A SIMULATION POLICY ANALYSIS OF THE WESTERN. NIGERIAN COCOA INDUSTRY , Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY KWONG -YUAN« CHUNG 1973 H T Michigan State University This is to certify that the thesis entitled A SIMULATION POLICY ANALYSIS OF THE WESTERN NIGERIAN COCOA INDUSTRY presented by Kwong-Yuan Chong has been accepted towards fulfillment of the requirements for Ph.D. degree in Agricultural Economics (Latina Major professor Date g... ’9’ 7 5 0-7639 r i... V LIBRARY ‘ A!" - . .' ‘ ammua av * HOAE & SDNS’ BIIIIK BINUERY INC. I LIBRARY BINDERS i gammy, mum! ‘17.} I ABSTRACT A SIMULATION POLICY ANALYSIS OF THE WESTERN NIGERIAN COCOA INDUSTRY By Kwonnguan Chong Since cocoa is the major source of Western Nigeria's income, employment, revenue, and foreign exchange, the industry is very im- portant to the regional economy. Over 95 percent of Nigerian cocoa production, covering a total area of approximately 1.6 million acres-- cultivated by over 400,000 households, is located in the Western State. Nigerian cocoa production, which is relatively labor-intensive, is almost exclusively a smallholder enterprise. All the cocoa pro- duced is sold to the Nigerian Cocoa Marketing Board, a statutory monop- sonist. However, the producer prices farmers receive are generally less than the world prices. Between the world prices and producer prices, the government collects export duties, producer sales tax, and the marketing board collects a trading surplus tax. Additionally, farmers also pay for the Operational and handling costs involved in the sale of their output. The total differences in some years may amount to as much as 50 percent of the world price. Hence, most economists recommend the increase in the cocoa producer prices which may, in turn, increase the cocoa output and output capacity. Kwong-Yuan Chong Furthermore, since the yield of many of the existing Amelonado cocoa trees is relatively low when compared to the recommended higher- yielding Upper Amazon species, the Western Nigerian government is encouraging farmers to grow more of the latter. In addition, the government is encouraging farmers to grow the higher-yielding Upper Amazon cocoa trees in land presently in bush or food. The primary purpose of this study was to adapt components of the Nigerian Agricultural System Simulation Model developed at Michigan State University to analyze the proposed revamping of the Nigerian cocoa producer pricing policy, and the government-initiated cocoa new planting and replanting production campaigns. Specifically, the system approach accounted for the dynamic interactions and feed- back effects that might occur within the economy as a result of the prOposed price-income changes. The cocoa system simulation model has four major components which (1) allocated land use according to the farmers' perceived profitabilities of cocoa and food subject to the land, labor, and capital constraints; (2) calculated the yield and output of cocoa and food, and their reSpective producer and market prices; (3) provided the instrumental linkages for the government revenue, marketing board trading surplus, and production campaign policies; and (4) generated the performance criteria to evaluate the impact of alternative programs on the cocoa economy through time. The three major sets of assumptions investigated were (1) alternative world cocoa prices, (2) alternative government revenue and marketing board producer pricing policies, and (3) proposed government cocoa planting and replanting production campaigns. Four Kwong-Yuan Chong world price functions, representing the moderate (most likely), high, low, and cyclical price projections were used. Alternative producer price policies and production campaigns were combined into five basic policy options. They were (a) status quo producer pricing policy with no government-initiated production campaign, (b) status quo pricing policy with replanting and new planting production cam- paigns, (c) a "dramatic" producer price increase with production campaigns, (d) a more gradual producer price increase with production campaigns, and (e) option "b" with an added predetermined minimum producer price guarantee, supported by previously accumulated marketing board trading surpluses. The results of the cocoa policy experiments were discussed in terms of the projected time paths (from 1970 to 1995) of six of the more important performance indices incorporated in the model. They were (1) total output of cocoa, (2) total and compositional (tradi- tional and modern) acreages of cocoa trees, (3) foreign exchange generated from cocoa exports, (4) capitalized agricultural land value of the cocoa-food ecological zone, (5) disposable agricultural income per capita, and (6) accumulated government revenue and marketing board trading surpluses. In general the study demonstrated that (l) the projected outcomes with the government production campaigns were all greater than the base run which assumed no replanting and new planting production campaigns; and (2) the projected outcomes under the various producer pricing alternatives were also significantly different. However, because of the model's present agricultural Kwong-Yuan Chong land allocation and harvest response mechanisms, the time-paths of the cocoa outputs tend to cluster. Nevertheless, the relative dif- ferences in the time paths of various performance indices provided a more comprehensive basis for selecting the most efficacious cocoa producer pricing policy. The study also demonstrated that the system simulation approach with a computerized model of the economy which incorporated information from diverse sources, and accounted eXplicitly for the dynamic interactions and feedbacks that might occur, can be a very useful methodological tool for policy analysis. A SIMULATION POLICY ANALYSIS OF THE WESTERN NIGERIAN COCOA INDUSTRY By Kwong-Yuan Chong A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultural Economics 1973 (gift To My Parents and M, my gratitude ii ACKNOWLEDGMENTS The author thanks Glenn Johnson, project director and thesis supervisor for his guidance and encouragement, Thomas Manetsch, Michael Abkin and Marvin Hayenga, colleagues and friends in the Michigan State University Agricultural Sector System Simulation Team for their critical yet perceptive comments on earlier drafts of the thesis; Carl Eicher, major professor, Anthony Koo, Everett Rogers (who together with Johnson and Manetsch constituted the author's doctoral guidance committee) for their composite intellectual influences; Ken Fettig and Pauline Sondag for their editorial assistance; Judy Pardee, Lois Robertson, Kathy Kohl and Connie Schweifler for typing earlier drafts of the thesis; my friends, Robert and Jeanne Ranger for pro- viding a home away from home; and finally, the United States Agency for International Development (under contracts AID/csd-lSS7 and AID/csd-led) for general financial support. iii TABLE OF CONTENTS LIST OF TABLES . LIST OF FIGURES Chapter I. SCOPE AND NATURE OF STUDY . Introduction Features and Limitations of the Econometric Approach for Cocoa Policy Analysis . Features and Limitations of the Partial Budget and Project Appraisal Approach for Cocoa Policy Analysis . General System Simulation Approach as a Tool for Cocoa Policy Analysis Research Objectives and Procedure Thesis Outline . II. PROBLEM SETTING AND ECONOMIC FRAMEWORK OF STUDY Introduction Goals of Nigerian Agriculture . Importance of Cocoa to Western Nigeria . Nature of Cocoa Production in Western Nigeria Agricultural Price Policy, Production Campaigns and Cocoa Development . Comparison Between Cocoa Producer Price Policy and Production Campaigns . III. THE WESTERN NIGERIAN COCOA SYSTEM SIMULATION MODEL . Introduction The Southern Regional Model with Special Reference to the Western Nigerian Cocoa-Food Ecological Zone iv Page vi vii 12 19 20 20 20 21 22 27 36 38 38 42 Chapter Page 1. Land Allocation and Modernization Components . . . . . . . . 46 2. Allocative Decisions of Land Use. Components . . . . . . 51 3. Agricultural Production and Marketing Component . . . . . . . . . 61 4. Price Generation Component . . . . . . . 67 5. Policy Entries Component . . . . . . . . 69 6 Aggregative Sector Accounting Component . . . . . . . . . . 71 Model Validation and Testing . . . . . . . . 72 IV. POLICY EXPERIMENTS ON THE WESTERN NIGERIAN COCOA INDUSTRY: RESULTS AND INFERENCES . . . . . . 77 Introduction . . . . . . . . . 77 Expected World Cocoa Prices . . . . . . . 78 Government Revenue, Marketing Board Producer Pricing Policies and Production Campaigns . . . . . . 80 A. Results of Policy Experiment Assuming Moderate (Most Likely) World Cocoa Prices . . 83 B. Results of Policy Experiment Assuming High WOrld Cocoa Prices . . . . 100 C. Results of Policy Experiment Assuming Low World Cocoa Prices . . . . . . 102 D. Results of Policy Experiment Assuming Cyclical World Cocoa Prices . . . . . 117 Discussion of Simulation Results from Cocoa Policy Experiments . . . . . . . 122 Effects of Cocoa Policy Experiments on Food Production . . . . . . . . . . . 130 V. SUMMARY AND CONCLUSIONS . . . . . . . . . . . 136 Introduction . . . . . 136 Policy Implications from the Simulation Experiments on the Western Nigerian Cocoa Economy . . . . . 140 Conclusions from the Use .of System Simulation Approach for Cocoa Development Planning . . . 143 Further Research on the Nigerian Agricultural Simulation Model for Policy Analysis . . . . 146 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . 152 Table 2.1. 2.2. 3.1. 3.2. LIST OF TABLES World Price, Producer Price, Export Quantity and Cocoa Earnings, Nigeria, 1953-1967 . Export Duties, Produce Sales Tax, Trading Surplus and Administrative Expenses of Western Nigerian Cocoa Export, 1953-67. Average Annual Yield and Input Requirements for Traditional and Modern Cocoa Production Per Acre by Age-Cohort, Nigeria Time-Series Tracking of Observed and Simulated Data of Nigerian Cocoa Exports, Palm Oil Exports, Rubber Exports, Market Food Prices (South), 1953-1965 vi Page 23 24 49 7S Figure 3.1. 3.2. 3.3. 3.4. 3.5. 3.6. 4.1. 4.A1. 4.A2. 4.A3. LIST OF FIGURES Interacting Submodels of the Nigerian Agricultural Simulation Model The Four Crop-Regions of the Southern Regional Model . . . . . . . Major Sectors and Interactions of the Southern Nigerian Agricultural Model Building Blocks of the Southern Regional Model . Diagram Showing Profitability Response Function in Agricultural Land Reallocation Decisions Diagram Showing Abandonment Response Function in Agricultural Land Abandonment Decisions Four Illustrative Sets of World Cocoa Prices Used in Policy Experiments on the Western Nigerian Cocoa Economy, 1970-1995 Market and Producer Prices of Nigerian Cocoa Under Indicated Policy Alternatives Assuming Moderate (Most Likely) World Prices, 1970-1995 Total Projected Nigerian Cocoa Output Under Indicated Policy Alternatives Assuming Moderate (Most Likely) World Cocoa Prices, 1970-1995 . . . . . . . Foreign Exchange Generated from Nigerian Cocoa Exports Under Indicated Policy Alternatives Assuming Moderate (Most Likely) World Cocoa Prices, 1970-1995 . . . . vii Page 40 41 43 44 57 62 79 84 85 93 Figure 4.A4. Agricultural Disposable Income Per Capita in Nigerian Cocoa Sector Under Indicated Policy Alternatives Assuming Moderate (Most Likely) World Cocoa Prices, 1970-1995 . . Total, Traditional and Modern Nigerian Cocoa Acreages Under Indicated Policy Alternatives Assuming Moderate (Most Likely) World Cocoa Prices, 1970- 1995 . . . Capitalized Agricultural Land Value in Nigerian Cocoa-Food EcolOgical Zone Under Indicated Policy Alternatives Assuming Moderate (Most Likely) World Cocoa Prices, 1970-1995 Accumulated Government Revenue and Marketing Board Trading Surpluses from Nigerian Cocoa Production Under Indicated Policy Alternatives Assuming Moderate GWost Likely) World Cocoa Prices, 1970-1995 . . . . . Market and Producer Prices of Nigerian Cocoa Under Indicated Policy Alternatives Assuming High WOrld Cocoa Prices, 1970-1995 Total Projected Nigerian Cocoa Output Under Indicated Policy Alternatives Assuming High World Cocoa Prices, 1970-1995 . Foreign Exchange Generated from Nigerian Cocoa Exports Under Indicated Policy Alternatives Assuming High World Cocoa Prices, 1970-1995 Agricultural Disposable Income Per Capita in Nigerian Cocoa Sector Under Indicated Policy Alternatives Assuming High World Cocoa Prices, 1970-1995 . Total, Traditional and Modern Nigerian Cocoa Acreages Under Indicated Policy Alternatives Assuming High World Cocoa Prices, 1970-1995 Capitalized Agricultural Land Value in Nigerian Cocoa-Food Ecological Zone Under Indicated Policy Alternatives Assuming High World Cocoa Prices, 1970-1995 . . . . . viii Page 94 95 97 98 101 103 104 105 106 107 Figure 4.87. Accumulated Government Revenue and Marketing Board Trading Surpluses from Nigerian Cocoa Production Under Indicated Policy Alternatives Assuming High World Cocoa Prices of 1970-1995 . Market and Producer Prices of Nigerian Cocoa Under Indicated Policy Alternatives Assuming Low World Cocoa Prices, 1970—1995 . Total Projected Nigerian Cocoa Output Under Indicated Policy Alternatives Assuming Low World Cocoa Prices, 1970-1995 Foreign Exchange Generated from Nigerian Cocoa Exports Under Indicated Policy Alternatives Assuming Low World Cocoa Prices, 1970-1995 . Agricultural Disposable Income Per Capita in Nigerian Cocoa Sector Under Indicated Policy Alternatives Assuming Low World Cocoa Prices, 1970-1995 . Total, Traditional and Modern Nigerian Cocoa Acreages Under Indicated Policy Alternatives Assuming Low World Cocoa Prices, 1970-1995 . Capitalized Agricultural Land Value in Nigerian Cocoa-Food Ecological Zone Under Indicated Policy Alternatives Assuming Low World Cocoa Prices, 1970-1995 Accumulated Government Revenue and Marketing Board Trading Surpluses from Nigerian Cocoa P Production Under Indicated Policy Alternatives Assuming Low World Cocoa Prices, 1970—1995 . Market and Producer Prices of Nigerian Cocoa Under Indicated Policy Alternatives Assuming World Cyclical Prices, 197041995 Total Projected Nigerian Cocoa Output Under Indicated Policy Alternatives Assuming Cyclical World Cocoa Prices, 1970-1995 . Foreign Exchange Generated from Nigerian Cocoa Exports Under Indicated Policy Alternatives Assuming Cyclical World Cocoa Prices, 1970-1995 ix Page 108 109 111 112 113 114 115 116 118 120 121 Figure 4.D4. 4.D5. 4.D6. 4.D7. 4.2. 4.3. 4.4. Agricultural Disposable Income Per Capita in Nigerian Cocoa Sector Under Indicated Policy Alternatives Assuming Cyclical World Cocoa Prices, 1970-1995 . Total, Traditional and Modern Nigerian Cocoa Acreages Under Indicated Policy Alternatives Assuming Low World Cocoa Prices, 1970-1995 Capitalized Agricultural Land Value in Nigerian Cocoa-Food Ecological Zone Under Indicated Policy Alternatives Assuming Cyclical World Cocoa Price Functions, 1970—1995 . Accumulated Government Revenue and Marketing Board Trading Surpluses from Nigerian Cocoa Production Under Indicated Policy Alternatives Assuming Cyclical World Cocoa Prices, 1970-1995 . Total Projected Nigerian Cocoa Output Under Indicated Alternative World Cocoa Prices Assuming Low Tax Rates and Production Campaigns, 1970-1995 . Total, Subsistence, Cash Food Land in Nigerian Cocoa EcolOgical Zone Under Indicated Cocoa Policy Alternatives Assuming Cyclical World Cocoa Prices, 1970-1995 . . Market and Producer Prices of Food Composite in Nigerian Cocoa Ecological Zone Under Indicated Cocoa Policy Alternatives Assuming Cyclical World Cocoa Prices, 1970-1995 . Page 123 124 125 126 127 131 132 CHAPTER I SCOPE AND NATURE OF STUDY Introduction Most economists and planners concerned with the Western Nigerian cocoa economy generally agree that (l) cocoa farmers have been responsive in their output to changes in cocoa producer prices, (2) continued low producer prices might be counterproductive, if not adverse to, the cocoa industry, and (3) increases in producer prices and government production campaigns would increase both cocoa output and output capacity and government revenue in the long term. The loss in revenue from the producer tax decrease may be offset by increases in tax base, with the increase in output, income and asset values resulting from the producer price increases and production campaigns (Johnson §£_al,, 1969). However, most of the Nigerian cocoa studies did not provide a comprehensive, dynamic and analytical basis that would allow policymakers and researchers to interact and evaluate policy alternatives in the larger context of how the sector operates through time. For discussion, these Nigerian cocoa studies are divided broadly into (1) econometric, or more Specifically, the statistically estimated simultaneous equations, and (2) the partial budgeting and project appraisal categories. The first category includes such studies by Bauer and Yamey (1959), Stern (1962), Sanders (1968), Ady (1968), Okurume (1970), Olayemi (1970), Oni (1970), and Olatunbosun and Olayide (1971). The second category includes chapters on cocoa development from the Food and Agriculture Organ— ization Study (1966), the Consortium for the Study of Nigerian Rural Development (CSNRD) Report (1969), and the Cocoa Research Institute of Nigeria Report (1971) for the National Agricultural Development Seminar. The features and limitations of the two approaches will be examined, compared and contrasted with the general system simula- tion approach in which we propose to study the major economic policy issues confronting the Western Nigerian economy. Features and Limitations of the Econometric Approach for Cocoa POIicy Analysis The econometric approach is essentially a procedure for esti— mating the coefficients of an equation or a set of simultaneous equations based on empirical observations of the economic phenomena (Johnston, 1964). The estimated coefficients, in turn, provide an analytical basis for testing hypotheses, policy evaluation and pre- diction. The empirically-based approach, therefore, depends heavily on the availability of reliable time series and cross-sectional data in order to statistically estimate the coefficient matrices. Most of the past econometric studies on