.———V - —_ W d‘vl—‘l_~ .“a‘wo 40-» wgo...‘*“-.--" - »'-"---~~~O— -4...~c-. 7-- .o—. 7-..“. -“_ _ ANALYSIS OF SUPPLY AND DEMAND 0F 1. l ' AND SORGHUM IN UPPER-YDLTA Thesis for the Degree of M; S.- MICHIGAN STATE UNIVERSITY H. PAMATEBA DIENDERE 1984 I.'0~I ' 0 I u C .‘ufdfoglflzl)!}.'I)DQ;‘.‘.!\O.. I . o . MILLET ...... MICH. STATE UNIV. N N PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINE-3 return on or before date due. DATE DUE MTE DUE DATE DUE MICH. STATE UNIV. W— . 0 NW,“ “ S .4 itiv JR} J 1/98 chlRC/DctoDmpfi-p.“ ABSTRACT ANALYSIS OF SUPPLY AND DEMAND 0F MILLET AND SORGHUM IN UPPER-VOLTA By H. Pamateba Diendere Upper Volta desires to be self-sufficient in cereal production; but, since independence it has relied upon imports and food aid to meet its food grain requirements of mostly millet and sorghum. This study's main objective is to analyze the millet and sorghum sector of Upper Volta, and to identify and estimate elasticities of the most important variables affecting its supply and demand. To achieve this study's objective, an econometric model was speci- fied and estimated. Then the equations were combined and solved simul- taneously to make forecasts of millet and sorghum production and demand under different alternatives. The analysis finds that: (1) annual increases in millet and sorghum production within the range of data did not keep pace with population growth; (2) increases in millet and sorghum production are hampered by a low level use of fertilizer, poor rainfall, and lack of suitable land; (3) millet and sorghum supply and demand are only slightly responsive to prices: the own price elasticities of supply and demand are very low; and (4) self-sufficiency in millet and sorghum by l988 would require a 5 percent increase in yields. ANALYSIS OF SUPPLY AND DEMAND 0F MILLET AND SORGHUM IN UPPER-VOLTA BY H. Pamateba Diendere A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Economics 1984 To my wife Francoise in appreciation for her love ii ACKNOWLEDGEMENTS I appreciate Dr. L. Robison, my major advisor, for his friendship. guidance and encouragement throughout this study. Dr. Robison has blue- penciled my manuscript at several stages. At the beginning of my pro- ject, I was scared to death with his blue pencil. By the end, I was un- happy for not having enough blue marks in my drafts. I accepted all the changes he recommended. Special thanks are also due to Dr. S. Thompson for his valuable suggestions with the manuscript. Dr. B. Brown for his participation on my oral examination committee, Dr. M. Koelling and P. Bonnar for their help in the revision of the manuscript, and Dr. L. V. Manderscheid for his assistance during my graduate studies. Special acknowledgement also to Ms. Debbie Greer who typed the first draft of the manuscript and also the final copy. Finally, special thanks to my parents Nabyoure and Paougba for their love and encouragement. And most of all, I express my appreciation to my loving wife Francoise for her support and tolerant attitude. Her contributions to the success of this work cannot be overstated. iii TABLE OF CONTENTS CHAPTER I. INTRODUCTION .................... 1.l Importance of the Study ................ 1.2 Objectives of the Study ................ l.3 Plan of the Study ................... CHAPTER II. THE AGRICULTURAL SECTOR .............. 2.1 Physical Characteristics of Upper Volta ........ 2.1.l Location and Climate .............. 2.1.2 Population ................... 2.2 Socio-Economic Characteristics ............ 2.2.1 The Government Policy Formulation ....... 2.2.2 Income Distribution .............. 2.2.3 Land Tenure and Use .............. 2.3 Role of Agriculture in the National Economy ...... 2.3.1 Contribution to the Gross National Product .................... 2.3.2 Agriculture in Exports ............. 2.3.3 Agriculture in Imports ............. CHAPTER III. PRODUCTION AND MARKETING ............. 3.1 Production ...................... 3.l.l Evolution of the Voltan's Self-Sufficiency Level in Staple Food .............. 3.l.2 Actual Supply and Demand Levels of Millet and Sorghum ............... 3.1.3 The Main Determinants of Millet/Sorghum Production ................... 3.l.3.l Acreage ................ .3.2 Yields ................ 3.l.4 l Fertilizer .............. 2 Pesticides .............. .3 Seeds ................. o rces of Credit ............... 5 1 Institutional Money Lenders ...... 3.l.5.l.l Public Institutions . . . . 3.1.5.1.2 Interest Rates ....... 3.l.5.l.3 Private Institutions. . . . 3.1.5.2 Individual Money Lenders ....... 3.1.6 Government Extension and Research on Millet and Sorghum ............... 3.1.6.1 Government Organizations and the Nature of Research in Millet/ Sorghum ................ 3.1.5 iv v 0.! CD dO‘OOO’iODCflU'IU'I (JON—I'd —l-—l .p._a_a .a—J mm 3.1.6.1.1 IRAT ............ 3.1.6.1.2 ICRISAT .......... 3.2 Marketing of Millet and Sorghum ............ 3.2.l Marketing Channels ............... 3.2.2.1 The Public Marketing Circuit ..... 3.2.1.2 The Private Marketing Channels . . . . 2 Marketing Services ............... 3.2.2.1 Transportation ............ 3 Storage .................... Risk Bearing .................. 5 Pricing Policy ................. CHAPTER V. 4.1 So EARCH PROCEDURE ................. rces and Limitations of Data ............ 1 Production Data ................ 2 Price Data ................... 3 Population and Consumption Data ........ .4 Weather Data .................. o l 2. 2. 2. 2. R u .1. .1. .L .1 4.2 th d of Analysis .................. .2 E u l. 1. l 1 a 2. 2. hgbih-bf Estimation Techniques ............. CHAPTER V 5.1 MPIRICAL RESULTS .................. Eq ations Formulation and Estimation ......... 5.1 The Demand Equation .............. 5. 2 The Supply Model ................ 5. .3 The General Model ............... 5. .4 Estimation ................... 5.2 Ev luation ...................... 5.1 Economic Evaluation ofthe Model ........ 5. 2 Statistical Evaluation ............. 5.2.3 Econometric Evaluation ............. 5 3 Elasticities ..................... 5.4 Summary of Empirical Results ........ . . . . . CHAPTER VI. SUPPLY AND DEMAND PROJECTION OF MILLET/SORGHUM. . . 6.1 General Methodology .................. 6.2 Demand Forecast Using the Regression Model ...... 6.2.1 Population Projection ............. 6.2.2 Consumption Projection ............. 6.2.3 Income Projection ............... 6.2.4 Projection of Prices .............. Demand Projection Using the FAO-OECD Method ...... Forecast of Supply .................. Supply and Demand Balances .............. CHAPTER VII. SUMMARY AND POLICY STATEMENTS ........... 050505 U190) CHAPTER VIII. CONCLUSION AND NEEDS FOR FURTHER RESEARCH . . . . 8.1 Conclusion ...................... 8.2 Further Research Needs ................ Page APPENDIX I. DATA USED IN THE STUDY ............... 100 APPENDIX II. ACRONYMS AND CURRENCY EQUIVALENTS ......... 103 APPENDIX III. OVERTIME REGIONAL EVOLUTION OF CEREALS PRODUCTION (1973-1980/1981) ................... 