénegel. Paris: Ministére des Relations Extérieures, Coopération et Développement, 1984. Zuckerman, P. "Different Smallholder Types and Their Development Needs," Deurnel of Agrieulteral Eeoeemies 28 (1977): 119-27. -------- . "Simulating the Decision Making Process of a Nigerian Smallholder," anadi Jo rnal A ri ltural onomics 27 (1979): 17-26. Zusman, P. and A. Amiad. "Simulation. A Tool for Farm Planning Under Conditions of Weather Uncertainty,” Qee:eel_ef_£erm_£eeeemies 47 (1965): 574-94.