the Nigerian cocoa economy were motivated by the now passe neoclassic agricultural economic problem of estimating the responsiveness in output of farmers in developing countries with producer price changes. However, the economic thinking of many Nigerian government officials in the 19505 and early 19605 was that the farmers' price elasticity of out- put was zero. The zero price elasticity assumption therefore pro- vided a comfortable rationale for the government low producer price policy. Hence, these studies were able to show that the price elasticities of the cocoa farmers were positive, thus challenging the rationale for the Marketing Board's low producer price policy. These studies, however, were not able to relate the short-term and long-term output responses of cocoa production and their inter- relations with food price and food production in the regional economy. Many issues remain unanswered in these studies. For example, what was the economic rationale that governed the farmers' decision to allocate their resources between the short-term and long-term returns? When cocoa producer prices were high, did producers attempt to maximize the short-term output by allocating all their labor to harvest the biologically available output in that year? Or did they also attempt to maximize their long-term income potential by expand- ing their cocoa acreages? Conversely, what did farmers do when their prices were low? Did they increase their short-term output in order to maintain their yearly expected income? Or did the farmers decrease their output because of the lower prices? What about food production? Did they increase food production when the cocoa producer prices were low in order to offset the income decline? In order to estimate the coefficients of the postulated economic phenomena, most of these studies assumed that (1) the cocoa producers were profit-maximizers, and (2) the economic forces that linked the various components in the cocoa economy were self- equilibrating. However, we have to challenge the validity of these two assumptions if their findings are to be used for policy analysis and prognostication. First of all, Nigerian cocoa farmers are not really profit maximizers in the neoclassical sense. Their production decisions are determined by the complex interplay among their personal motives, managerial capacity, resource endowments, ability to command addi- tional resources; and in the future, the proposed changes in the marketing board producer pricing policy and the program features of the production campaigns. Hence, the final outcome of the farmers' decision to produce food and cocoa depends on their skill and edu- cational level, psychological aversions for change and risk, the complementaries of other inputs, the physical and financial "lumpi- ness" of the new technology, and the existing institutional and administrative constraints.1 Secondly, the Nigerian cocoa economy would most probably not be in equilibrium since the very purpose of economic development is to set in motion disequilibrating economic forces to transform its underlying structure. Hence, the endogenous variables estimated from 1See Johnson (1972) for a critique of the conventional theorizing of firm behavior which most of these econometric studies of supply responsiveness are built on. past time series data from the cocoa studies cannot be extrapolated linearly into the future time periods. The anticipated economic changes and conditions of the Nigerian economy have to be accounted fOr and modeled explicitly for policy analysis as much as possible. Finally, the capital stock adjustment and the Nerlovian lagged reSponse models used in the cocoa econometric studies were restrictive in explaining the cocoa farmers' decision to invest and disinvest in their cocoa trees. Perennials are partially fixed assets whose value at any one time lies between their establishment cost and salvage value. The revenue generated from cocoa production flows from the stock value of the trees. The asset value of an acre of cocoa trees (without considering location or real estate potential of the land) is highly correlated with the income-generating capacity from the trees, even though the correSpondence is not perfect. For example, in the first four years while the trees are gestating, establishment cost is also increasing. The asset value of the trees is very low and may even be negative should the land be converted for other agricultural or residential purposes. Furthermore, the young trees are also more vulnerable to wind, flood, and disease damage. However, once the trees are established, their asset value appreciates. Corresponding to the potential income flow, the asset value remains relatively high throughout the maximal production stage and then begins to decline as the trees grow older. The asset value of the trees may also change externally with changes in output prices or production costs. However, the farmers do not necessarily contract and expand their cocoa acreages because of the relative inelasticities of transformation between cocoa and other competing cr0ps. Even if the producer price of cocoa declines to less than the minimum average variable cost, farmers would probably continue in production as long as the marginal revenue from the cocoa production is greater than the salvage price of their fixed inputs, and the acquisition price of their variable inputs. In cocoa pro- duction, the fixed input is their household labor (whose opportunity cost is very low), and the variable inputs are the harvesting materials. However, if producer prices are persistently low and are expected to remain low in the foreseeable future, some of the cocoa farmers may abandon their less productive cocoa acreages or convert their land use. Thus we see that the investment and dis- investment decisions for cocoa production are not completely sym- metrical or reversible. The change in their output capacity depends on the relevant price range and magnitude, direction and time duration of the producers' price changes. However, one of the fundamental methodological problems faced by the cocoa econometric studies is the nonavailability of time-series and cross-sectional data of the Nigerian cocoa economy to establish the coefficients. Although most of the econometric models were conceptualized in nonlinear and interactive terms, with dynamic feedbacks, the final estimations were linear and additive, because of the data constraints. Features and Limitations of the Partial Budget and Project Appraisal Approach for Cocoa Policy Analysis The cocoa project appraisals are generally conducted in rigid geOgraphic, time, and program terms. Some examples are: the assess- ment of the financial feasibility of a five-year cocoa replanting program in Ife, a subsidized fertilizer distribution scheme for the Western State cocoa farmers, or a government loan program for cocoa spray equipment. The usual criteria used to determine the project's feasibility are the net present value of future returns, the internal rate of return, cost-benefit ratio or the payoff period of the initial investment. These decision criteria can be calculated in terms of one specific input (which is generally the most limiting input) or the total project (Prest and Turvey, 1969). Recently, two major developments have been made to strengthen and improve the project appraisal approach for policy analysis. The first is the expansion of the criteria used to determine the effica- cies of the project, by including secondary and other indirect costs and benefits that might occur outside of the cocoa sector as a result of the pragram (Gittinger, 1972). For example, there is the addi- tional employment that may be generated in agriculture-related industries, such as domestic fertilizer plants, resulting from the expansion of the cocoa economy. Unfortunately, many of the indirect costs and benefits may not be tangible, quantifiable or expressible in monetary terms, and their inclusion for project evaluation may still be arbitrary. The second development is to express the values of the crucial factors along with their probability distribution of occurrences in order to arrive at a statistically more complete picture of the anticipated outcome (Reutlinger, 1971). Since project appraisals are generally conducted and prepared as feasibility studies for funding and administrative purposes, the approach has a strong administrative and accounting bias. These studies typically are very concerned with calculating the financial returns made by both the private and public sectors, the repayment capacities of the project, the general impact on the economy, the personnel and 10gistics requirements, and the program phases and time table. Such findings on the cocoa economy are obviously of immense interest and concern to the Nigerian Government and inter- national loan agencies like the World Bank, which monitors the progress of the project or loan program. However, the approach is quite mechanistic in projecting the consequences of the program. Using the principles of partial budget- ing, outcomes of alternative programs are projected under different predetermined rates of program expansion and price assumptions. Little attempt is made to capture the motivating mechanisms of the change processes, or the interactions or the positive and negative feedback effects that might modify the postulated consequences as time proceeds. At best, the initial projected results are sometimes re-adjusted to reflect some anticipated, intuitive contingencies. These re-adjustments, however, are generally ad_h23_using arbitrary discount factors. The initial projected total output of cocoa, for example, may be reduced by 10 percent across-the—board to reflect the less than Optimal agronomic conditions. Moreover, such a mechanistic approach may not allow any rigorous analysis or interaction between researchers and policymakers. Hence, it would indeed be very useful if all the intuitive knowledge and qualitative judgment of the subject matter experts would also be incorporated clearly and con- sistently into a joint analytical framework with the underlying assumptions stated explicitly. As we shall see, the methodological orientation of the system simulation approach is to provide a systematic framework to make use of such information. General System Simulation Approach as a Tool for Cocoa Policy Analysis To address ourselves to some of the methodological and policy problems, we propose to use the system simulation approach as a tool for developmental planning and policy analysis of the cocoa economy. The system simulation approach, following the principles of scientific method and problem-solving research, is a formalized process which begins with the identification of the problem under investigation, and ends with the evaluation of feasible alternative solutions.2 This approach generally includes a mathematical model which enables researchers to express the socio-economic phenomena more precisely for analysis. Once the relevant system with its structural com- ponents and functional linkages is identified and the system's values are specified, its validity can be tested, and experiments using alternative policies can be conducted to draw inferences on the impact of these policies. 2For more detailed discussion of the philosophy of the general system simulation approach for problem-solving research, see Manetsch g£_gl. (1971), Abkin (1972), and Forrester (1972). 10 Although the economy can be modeled mathematically by using ordinary and partial differential equations containing linear and nonlinear terms, explicit solutions for the economic model are often very difficult, if not impossible, with the increasing number and complexity of the differential equations. Instead, researchers (aided by large-scale digital computers) have turned to simulation as a possible means of generating numerical solutions that may pro- vide policymakers with information about the likely consequences of alternative economic developmental strategies. Conceptually, a simulation model of an economic system can be viewed in the following general mathematical form: ¢(t+1)=F[¢(t).a(t),B(t),Y(t)] “(t1 = G[¢(t1.0(t).8(t),Y(t)] where: W(t) a set of variables defining the state of the simulated system at any given time. State variables usually involve the level of a variable at a given time and may include such quantities as production capacities, prices, population by subgroups, levels of technology, 812C. H(t) a set of output variables measuring the system's simulated performances, such performance criteria as output, foreign exchange generated from exports, income, etc. a(t) a set of parameters defining the structure of the system. These usually involve rates of change of variables between levels and input-output coefficients, such as technical coefficients, reSponse coefficients, price elasticities, etc. 8(t) a set of environmental variables, such as world prices, etc. a set of policy instruments, such as tax policies, production campaigns, investment alternatives, etc. Y(t) 11 The state equation illustrates how the state of the system U at t+l is a function of the state at time t and of the values of a,E3and Y at time t. This is a general representation of the dif- ference equation formulation of the system model which describes the state of the system and subsequent performance at discrete points in time. The output equation generates the performance criteria H used to evaluate the performance of the system over time under various policy alternatives. There are three distinguishing features of the general system simulation approach particularly useful for the policy analysis of the Nigerian cocoa economy. First, it is a generalized approach which makes use of all available primary information and calculated findings including estimated coefficients and parameters from econometrics, partial budget and project analysis, qualitative judg- ment and insights of subject matter experts, and descriptive work about the cocoa industry from other social science disciplines. Since the research and model-building process is iterative and flexible, new information can easily be incorporated as it becomes available, and the structure of the model modified accordingly. Second, in the system approach, the functional relations can be nonlinear, and dynamic with discrete or continuous lags and feed- backs, discontinuous and asymmetric according to the theoretical postulates or empirical findings. In contrast, because of the com- puting and estimation procedures, most of the econometric relations are generally assumed to be causal, linear, and additive. Hopefully, 12 the flexible mathematical mode used in system simulation will model reality better (Manetsch and Park, 1972). Third, the approach does not have to assume (but does not preclude) any profit or utility maximizing producers and consumers, or any self-equilibrating economic adjustments. It does not neces- sarily involve a unique set of Optimizing solutions based upon a common objective or a predetermined singular goal, which does not in reality exist. In contrast, the approach is more guided by the problem under investigation. The system simulation approach provides, basically, a conditional feedback framework in which various dynamic interactions and anticipated consequences under alternative policies and programs can be projected through time and thus evaluated. The projected time paths of some of the more important performance indices can therefore provide a composite basis for evaluating alternative strategies for the Western Nigerian cocoa economy. Research Objectives and Procedure Background of Study The present study has two progenitors, both headquartered at Michigan State University under the directorship of Glenn Johnson. The first is the Consortium for the Study of Nigerian Rural Develop- ment (CSNRD), whose policy—oriented research on Nigerian agriculture (Johnson g£_al:, 1969), took the monumental Food and Agricultural Organization study, entitled Agricultural DevelOpment in Nigeria 1965-80, as a point of departure. The second is the MSU Agricultural Sector Simulation Team which was motivated to develop a generalized 13 system simulation methodology for agricultural policy analysis (Manetsch g£_al:, 1971). In both the studies, the Western Nigerian cocoa industry was treated as one of the major income and revenue- generating agricultural sectors of the Nigerian economy. The Consortium approach, using pen and pencil (and occasion- ally, desk calculator!) projections, studied the impact of alterna- tive policies on the future development of the cocoa economy-~in conjunction with the other proposed national and regional agricul- tural policies and programs. While the CSNRD approach provided a very useful analytical framework for Nigerian policymakers, many research questions remain unanswered. To facilitate computations, simple (and perhaps, simplistic) assumptions were made about the technological coefficients (e.g. cocoa yield, input requirements, total acreages), costs and returns of cocoa production, and expected world cocoa prices. Hence, based on their averaged values, the future aggregative impact under alternative cocoa policies and pro- grams were projected to the next fifteen years, with 1970, 1975, and 1985 used as benchmark dates. Because of the time-consuming and tedious nature of the pen and pencil calculations, the study did not: (1) explore the outcome of the proposed policy alternatives under different technological data and farmers' behavioral assumptions, and (2) the projected time paths of other performance indices whose composite outcome might also interest policymakers. Furthermore, as discussed earlier, little attempt was made to show the motivating mechanisms, interactions, and feedbacks of the change processes that might modify the initially projected consequences. For example, 14 CSNRD's cocoa analysis was based on only op§_set of future world cocoa prices (assumed to decline £10 per long ton stepwise, every five years). Hence, it is eminently conceivable that a different set of world cocoa prices might stimulate measurably different short-term harvest reSponses and long-term output capacity responses among the cocoa farmers, thus affecting the reported outcome. (More- over, as we shall see, the interactions between the farmers' short- term and long-term output responses may further modify the projected outcome.) Nevertheless, in fairness to the CSNRD approach, many of these methodological issues may have been taken into account implicitly. As a result of some of these methodological difficulties, interest was generated by some economists and system scientists in using a computerized system simulation approach for agricultural planning. It was hoped that the approach would provide a more dynamic, rigorous and integrative approach for planning--in contrast to the §d_hgg_pen and pencil, "common sense" approach used in CSNRD. Accordingly, an interdisciplinary research team of system scientists and agricultural economists was assembled at Michigan State Univer- sity (of which the author is a members). The research Objective was 3As a member of the agricultural simulation team, the author assisted in the design, refinement, and validation Of the structural and functional linkages of the Southern Nigerian Agricultural Submodel. This involved developing the basic analytical structure and general information and data of Southern Nigerian agriculture which described the behavioral structure of the economy (Manetsch g£_al., 1971). The author also had the major responsibility for writing Chapter II of the report, which provided the overall ecological, economic, and political problem setting Of the study. 