105 BIBLIOGRAPHY .......................... 106 vi 0'5 03 05 U1 0) 0d 0 LIST OF TABLES Agriculture's Importance in the Economy of Upper Volta, 1960-1979 ................. Agriculture's Contribution to Export Earnings, 1969-1979 (millions of current CFA) ........... Share of Agriculture in Imports (values in millions CFA) ...................... Evolution of the Voltan Level of Self-Sufficiency in Staple Foods ..................... Regional Characteristics of Cereals Production and Demand (Campaign 1978) ............... Relative Importance of Millet/Sorghum in Cereal Production (1978-1979-1980 Average) ........... Ofnacer Grain Purchases, 1978/79 ............ Summary of Equations Used in the Study ......... Projection of Prices, Consumption and Fertilizer Use . . Projection of Acreage and Yield ............. Supply and Demand Balances Using the Regression Model (1000 t) ..................... Supply and Demand Balances Using the FAO Model (1000 tons). . . . ................... Millet and Sorghum Production UnderAlternative Rainfall Price, and Yield Scenarios (Simulation Results) ........................ vii 12 13 15 17 26 27 43 67 86 87 88 89 Figure 0) O U" (A) (A) (A) on N C O O 0 (11 0" 0'1 0" 01 O 0 O O I U" h 0) N LIST OF FIGURES Actual Land Use in Upper Volta ............. Rainfall in Upper Volta ................. Yield Area and Production of Millet and Sorghum ..... Millet and Sorghum Productions ............. Population, and Millet-Sorghum Production ........ Official Producer, Consumer Prices and Market Retail Prices ...................... Areas Cultivated .................... Flow Chart of Legal Grain Marketing Channel in Upper Volta ....................... Flow Chart of Traditional Grain Marketing Channel in Upper Volta ..................... Actual and Estimated Acreage of Millet ......... Actual and Estimated Acreage of Sorghum ......... Actual and Estimated Yield of Sorghum .......... Actual and Estimated Yield of Millet .......... Actual and Estimated Market Price of Millet and Sorghum ......................... Actual and Estimated Production of Millet and Sorghum ......................... viii Page 7 8 20 21 23 24 30 42 45 68 69 7O 71 72 73 CHAPTER I INTRODUCTION 1.1 Importance of the Study Upper Volta desires to be self-sufficient in the production of cereals. But ever since the favorable harvest of 1969/70, the imbalance between the rate of growth of agricultural production and Upper Volta's population growth rate has forced the country to rely increasingly upon imports and food aid to meet its food grain requirements of mostly millet and sorghum. This import requirement is in contrast to Upper Volta's objective of becoming self-sufficient in staple foods. For some policy makers, the main difficulties to overcome in order to reach self-sufficiency in food crops production are: low soil fertility, unimproved or inappropriate technology, ineffective extension services, wide spread use of low yielding varieties by farmers, and inappropriate incentives created by government policies. To aid in the formulation and implementation of appropriate policies, programs and projects designed to obtain food self-sufficiency in Upper Volta, this paper proposes to answer several questions related to pro- duction. The questions to be answered include: 1) What regions are the country's best millet and sorghum producers? 2) What factors explain why some areas are self-sufficient in food? 3) What proportion of all cereals produced in Upper Volta is millet and sorghum? 4) What factors limit increased productivity in cereals? It can be argued that at T‘g. ".., 2 planting time, most farmers compare the advantage and costs of produc- ing cash versus food crops to make planting decisions. This study asks: to what degree is this statement true for Voltan farmers whose produc- tion is more oriented toward home consumption and who are faced year after year with the risk of drought and low producer prices imposed by the government? In other words, it asks what variables do Voltan farmers respond to in making their planting decisions? Also of interest is the nature of farmers' responses to changes in prices. What are factors to be taken into consideration in the evaluation of millet and sorghum prices and how do these influence the production and marketing activities of farmers? Is self-sufficiency in staple food a realistic target? If so, when can it be achieved? 1.2 Objectives of the Study In the above section, many important questions were raised about Voltan agriculture. These questions are complex and in some cases the answers require data not available. As a result, some questions may re- main unanswered empirically. Otherwise, this study intends to answer all questions raised except where the data do not permit. The following are the objectives of this study: 1. To analyze the actual situation prevailing in the millet and sorghum industry. 2. To determine what relevant variables affect supply and demand, and to use these discovered variables and their interrelationships for policy analysis, formulation and implementation. 3. To estimate the supply demand elasticities of the respective variables identified in the previous objective. 4. To predict future supply and demand conditions under different alternative scenarios including: a. alternative growth rate of GNP; b. alternative technological package; c. alternative prices. l.3 Plan of the Study The study is quantitative as well as qualitative in nature. The reason for this is that some of the questions raised, although impor- tant, could not be verified on a quantitative basis, due to lack of data. In such cases, all available information has been pooled together in an attempt to give some answers based from qualitative analysis. <3 1w ‘ The remainder of this study is divided into six parts. Chapter two 5;);155 provides a background information about Upper Volta and the role of ff“' V1 agriculture in the national economy. Chapter three reviews the present ifgij‘fi food supply (production) and marketing situation during the period :- :5; ' 1961-1981. It gives a description of the main factors that determine millet and sorghum production, and the effects of official marketing and pricing policies on the grain subsector. In chapter four, a dis- cussion of the sources and limitations of the data used in the study is presented in addition to providing an analytical framework for the evaluation of the empirical results presented in the following chapter. It emphasizes the need to improve the statistical data relating to Voltan agriculture to help make coherent cereal policies for the country. Chapter five is divided into two parts. The first part identifies vari- ables needed to estimate demand, price, yield and acreage fOr millet and sorghum. The second part of chapter five presents the estimated results and simulations as well as and the estimated elasticity measures of interest in millet and sorghum production equations. Chapter six projects cereal supplies and requirements for the period 1982 to 1990, using the equations described in chapter five. These forecasts are based on a number of assumptions that are discussed throughout the chapter. The two last chapters, chapter seven and eight, conclude the study with a summary and policy statements and some guidelines for future research. They identify some relevant questions whose correct answers would provide more information needed in diagnosing the problems of the agriculture subsector. The data used in this study are presented in Appendix 1. CHAPTER II THE AGRICULTURAL SECTOR OF UPPER VOLTA This chapter presents a brief review of background information about Upper Volta and the role of agriculture in the Voltan economy. 