15 to build and test the usefulness of a system simulation model as a tool for economic planning. Nigeria was chosen as the case country because of the reservoir of information and expertise at Michigan State University--thanks to the CSNRD experience. The research, however, is also motivated to be Of assistance in the agricultural economic planning of other developing countries that share similar ecological and institutional features, such as the rubber industry in Malaysia, cattle industry in Colombia (Posada, 1973), and the total agricultural sector of South Korea (Rossmiller g£_§l;, 1973). In the report by Manetsch g£_al: (1971), our primary concern was in validating the usefulness of the system simulation approach and the Nigerian Simulation Model for developmental planning. The usefulness of the approach was further extended when the Model was used to analyze specific policy issues concerning the likely conse- quences of alternative Nigerian agricultural developmental strategies (Olayide, Abkin, and Johnson, 1972). Thus, the present study can be viewed as part of the continuing process of the model's development, validation, and application, of the system simulation approach as a tool of planning and policy analysis by focusing Specifically on the cocoa sector of South Nigeria. In addition to the present study's sector-specific focus, there are two new features which were not considered in previous studies. First, world cocoa prices in the previous studies (for the purpose of validating the usefulness of the system simulation approach for developmental planning) were assumed to be constant throughout the planning horizon of the analysis. In this study, we shall 16 interject a little more realism by using alternative world cocoa prices obtained from other studies to analyze their projected conse- quences in the cocoa economy. Second, we introduce for the first time a rudimentary producer price guarantee (along the recommendation of CSNRD) whereby farmers would be paid a predetermined floor price should their price, after accounting for the taxes and handling costs, go below the level. The cocoa producer price support program would be financed by previously accumulated Marketing Board trading surpluses, or if necessary, from other outside sources. It is postulated that such a producer price guarantee feature would be eSpecially helpful if future world cocoa prices are expected to be low and fluctuating. The price guarantee may thus stabilize the cocoa farmers' income, and perhaps increase their income as well. Objective of Thesis The main purpose of the present study is, therefore, to adapt components of the Nigerian Agricultural Simulation Model to analyze the effects on the Western Nigerian cocoa economy of (1) alternative world cocoa prices, (2) prOposed revamping of cocoa pro- ducer pricing policies, and (3) government-initiated cocoa new planting and replanting production campaigns to assist the farmers expand their output capacity. Specifically, the policy experiments conducted on the model shall consider the effects of (1) four sets of world cocoa prices, and (2) five different combinations of producer pricing policies and production campaigns. Based on the Bateman (1971) study of the 17 world cocoa market, three sets of world prices are used--representing the moderate (most likely), high, and low expectations. The fourth set, a cyclical world price function, has been constructed by the author to evaluate its impact on the model's outcome. It is hoped that the four sets of world prices will capture all the relevant world cocoa price behavior germane to the policy analysis of the proposed cocoa producer pricing changes and production campaigns. The first of the five policy options is a base run which approximates the present policy. It has a relatively low government revenue tax of 10 percent on the prevailing world price, a 20 percent marketing board trading surplus tax on the market price, and no government-initiated production campaign. The next three policy Options compare the effects of the proposed production campaigns with varying producer pricing features. In Run 2, the government initiates the production campaigns with the same tax rates as the base run. The benefits of the production campaigns are Obtained by comparing the projected outcome of Run 2 with Run 1. In Runs 3 and 4, the tax rates are assumed to be slightly higher than the base run at 20 and 30 percent of the respective world and market prices. However, in Run 3, the higher tax rates are cut Off the following year, whereas in Run 4, the taxes are phased out linearly over the next five years. The purpose of these two runs is to compare the projected outcome on the cocoa economy Of a "dramatic” producer price increase under Run 3 with a more gradual producer price increase under Run 4, in conjunction with the production 18 campaigns. Because of the interactions between the short-term and long-term output responses, under alternative producer pricing policies, the projected outcome may differ. Finally, Run 5 compares the effects of the rudimentary pro- ducer price guarantee feature. The rest Of the policy features of Run 5 are identical to Run 2. The study hypothesizes that the out- come Of the production campaigns with varying producer pricing features under Runs 2, 3, 4, and 5 would differ accordingly. Their differential income, in turn, would have interesting policy implica— tions. Among the more important performance indices used to evaluate the projected impact in the cocoa economy (from 1970 to 1995) are: (l) in the producing subsector: the annual total output of cocoa from existing, replanted and newly planted trees, the foreign ex- change generated from cocoa exports, the annual agricultural dis— posable income per capita, and the capitalized value of the agricul- tural land; (2) in the food subsector: the amount of food produced and its producer and market prices; and (3) in the government and marketing board subsector: the accumulated revenue and trading surpluses collected from the marketed and exported cocoa. Based on the policy experiments using the four sets of world prices and the five combinations of producer pricing and production campaign policies, the study will draw some limited policy implications for the Nigerian cocoa economy. 19 Thesis Outline Chapter II provides the problem setting Of the study. It discusses the roles and objectives of Nigerian agriculture, the nature of Nigerian cocoa production and the postulated production, consumption and revenue effects that may result from the proposed revamping Of the cocoa pricing policies and production campaigns. In Chapter III, we present a description of the major components of the Southern Regional Submodel of the Nigerian Agri- cultural Simulation Model (hereafter called the Western Nigerian Cocoa Simulation Model) which are used to conduct policy experiments on the Nigerian cocoa economy. Chapter IV presents the simulation policy results under the various combinations of world cocoa prices, proposed government tax and marketing board producer pricing policies, and the new planting and replanting production campaigns to expand cocoa output capacity. Based on the results of the policy experiments, this chapter shall discuss some limited policy implications for the Western Nigerian cocoa economy. In Chapter V, the major methodological and policy conclusions of the study are summarized and possible extensions of the model are presented. CHAPTER II PROBLEM SETTING AND ECONOMIC FRAMEWORK OF STUDY Introduction In order to better appreciate the problem setting of the study, this chapter discusses first the general objectives of Nigerian agriculture and the nature of cocoa production in Western Nigeria. Secondly, the chapter discusses the postulated production, consumption and revenue effects of the proposed cocoa price-income changes; the interrelations between farmer reSponses and the proposed government revenue and marketing board, producer-pricing changes; and the govern- ment initiated production campaigns to increase cocoa output capacity. Goals of Nigerian Agriculture According to Oluwasanmi (1966), CSNRD (1969) and others, there are three basic roles and goals for Nigerian agriculture. First, Nigerians have to be fed adequately and nutritionally both in the rural and urban sectors. The solution to the food problem depends crucially on the interplay among the effective demand of the consumers, the responsive supply of the food producers, and the adequacy of the marketing channels and food distribution system. In order to effec- tively demand and purchase food, the general population must maintain 20 21 an adequate income. The farm-gate price Of food, in turn, must be sufficiently high to encourage producers to meet the market demand. Finally, the pricing mechanisms and market structure must be such that any long-term changes in food prices and/or output are passed on efficiently to the producers or the consumers without the various intermediaries retaining a disproportionate share of the benefits. Second, in the next decade or so, agriculture will probably be the chief sector for providing employment opportunities and an adequate income to most of the country's population and labor force. The industrial and service sectors in national development will still be limited because of their relatively high investment requirement and low labor absorption capacity. Third, in the longer run, agriculture must also be one of the major sources of revenue and resources for the transformation of the country's economic structure, despite the increasing significance of other economic activities, such as the Nigerian petroleum industry. Importance of Cocoa to Western Nigeria The cocoa industry is very important to the Western Nigerian economy since it is the major source of its income, employment, revenue, and foreign exchange. In the last ten years, agriculture accounted for over 65 percent of the gross national product--and cocoa contributed 20 percent of that amount. The other major Nigerian agricultural crops, which are also exported, are: ground nuts (grown in the North), rubber (Midwest), and Oil palm (through- out the South). 22 Most of the cocoa produced in Nigeria comes from the Western State (with the exception of Egbada, Oyo, and Okitipupa Divisions) and Afenami Division of the Midwestern State. Cocoa is cultivated by over 400,000 households covering a total area of approximately 1.6 million acres, or nearly 60 percent of the total cultivated land. Despite the increasing importance of other nonagricultural sources of income and employment, notably in the urban sectors, cocoa production and its marketing still provide for a substantial propor— tion of the farmers in the region (estimated to be between 30 and 45 percent of the total agricultural population), the chief source of income and employment. In 1967, cocoa accounted for over 22 percent of Nigeria's total exports, amounting to £54.7 million. The eXport earnings from palm oil and palm kernels totaled E13.8 million; groundnuts $46.9 million; and rubber E6.5 million. The annual average output of cocoa from 1963-67 is about 212,000 long tons (see Table 2.1). Total ex- port duties, producer sales tax and marketing board trading surplus collected from the cocoa sector in 1967 amounted to E7 million, E.9 million and E12.7 million, reSpectively (see Table 2.2). Nature of Cocoa Production in Western Nigeria Nigerian cocoa is produced primarily by smallholders whose typical landholding consists of about 3 acres of cocoa and 2 acres Of food. Although there is now increasing evidence and concern that the distribution of landholding size and income level among the cocoa farmers is skewed (Essang, 1971),the distribution is not as 23 .flaamsv oeaaaao use csmoncauafio ecu .floomav Humane“ .noamag saoeemws ace coaoam "mouusom N.o~ n.vm m.vv~ 0.0m o.va nomH m.m m.w~ N.omH o.mo o.omH coma m.m~ n.mv m.mm~ o.o~H o.o~H momfi n.mH H.ov w.omH o.oHa o.onH coma 0.5H «.mm o.vnH o.moH o.moH momH m.mH v.mm n.va o.ooH o.mmH Noma v.mH n.mm m.mmH o.NHH o.nNH Homa o.H~ w.om N.va o.ooH o.wom coma v.mm m.wm w.NvH o.omH o.~n~ mmmH u.m~ n.0m n.nw o.omH 0.00m mmma m.o~ o.vN m.mmH o.omH o.wo~ nmmfi m.nH o.v~ ~.BHH o.oom o.mo~ omma n.mH N.om v.ww o.oo~ 0.0mm mmmH m.o~ m.mm e.wm o.onH o.~mm «mmfi o.o~ m.v~ n.voH o.onH o.ovm mmmfl :OWHHwE z u co» wcoH vammsonh cow m:OH\a 2 now w:OH\a z mmcfichmm whoaxm Hmuoe mo unmouom mmcficumm uuomxm a“ xuwucmso unoaxm oowna Hooaponm pawn; vane: nmo> .noma-mmma .mfiuomMz .mwcwchmm mooou ecu xufiucwso anomxm .Oowha hoospoua .Ouwha eflhoz--.fi.~ mamwumhumflcfisp< can mDHthm mcwpmhh .xmh mOHHm auspOHa .mOwuso uhomxmnn.m.m m0mo¢ 03% 3:30:05 05005 .3:— AEhU .0300: 05300030003 Haunu0=thu< 080.00g: 23 mo «000550.010. unwuouuoucuiéé 0.5»: - ton—um 5.00.0052 «So-U acn— l‘ cup-.005 3:70 aca— 00:150..” Stuck 33¢ 2.05 «0330-.» occa0c~:_..on 08.2.33. ’82 W .1 J vouch ~31.— 33590< 1.03012 has. I and 8.0955000 .— 09.8 6.00)..“ .0095. 103103.118! defies-.900. ( . 0...... 3305 0095300— 32.:- noxioul 31:31: 3:— .1335 9.8 .1090 00:32.00 0‘80 L 2...... XX on not-nos 00305.01 .30 c.3230 no nag-80 one 2:03.003 .02.!— 0oA¢ — van. rut-8.31 .couuuavohnu .8309ng 13316:.< ~4..3—q.u€l< tech :3 took 5.00 OUT-nu 1615.9 coach 1020 opt... . >1: , coo..— JOOU , 1.03.938!— OOOL £010 filflfiHIL a... .Illauull. i=5 ., 030...! 05.32. Stable-KOO inaugumuu< .530“. 5.54.3053 =3. 0.. :5 u..—I< 18h £040 k 0.058.020 52:30 0:86....» 51:92 .00.: Ian. 1 I sou-xi all-lug .3901.— 10330u< 41 .030: 02833. 00.350.00.01 05 mo "008330-995 use..— ofi--.~.m unnum— O.- 0: 005300.? 30. 0‘05 0303...! cook 90:010.! 0.3 0:02.08... 0.80 a.“ 08 1.3.3 0.3. .3053an 0.— .030 .050 03 H - / 08: 400.8 030.3.- 03005 3004.. no! 51...- .3000000l 53.041 5—da— ‘Oo 95.93300— 4 b 0 103130Q< o 0. 190.- 40.0 09:0..— goo—.2 00A..- 0’83..- 30003000— 20000000.! 0333- ‘- usgioul / v1.5.0 03 0033000 Isa-3.0301 via..— 16. is: < 34 ecu-non. 03— '50000000— 03.3.0.3 , \ a. . Ouunaou. vi- 03991000— quota 08:3... 0.00:3. nest-“0900m— L .0030: icon—Hugh 0050 0.0.00— 00:06 44 425: H2833. Eoguzow 23 mo 3.008 coutu— ucwvdflsm--.q.m ouauwu 313-100 «coal-2'5 6 32.3 .390. fiuggutu< .n oOZCt-b 3.3000- ?nfiunz .N .08 13.890. 5 333.3, 32.88% .- OEZQOUU< FHSDD Ava—U; 02¢ (ASHES—U 36900“; bang-IOU .3358 325-02: 5335.30 5.. .033...» §09< .35... 3 953...! nag-accounu actuatohm .33» 0.3. ul— 3:36.500 conic-mobs- uo ~23..— Jtu'év'fiuo'h' UEflv-G; GZ< Gaga—nu .ZO—PUDOOIQ J<¢BJDU~IO< coo—Ln— .888 «34.3300 30.5.2.3 can... 00:.— .xl .03.»; .3939:— vouauga he: goof:— non—{oun— ooofa nan-Doha .032— «01.33 novann 110’ gauge HUS—n.— o o a o o - N n ' n .03:— 2000 gonna. 3:33! A 00: .33> v.34 . oil!- .u. in. near—m vilihg< 3:359 2.35.02: uo— vans—OD .v 603029—033 .n 0602.93? 1.34 .~ - 03:1. 59:— olun?» woe-5003.0 .— Quin-950 60333000: “loanvfld ZO—Haoaaoe efloemoeee omcoamom HHmEm puma omcommom omhmg mama omcommom o.H omcoamom xufifiwnmpmeHQ 58 Moreover, cocoa replanting not only deprives the farmers of their major income source, but also requires additional financial outlay. Hence, if the farmers have to borrow externally for the new produc- tion alternative (either from private or public sources) they would be reluctant to do so since the total investment costs must now include the interests and other carrying charges, and the additional risks from insolvency or foreclosure of assets that are offered as collateral to secure the loan. The cash inflow and outflow generated for food production is quite different from cocoa production. The financial payoff period or turnaround time of food production occurs typically within the year, whereas the payoff period of trees generated from the cash inflow derived from their output does not occur until some years after the trees' gestation stage. Furthermore, the elasticities of transformation among food crops as annuals or biennials are much greater than between food crop and cocoa, because of the partial fixity or the less liquid nature of the latter's asset structure. In the extreme case, should output prices decline, and farmers decide to grow another commodity, the salvage value of the trees may not only be very low, it may be negative since additional cost is neces- sary to remove the trees. Finally, farmers typically tend to be more realistic (and skepti- cal!) in evaluating the relative profitabilities by imputing their own discount functions to further reduce the calculated returns. This healthy pessimism is based on past unfavorable experience with the government extension service. Since they have to bear the major burden of loss should the productive alternative recommended by the 59 government fail, they are naturally more cautious and financially con- servative. The response threshold values between land presently in traditional cocoa production and modern (replanted) cocoa is 0.4. This means that before the farmers would replant their present trees with the higher-yielding trees, the expected returns from the re- planted trees have to be 40 percent greater than the expected income from the present trees. The response threshold value between the present food land and modern cocoa (new planting) is 0.4. Capital and Credit Constraints In order to compute the final land use allocation transition rates, the reSponse functions are constrained by the availability of capital necessary for the production alternatives. The two major sources or potential sources of capital in the cocoa industry for agricultural expenditures are (1) savings from disposable income (after accounting for consumption) within the sector that could be channeled into investment, and (2) the increased credit or collateral value from the increased capitalized value of cultivated land, which in turn may increase the flow of capital, public or private, from outside the cocoa industry. However, there are two major caveats in this conceptualization for the Nigerian context. First, there may not be any financial institutions and intermediaries that can channel the capital flow or appreciably mobilize the untapped financial resources inside or out- side the cocoa-food zone for agricultural expansion. Although the sectorial aggregate accounts may show the income (after accounting for production and consumption) to be positive, because of the skewed 60 distribution of income and asset ownership within the sector, and the outside linkages some of the larger farmers may have with the urban sector, the net savings may be channeled into investments outside of agriculture, rather than inside the sector. These investments, such as urban housing, have higher and quicker financial returns. Hence, the potential capital source within the private sector of the economy may not be effectively available to the majority of the farmers for agricultural expansion. Second, even if these financial institutions or mechanisms exist, they are rudimentary because of the present limited scope of their Operations. For example, the actual credit extended to cocoa farmers may be far below their collateral level; or conversely, the costs of borrowing may be exorbitant and not reflect the attendant risks of such financial transactions. The demand for capital in any one year for agricultural ex- pansion is the annual establishment costs required to maintain the crop before production. The total demand for capital for each alternative is then compared with the total capital available from the two above sources, to determine the effects of the capital con- straint. This finally determines the amount of land allocated and reallocated for agricultural expansion. Asymmetric Response In this model, we have also built the mechanisms that would determine the rate of abandonment of cocoa land that may result from adverse cocoa producer prices. The abandonment would begin to take 61 place if the current returns are negative, and the remaining ex- pected profitability from the trees is below an abandonment threshold level (see Figure 3.6). The abandonment threshold level of the cocoa trees (analogous to the investment threshold level) depends on the farmers behavioral characteristics, financial attributes, and the alternative uses of the present resources of land, capital, and labor. The abandonment rate would increase up to a predetermined maximum as the profitability continues to fall. The values of the threshold and abandonment points (which attempt to capture by proxy the far- mers' investment and disinvestment decisions) are not symmetrical. Their values, together with the values of the positive response and abandonment response rate, depend on the computed relative profit- ability of the commodities. 3. Agricultural Production and Marketing Component This component determines (l) the amount of food that is produced and consumed within the household and sold and purchased in the market; (2) the amount of cocoa that is produced, marketed, exported; and (3) the agricultural input requirements and some of the major economic performance indices used to evaluate the cocoa economy. Determination of the Subsistence Level of Food Production The amount of food produced and consumed in the sector de- pends on the interactions between the effective demand and responsive supply of the food market. Nigerian food prices fluctuate daily and 62 .mconfiooo pcoscovcmn< was; HmHSufisofiHw< cw :ofiHUGSm omconmom aquacovcmn< mcfizonm Euummfio--.o.m enema» xuflfianmuMMORm vfiocmohzk o>wumfiom ucoacowawn< oumm omfioammm ucoecoecmn< .\ llllll'lllll ovum ucofidovcan< asefixaz 63 seasonally because of the weather's effects on production, and in the longer term because of the ineffectiveness of the cash food market with its many intermediaries between the producers and con- sumers. However, in this study, we are more concerned with the annual food price behavior. The total staple food produced in the cocoa-food sector can be divided into the subsistence and the cash food prOportions. We will assume that 80 percent of the food pro- duced on farms is also consumed on farms. The rest of the food is sold and purchased in the cash food market. The degree the farmers would rely on the cash food market is determined by the agricultural population's total demand for calories, the stability of cash food prices, the food price level, and the farmer's total cash income from cocoa. For example, if annual cash food prices fluctuates a great deal, it will increase the farmers reliance on their own production to meet their consump- tion. In addition, the reliance also depends on the ratio between the value of the cash food consumed by the agricultural pOpulation, and the total cash income farmers obtain from cocoa. Thus, de- creases in either the market price of food or costs of agricultural production, or increases in cocoa output or cocoa price would in- crease farmers reliance on the cash food market. Annual Average Yield of Output The three factors that determine the annual average yield of food and cocoa are (1) the secular trend effect as past experience, improves farmer production methods; (2) the intermediate change in 64 yield potential resulting from adoption of new, improved, culti- vational and managerial practices in the preceding years, and (3) farmers' short-term harvest response to the prevailing price. The inclusion of the first factor is an attempt to model the often-observed phenomenon where yields from established trees in- crease in time as farmers, learning from one another and personal experience, improve on their cultivational methods, without neces- sarily adopting any new technology. The yield would thus move towards its biological potential as a function of a time-variant learning curve. Although the effect of the time-variant output in- crease is very small, we have included its contribution in computing the annual average yield of cocoa and food. When farmers rehabilitate their present cocoa trees by using improved cultivational methods, the output increase is often lagged. For example, the use of fertilizer on cocoa trees does not increase cocoa output in the same year, but a year later, after which the output-increase resulting from the initial fertilizer use begins to decrease unless fertilizer is applied continuously. The most significant factor that determines the annual output from cocoa trees is the short-term output response. This is obtained by multiplying the ratio between the prevailing producer price and the normal expected price (assumed to be the exponential average of the past ten years' producer prices), by the exponential value of the harvest response elasticity. In this study, the harvest elasticity of cocoa is assumed to be 0.05. If the prevailing pro- ducer price (the numerator of the ratio), is higher than the normal 65 price (the denominator of the ratio), the price ratio will be greater than one. Hence, the short-term output reSponse, which is obtained by multiplying the exponential value of the harvest elasticity, is positive. If the prevailing producer price is much higher than the normal price, the price ratio is also larger. The consequent short- term reSponse for that year is likewise higher. However, if subse- quent producer prices continue to remain high, the value of the normal producer price in the denominator also increases, which in turn, decreases the price ratio. The decrease in the price ratio results in a relative cutback of the short-term output reSponse even though the producer price may still be increasing, as farmers regard the higher price as "normal." Food and Cocoa Production In computing food production, the component first calculates the food land necessary to meet the subsistence demand of the agri- cultural population. The remaining food land goes for cash food production. Total annual food production is the sum of the product of the annual average food yield per acre and the subsistence and cash food acreages, and the amount of staple food intercropped be— tween gestating cocoa trees. Total cocoa production is the sum of the product of the average yield of the various productive charts of the four cocoa substreams, and their respective acreages. Food and Cocoa Marketing, The amount of food sold in the cash food market is obtained by subtracting the subsistence amount consumed on farms from the 66 total food output (after accounting for the food loss due to spoilage and waste). In this study, we shall assume the market loss factor for food to be 5 percent, and cocoa to be 20 percent. Since the domestic consumption of cocoa is very low, we assume that all the cocoa produced (after accounting for the field loss) is sold. We further assume that all cocoa sold is also exported within the same year, with no government inventory. Input Accounting_ The component also computes and accounts for all the input requirements--land, labor, and material, and their respective costs for food and cocoa production and marketing. In this study, we assume that the supply functions of all the purchased inputs are perfectly elastic and the inputs are available at an average constant per unit cost, with no constraints. This may not necessarily be true in reality. The derived supply functions of the various inputs may vary according to the amount produced and, with a time lag, the demand's interaction. The supply functions of the inputs may even be discontinuous if there are any institutional constraints or market barriers. Output Accounting Finally, the component computes the total annual income generated from food and cocoa, and the capital formation accrued to the agricultural land. The total income from food is obtained by adding the income in kind for the food produced and consumed on the farm, and the 67 income from the food sold at the farm—gate level in the cash food market. Since all the cocoa output is marketed, the total income from cocoa is obtained from multiplying output by the producer price in the prevailing year. The capitalized value of agricultural land is obtained by dividing the annual average returns in an acre of land by the pre- vailing interest rate, which we assume to be 6 percent throughout the planning horizon. The total capitalized value in the ecological zone is the sum of the values of the total acreages. The approach used to obtain the capitalized value may be an oversimplification, since it does not take into explicit account the differential income- generating capacity of the cocoa trees of varying productive stages. Nevertheless, it provides a very useful quantitative basis to account for the capital formation in agricultural land which may result from the expansion of cocoa and food acreages. In turn, the increase in the capitalized value of agricultural land not only increases the 'wealth level' of farmers but also the collateral value of their assets, enabling them to borrow more capital should they need it for further agricultural expansion. It should also be mentioned that the capitalized value of an acre of agricultural land can be in- creased by the increase in output, output price and decrease in the cost of production. Furthermore, the change in the interest rate in the economy affects the capitalized land value. 4. Price Generation Component This component calculates (l) cocoa producer prices from world prices which are specified externally at the beginning of the 68 policy experiment, and (2) the food producer and market prices which are determined endogenously throughout the simulated time period of the planning horizon. Cocoa Prices The three relevant price functions in this study of the cocoa industry are (1) the world price which the economy gets for its exports, (2) the market price which the marketing board gets after accounting for the export duties and produce sales tax, and (3) the producers' price which the farmers get after accounting for the marketing board trading surplus and the administrative and handling costs from the marketing price. In this study, the world price function of cocoa is Specified exogenously at the onset. Ideally, the world price function that Nigeria faces should include the dynamic interactions of Nigeria's output in the world market because a relative increase in that out- put may create an excess supply, thus decreasing the long-term world price. The annual foreign exchange generated from cocoa exports is obtained by multiplying the annual export by the annual cocoa world price. To simplify the analysis, we combine the export duties and producer sales tax as government revenue, and further assume that the revenue, marketing board trading surplus and the handling charges are predetermined proportions of the world price and market price of cocoa. According to the government revenue and marketing board stabilization policies, these proportions could also vary in rates and time duration. 69 Food Prices The market price of food in the cocoa sector is conceptual- ized to depend on the total supply and demand of the cash food market in South Nigeria. It is calculated by using the price elas- ticity (assumed to be -O.3) for the demand of food in all of the region. On the other hand, the producer price of food is obtained by subtracting the 70 percent assumed to be detained by the various market intermediaries from the market price of food. We have not attempted to model in detail a sector-specific food price determina- tion mechanism. 5. Policy Entries Component The model is built to provide a quantitative basis for pro- jecting the likely consequences of (l) the government production campaigns to encourage cocoa farmers to replant their relatively low—yielding trees with higher-yielding trees, and newly plant the higher—yielding trees in land now in food or bush; and (2) changing the government revenue and marketing board producer pricing policies. Production Campaigns The three basic analytical features in the production campaign executive routine are (l) a time profile to show when the prOgram is first initiated, begins to reach its maximum, phases out, and terminates; (2) a budget profile to show the annual budget and total budget appropriations; and (3) an intra-budget allocation for administrative overhead and technical assistance and personnel, 7O subsidy for inputs, and direct cash grant to the farmers in the production campaign package. Government Revenue and Marketing_ Board Pricing_Policies As previously discussed, the total taxes the government and marketing board collect from the cocoa producers are the export duties and producer sales tax and marketing board trading sur- pluses. These we assume are predetermined proportions of the reSpective prevailing world price and market price. The annual government revenue and marketing board trading surplus are then accumulated for comparison. These prOportions, however, may be varied, phased out, or cut off according to the government fiscal policies. In this study, we introduce a rudimentary guaranteed pro- ducer floor price feature (recommended by CSNRD), in which cocoa farmers are paid at least the floor price should the producer price (after accounting for the revenue tax and the accumulated trading surpluses) go below that level. The guaranteed producer price of cocoa is supported by previously accumulated trading surpluses. As we shall see, this guaranteed price is especially relevant if future world prices of cocoa are expected to be low and fluctuating. The producer price guarantee Operating from the accumulated marketing surpluses may thus stabilize the farmers' annual income from the fluctuations of the world price of cocoa, and increase their income in the years when the world price of cocoa is low. 71 6. Aggregative Sector Accounting Component This component determines the total receipts, costs of pro- duction, diSposable income, consumption, savings, investment and credit of the cocoa-food sector. In our description, we shall concen- trate primarily on arriving at the annual agricultural disposable income per capita of the sector. This performance index is used to evaluate the outcome of the cocoa policy experiments. Total Receipts and Expenditures The annual total receipts in the zone are obtained by adding the income from the production of cocoa, the income in kind and cash from food, and the investment in the form of loans coming from outside the sector. The total expenditure by the sector includes the total cost of production for the use of chemicals, biolOgical materials, and the amortized cost of equipment, debt services and interest pay- ments as well as cash food expenditures. Agricultural Disposable lncomepper Capita in Cocoa Sector Total disposable income is obtained by subtracting total receipts from total expenditures. Disposable income is further con- strained to cover at least the other major nonfood expenditures such as housing, poll tax, children's school fees and festivals. Any shortage of income is made up by external borrowing. Finally, sectorial per capita disposable income is obtained by dividing the total disposable income by the population which is assumed to increase 72 by 2.2 percent per annum.3 It is obvious that per capita income would be affected directly by the change in the rate of the population in- crease. In this study, we are more interested in the relative changes of the projected outcome under alternative government cocoa policies than in the absolute values of the projected outcome. Thus, a dif— ferent rate of population increase would not appreciably affect the relative order and values of the projected outcome unless the actual population rate increase is drastically different from the assumed rate. Such a drastic increase would affect the fundamental, under- lying conceptual framework of the model and analysis. Model Validation and Testing Before we present the cocoa policy experiments and their simu- lated results in the following chapter, we shall discuss briefly the validation tests conducted by the Nigerian Agricultural Simulation Team. These tests determine how well the system model simulated the relevant behavior of the real system. In addition, we should view the present phase of policy experiments conducted on the cocoa economy as part of the continuing process of the validation of the system model, and its usefulness as a tool of policy analysis and develop- mental planning. The projected results of the model (using alterna- tive policies) can be compared in terms of the reasonableness with received economic theory, other empirical observations and the judg- ments of experts on the Nigerian agricultural economy. 