2.1 Physical Characteristics of Upper Volta 2.1.1 Location and Climate Upper Volta is a small country located in the West Sahelian part of Africa. It is a landlocked country that lies above the equator between the latitude of 100 and 150 North and longitude 6°W and 2°East. It is bounded by Mali to the North and West, by Niger to the East and by four coastal countries, Ivory Coast, Ghana, Togo, and Benin to the South. Upper Volta occupies a surface area of 274,000 km2 (106,560 miz) and is divided into three major zones of vegetation each depicting a specific pattern of rainfall: l) the Sahelian zone; 2) the Sudano- Guinean zone; and 3) the Sudano-Sahelian zone. The Sahelian zone, which covers the northern part of the country, receives less than 600 mm of rainfall per year. The Sudano-Guinean zone in the Centre, the largest climatic zone, receives between 600 and 900 mm rainfall. The Sudano Sahelian zone, the most productive zone receives over 1000 mm each year. More than 80% of the Voltaic farmers live in these last two agroclimatic zones. Together these two regions represent 89 percent of the 8,915,000 ha considered to be arable. 5 There are two distinct seasons in the vegetation zones of Upper Volta, a short rainy season from June to October and a long dry season from November to May. Duration of the rainy season varies from one zone to another. However, its average duration for the whole country is about 3.5 months. 2.1.2 Population The official 1982 Upper Volta population estimate is 6.8 million people and is increasing at a rate of 2.6 percent annually. However, not all of Upper Volta's population live in the country. According to a World Bank estimate, as many as .7 million persons live and work in the neighboring Coastal countries of Ghana, Ivory Coast, T090 and Benin. Upper Volta's imigration rate, the highest among all Sahelian 1 countries, is estimated by the World Bank 1960-1980. to be 1.7% during the period The population living in Upper Volta is located mostly in rural areas with only 8 percent living in urban areas. Population density varies widely from region to region. The Mossi Plateau is the most 2 2 populated zone. Its density averages 40 persons/km against 12/km for the rest of the country. The total average density is 23 persons/kmz. 2.2 Socio-Economic Characteristics 2.2.1 The Government Policy Formulation Upper Volta is divided into 10 main administrative regions in which are located 11 regional development offices (0RD's). The 0RDs are 1World Bank: Upper Volta agricultural issues study; WB report 3296 UV Oct. 1982. U) x L (U Q. 05 - 1117-7-33 AK}: -.nt: '5 .52122- 53% m ff!ttt:t:0.°£oz-I_ ‘- CU J‘J‘fffl‘frrAN-l- ._' O 0 9-: S- (U 'U 0 x 0r- U '4- C 0r- 44 U) U) 'U U) OJ --------- (Us-- 'l— ----- . .C.. ,/ ----- .1; '. // .. t /l/ —-—-—— m o. O s. 0 OJ L) (D c o. d) O +9 S- m u 0r- 0).: .o m 3 cu mu —.-. :0°. _-_- no 0.-.. i... Z? O U >.' a: O > H "Etude sur les engrais en Afrique de l'Ouest" Vol. 4 UPPER VOLTA, 1976. Adapted from IFDC Source Figure 2.1 Actual Land Use in Upper Volta uppo> Lona: cw p—wmcvmm N.~ meamwm .onmp . mung: e .Po> =umm=o.P mu macwsm< cm mmmgucw mmp Lam muzpm= Honda seem umugmu< "mugaom “one p.2ucou-um.m ugh atsa=z _. 88 5:28 .23 2 Sac mcwmomumuauz 9:. 33.2.5.2...5 o 8% 32838 2» i255 .. 3% 533 2:26.23“. h 310 0058 Us... 23.....3 0 Eco poLucouévLoz of. . 95! m aaupo> x00; 93 mo ago .5552 v $9.33 ea 95 935:? m 88 3228-3»: 2: .5 fin...“ “ 8.3 :55 J responsible for all development programs in the regions and report their activities to the General Secretary of the Ministry of Rural Development. The government's policy toward agriculture is centralized at the Ministry of Rural Development which is implemented at the 0RDs and other levels. To date, the main policy objectives of the government toward agriculture have been: 1) to integrate rainfed agriculture with live- stock production; 2) to increase the development of swampland and irri- gated agriculture; 3) to relocate the population from infertile soils of the overpopulated Central Plateau to the west and southwest to the new and fertile lands of the Volta Valley; 4) to achieve self-sufficiency in food crops by increasing cereal production; and 5) to encourage agricultural diversification to increase export opportunities. 2.2.2 Income Distribution Income is unequally distributed in the country in particular between “ farm and nonfarm workers. Income in agriculture is generally low. It has been estimated that revenue of those working in the agricultural sector is 10 percent lower than that of nonagricultural workers. Accord- ing to the FAO the average per capita revenue grew at a rate of 4 percent annually between 1960 and 1975 from 8,463 CFA to 15,551 CFA respectively (US $34 and US $73)]. Within the same period, the nonagricultural per capita income increased at a rate of 7.5 percent, increasing from 29,174 CFA in 1960 to 76,090 CFA in 1970 (US $31 and US $355 equivalent respec- tively) while per capita average agricultural revenue grew only at a rate of .4 percent, or ten times lower than the national average. 1See Appendix 2 for exchange and information. 10 2.2.3 Land Tenure and Use There are no established official land tenure rights. But there are customary rules of land tenure expressed through families and lineage. The rights on a piece of land are established by cultivating the same land year after year, after receiving the land from traditional authori- ties such as the Tengsoba (chief of lands) in the Mossi tribe. The land can be passed from father to son, and, in a lesser degree. from son to F grandson. The household chief has all rights to lands which are culti- vated by his household members. He can decide to lease it or donate its control to friends or family. A lifetime right to the land may be allo- cated to the user but he cannot sell or trade it. In fact, there is no market for land in Upper Volta except in some towns for irrigated lands. In these cases, administrative rules of land tenure allowing private if ownership prevail. In many situations, this is not done without con- L; g flicts since the sets of rules governing the land control are directed .5 toward essentially the same public. S. 5” One can characterize Voltaic agriculture as land scarce. Not only is the total arable land area (33% of total area) small relative to the population, but about 5% of this area is not occupied as well because of endemic diseases and lack of drinking water. This raises the pressure on land, mainly in the Central region, a region densely populated which forces people to settle and cultivate on marginal lands. This fact helps explain the low productivity of agriculture and the high imigra- tion rate which characterize the country. 