3For more information, see Chapter VI, "The Population Model," in Manetsch 33.21: (1971). 73 There are three major overlapping and reinforcing ways to validate the model. The first method (alluded to in the above dis- cussion) compares the structure of the model and its simulated out- come using alternative assumptions about its behavior from experts on the economy and other published secondary sources. However, this process may be too personalized, judgmental, intuitive, and rather difficult to replicate. The second approach is more rigorous and concrete. Behavior predicted by the model under various policy conditions can be com- pared with what actually occurs as real time passes under the same conditions. However, this approach is not very useful for policy planning, since the purpose of the model is to project the possible consequences of the proposed policy alternatives before they occur! Nevertheless, when real time passes, the model can be tuned and up- dated as an on-going process by comparing the simulated results with the real world data. A complementary approach is to track the simu- lated results with historical data from the real world which are not used in the model-building process. The model can be tuned by track- ing one or more time series of past behavior by adjusting (through sensitivity tests) the values of certain system parameters, or in some instances, restructuralizing the computing mechanisms that govern the modeled behavior. Time-series tracking and sensitivity tests, as an interwoven process of model validation, require an understanding of the real system and the simulation model in order to zero-in on the meaningful parameters and/or relevant structure for adjustments. This allows the simulated behavior to conform to experienced behavior. 74 Four sets of time series (exports of cocoa, palm oil, rubber, and market food prices) from 1953 to 1965 were used to tune the Southern model with the goodness-of-fit measure. Goodness of fit was one of the many possible criterias used to determine the closeness between simulated and the observed real world data (see Table 3.2). The goodness-of—fit is measured by the squared normalized deviations between the observed real world data and the simulated value in each year of the four time series. Hence, the closer to zero the squared deviations, the better the fit between the observed real world data and the simulated data. In addition, the author also compared the model's projected total cocoa acreages from 1966 to 1995, using the values of the model reported in Manetsch 53.31: (1971) with the projected acreages con- tained in the FAO study of the Nigerian cocoa economy. It was dis- covered that the simulated projection of the total acreages (assuming the same level of government expenditure on the production campaigns) was unreasonably high at 4.2 million acres in 1995. Hence, by ad- justing the value of some of the model's parameters, the projected cocoa acreages in 1995 were reduced to approximately 2.7 million acres, a much more reasonable estimate. The readjustment was accom- plished primarily by reducing the value of the acreage presently in bush that is available for new cocoa planting. The effect of the reduction of the value is to set an upper limit in the rate of cocoa acreage expansion. Finally, sensitivity tests, which identify the model's parameters whose outcomes are most sensitive to their value changes, 7S .fluaosv eaxo< "ooeeom . . . monmscm aomomm. Hammwmo eoomwN oomoam we saw «Heoflfio. oomamflo. .~e~.oeH .wmo.~mH .Nmo.mm~ .ooo.omm .mom.mme .oo~.Hem mom“ mmnefifio. oomNHHo. .meo.mm~ .mme.HoH .mao.mm~ .ooH.oom .Hoa.eme .om~.Hee eomH mNNNHHo. commoHo. .Nma.mmfi .Hme.HeH .wmm.mm~ .oe~.~w~ .emo.mHe .ooo.~mn moma aHmoHHo. oomamHo. .oflm.a~3 .owm.mmH .eam.~o~ .ofiw.mom .efia.~om .omo.ome Nomfl «memofio. oommmfio. .aas.o- .,.eam.mNH .om~.mflm .omo.wom .amu.oam .eom.HHe Homfl emmmofio. ooomNHo. .ama.m- .mmfl.w~H .eme.mam .o~a.ofie .aHe.mmm .eao.Nmm coma oaowoflo. oomeoHo. .me.oHH .wmm.mHH .wwa.e~m .oao.oom .mao.mmm .Nam.oan amma weamoHo. oooaflmoo. o.omm.om o.Hom.Nm .Nmm.eem .wmm.me .Haa.mom .Hmm.omH «was oeeHoHo. oomowflo. ~.eom.am o.~wm.ew .omm.~em .mm~.~5m .aem.om~ .Neo.mom ammfi oooooHo. ooaHeHo. a.m~m.ma o.eme.mw .emo.~6m .omm.eae .m~e.ma~ .mam.~o~ omaa ammooflo. oomouHo. o.oom.ea o.~mo.wo .emm.oam .ooo.woe .mma.ao~ .meo.moH mmmfi mmaooHo. oomwan. w.~n~.ao c.6Hw.oe .Amm.mmm .ooo.eoe .ooo.mm~ .mmm.o- emmH osmooflo. ooooofio. m.aH~.am o.-o.ae .owm.ome .mao.fime .mmm.me~ .moe.en~ mmoa woumHsefiw mama woumasefim mama woumfisefim mung woumfizeflm mung m.ofl\av fi.es\.mes .mooeov n.es\.mofl .msoeov A.es\.mea .meoeov e66» npsom mo muuonxm Magnum muuomxm Hwo Efima muuonxm mooou mowed coon poxnmz .moofi-mmma .neosomv mousse oooe oozes: .moeooxm Hones“ .moeoexm Hfio same .mpuoaxm mooou cmfiuomwz mo mama woumfizswm use vo>Homno mo wcwxoanp mowuomuoafibuu.m.m mam

O c: 2 0.) E 200 L 2 Revenue 5 53 Revenue 2 '8 Revenue 3,4 g Surplus 3,4 Revenue 1 '3 100 5 E :3 U O < / _‘r Surplus S M. 1 l .1 l “v 1970 1975 1980 1985 1990 1995 Figure 4.A7.--Accumulated Government Revenue and Marketing Board Trading Surpluses from Nigerian Cocoa Production under Indicated Policy Alternatives Assuming Moderate (Most Likely) World Cocoa Prices, 1970-1995. 99 Runs 2 and S, on the other hand, are relatively close since their projected outputs are very close and their government revenue rates are the same. However, we should interpret the results of the projected agricultural income per capita, and government accumulated revenue and trading surpluses with caution. Government revenue and trading surpluses under the various policy runs in the simulation experiment are simply accumulated. No attempt is made to show the dynamic, multiplier effects on the total economy that may result from the government's reinvestment of the accumulated funds. In reality, the government would probably invest the accumulated funds on projects that may directly, or indirectly affect the welfare of the farmers. For example, the government could invest the accumulated funds gener- ated from Runs 1, 2, and S, in public service projects, such as health and education, which could increase the welfare of the farmers, without necessarily increasing their income level. However, the present model is not able to show the nonmonetary benefits that may result from these investments. Hence, there is an upward income bias in the projected income of Runs 3 and 4, when compared to the projected income of Runs 1, 2, and S, where producer taxes prevailed throughout the planning horizon. Although it is not within the scope of the thesis to discuss the merits of alternative agricultural taxes, it should be pointed out that the loss in government revenue from the producer tax cut- off or phase-out can be compensated by imposing alternative forms of taxes on the farmers. For example, the government can also collect 100 taxes from the farmers increased income, asset land value, or from their increased purchase of producer and consumer goods that result from the increased income which has, in turn, resulted from the in- crease in producer prices augmented by the subsequent increases in production and productive capacity. However, because of the distri- butional differences and inter—temporal trade-offs, such revenue compensations are not equivalent to one another (Hicks, 1969). Results of Policy Experiment Assuming_ High World Cocoa Prices In this set of runs, future world cocoa prices are assumed to be higher than in the previous set. The price increase is more rapid and it remains at a higher level than under the moderate price assumption. As a result, the absolute producer prices under the five pricing policy options are also higher (see Figure 4.81). Conse- quently, the farmers' short-term and long-term output responses differ from the responses under the moderate world cocoa price assumption. In general, because of the higher producer prices, the increase in the annual amortized returns the farmers expect to get by growing more cocoa trees is greater than the increase in the annual amortized returns they expect to get from their existing trees. Hence, the effects of the higher producer prices on the long-term output response are potentially greater than the short-term output response. Subject to meeting the other production requirements, cocoa farmers would allocate more of their resources for cocoa acreages expansion rather than increasing their short-term harvest output. 336 b 280 224 168 World, Market, Producer Prices of Cocoa (B/long ton) 112 L 1970 101 Run Definitions Run 1 Low tax rates and no production campaigns. Run 2 Low tax rates and production campaigns. Run 3 High tax rates cut off in 1971, and production campaigns. Run 4 High tax rates phased out by 1976, and production campaigns. Run 5 Low tax rates with producer floor price guarantee, Market lPrice 3 and production campaigns. Market Price 4 Producer Price 3 J \ Market Market World Price Price 3,4 Price 1,2,5 Producer Price 3,4 ‘— Producer Price 4 l 1975 Producer Price 5 Guaranteed Producer Floor Price 1 1980 1 1985 __—‘ Producer Price 1,2 1 1990 1 1995 Figure 4.Bl.--Market and Producer Prices of Nigerian Cocoa Under Indicated Policy Alternatives Assuming High World Cocoa Prices, 1970-1995. 102 Specifically, the long-term response of Run 3 where the pro- ducer price was increased dramatically with the tax cut-off is greater than Run 4 where the taxes were gradually phased out. Con- sequently, the projected annual output of Run 4 between 1970 to 1983 is greater than in Run 3 because farmers in Run 3 are allocating more of their resources to grow more cocoa trees rather than to further increase their output from the existing lower-yielding trees. How- ever, by 1983 when the trees that are grown in the 19705 come into production, the projected output of Run 3 (which also has the highest total cocoa acreages) surpasses Run 4. Thus we see that the effects of the greater long-term output responses in Run 3 are lagged and do not manifest themselves until many years later when the new cocoa trees come into production. In the intervening years, the total annual output is lower than what it would be if the farmers had not responded to the higher producer prices by expanding their output capacity. Because of the guaranteed producer floor price of Run 5, its projected output is higher than in Run 2. For the same reasons discussed previously, the relative ranking of the other performance indices correSpond to the ranking of the projected output (see Figures 4.82 to 4.87. Results of Policy Experiment Assuming_ Low World Cocoa Prices In contrast with the previous world price situation, a set of lower world prices is assumed. Consequently the values of other performance indices are also reduced. As seen in Figure 4.Cl, the producer prices in Runs 3 and 4, after 1985, are slightly below the 103 Total Cocoa Output (thousand long tons) Run Definitions Run 1 Low tax rates and no production campaign. Run 2 Low tax rates and production campaigns. Run 3 High tax rates cut off in 1971, and Output 3 production campaigns. 700 ” Run 4 High tax rates phased out by 1976, and production campaigns. Out t 4 Run 5 Low tax rates with producer floor 1",_ Ofigput 5 ' price guarantee, and production campaigns. 600 ~ 500 p 400 ‘ 300- r‘g=§é=’ ‘_“““~——- Output 1 200%- 100‘ 1 1 3 1 1 1970 1975 1980 1985 1990 1995 Figure 4.82.--Total Projected Nigerian Cocoa Output Under Indicated Policy Alternatives Assuming High World Cocoa Prices, 1970-1995. 104 Run Definitions Run 1 Low tax rates and no production campaign. 300 _. Run 2 Low tax rates and production campaigns. r5 Run 3 High tax rates cut off in 1971, and production 5 campaigns. 2: Run 4 High tax rates phased out by 1976, and production I: campaigns. E Run 5 Low tax rates with producer floor price guarantee, '" and production campaigns. 5, 250 i— Run 3 (I) .p S &' Run 4 Lu m 8 200 - Run 5 o c: 5 Run 2 H “.4 'u 0 g 150.— H o c Q) t: o 00 5 .5 100.— x LL} : CO -d a _ A, ’——Rfifi.l L2 7/ '3 50 a c 5 Jr 1 .1 l L 1970 1975 1980 1985 1990 1995 . Figure 4.B3.--Foreign Exchange Generated from Nigerian Cocoa Exports Under Indicated Policy Alternatives Assuming High World Cocoa Prices, 1970-1995. 105 Run Definitions Run 1 Low tax rates and no production campaign. Run 2 Low tax rates and production campaigns. Run 3 High tax rates cut off in 1971, and production campaigns. 120" Run 4 High tax rates phased out by 1976, and production campaigns. Run 5 Low tax rates with producer floor Run 3 price guarantee, and 3; production campaigns. Run 4 E 100 F (6 p "-4 CL (6 L) Run 5 H 3 u —\ 80 2 Run 2 O U E 2 .D O o. . m -r-4 :2 '3 H a S 40 L- 8 Run 1 "-4 5.. 00 <{, 20 F l l l l l 1970 1975 1980 1985 1990 1995 Figure 4.84.--Agricultural Disposable Income Per Capita in Nigerian Cocoa Sector Under Indicated Policy Alternatives Assuming High World Cocoa Prices, 1970-1995. 106 Run Definitions Run 1 Low tax rates and no production campaign. Run 2 Low tax rates and production campaigns. Run 3 High tax rates cut off in 1971, and production campaigns. Run 4 High tax rates phased out by 1976, and Total 3 production campaigns. Run 5 Low tax rates with producer floor price guarantee, and production campaigns. 3.0" ‘— Total 4 “Total 5 ,5 “Total 2 2’3 a 8 2.5 -- n: O o: E 8 2.0 h- d) In "‘ / Total 1 (U 0 U 8 Traditional 2,3,4,s E 1.5 p- -o g +— Modem 4 J "Modern 5 2 Modern 2 O S 1.0 h— or-c -o m a E— F: (U p O F 0.5 *- l 1 1 l L 1970 1975 1980 1985 1990 1995 Figure 4.BS.--Total, Traditional and Modern Nigerian Cocoa Acreages Under Indicated Policy Alternatives Assuming High World Cocoa Prices, 1970-1995. 107 Run Definitions Run 1 Low tax rates and no production campaign. Run 2 Low tax rates and production campaigns. Run 3 High tax rates cut off in 1971, and production campaigns. 1200 -’ Run 4 High tax rates phased out by 1976, and Run 3 production campaigns. r\ Run 5 Low tax rates with producer floor 5 price guarantee, and Run 4 :3 production campaigns. H I-- "-4 E 1000 Run 5 an E Q) :3 To > 800b Run 2 'o t: «3 .—I a a 3 g 600” U "-4 H b0 <3: 'o 3 an 1 W4 '3 400;“ p "-4 f? / U 200 l 1 l l 1 1970 1975 1980 1985 1990 1995 Figure 4.86.--Capitalized Agricultural Land Value in Nigerian Cocoa-Food Ecological Zone Under Indicated Policy Alternatives Assuming High World Cocoa Prices, 1970-1995. 108 Run Definitions Run 1 Low tax rates and no production campaign. Run 2 Low tax rates and production campaigns. r‘ 600 L. Run 3 High tax rates cut off in 1971, and production : campaigns. S 2 :3 Run 4 High tax rates phased out by 1976, and urplus :: production campaigns. E Run 5 Low tax rates with producer floor price .m guarantee, and cocoa production a; 500 __ campaigns. 2n 0) U) ,3 E- Surplus 5 3 (I) g? 400 >- "-4 'u (U a an '2 Surplus 1 m . g 300 i- 5 5 Revenue 2,5 as p I: 2% E 200 '- Revenue 1 3 o L: B Surplus 3,4 3 Revenue 3,4 SE} 100 h 5 U 0 < ‘45:! __ 1 1 1 1 1970 1975 1980 1985 1990 1995 Figure 4.B7.--Accumulated Government Revenue and Marketing Board Trading Surpluses from Nigerian Cocoa Production under Indi- cated Policy Alternatives Assuming High World Cocoa Prices of 1970-1995. 336 280 224 168 World, Market, Producer Prices of Cocoa (k/long ton) L . , 109 Run Definitions Run 1 Low tax rates and no production campaigns. Run 2 Low tax rates and production campaigns. Run 3 High tax rates cut off in 1971, and production campaigns. Run 4 High tax rates phased out by 1976, and production campaigns. Run 5 Low tax rates with producer floor price guarantee, and production campaigns. Market Market Price 3 Price 4 World Price Producer Price 3 Market Price 3,4 'K‘~Market Price 1,2,5 ‘- Producer Price 4 Producer Price 5 Guaranteed Producer Floor Price'K\producer Price 3 4 112 Producer Price 1,2 1970 1975 1980 1985 1990 1995 Figure 4.Cl.--Market and Producer Prices of Nigerian Cocoa Under Indicated Policy Alternatives Assuming Low World Cocoa Prices, 1970-1995. 