11 2.3 Role of Agriculture in the National Economy 2.3.1 Contribution to the Gross National Product Despite all the structural problems (insufficient rainfall, ineffi- cient marketing systems), the agricultural sector contributes up to 40 percent of the gross domestic product (see Table 2.1). According to the World Bank, the largest share. that is, 25% of the total agricul- tural contribution is supplied by crop cultivation and the remainder y (15%) by livestock. Table 2.1 shows that from 1965 to 1979, the rela- tive contribution of the entire sector has consistently declined as a proportion of GDP, from 59 percent in 1965 to 38 percent in 1979. Dur- ing the same period, both the GNP and the GDP steadily increased but with a relatively low per annual growth rate (4% in real terms). This might be an indication of economic growth provided by other sectors in I. the economy whose contribution is growing at a much faster rate. 2.3.2 Agriculture in Exports Table 2.1 shows that the agricultural sector has contributed a large share to the total export revenue. About 33 to 50 percent of all export revenues come from the sale of livestock. This is not surprising because, after all, Upper Volta is primarily an agricultural country with more than 90 percent of the population engaged in farming. But in 1960, agricultural exports contributed 15 percent of the total export revenue. In 1977, agricultural exports were 60 percent of total exports, in percentage terms, four times its 1960 level. Accounting for growth in agricultural exports have been the steady growth since 1960 of cotton and groundnuts. In 1960, cotton's share of total export revenue was 2 percent. In 1977, cotton earned 41 percent 12 ) TABLE 2.l Agriculture's Importance in the Economy of Upper Volta, 1960-1979 Contri- bution of Agri- GNPa GDPb culture (million (million to GDP Year CFA) CFA), ,_(%) 1965 71,800 61,700 59 1966 74,300 64,700 60 ,1 1967 77,300 70,000 53 “ 1968 79,500 72,000 50 1969 86,700 78,400 51 1970 89,400 79,300 48 1971 99,200 87,300 47 1972 103,600 89,800 46 1973 103,500 88,500 43 1974 122,100 104,500 45 1975 134,600 114,600 43 1976 146,900 122,400 39 1977 178,600 147,300 40 1978 199,400 165,200 38 1979 222,900 184,800 38 aGNP in current market prices. bGDP in current factor costs. ;: Ft Source: World Bank: Upper Volta,.AgLicultural_Ls§uesJitudy, Report #3296 UV, October 1982, table 1.1, p. 198. 13 TABLE 2.2 Agriculture's Contribution to Export Earnings, 1969- 1979 (millions of current CFA)* Value Contribu- Total of tion of Exports Agri- Agricul- Value culture ture in Exports Exports (CFA) (CFA) (%) 1960 1056 160 15 1961 812 105 13 1962 1654 312 19 1963 2769 362 13 1964 3087 497 16 1965 3432 782 23 1966 3869 888 23 1967 4429 1514 34 1968 5290 2063 39 1969 5329 2637 49 1970 5055 2894 57 1971 4408 2123 48 1972 5141 2169 42 1973 5598 2497 45 1974 8702 4519 52 1975 9368 5014 54 1976 12690 10004 79 1977 13614 8208 60 1978 9524 4750 50 1979 16240 9356 58 *Values are FOB in current prices. Source: World Bank: Upper Volta: Agricultural Issues Study, Report No. 3296, table 1.6 p. 203. 14 of all export revenues. Cereals are mainly consumed domestically and represent a small portion of agricultural exports. They are the small- est component of all agricultural export revenues. 2.3.3 Agriculture in Imports Upper Volta has been an importer of some few food items, mainly cereals and flour. The share of food grain imports including food aid, to total imports increased from 2 percent in 1960 to 12 percent in 1978. It represents the main component of all agricultural imports. The data indicate that substantial imports of grains started in 1975 following the drought period. 15 TABLE 2.3 Share of Agriculture in Imports (values in millions CFA) Portion of Agri- Culture All Imports Agri- in Total Cereal culture Total Year Imports Imports Imports_ Imports f 1960 2715 - - - 1961 6992 153 557 8 1962 8551 339 742 9 1963 9382 333 748 8 1964 9484 394 903 9 1965 9169 456 916 10 1966 9293 669 1067 11 1967 8970 600 1024 11 1968 10119 458 1100 11 1969 12450 834 1412 11 r"' 1 1970 12963 693 1259 10 I“ ”” 7 1971 13890 712 828 6 y: x- .1 1972 15312 677 1635 11 'é': 1973 21690 1260 2034 9 ;; g 1974 34664 4383 5638 16 :f “ 1975 32386 2239 3308 10 ;';§”, 1976 34423 2115 2369 7 f_;; 1977 51357 3989 4404 9 f v f 1978 51083 6057 6897 14 1979 63916 6285 7810 15 Source: WB, Upper Volta, Agricultural Issues Study, Report No. 3296 UV October 1982, pg. 204. CILSS/CTUB du Sahel, DEvelopment des Cultures pluviales en Haute Volta, Sept. 1982, pg. 60. CHAPTER III PRODUCTION AND MARKETING This chapter presents a brief review of the millet and sorghum industry in Upper Volta. A detailed discussion of the nature of millet and sorghum production, marketing, storage, and evaluation of some of the government policies in the cereals subsector are the main parts of this review. 3.1 Production 3.1.1 Evolution of the Voltan's Self-Sufficiency Level in Staple Food Table 3.1 provides three indicators of the Voltan level of self- sufficiency in cereals. The first in column 6, is the rate of self- sufficiency in millet and sorghum. It is the ratio of production to net availability of millet and sorghum for consumption expressed in percent terms. This ratio varies substantially from year to year. The highest level was in 1965 (98.6%) and the lowest in 1974 (90.8%) during the last year of the drought period (1972-74). The data show that the ratio dur- ing the 1960's averaged over 97.5 percent; whereas, in the last ten years, including the drought period, it averaged 94.6 percent which re- presents a decline of 3% over the previous decade. The second indicator, the rate of self-sufficiency in all cereals shown in column 5 of Table 3.1 indicates that over the past 20 years, the average rate of self-sufficiency in all cereals has been less than 16 17 .mmrgmm xoonmumgh ucm xoongmm> cowgozuogm om mumumc o.em~ m.mm m.~m coo. empp mmm mo—p Pm o.pmp m.~m N.Fm m~F_ NGNF peep mn—p om m.mmp ~.~m m.Pm ooPp mmmp muop ncpp as m.~np m.mm “.mm mmop mopp com mmop mu n.mo~ o.mm “.mm mnpp mump mmpp ompp mu m.omp m.nm m.mo wmo— mm—P coop wwop ms m.uo~ m.nm m.oo «opp momp muop oo_P m“ ~.mcp w.om u.mm mom mom emu pr em m.emp m.mm P.mo Pmo Pmop NFm pom ms m.mm~ o.mm m.¢m mpop NFFP mum pmop Nu ~.mo~ m.nm m.mm mmm mmo— mow mam .m>m mumomu ~.Nm— m.om m.mo Fsm coop Fem m_o— Pu m.mm~ m.om w.¢m omm noop mum Fpop o“ m.m- m.nm ~.om cum mop— mow poop mo m.N—~ m.