110 guaranteed floor price of £168. Since the relative ranking of the performance indices in this set of runs is basically similar to the ranking derived from the moderate world price set, we shall highlight the more interesting simulation results. From Figure 4.C2, we see that because of the higher initial producer price of Run 5 (with the producer price guarantee) in 1970, the short-term output response is such that the projected output of Run 5 in that year is also the highest. However, in subsequent years when the producer prices of Runs 3 and 4 (with the tax cut—off and phase-out features) are higher, the projected output of Run 5 is lower than in Runs 3 and 4. As in previous situations the long term responses in capacity expansion of the various runs, interacting with the government production campaigns, depend on their respective producer prices. For example, because the producer price of Run 5 is highest in the initial years, most of the trees planted are of the traditional variety. Likewise, since the producer price increase in Run 3 also occurs in the initial years, the expanded output capacity consists more of the Amelonado trees. On the other hand, because the peak of the production campaigns coincide with the gradual producer price increases of Run 4, most of the new acreages consist of the Upper Amazon trees (see Figure 4.CS). The differences in the compositional acreages in turn manifest themselves in subse- quent years with the relative ranking of the projected output of the various policy runs. Hence, deSpite the slightly higher producer prices of Run 5 from 1985 upwards, its projected annual cocoa output is lower than in Runs 3 and 4. 700 600 ’u? 8 H to g 500 H ‘U 5 01 5 2 :5 400 U a O. H :3 c> d 3 300 O U ,4 d p O f. 200 100 Policy 111 Run Definitions Run 1 Low tax rates and no production campaign. Run 2 Low tax rates and production campaigns. Run 3 High tax rates cut off in 1971, and production campaigns. Run 4 High tax rates phased out by 1976, and production campaigns. P Run 5 Low tax rates with producer floor price guarantee, and production campaigns. Output 4 Output 3 _ Output 5 Output 2 x Output 1 p l l ;l l J 1970 1975 1980 1985 1990 1995 Figure 4.C2.--Tota1 Projected Nigerian Cocoa Output Under Indicated Alternatives Assuming Low World Cocoa Prices, 1970-1995. 112 Run Definitions Run 1 Low tax rates and no production campaign. 300 Run 2 Low tax rates and production campaigns. " Run 3 High tax rates cut off in 1971, and production campaigns. Run 4 High tax rates phased out by 1976, and production campaigns. Run 5 Low tax rates with producer floor price 250'- 200'- 150!- 100 p- 50 P' Annual Foreign Exchange Generated from Cocoa Exports (N a million) 1970 guarantee, and production campaigns. Run 1 1 ' l 1 1 4 1975 1980 1985 1990 1995 Figure 4.C3.--Foreign Exchange Generated from Nigerian Cocoa Exports Under Indicated Policy Alternatives Assuming Low World Cocoa Prices, 1970-1995. 113 Run Definitions Run 1 Low tax rates and no production campaign. Run 2 Low tax rates and production campaigns. Run 3 High tax rates cut off in 1971, and production 120'- campaigns. Run 4 High tax rates phased out by 1976, and production campaigns. Run 5 Low tax rates with producer floor price guarantee, and production campaigns. 100}- 30+- Agricultural Disposable Income Per Capita (N b ) Run 5 Run 4 Run 3 Run 2 I 40L. Run 1 20 - L l L 1 1 1970 1975 1980 1985 1990 1995 Figure 4.C4.--Agricu1tural Disposable Income Per Capita in Nigerian Cocoa Sector Under Indicated Policy Alternatives Assuming Low World Cocoa Prices, 1970-1995. 3.0 _' 5’ u: I F‘ c: If Total, Traditional, Modern Cocoa Trees (million acres) H H E: 01 l O 9" 1970 Run Run Run Run Run Run Definitions 114 1 Low tax rates and no production campaign. 2 Low tax rates and production campaigns. 3 High tax rates cut off in 1971, and production campaigns. 4 High tax rates phased out by 1976, and 5 Low tax rates with producer floor price guarantee, and production I 1975 production campaigns. campaigns. I 1980 _! 1985 l 1990 Total 3 ./ i_——————-'To \ J 1995 Total 4 Total 5 ‘\ Total 2 tal 1 Traditional 5 Traditional 3 ;; Traditional 4 Traditional 2 Modern 4 Modern 3 Modern 5 Modern 2 Figure 4.CS.--Total, Traditional and Modern Nigerian Cocoa Acreages Under Indicated Policy Alternatives Assuming Low World Cocoa Prices, 1970-1995. 115 Run Definitions Run 1 Low tax rates and no production campaign. Run 2 Low tax rates and production campaigns. 120d-' Run 3 High tax rates cut off in 1971, and production campaigns. Run 4 High tax rates phased out by 1976, and production campaigns. Run 5 Low tax rates with producer floor price guarantee, and production campaigns. 1000" Run 5 Run 4 Run 3 80*- Run 2 600‘- 4ooH.. / Capitalized Agricultural Land Value (N'E million) V 200 u- l l I l J 1970 1975 1980 1985 1990 1995 Figure 4.C6.--Capita1ized Agricultural Land Value in Nigerian Cocoa-Food Ecological Zone Under Indicated Policy Alternatives Assuming Low World Cocoa Prices, 1970-1995. 116 Run Definitions Run 1 Low tax rates and no production campaign. Run 2 Low tax rates and production campaigns. Run 3 High tax rates cut off in 1971, and production campaigns. 600r- Run 4 High tax rates phased out by 1976, and production campaigns. Run 5 Low tax rates with producer floor price guarantee, and production campaigns. 500‘- Surplus 2 400 I- 300 _ Surplus l ,————- Revenue 5 00 F’ 2 ‘I‘\“~ Revenue 2 Accumulated Government Revenue and Trading Surpluses (N i million) 'evenue l 100 !- Revenue 3,4 Surplus 3,4 // .4: L 1 1 1 L 7' 198 1970 1975 0 1985 1990 1995 Surplus 5 Figure 4.C7.--Accumu1ated Government Revenue and Marketing Board Trading Surpluses from Nigerian Cocoa Production Under Indicated Policy Alternatives Assuming Low World Cocoa Prices, 1970-1995. 117 The ranking of the accumulated government revenue and market— ing board trading surpluses for the various runs depends on their tax rates and projected annual output. Although the government revenue rates of Runs 2 and 5 are the same, the accumulated revenue of Run 5 is higher than in Run 2, because of the latter's higher projected output. Due to the low world price assumption used in this set of policy experiments, the producer price of all the five runs are also very low. Consequently, the trading surpluses of Run 5 (with the price guarantee) are negative, and the deficit increases with time because of the continuing subsidy necessary to maintain the guarantee. As in the preceding experiments, the values of the other performance indices of Run 5 are higher than in Run 2, which does not have the producer price guarantee. D. Results of Policy Experiment Assuming! chlical World Cocoa Prices Finally, the effects of a set of cyclical world cocoa prices on the economy are discussed in terms of the simulated time paths of the six performance indices presented from Figures 4.D2 to 4.07. The world, market and producer prices in this policy experiment are shown in Figure 4.01. The relative ranking of the projected curve output can be best explained by recalling the major factors that determine the projected annual output. They are (1) total and compositional (traditional and modern) cocoa acreages of producing trees, (2) the past producer price trend which determine the exponentially—averaged "normal" price, and (3) prevailing producer prices. The ratio of the last two factors multiplied by the farmers' short-term exponential 118 Run Definitions Run 1 Low tax rates and no production campaign. Run 2 Low tax rates and production campaigns. Run 3 High tax rates cut off in 1971, and production campaigns. Run 4 High tax rates phased out by 1976, and production campaigns. Run 5 Low tax rates with producer floor price guarantee, and 336r- production campaigns. Market Market Price 4 Price 3 World Price 280- Market Price 3,4 “Producer Price 3 224 ' / Producer Market Price 1,2,5 Price 4 Producer 'rice 3,4 Guaranteed Producer Floor Price 168 /\ i;— ! ‘\\\\\\\~./’/////7 ‘\\\\\\\\::‘:i::::/:5itg S I 1‘ 112 Producer Price 1,2 World, Market, Producer Prices of Cocoa (h/long ton) J, I l J l .1 l 1970 1975 1980 1985 1990 1995 Figure 4.Dl.--Market and Producer Prices of Nigerian Cocoa Under Indicated Policy Alternatives Assuming World Cyclical Prices, 1970-1995. 119 harvest elasticity determine the annual output from the producing cocoa trees. On the other hand, the long-term responses in output capacity expansion (which determine the total and compositional acreages) depend on (1) the prevailing producer prices, (2) expected producer prices and output (which determine the relative profitability of cocoa production), and (3) program features of the production campaign. Because of the cyclical nature of this set of world prices, the total output effects depend crucially on the interactions between the short-term harvest responses and the long-term, expansion re- sponses. As evident in Figure 4.D2, the outputs of Runs 2 to S (with the production campaigns) are in all cases higher than Run 1 (without the production campaign). However, the order of their projected out- put depends on their reSpective producer prices, which in turn determine their annual short-term and long-term output responses. Because of the high initial producer price guarantee feature, the short-term response in Run 5 causes its projected output to be the highest. However, the projected outputs of Runs 3 and 4 with the higher producer prices (and therefore higher short—term output re- sponses) soon surpass Run 5. As in the other policy experiments, the effects of the long-term output capacity responses depend on the respective producer pricing policies, and their interactions with the production campaigns. The interactions result in the differences in the total and compositional acreages of the cocoa trees among the runs (see Figure 4.DS). Since the ranking of the projected time paths of the other performance indices are similar to the ranking 120 Bun Definitions Run 1 Low tax rates and no production campaign. Run 2 Low tax rates and production campaigns. Run 3 High tax rates cut off in 1971, and production campaigns. Run 4 High tax rates phased out by 1976, and production campaigns. 700 "Run 5 Low tax rates with producer floor price guarantee, and production campaigns. Run 3 4 un 4'” Run 2 600 ~ \Run 5 Z? 5 // “ // E“ 500 b .9. -u 5 m .‘J 0 g 400 - H a ‘1- U 8 8 300 b 0 O U '3 _—_' ‘ Run 1 ‘J x O a- 100 - J 1 i l 4 1970 1975 1980 1985 1990 1995 Figure 4.DZ.--Total Projected Nigerian Cocoa Output Under Indicated Policy Alternatives Assuming Cyclical World Cocoa Prices, 1970-1995. Annual Foreign Exchange Generated from Cocoa Exports (N g million) 250 200 150 100 50 121 Run Definitions Run 1 Low tax rates and no production campaign. Run 2 Low tax rates and production campaigns. Run 3 High tax rates cut off in 1971, and production campaigns. Run 4 High tax rates phased out by 1976, and production campaigns. Run 5 Low tax rates with producer floor price guarantee, and production campaigns. l I 1 1 1 1970 1975 1980 1985 1990 1995 Figure 4.D3.--Foreign Exchange Generated from Nigerian Cocoa Exports Under Indicated Policy Alternatives Assuming Cyclical World Cocoa Prices, 1970-1995. 122 under the moderate world cocoa price assumption, they will not he discussed. Under the cyclical world price assumption, the stabilization role of the rudimentary producer floor price guarantee of Run 5 is illustrated most dramatically in Figure 4.D4. The annual projected agricultural income per capita of Run 5 increases very smoothly. The projected income is determined primarily by the increase in output with the fluctuations in producer prices minimized. In contrast, the increases in income of the other runs fluctuate according to their prevailing producer prices. However, the higher and more stabilized income of Run 5, compared to Run 2 (which does not have the floor price guaranteed) is offset by its lower accumulated trading surpluses as shown in Figure 4.D7. Discussion of Simulation Results from Cocoa Policy Experiments There are five major inferences to be drawn from the results of the cocoa policy experiments we have conducted. First, the values of the performance indices depend crucially on the world cocoa price assumptions. For example, in Figure 4.2, which shows the projected outputs of Run 2 under the four world prices, it is clear that the output under the high world price assumption is greater than the output under the moderate and low world price assumptions. Since the average price of the cyclical world prices is also greater than the average price of the low world prices, its projected output is thus higher. 123 Run Definitions Run 1 Low tax rates and no production campaign. Run 2 Low tax rates and production campaigns. Run 3 High tax rates cut off in 1971, and 120 _. production campaigns. Run 4 High tax rates phased out by 1976, and production campaigns. Run 4 r‘ Run 5 Low tax rates with producer f cor .w price guarantee, and 2: , production campaigns. Run 3 V 100 - m 4.) "-4 3‘ Run 5 c) H Run 2 0 O- 0 80 F‘ E O U E H '3 8 60 - Q. U} «4 c: H Run 1 m H z p '3 4o - U "-1 a on < 20 — L l l l l 1970 1975 1980 1985 1990 1995 Figure 4.D4.--Agricultural DiSposable Income Per Capita in Nigerian Cocoa Sector Under Indicated Policy Alternatives Assuming Cyclical WOrld Cocoa Prices, 1970-1995. Total, Traditional, Modern Cocoa Trees (million acres) 124 Run Definitions Run 1 Low tax rates and no production campaign. Run 2 Low tax rates and production campaigns. Run 3 High tax rates cut off in 1971, and 3.0'- production campaigns. 0.5 Run 4 High tax rates phased out by 1976, and production campaigns. Run 5 Low tax rates with producer floor price guarantee, and production ////,(f:, campaigns. //;,/"’(’ Total 3 Total 4 Total Total 2 #_-_ ‘_____.—————* Total 1 Traditional 5 Traditional Traditional 4 Traditional 2 Modern odern J 1 1 1 1 1970 1975 1980 1985 1990 1995 Figure 4.D5.--Tota1, Traditional and Modern Nigerian Cocoa Acreages Under Indicated Policy Alternatives Assuming Low World Cocoa Prices, 1970-1995. 125 Run Definitions Run 1 Low tax rates and no production campaign. Capitalized Agricultural Land Value (N k million) Run 2 Low tax rates and production campaigns. Run 3 High tax rates cut off in 1971, and R 4 1200 _' production campaigns. u" Run 4 High tax rates phased out by 1976, and production campaigns. Run 5 Low tax rates with producer floor price guarantee, and production campaigns. ‘M3 1000 *- Run 5 Run 2 300 l- 600.- / Run 1 400.- ./ zoo *- 1 1 .1 1 J 1970 1975 1980 1985 1990 ' 1995 Figure 4.D6.--Capitalized Agricultural Land Value in Nigerian Cocoa-Food Ecological Zone Under Indicated Policy Alternatives Assuming Cyclical World Cocoa Price Functions, 1970-1995. 600 500 400 300 200 100 Accumulated Government Revenue and Trading Surpluses (N a million) 1970 Run 126 Definitions Run Run Run Run Run 1 Low tax rates and no production campaign. 2 Low tax rates and production campaigns. 3 High tax rates cut off in 1971, and production campaigns. 4 High tax rates phased out by 1976, and production campaigns. 5 Low tax rates with producer floor price guarantee, and production campaigns. Surplus 2 Surplus 1 Revenue 2,5 Revenue 1 Revenue 3,4 Surplus 3,4 Surplus 5 ” 1 1, 1 1 1 1975 1980 1985 1990 1995 Figure 4.D7.--Accumulated Government Revenue and Marketing Board Trading Surpluses from Nigerian Cocoa Production Under Indicated Policy Alternatives Assuming Cyclical World Cocoa Prices, 1970-1995. 127 Run Definitions 800 _- A Moderate (Most Likely) World Cocoa Prices B High World Cocoa Prices C Low World Cocoa Prices Output 8 Output A D Cyclical World Cocoa Prices 700 b ‘fa”" Output D ';~600 . ’ I: O H N I: o H .6 500 r 3 Output C o .c :1 ,, 400 “ s . O. u 8 a //// 8 300 - / O L, t '3 a.) O as 200 ' 100 P 1 L 1 1 1 -1970 1975 1980 1985 1990 1995 Figure 4.2.--Total Projected Nigerian Cocoa Output Under Indicated Alternative World Cocoa Prices Assuming Low Tax Rates and Production Campaigns, 1970-1995. 128 Second, the ranking of the projected cocoa output also depends on world prices over the planning horizon. For example, because of the higher world prices of "B," which remain continuously high, the output of Run 3 with the tax cut-off feature remains higher than Run 4 where the taxes are phased out (see Figure 4.82). On the other hand, in the low world price situation of "C” where the prices after 1985 remain relatively lower than their initial level, the projected output of Run 4 is higher than Run 3 (see Figure 4.C2). Third, given the assumptions of the model's mechanisms, the influences of short-term output responses from a dramatic producer price increase are dissipated very quickly, as the farmers begin to regard the higher producer prices as "normal." Hence, the short—term output effect of a gradual producer price increase is cumulatively greater than that of the dramatic price increase. The projected annual output of Run 4 with the tax phase-off feature is generally higher than in Run 3 even though producer prices after 1976 are the same. Fourth, the benefits of long-term output reSponses in cocoa acreage expansion can be further reinforced if they also coincide with the peak of the production campaigns. Otherwise, the farmers given the higher producer prices may prematurely expand their cocoa acreages without taking full advantage of the communication effects of the production campaigns. If the producer price increases coin- cide with the peak of the production campaigns, more of the new acreages would consist of the higher-yielding Upper Amazon trees promoted by the campaigns. Or conversely, if the full effects of the 129 production campaigns were to occur concurrently with a favorable producer price increase, an even greater proportion of the acreage expansion in both new planting and replanting would consist of the higher-yielding variety. Thus, the effects of the producer price increase and the production campaigns are highly complementary in encouraging the farmers to cultivate the higher-yielding Upper Amazon cocoa trees as they expand their cocoa acreages. Finally, each of the five policy runs have projected differ- ent outcomes at different time-phases of the planning horizon. Their projected differences can in turn provide policymakers with a basis for selecting the preferred policy Option with its tradeoffs accord- ing to their perceived objectives of the economy. For example, if the perceived overriding objective for the cocoa sector for the next twenty-five years is to maximize farmers' income and the foreign exchange generated from cocoa exports, Run 4 (which has the tax phase-out and production campaign features) may be the preferred option. On the other hand, if the perceived paramount objective is to maximize government revenue and marketing board trading surpluses, Run 2, which gives the highest accumulated funds is the logical policy Option over Runs 3, 4, and 5, even though the consequent projected output, foreign exchange, and farmers income of Run 2 are lower. However, if the objective is to ensure the cocoa farmers a steady and stable income increase, Run 5 (with the guaranteed pro- ducer floor price feature) is recommended, especially if the world price is expected to behave cyclically as we have seen in Figure 4.D4. Hence, in this particular situation, the tradeoff is between a loss 130 in the accumulated government revenue and marketing board surpluses, and an increase in output, foreign exchange generated from cocoa exports, and personal income of the farmers in the cocoa economy. Effects of Cocoa Policy Experiments on Food Production A secondary objective of the study is to eXplore the impact on food production in Western Nigeria which may result from the prOposed cocoa producer price changes and the government production campaigns to expand cocoa output capacity. Given the model's present structure and its underlying assumptions, the study is not able to show convincingly the cocoa policy effects on food production in Western Nigeria. Figures 4.3 and 4.4 show the total subsistence and cash food acreages of the sector, and the market and the producer prices of food composite for all of Southern Nigeria under Runs 1, 2, and 4, from 1970 to 1995 (assuming the cyclical world price function of cocoa). As discussed in Chapter III, the total demand for staple food in the region (expressed generically in caloric value) is met by subsistence and cash food production. The general increase in the total food land in all three runs is caused by the increase in the rural population. In Run 1, while the total food acreages in- crease, cash food land remains relatively constant. Although total food land acreage in Run 2 is similar to that in Run 1, Run 2's increase in cash food land after 1980 is offset by the decrease in the subsistence food land. Hence, we may infer that the expansion of the cocoa acreages resulting from the production campaigns did 131 Run Definitions Total 12 Run 1 Low tax rates and no cocoa production campaign. a Run 2 Low tax rates and cocoa production campaigns. Run 4 High tax rates phased out by 1976, Total 4 and cocoa production campaigns. 