~m m.mm emm mmop com mmop we N.NFN N.em F.mm omm mmop cam omop no ¢.m_~ o.~m m.¢m mom “mop 0mm Npop mm m.mm~ m.wm w.om mmop mmm_ mmop nwpp mo m.oo~ F.mm m.mm Pan mmm mun Pom em w.mmp m.wm e.em an mum own cum mo m.~mp m.om m.~m mum mmn mom pom Nm\Pmmp Auxq .INNV «xv .Ah coopv «H ooo_v Ah ooo_v A» ooo_v mcam> upwamu 8 Foo m Foo cowu853m cowpasam cowuusc cowuuzc so; N Foo F Foo Icou com Icoo Com Ice; Ice; upmmemo Eszmcomx mpmmgmu cw apanVm>< mpnmppm>< Escugom Pamgmo we cow» pm—sz c? zucmwopm Ezcmgom ~mmgmu vcu rosuoga xocwvowmwzm Imsmuepwm new umppvz epmm Lo mpmm So mama umppvz m m m . .w m m P coca mpnapm cw sucowurteam-tpom to Fw>ms cappo> 6;“ Lo eomp=.o>u P.m msmmzo: .mpmasox .aomsouzox .mzmx .mmcmum> .m one Go ago as» mucmcmgaeou Agmucmuv cowmmm ngpcmu maps .Esccm Lma ox o~_ ma on AonPV owm mm mgamvm upwamu emu cowuaszmzou mpmmemu .FFP .n m mpnmh ”=mmrmnmp Fozccm ugoagmm=uhmm.~ mu omo sage new "mm .ma m mpnmh Ammmpv oumo\mmsfiu sate mom .ma m a_nah Awkm_v o uouH Eoge gospsm mgp an copvasou mama "mugaom mom mmop mwom memw o.mp meow Peach 44 Pm om_ 042 o.k cam Fmemm 45 mo, map o_N o.o one “mam com ooe m_m owe m.4_ Nmmp puma: omm FNm. mmpp mFm_ o.mm “Fem weapcao aw=03 coonwflmcoa ooon Am; oooaqr,fla; ooo_v «max\ Amyn. meommmm ceaeao eowpuauoca mme< ame< Davy ooo_v Pompom Pmmgmu mpmmgmu umgm>+pp=u auwmcmo cowpmpsaom cowuap , Isaoa Amnmp cmvmasmuv cameos new covpozuoga mpmmgou mo muvumvgmuumgmgu pccovmmm N.m m4m xumpm mmm mm ommpmm omwmmv mm mommne opon~m emu: .F Q»: m 23.50 ma: m meLmu m Pmewu qmwz m Fame-mu page» & Payee Peach a pmuop mm ma: mm ma: mcowamm Aa;\quw I~Hooo~q cowpusuoca Am; ooon amc< cme> Ammmgm>< ommpumnmprmnmpv covuuauoga megmo cw Ezzmgom\¢m—sz yo mucwusoaem o>vpmpmm m.m m4m<fi 28 Farmers' allocation of land to millet/sorghum production is also affected by the availability of family labor, rainfall, plant disease, insect attack and more importantly, by the quality of the soils. There are no reliable estimates of the quantities of labor input used in the production of millet and sorghum. According to a World Bank document, traditional cultivation of sorghum requires 70 man days per ha in the center, 85 and 75 man days per ha in the southwest and east, respectively. In general, labor is not a limiting factor at the beginning of the season because seeding is not a very complicated job. As a result, large areas are quickly planted. However, total labor available is greatly reduced during weeding time and particularly during the first weedings. During that period, women labor is focused in their groundnut and okra fields. As a result, the weeding task in the millet and sor- E; . ghum fields is not done which depresses yields. j f 3.1.3.2 Yields There are many factors that determine yields: social, economic, and biotechnical. Biotechnical factors include soil, climate, date of planting, insect attacks, plant diseases, etc. The economic incentives affecting millet and sorghum yields are the official purchase price of millet/sorghum, i.e., its relative pro- fitability as compared to other food and cash crops. Also important is the relationship between the farmer's household and the village community, his choice of production technology, and the nature and the availability Iof'hired labor. All these factors play important roles in the farmer's .Yield. a'T-il. - 29 The average yield of millet/sorghum is estimated to be between 400 and 450 kg/ha (Euro Consult 1980). This appears quite low when compared to the 900 kg/ha yields of the research stations.“ Figures 3.1 and 3.2 and the time series data on yields on Appendix 1 indicate a declining trend for millet/sorghum yields with no change in total area planted (see Figure 3.5) reflecting the decreasing availability of arable land and the poor fertility of soils. The "Sous Commission de la Production Vegetale" (1980) reports that the low yields per hec- tare of millet/sorghum are mainly due to the unavailability of varieties which are well adapted to the different soils and Climates. But this is not the central issue. As long as the research centres and experi- ment stations are isolated from farmers' reach, and as long as the price of fertilizer, spray materials and traction equipment remain high, the resource allocations of farmers will remain the same with no im- provement in millet/sorghum production. Thus economic, social and in- stitutional factors are as important as the biotechnical factors in de- termining yields. Market difficulties, lack of adequate storage policies, and lack of coordination at the government level for establishing long term and durable cereal expansion policies are other elements that affect yields. 3.l.4 Supply of Agricultural Inputs 3.1.4.1 Fertilizer In Upper Volta, there are two kinds of fertilizers that can be applied to cereals: chemical fertilizers and natural rock phosphate. Chemical fertilizers are imported. Consequently, they are available only in the two major towns, Ouagadougou and Bobo-Dioulasso. The 3000 2000 1500 1000 01 L3 0 30 _"" COMBINED SORGHUM~MIILET GREG """ SORGHUM HREH "’ HILLET FREQ "“ TOIRL CERERL GREG "—-'COTICN 9ND FEHNUT RREQS 1/“\ , \ s a ‘ ._ I ~ I ‘ I‘. .............. \ ’ .‘ I I'rs ‘ ’ , ‘l O‘ ’, " I \\l ' \ II _________ \ P‘_ -—" ..... ------- f If \‘ /l \\ / \ ‘vv” " ‘gJ’ \./ \_ ,_--——-~—-'”'"""‘“—-\/"~\ /’T‘—'—"—'\ /— I I I I I I 1 I I I 2 4 8 8 10 12 I4 15 18 20 22 Figure 3.5 Areas Cultivated 31 natural rock phosphates are produced at Kodjari in the eastern part of the country. Until 1978, Sofitex (government service responsible for cotton production and marketing) was the sole importer and distributor of chemical fertilizers in the country. Fertilizers are brought into each cotton zone at the Sous Prefecture level and then, to the village level by the 0RD's extension agents. In 1978, Sosuhv (the government society responsible for sugar production) joined Sofitex in fertilizers impor- tation, therefore increasing its use and availability in the country. The World Bank (1982) estimates that total imports of fertilizers grew from 500 to 16,000 T between 1965 and 1978. While chemical fertilizers give the farmer the opportunity to increase yields by increasing the fertility of the soils, SDP/0RD (1980) argues that the use of chemical fertilizer alone is not beneficial and is even disastrous in the long term. The use of manufactured fertilizer for mi llet/sorghum production should be coupled with manure. They found that a combined use of fer- tilizer, manure, animal traction and erosion control can double the yields for sorghum. However, use of fertilizer on millet and sorghum fields is quite uncommon. Only 5% (mainly cotton fields) of the total cultivated lands I‘eceive chemical fertilizers at a low level usage (6 kg/ha for food cm13$). The high price of fertilizers, about 35,000 CFA per metric ton, Puts it out of reach for most cereal farmers. 3.1.4.2 Pesticides Pesticides (endrine) are provided by Sofitex and also by the §_e_r_‘- VJQLNationale des Semences. Farmers are familiar with this product. 