1.6 1.41- ’u? g ubsistence 1 g r. : f :3 1.2 1. :: ubsistence 2 L E -u (i g 1.0'- r” 'Subsistence 4 —o O 0 * U... 'c h g .8 L) a.“ 0 5‘3 4" .6 - U) 0H U) .D 3 U) . Cash 4 ‘g .4 _ __"”, Ed / caSh 2 I \ .2 I- CaSh 1 1 11 1 - 1 1 1970 1975 1980 1985 1990 1995 Figure 4.3.-~Total, Subsistence, Cash Food Land in Nigerian Cocoa Ecological Zone Under Indicated Cocoa Policy Alternatives Assuming Cyclical World Cocoa Prices, 1970-1995. 132 016 Run Definitions Run 1 Low tax rates and no cocoa production campaigns. Run 2 Low tax rates and cocoa production campaigns. ;; -014 Run 4 High tax rates phased out by 1976, and g cocoa production campaigns. 9.. :3 Market Food 2: .012 Prices 4,2,1 p "-1 U) 0 D. E 8 ,0 .010 O 0 LL. ‘H O m .008 8 A "-1 H °‘ Producer Food 3 Prices 4,2,1 .x g; .006 2: "D a 1... 3 004 s . —u C 5-4 0‘ .002 1 1 1 1 1 1970 1975 1980 1985 1990 1995 Figure 4.4.-~Market and Producer Prices of Food Composite in Nigerian Cocoa Ecological Zone Under Indicated Cocoa Policy Alterna- tives Assuming Cyclical World Cocoa Prices, 1970-1995. 133 not increase appreciably the expansion rate of total food land. Instead, a greater proportion of the staple food consumed by the farmers is now obtained from the cash food market. The increased dependence on the cash food market of Run 2 (with the cocoa production campaigns) results from the increased income generated from the greater cocoa output. In Run 4 (which has the highest projected cocoa output and acreages), total food land increases at a lower rate after 1978 than in Runs 1 and 2. The projected cash food land of Run 4 throughout the planning horizon is also highest and subsistence food land the lowest. From Figure 4.4, we see that the annual producer and market prices of the staple food composite in the sector are very small among the three policy runs. According to the model, the market price of food depends on the total supply and demand of food in all of South Nigeria. Hence, policy experiments focusing on one par- ticular sector in the region does not appreciably affect the food pricing mechanisms. (In contrast, when policy experiments were conducted on the total region, the projected market food prices under alternative policies were markedly different [Manetsch s£_al:, 1971]). The increasing food prices in all three runs are caused by the increasing population in the sector. Although the behavior of food prices is consistent with the model's assumptions, we must interpret the results with caution. Let us recall the assumptions we built into the food price determination mechanisms: (1) food prices are adjusted annual averages without considering explicitly the considerable intra- year or seasonal fluctuations; (2) the difference between the producer 134 food price at the farm gate and the market food price is a constant markdown from the market price of fecd; and (3) consumer preferences and the cultivational technology of the food composite are assumed to be constant throughout the planning horizon. The last assumption is tenuous in reality. Obviously, within the 25-year planning horizon, the population's preferences for staple food along with their relative prices would change. The changes in consumer preference would probably affect the production of the food crops. Likewise, the production fUnctions of the various staple foods would change. For example, although Western Nigerians prefer yam over cassava, the lower labor requirement and higher output for cassava, and hence, its higher caloric value per unit cost would certainly increase its production. This, in turn, would increase the use of cassava as a food staple. In addition, future increases in income from increased cocoa production may also affect the farmers' total and compositional demand for staple foods, just as it would probably raise their demand for nonstaple foods and consumer durables and nondurables. Nevertheless, we can conclude from the policy experiments that the expansion of cocoa acreages would dampen the expansion of total food land in the region. Since we have also assumed that cocoa farmers would always feed themselves first from their food production, the relative decline in total food land in the cocoa-food zone would therefore affect primarily the nonagricultural pOpulation. Hence, unless the relative decline in production is met by increased food production from the Oyo Province in the West, or food shipment from 13S Midwestern and Northern Nigeria, annual staple food consumption by the lower income segments of the nonagricultural population would probably decline in the future. CHAPTER V SUMMARY AND CONCLUSIONS Introduction Since cocoa is the major source of Western Nigeria's income, employment, revenue, and foreign exchange, the industry is very important to the regional economy. Over 95 percent of Nigerian cocoa production, covering a total area of approximately 1.6 million acres-- cultivated by over 400,000 households, is located in the Western State. Nigerian cocoa production, which is relatively labor-intensive, is almost exclusively a smallholder enterprise. The typical landhold- ing consists of about three acres of cocoa providing a cash income, and two acres of food cultivated primarily for household consumption (Okurume, 1970). All the cocoa produced is sold to the Nigerian Cocoa Marketing Board, a statutory monopsonist. However, the producer prices farmers receive are generally less than the world prices. Between the world prices and producer prices, the government collects export duties, producer sales tax, and the marketing board collects a trading sur- plus tax. Additionally, farmers also pay for the Operational and handling costs involved in the sale of their output. The total dif- ferences in some years may amount to as much as 50 percent of the 136 137 world price. Hence, most economists recommended the increase in the cocoa producer prices. Moreover, since the cocoa farmers have been responsive in their output to producer price changes, any increase in their prices would further increase their output and output capacity. The loss in government revenue due to the initial commodity tax decrease may be Offset by the increased tax base from (1) in- creased cocoa production and productive capacity, and (2) indirectly by the increase in the producers' income, and the asset value of cocoa land. Thus, taxing the increased tax base may recuperate the initial tax loss. Furthermore, since the yield of many Of the existing Amelonado cocoa trees is relatively low when compared to the recommended higher- yielding Upper Amazon Species, the Western Nigerian government is encouraging farmers to grow more Of the latter (Johnson g£_al:, 1969). In addition, the government is encouraging farmers to grow the higher- yielding Upper Amazon cocoa trees in land presently in bush or food. The primary purpose Of this study was to adapt components of the Nigerian Agricultural Simulation Model developed at Michigan State University in order to analyze the proposed revamping of the Nigerian cocoa producer pricing policy, and the government-initiated cocoa production campaigns. The study can also be viewed as part of the continuing process to further validate and improve the operational usefulness of the system simulation approach and the model for agricultural planning. To accomplish these objectives, the study focused specifically on the Nigerian cocoa sector, one of the four ecological zones in Southern Nigeria. 138 In addition, more realistic alternative world cocoa prices were used. (In previous studies [Manetsch e£_al:, 1970; Olayide, Abkin, and Johnson, 1971; and Abkin, 1972], for the purpose of model testing, future world commodity prices were assumed to be constant.) The study also introduced and tested the usefulness of a guaranteed producers' floor price, whereby farmers would be paid a predetermined .I Vial-"'- minimum, should their prices (after accounting for the various taxes and handling costs) go below the level. The price support program would be financed from previously accumulated trading surpluses, and if necessary, from outside sources. The Western Nigerian Cocoa System Simulation Model To explore the ramifications of the proposed government cocoa price-income strategy and production campaigns and their resultant interactions and feedback effects, the study adapted components of the Michigan State University Nigerian Agricultural Simulation Model to analyze their impact on the Western Nigerian cocoa economy. The cocoa simulation model has four major components which (1) allocated land use according to the farmers' perceived profitabilities of cocoa and food subject to the land, labor, and capital constraints, (2) cal- culated the yield and output of cocoa and food, and their respective producer and market prices, (3) provided the instrumental linkages for the government revenue, marketing board trading surplus, and production campaign policies, and (4) generated the performance criteria to evaluate the impact of alternative programs on the cocoa economy through time. 139 Cocoa Simulation Policy_ Experiments Conducted The three major sets of assumptions investigated were (1) alternative world cocoa prices, (2) alternative revenue and marketing board producer pricing policies and, (3) proposed govern- ment cocoa planting and replanting production campaigns. The four world price functions used in the study represented the moderate (most likely), high, low, and cyclical price projections that would be relevant to the analysis of the Nigerian cocoa economy. The first Of the five policy Options (which combined alter- native producer pricing features and the proposed production cam- paigns) is Run 1, the base run which approximated the status quo policy. It assumed a relatively low government revenue tax of 10 per- cent of the prevailing world cocoa price, and a 20 percent marketing board trading surplus tax on the market price, and no government production campaign. Run 2 (which assumed the tax features of the base run and government replanting and new planting production cam- paigns) examined the effects of the campaign on the economy. Runs 3, 4, and 5 compared the effects of the production campaigns with vary- ing tax features. The tax rates of Runs 3 and 4 were assumed to be 20 and 30 percent of the world and market prices. However, the taxes of Run 3 were cut Off in the following year, whereas in Run 4, the taxes were phased out linearly over the following five years. Run 5 (with the same tax features and production campaigns as in Run 2) compared the effects of a guaranteed floor price feature supported by previously accumulated marketing board trading surpluses. 140 The results of the cocoa policy experiments were discussed in terms of the projected time paths (from 1970 to 1995) of six of the more important performance indices incorporated in the model. They were (1) total annual output of cocoa, (2) total and compositional (traditional and modern) acreages of cocoa trees, (3) foreign exchange generated from cocoa exports, (4) capitalized agricultural land value Of the cocoa-food ecological zone, (5) disposable agricultural income per capita, and (6) accumulated government revenue and marketing board trading surpluses. Policy Implications from the Simulation Experiments on the Western Nigerian Cocoa Economy There are five major policy implications from the simulation experiments which agree substantially with the findings of CSNRD's study Of the Nigerian cocoa industry. First, the total benefits to the Nigerian cocoa economy depend importantly on the world price Nigeria receives for her export. Obviously, the higher the world price of cocoa, the more it benefits the economy. It is therefore very important for the government to secure the highest world price for her exports. Second, investments in government production campaigns which encourage cocoa farmers to cultivate higher-yielding Upper Amazon trees are justifiable since the projected output, foreign exchange, farmers' income, and accumulated government revenue and marketing board trading surpluses functions were in all cases (regardless of their tax features) higher than Run 1 (without the production cam- paigns). The modernization Of the Nigerian cocoa economy is defensible, 141 even if world cocoa prices should decline subsequently--either re— sulting directly from an excess capacity because of the slower growth rates of demand, or from other factors outside the direct control of Nigeria and other producing countries. Third, based on the farmers' positive supply responses, increases in cocoa producer prices raise both the short-term harvest output and the long-term output capacity reSponses and hence, far- mers' disposable income as well as the foreign exchange earnings from additional cocoa exports. In turn, increases in cocoa exports and farmers' income would most probably have their spill-over benefits in the other sectors of the economy. Increases in fOreign exchange earnings increase Nigeria's capacity to import. Increases in output and output capacity would also increase the demand for agricultural productive goods and services, viz., more fertilizer, seedlings, tools, and machinery. Similarly, increases in farmers' income also increase their effective demand for other durable and nondurable consumer goods, in addition to the productive goods mentioned earlier. Hence, the increase in the profitability of cocoa production resulting from higher producer prices, and the decrease in the farmers' cost of production campaigns reinforced the expansion of the agricultural land. The expansion of the cocoa land and food land by the farmers is thus a very important means of private capital formation. Any increased loss in government revenue that might occur from the producer tax decrease could be compensated by taxing the farmers' increased income and the asset value of their cocoa land, 142 or indirectly by taxing the increased purchase Of their producer and consumer goods. However, the effects of alternative taxes on the economy do not necessarily correspond to one another. Besides the administrative problems Of enforcement, costs of collection, distribu- tion and equity impact, there is also the intertemporal problem and trade-off between present revenue and future revenue of the alterna- tive methods. Although it may be argued that the future tax base resulting from the proposed tax decrease may be greater, the increased taxes from the increased output would not be collected until the trees come into production. Hence, if government revenue is needed for other development programs, it is highly unlikely that the government would reduce cocoa producer taxes unless the revenue loss is replaced by other sources outside Of the cocoa sector, such as royalties from the petroleum industry. Fourth, the introduction of an annual guaranteed producer floor price would stabilize the farmers' income, especially if world cocoa prices are expected to behave cyclically. The producer price support also increases farmers' income, cocoa acreages, capitalized agricultural land value as well as the foreign exchange generated from cocoa exports in the years when cocoa prices are expected to be relatively low. Finally, based on the policy experiments, the projected time paths of the various performance indices under alternative cocoa pricing policies (with their varying tax features) differed discern- ibly. Their projected differences, in turn, can provide a more comprehensive basis for policymakers to select the most efficacious “711—1"; . 