1!: u ICC—Pa. .r M. fin‘wfl“ P 32 In fact, they use it in their houses to kill insects as well. Like in the case of fertilizer, the use of pesticides on millet/sorghum fields is not widespread. They are mainly used in cotton fields.- Their cost is high and there is a limited quantity of spraying material available, but their contribution to increased yields is well documented. However, the lack of extension agents to show farmers how to use pesticides and the low level of education of the farmers and the extension agent are . constraints to increased production through insecticide use. 3.1.4.3 Seeds The 0RD's agents are responsible for the distribution of millet and sorghum seeds to farmers. Improved seeds are provided to 0RD's by the Service National Semen- 5: F I cier based at Ouagadougou which is responsible for the production and marketing of some selected seeds. The remaining seed needs are placed under the responsibility of each 0RD to produce them. But the ORDs', faced with financial difficulties and lack of qualified personnel, have been unable to provide guaranteed high quality improved seeds to farmers. In addition, farmers could not get the product on time due to 0RD's bad management. Furthermore, although the price of improved seed is not too high in absolute terms, it was quite high in relative terms since the product is sold to farmers in bags of five kg, making it unaffordable by the average farmer. Considering the risk involved and Without proof that the new variety is better than the local ones, farmers are reluctant to adopt new im- IDroved seeds. According to the World Bank (1982), from 1975 to 1979, the annual average production of certified seeds amounted to 210/metric tons with only 12% reserved to millet and sorghum. (llffl 151414511 :- Paihufi.‘ b.» . Anna...- 33 3.1.5 Sources of Credit The purpose of this section is to describe how and where millet and sorghum farmers get most of their credit. Credit sources can be divided into two groups: institutional and individual money lenders. 3.1.5.1 Institutional Money Lenders 3.l.5.l.l Public Institutions Farmers receive most of their credit from the different 0RDs' and .A‘YI '8 A from the Rural Development Fund (FDR). Both ORDs and FOR receive their funds from the National Development Bank (BND). This was a joint venture where the 0RDs serve as intermediaries between the Bank and the pre-cooperative or village groups of young men and women. Under ‘the joint venture, most of the credit offered to farmers was for short 55 fl‘ ‘ 'term input purchases. The 0RD's were given the role to select potential 'farmers eligible for loans, to process the money or inputs to farmers. “ ,‘ 'to give the necessary technical support, and to collect the repayments for the Bank. At the 0RD level, this job carried out in the field by the extension workers in addition to their usual tasks. The FDR has a role similar to that of the Bank but with emphasis on animal traction equipment and low lands development. The BND/0RD joint venture ended in 1978. Various problems contri- buted to the collapse: a lack of qualified personnel to handle the loan operation at the 0RD level, use of inappropriate criteria of eligibility for credit, confusion of roles between the extension worker and the cre- dit agent, use of reimbursement by some 0RDs to finance their own operat- 109 eXpenses, and embezzlements. In 1980, the BND was replaced by the CNCA. The 0RD continues to play the same role as before. However, in .1 14*! 6514:4an . u. . 2:.“ a 34 each 0RD, a CNCA - financed controller was attached to strengthen the 0RD's personnel but nothing was introduced to increase the volume of credit destined to farmers. 3.1.5.1.2 Interest Rates During the BND/0RD joint venture, the share of agricultural credit amounted only to 5 to 10% of the total credit granted as compared to 30 to 35% for the industrial sector. In 1975. this share was estimated ”‘ to be approximately 5% for the agricultural sector and 95% for the in- dustrial sector. During the joint-venture, the interest rate Charged by the BND to the 0RDs for both short and medium term credit wad 5.5%. A maximum of 12 and 24 month repayment period for short and medium term credit respec- §'I . I tively were allowed to the 0RD for the full repayment of the loan. Any (3RD with total debts amounting to 20% or more is no longer eligible for 3[; a loan. 5; Interest rates Charged individual farmers are higher and more com- plex. In the case of equipment (or medium term credit), if the farmer pays 50% of the total credit amount, the credit is granted to him at 11 percent interest rate on the remaining amount to be paid within 12 months. With 35% cash payment, the interest rate is 8% charged on the remaining amount to be paid in 24 months. During the agricultural cam- Paign 1981-1982, the first operational year of the CNCA, the interest rates charged to individual farmers averaged 13% for both medium and short term. This is an increase of 62 percent over the eight percent levied during the BND/0RD joint venture. I l r ‘ *‘g t‘ ‘I1ISI‘ . . a gnu.“ J Inn..- 35 While the CNCA represents an improvement in the agricultural cred- it system, its higher interest rate charges limit the demand for cred- it at the farm level. 3.l.5.l.3 Private Institutions Generally speaking, there are no private lending institutions, but there are external donors such as government organizations including IDA, USAID, FED, Counseil de l'Entente and also by some nongovernmental organizations such as Oxfam, Club 4C etc. These donors, through the 0RDs' intervene in specific projects that do not often benefit the poorer farmers. Their credit is mainly oriented toward animal traction equipment, to creation of cereal banks, infrastructure, and financial ;:'-‘ supports to some larger businesses. Between 1972 and 1980, the total external financing for agricul- tural development amounted to 51,714 millions CFA or approximately 6,464 3‘ millions CFA on average per year (W8 82). However, more than 70% went to infrastructure and financial assistance. 3.1.5.2 Individual Money Lenders Individual money lending is a common practice among millet/sorghum farmers. It involves small amounts of money but plans an important role in agriculture. According to Tapsoba,1 an average farmer in the Eastern region borrows 1,500 CFA per year. Private grain traders are also a source of credit for millet/sorghum producers. However, their credits are used only indirectly for production lTapsoba E.K. (1981), An Economic and Institutional Analysis of For- Irml and Informal Credit in Eastern Upper Volta: 'Empirical EVidence andT Eiglicy Implications. Ph.D. Dissertation, Michigan State University, 1981. I I, “ Ir: I t.. d..."...~ a. u p‘ufi‘ n! 1| 36 purposes. Most of their transactions occur during the "hungry season," August-September. At that time, the granaries are empty. Private traders operating through selected farmers provide cash to the needy households. The repayment is most often made in kind (grains of millet/ sorghum) at the disadvantage of the borrower. Interest rates charged by private traders are very high. They can vary between 20 to 100 percent.1 The second group of individual credit lenders are the farmers themselves. Within the village, credits are exchanged between individ- ual farmers.. These credits are in kind or in cash. Tapsoba reports that in the Eastern 0RD, only 34% of all individual money lending trans- actions provided by farmers are in kind. Over 90% of borrowing in kind .I'f.‘ I is in the form of millet and sorghum. In kind borrowing of millet and - ' sorghum is for consumption purposes as well as for planting purposes. #1: 1. Cash borrowings, much more important than in kind borrowing, is 5} 5 1 mainly used to meet social obligations such as funerals, weddings, (I? family expenditures, or other social events. They are seldom used for production purposes. Neither the terms of repayment nor the interest rates are known. Although in kind borrowings are often due at harvest time. In summary, Voltan farmers can get credit from three main sources: the 0RDs, the private grain traders, and from other farmers. Credit pro- vided by the 0RDs are the most important of the three. These credits concern both medium-term credit for animals and equipment for animal traction and short term credit as well. 1Societe Africaine d'Etude et de Developpement (SAED); "1e Credit agricole"; bulletin No. 23, 1976. - 11rd .- .E1. .1 . rd 2C3.» .. 37 The short term credits are given to groups of producers or cooper- atives, but seldom to individuals directly. Short term credits are used to purchase inputs like seeds, fertilizer and insecticides. Upper Volta has no means to mobilize local financial resources for investment in agriculture (Vincent Barret et a1. 1980). The bulk of the credit to farmers is provided by outside sources using the 0RDs as intermediaries. At the 0RD level, the lack of qualified personnel to n manage the different loan programs, the difficult and confusing criteria for selection of farmers who will benefit from credit and the addition of credit collection and distribution functions to the already fully occupied extension agent, all make the system perform less adequately. Interest rates are high for farmers. Rates of collection are low. Em- bezzlements are frequent. Credit transactions made by private grain traders as well as cred- 4: w t'} its exchanged between farmers are also important in volume. Unfortunate- 3; 1y, they are not oriented toward production. A large part of these 14 transactions occur during the "hungry season" for consumption purpose. 3.1.6 Government Extension and Research on Millet and Sorghum There are three types of agricultural research in Upper Volta: crop improvement, agronomy and farm management. None of them is working well. The reason for this situation lies on three major arguments: 1) there is no clear national research policy; 2) there is lack of adap- tive research; and 3) there is lack of coordination. 1) Lack of Research Orientation Upper Volta has no established policy in agricultural research. There are however, many agencies involved in field or station work, each - LPdPM .10....4 . . . 2:... mail 38 doing what it wants to do. There are no national guidelines to follow in conducting their research. Financially independent, they plan and execute their own programs with varying degrees of concern for the re- commendations made to them. 2) Lack of Research Adaptability Research in millet/sorghum in Upper Volta is not well adapted. Crop research represents 80% of the total agricultural research fund. The objective is mainly to select seed varieties which are high yield- ing and resistant to drought and disease. Since researchers were free to do whatever they want, their research is not focused on real problems faced by farmers. As a result, their recommendations are most often inconsistent with the farmers' actual conditions. 3) Lack of Research Coordination There are 20 agencies and 5 major institutes that are engaged in crop production research in Upper Volta. Some evidence indicates many overlapping research projects in which some agencies are doing the same research in the same regions at the same time. As a result, some re- gions such as the southwestern and central regions are heavily emphasized while others like the eastern and northern regions are almost forgotten. Furthermore, research has long focused on cash crops, mainly cotton, and has neglected millet/sorghum. 3.1.6.1 Government Organizations and the Nature of Research in Millet/Sorghum Since 1981, research organizations are placed under the Institut Voltaique de Recherche Agronomique et Zootechnique in the Ministry of Ifigher Education. This institute replaces the former Institut de la - - .l-xfdklm «flirted... n 3L2}... h... 39 Recherche Agronomique (IRA) which was under the Ministry of Rural De- velopment. IRA still controls most of the resources and personnel, making the new arrangement less effective. As a result, agricultural research is extremely dependent upon foreign institutions. There are five of them with only two (IRAT and ICRISAT) working on millet/sorghum. 3.1.6.1.1 IRAT IRAT is a long—established institute in Voltan agricultural re- I search history. Its activities are broad and cover both fertilizers Y and cereal variety selection as well as new agricultural practices in— tended to improve farming conditions. Its work on millet has included the testing of a new short-cycle and disease-resistant variety at E. Linoghin and Saria (experimental stations located in the AVVs' zones). I With respect to sorghum, the institute is working on short-cycle '3 and high-yielding Senegalese varieties. Per hectare yields are expected to reach 7,500 kg. The 5,174, another tong-cycle, high-yielding and good-tasting variety, has been successfully tested. If promoted ade- quately, its adoption by farmers can improve Voltan agricultural pro- duction performance. 3.1.6.1.2 ICRISAT This institute is based at Kamboinse, near the capital city. Basic- ally, it is doing the same kind of research as IRAT. It tries to select resistant and high-yielding varieties of millet/sorghum and tries to find ways to cultivate them extensively without extensive soil damage. In summary, agricultural research in Upper Volta indicates a strong tendency towards production oriented research. Selections of high-yielding 40 varieties resistant to drought and disease is the main objective in the research for millet/sorghum. So farmers are not convinced that these new varieties are really superior to the local ones. It is time for researchers to go one step further and include mar- keting research. This will help to improve supply and demand analysis, projections, for better policy formulations. 3.2 Marketing of Millet and Sorghum In principle, farmers produce millet and sorghum for their own consumption. Sales occur only when there is an excess of production over own consumption. It is very difficult to determine the quantity which is marketed and that which is consumed at home. In 1974, the marketable surplus for all cereals was estimated to be 30,000 metric tons out of a total production of 1,004,300 tons.1 Out of the 22,400 T that was marketed, millet and sorghum amounted to 15,786 T, or about 71%. The CILSS estimates that on the average nearly 15% of the total production is marketed each year through official and private channels while about 80% is home-consumed. The remaining 5% are seeds and losses. 3.2.l Marketing Channels In Upper Volta, millet and sorghum are marketed either privately or in the public market. Knowledge is limited on the importance and the performance of the private channel. The public circuit is dominated by private traders who collect large volumes that allow them to manipulate prices. This prevents Ofnacer (the state cereal board) from realizing 1Figures are from: "Sous Commission de la Production Vegetale: _ggfinition d'une_politique cerealiere" p. 3, n.d. -—”-‘9 A 41 some of its basic objectives which are: 1) to provide farmers a stable and acceptable price for their product; 2) establish a stock in order to reduce seasonal price fluctuations and regional deficits; 3) estab- lish a security stock for eventual drought period; and 4) to transfer cereals from regions where there is overproduction to regions experienc- ing shortages. Figure 3.6 Shows marketing channels for millet and sor- ghum in both public and private sectors. 3.2.1.1 The Public Marketing Circuit Ofnacer created in 1971 has a general monopoly on the collection and marketing of millet and sorghum. How Ofnacer gets the grain from producers is explained in Figure 3.6. Authorized dealers and groups of young men are responsible for the collection of millet and sorghum 3 Z. . grains at the village level. Grain collected by the groups goes to the 0RDs which provide the money for that purpose. In 1979, there were 0 3,584 registered village groups in more than 7,000 villages. Today, 5* with the widespread creation of village cereal banks, a substantial part of the grain collected stays at the village level in these banks. Ofnacer agents also intervene in the grain collection at the village level. But to date, their effectiveness has not been demonstrated. In 1978/79, out of 15,300 metric tons of millet and sorghum collected, 7,000 tons was bought directly by Ofnacer's agents, 4,100 tons by village groups, 2,900 tons by unlicensed traders, and 1,300 tons by licensed traders as indicated in Table 3.4 below. Grain collected by the licensed traders, those who are legally en- titled to engage in the grain trade, can be sent to either the 0RDs or directly to Ofnacer. Grain purchases by 0RDs through the different 42 mapo> Loan: cw chcmgo aceawxgez apnea puma; mo pence sopu m.m ms:m_m .u.: .e»_o> some: .sowaoomo mcvpoxgmz erase "mupo> Loan: "ucrswcox .2 esoo ease emaamc< "mugzom muema< muuPL¢ zonhaznmzou zo_»=m~zhm~o bzuzmmaucza zonhumgaou zo~huaooma muu gmomceo prca rm; mu\mnm— .mommgogae cvmgo gmumcmo ¢.m m4m O 62 > O o3 < O o4 > 0 Equation 2: AMI = 80 + 81 LMPCE + 82 TOTAC + 83 LAMI + 84 L086 + e2 81 > O 82 > 0 83 > 0 84 < 0 Equation 3: YLDS = YO + Y] RAIN + 72 FERTM + Y3 A5 + Y4 T + 93 y]>0 y2>0 y3<0y4>0 Equation 4: YLDM = ”O + u1 RAIN + "2 T + e4 u1>0 u2>0 Equation 5: MPCE = 20 + A] POP + A2 GNPOP + A3 DMT + 14 D + As MPR-he5 ‘01 > 0 AZ > 0 A3 < 0 A4 5 > 0 Identity QMS = (YLDS * AS)/1000 + (YLDM * AMI)/1000 + Imports- Exports > O A The five equations together with the identity form our general model which is to be solved simultaneously, using the Gauss Scidel method and historical data from 1961 to 1981. 5.1.4 Estimation To test hypotheses about relationships between variables, the five equations included in the general model are first estimated using ordi- nary least squares (OLS). While this method is used for estimation pur- pose, some basic assumptions governing the specification of the model 66 have to be made in order for the method to provide BLUE estimates, ef- ficient prediction and forecast. In particular, the disturbance term is assumed to be normally distributed with a zero expected mean and constant variance, with no correlation between the error terms and the regressors. The empirical results of the estimated multiple regression equa- tions explaining price, areas cropped and yields of millet and sorghum are summarized in Table 5.1. In general the performances of the dif- ferent equations are satisfactory as can be seen from visual plots which compare the actual and the estimated values against time (see Figures 5.1 to 5.6). 5.2 Evaluation 5.2.1 Economic Evaluation of the Model Economic evaluation of the estimated equations refers mainly to the sign and magnitudes of the estimated parameters as postulated in section 5.1.3. 0n the demand side, the estimatedIJLS regression, equation 5, indi- cates that the actual market price of millet and sorghum, MPCE is direct- ly related to population (POP), to per capita income (GNPOP), and to the binary variable (D), but inversely related to total consumption (DMT). These results were expected. Inclusion of rice as a substitute in con- sumption of millet and sorghum was later rejected on economic grounds. The sign of MPR was inconsistent with our theoretical expectations. vIn addition, its coefficient was insignificant and could not be used to suggest a possible existence of complementarity between the two products. The sign of each included variable is also consistent with a priori 67' oewo> » a w Mose segmeem we eoeowuo - m< azuv eeeucw coweeu wen u eoezo eewuecwELeuee wnemwmauflwmmum H 3m werpwuwow eew—eee n zwmmw N 3.5 :35... I 55. eweewee> zzsee n o Ameeemeezuv uueeege peeewuee mmeem u ezu mew» u w Ameeemeeguv eewuepeeee a eon 3.2.8 eeeeeeee we 82.. Eewwwe I eee .358 we: we 3?... pace... I wee: Aezv pep—we we oueoeue new a ~2<4 «enumeeu u 3.3 So: we: we 2.2.2. .33 I :3 22:9: geese... we Be: I 8.; Anew eeueewe eewe page» a u .99 so u mm o w ago; uxce wzn u Am; ooon mum; mo mammcum u m< «a; coopv cgou do mommguo u zmou< Am; coopv Eagmcom mo mammgum u m< an; coopv couuou we mommgom u u< an; ooopv umPFVE we mmamgom u Hz< .uowgwn cowuumnoga mg» so» Lama wmmn a mp mep .Pmmp vcm Aucmzosu mgu Lopes mamma ozpv mmmp cmmzpma cowpmgwuwmcou Laue: mpnavgm> «no mo guzogm Pmcsamc on» pcmmmgnog xu;h .Eaccm awn wmmucmugma cw vmmmmgaxm .manme> comm mo cpzosm mo moan; umssmmm ugm mwmmzucmgoa cw spat och "mac emc mum omp mpp FNN e¢ moep nma camp mew mew why NPP mpm mv emmp mmm mwmp mme coo “up mo? mom me mmm— —mm mmmp “me mmm mu— oo— mom m.~¢ mmmp «em nmmp mmu omo mnp mop cap we NFmF 0mm owmp mme mum Pup oo_ pap we mmmp mum mwmp ome oco on— mm mm_ Pv comp Fma vwmp wee mmo mop mm omp Fe memp cpm mwmp eve one mop mm mNF oe Pump mom Nmmp eee mum vop om cup oe comp com Pmmp CS; QwJ $3 38.3 $8 av GE. 5 Ame 28> zog> mo4> zmou< o< w< m< m< Hz< cFaw> new aaaaau< co eoFbuanoca N.o msm