41—7. "14 "T mm. W 143 cocoa pricing strategy. The choice of the strategy depends on the policymakers' perceived Objective of the pricing policy. Hence, if the perceived overriding objective for the cocoa pricing policy is to maximize farmers' income and the foreign exchange generated from cocoa exports, then the policy which phased out all taxes would be pre- ferred. On the other hand, if the perceived Objective of the pricing policy is to maximize government revenue and marketing board trading surpluses, the status quo policy (which was projected to accumulate the highest revenue and trading surpluses) would be the choice. Finally, if the Objective is to ensure the cocoa farmers a steady increase in their income from cocoa production, the proposed pro- ducers' floor price feature discussed above can be incorporated, as part Of the cocoa pricing strategy. Conclusion from the Use Of the System Simulation Approach for Cocoa Development Planning, In using the system simulation approach, we were interested in providing an analytical framework in which researchers and policy- makers could interact while formulating alternative cocoa policies. Hence, we were guided more by the nature of the problem we attempted to solve rather than by the availability of relevant data to estimate statistically the coefficients Of the postulated economic phenomena, or by the constraints imposed by a computing algorithm. In this study, we were Specifically interested in evaluating the long-term economic impact of the proposed revamping of the cocoa producer price policy, and the government production campaigns to expand the Western Nigerian cocoa industry. 144 To this end, the computerized Nigerian Agricultural Simulation Model provided a very useful and convenient "laboratory" whereby the outcomes of various combinations of cocoa programs and policies were compared and analyzed. Based on the projected time paths Of the various performance indicies of the cocoa sector, the merits of various policy alternatives were discussed. This particular computing advantage of the computerized system simulation model (with its rela- tive low cost and quick computation turn-around time) should not be minimized, especially when compared to the tedious and time-consuming conventional pen and pencil projections used in CSNRD. However, the prolific ability of the computerized model to generate results under alternative assumptions, like the Opening of Pandora's box, still does not negate the researcher's responsibility to interpret the simulated results to policymakers in light of the model's limitations and underlying assumptions. At a more subtle epistemolOgical level, a major characteristic of the approach used in the study is simulation. Simulation is a formalized process of thinking through some of the interactions, feed- backs, effects and related policy implications. The fundamental methodological problem, therefore, was to determine what were the es- sential variables, interrelationships, processes, and boundary condi- tions of the Nigerian cocoa szggsmf-recognizing that the essence of scientific inquiry is to simplify complexity without being overly simplistic. 0n the other hand, the cocoa system simulation model would have been very cumbersome, if all the thought-Of-linkages and processes were included, without the investigators first discerning .a 145 which were the most important or significant. Moreover, since it is very difficult to validate (or refute) the postulated interrelation- ships if they became too complex, it was methodologically more defen- sible to build and validate the simpler relationships. 0n the other hand, we could have been too parsimonious, by selecting out a_priori most of these interrelationships, or by consigning the developmental : process to occur at a predetermined rate, thereby assuming the research problem away. Hence, the research was motivated toward the middle-ground between these two extreme positions. “- .—-_.1_._ The process of deveIOping a system simulation model of the Nigerian cocoa economy had led us to consider variables and inter- relationships which were less amenable for analysis by the more conventional techniques. Past cocoa econometric studies (constrained by data availability) were concerned primarily with establishing statistically past empirical relations of a handful of variables in order to forecast what might happen to the industry under proposed intervening policies. Such linear extrapolations, however, might not be valid for policy prognostication, since the underlying economic conditions in the future would have changed. Moreover, the very purpose of development is to set in motion forces that are not self-equilibrating, but interactive. Following the system concepts expounded by Forrester (1972) and Von Bertalanffy (1972), the model explored the implications of the interactions and positive and nega- tive feedback effects that might modify the final outcome. As a general conclusion, it is hoped that the computerized, system simulation model Of the Nigerian cocoa economy has captured 146 some of the more important dynamic and interactive relationships, under alternative government policies and data and behavioral assump- tions. In addition, the model's projected time-paths Of the various performance indices, may better assist policymakers in cocoa-plan- ning and policy analysis with the display of the composite outcome of each prOposed policy. E Further Research on the Nigerian Agricultural 1 Simulation Model for Policy Analysis There are six additional ways in which the Nigerian Agri- cultural Simulation Model can be extended for policy analysis.1 B First, it should provide a more comprehensive basis for analyzing government revenue and investment policies of the cocoa sector. The present model was not able, nor built to, compare the efficacy of cocoa producer tax as a means of generating government revenue, vis-a-vis alternative income, consumption, import, capital gain and land taxes. Likewise, the present model did not provide a compre- hensive basis to evaluate the impact and effectiveness of the govern- ment overall investment prOgram, including the reinvestment of the revenue generated from the cocoa industry. In the present model, we have attempted to capture the benefits that might accrue to the cocoa producers in the private sector by showing their increase in the . income (flow) and the increase in their capitalized land value (stock). However, in the policy Options where producer taxes pre- vailed, the government revenue and marketing board trading surpluses 1See also Abkin (1972) for possible improvements and exten- sions of the Nigerian Agricultural Simulation Model for policy analysis. 147 that resulted from the various tax policies were merely accumulated. NO attempt was made to show how the returns to the economy might have been greater if the accumulated revenue and trading surpluses had been "ploughed-back" either directly or indirectly to the cocoa sector. Obviously, the benefits derived from the government reinvest- ment Of the revenue and trading surpluses would depend on the nature ; Of the activity invested. The central unresolved analytical issue is: How do we determine who is the more efficient and effective allocator of the reinvestable surpluses that result from the government tax and invest- b ment policies? Is it the public sector and its agencies who, with the collected revenue may attempt to increase public and private capital formation through various means? These could include equity, participation in quasi-government corporations; financial grants to colleges, extension services, and research institutes; and directly to the agricultural producer sector, through cash, subsidy and low interest loans to the farmers. Or is it the cocoa producers in the private sector, who with the increase in income resulting from the increase in producer price, may invest further in their present agri- cultural production, Or other alternative ventures either within agriculture or outside of agriculture? Secondly, to discuss meaningfully the impact and distribu- tional costs and benefits of government investment, including the prOposed production campaigns and the revamping Of the producer pricing policy, the cocoa sector may have to be disaggregated by its geo~political divisions. Since the trees in each division have their 148 own genetic, cultivational, and age cohort characteristics, such disaggregation may indeed be helpful. For example, the trees in the Ibadan Division are generally Older than those in the Ife Divi- sion. Because Of the present rather intensive land utilization pattern, Ibadan cocoa farmers would probably have to replant their existing trees in order to increase their output capacity, whereas the Ife cocoa farmers could still annex neighboring bush land for cocoa cultivation. Since we are also interested in the differential output and consumption responses under the proposed producer price and resultant income changes, and the distributional impact of government invest- ment policies in the sector, it may be useful to further subcategorize the cocoa industry in each geopolitical division by farm size. Although there are other attributes by which we can divide the industry, farm size seems to be the best proxy variable especially since the model's present unit of analysis is an acre of agricultural land, and not the individual agricultural decision maker. If the unit were the latter, then it might be more useful to subcategorize the industry by the income levels or social characteristics of the farmers. The third model extension is to provide new policy entry points in fUture simulated time-periods Of the experiment. The present policy experiments on the cocoa economy were conducted with the government revenue and marketing board producer pricing policies throughout the planning horizon, Specified at the base year. The time paths of the various performance indicies were then 149 projected through simulated time. However, the present approach did not allow policymakers who (upon the assessment of the performance indices at the future time period) may want to introduce another policy in a future time period to counter or augment the consequences of the prevailing policy alternative. For example, a policymaker may find the market price of food in 1980, resulting from a crash #1 cocoa production campiagn in 1972 intolerably high. He can then {1_ experiment with alternative food production campaigns, initiated at ."e. different time periods between 1972 and 1980, that might reduce the high market price of food in 1980. ‘finm 9- . I i Fourth, closely related with the above, is the inclusion Of a semi-automatic decision-making interphase whereby the government tax policy in any one year depends on the interaction between the prevailing world price of cocoa and a predetermined tax rate schedule. Although the general trend of world cocoa prices can be specified, the actual price in any one year would probably occur randomly within a predetermined variance from that trend. Hence, to model the phenomena, cocoa world prices can be specified stochastically with a random component to simulate the uncertainty of the price in any one year. Should the simulated world price in any one year go below a predetermined low level, the government tax rates for that year could be reduced accordingly. Likewise, should the simulated world price in another year exceed a predetermined upper level, the tax rates for that year could be increased. The major benefit of the model extension is to help determine a more flexible governmental tax policy which llalll'fl'ivllflll‘llili Hill 150 would stabilize farmers' income as well as the government revenue, given the fluctuations of world prices. Fifth, as part of the continuing process to update and improve the substantive and informational base of the model, more field and empirical research must be conducted to verify some of the projected outcome anticipated in the present series Of simulation experiments. [a r Specifically, more concrete evidence is necessary to substantiate if (or refute) (1) the postulated effects and interrelations of the ( cocoa farmers' short-term harvest responses and their long-term output ; capacity expansion responses, given producer price changes; (2) the i communicational effects of the government production campaigns; and (3) the constraining effects of inputs (such as labor and credit) on the cocoa farmers' agricultural land allocative decisions. Finally, alternative formulation of the model's present land allocation and the annual harvest reSponse mechanisms may have to be made and tested to conform to the farmers' actual output response behavior. The projected annual cocoa outputs (assuming government- initiated production campaigns) under alternative producer price policies tend to cluster. On the other hand, economy theory suggests a wider divergence among the projected time paths. Specifically, the land allocation mechanism (which models the farmers' long-term output capacity expansion) may have to be modified such that the expansion rate Of cocoa acreages would be more sensitive to producer price changes. Likewise, the short-term harvest response may have to be modified such that the output increase resulting from producer price 151 increase would also depend on the farmers' past experience on price changes as well as their anticipated changes--given the biological yield capacity and composition of the cocoa trees. BI BLIOGRAPHY BIBLIOGRAPHY Abkin, Michael. "Policy Making for Economic Development: A System f: Simulation Model of the Agricultural Economy of Southern Nigeria." Unpublished Ph.D. dissertation, Michigan State 1 University, 1972. Ady, Peter. "Supply Functions in Tropical Agriculture." Bulletin of Oxford Universi§y_lnstitute Of Economics and Statistics (May, 1968). ‘1” Bateman, Merrill. Ghana Cocoa Prospects, 1971-1985. Unpublished manuscript, Brigham Young University, 1971. Bauer, P. T., and Yancy; B. S. “A Case Study of ReSponse to Price in an Underdeveloped Country." Economic Journal (December, 1959). Berry, Sara. "A Tree in Two Towns: Studies Of Cocoa Farming and Economic Development in Western Nigeria." Unpublished Ph.D. dissertation, Indiana University, 1968. Boulding, Kenneth. "Economics and General Systems." The Relevance of General Systems Theory. Edited by Ervin Laszlo. New York? George Braziller, 1972. Byerlee, Derek. "Agricultural Development and Urban Unemployment: A Simulation Analysis of the Nigerian Economy." Unpublished Ph.D. dissertation, Oregon State University, 1972. Cocoa Research Institute of Nigeria. National Agricultural Develop; ment Committee Report on Cocoa. Lagos, Nigeria: Federal Department of Agriculture, 1971. Eicher, Carl, and Liedholm, Carl, eds. Growth and Development of the Ni erian Econom . East Lansing, Mich.: MiEhigan State University Press, 1é70. Essang, Sunday Matthew. "The Distribution of Earnings in the Cocoa Economy of Western Nigeria: Implications for Development." Unpublished Ph.D. dissertation, Michigan State University, 1970. 152 153 Food and Agricultural Organization. ggricultural Development in Nigeria, 1965-1980. Rome, 19 . Forrester, Jay. World Dynamics. Cambridge: Wright-Allen Press, Inc., 1971. Gittinger, P. Economic Analysis of A ricultural Projects. Washington, D.C.: International Bank of Reconstruction and Development, 1972. Gusten, Rolf. Studies in the Steple Food Economy of Western Nigeria. Ibadan: Nigerian Institute in Social and Economic Research, a 1967. ‘ Helleiner, Gerald. Peasant Agriculture, Government and Economic Growth in Nigeria. Homewood, 111.: Richard D. Irwin, Inc., 1966. Hicks, Ursula. Development Finance: Plannipgand Control. New York: i Oxford University Press, 1968. Idachaba, Francis. "Marketing Board Taxation of Nigerian Agriculture: Some Implications." Unpublished Ph.D. dissertation, Michigan State University, 1972. Ilori, Christopher Oladosu. "Economic Study of Production and Distribution of Staple Foodcrops in Western Nigeria." Un- published Ph.D. dissertation, Stanford University, 1967. Johnson, Glenn. "Alternatives to the Neoclassical Theory of the Firm." American Journal of Agricultural Economics (May, 1972). "Factor Markets and Economic Development." Economic DevelOpment of Trepical Agriculture. Edited by W. W. M McPherson. Gainesville: University of Florida Press, 1967, Chapter 6. , and Zerby, Lewis. What Economdsts Do About Values: Case Studies of Their Answers to destions They Don't Dare Ask. East Lansing: Michigan State Agricultural Experiment Station, 1972. e2_gl, Strategies and Recommendations for Nigerian Rural Development, 1969-1985. East Lansing: MiChigan State University, 1969. Johnston, J. Econometric Methods. New York: McGraw—Hill Book Company, 1964. 154 Kellogg, Earl. "A Temporal and Spatial Model to Assist in Evaluating Investments in the Nigerian Beef Distribution System." Unpublished Ph.D. dissertation, Michigan State University, 1971. Kriesel, H. C. Cocoa Marketing in Nigeria. East Lansing: Consortium for the Study of Nigerian Rural DevelOpment, Michigan State University, December, 1968. Laszlo, Ervin, ed. The Relevance Of General Systems Theory. New York: George Braziller, 1972. Manetsch, T. et al. A Generalized Simulation Approach to Agri icultural Sector Analysis with Special Reference to Nigeria. East Lansing: Michigan State University, 1971. , and Park, G. System Analysis and Simulation with Appli- cation tO Economic and Social Systems. Preliminary edition. East Lansing: Michigan State University, 1972. Okurume, Godwin. "Foreign Trade and the Subsistence Sector in a Peasant Export Economy: A Case Study Of the Impact of Agricultural Export Expansion on Domestic Supplies in Western Nigeria." Unpublished Ph.D. dissertation, Yale University, 1970. Olatunboson, Dupe, and Olayide, S. O. The Effect of the Marketipg_ Boards on the Output and Income of Primary Producers. Inter- national Conference on the Mafketing Board System, Ibadan, Nigeria, 1971. Olayemi, Joseph Kayode. "Peasant Cocoa Production in Western Nigeria: Economic Analysis." Unpublished Ph.D. disserta- tion, McGill University, 1971. Olayide, 5.; Abkin, Michael; and Johnson, G. ”Agriculture and Growth of the Nigerian Economy: Results of Policy Simulation Experi- ments." Unpublished paper, Michigan State University, 1972. Oluwasanmi, H. A. A riculture on Ni erian Economic Develppment. London: Oxford University Press, 1966. Phillips, T. A. An Agricultural Notebook. Ikeja, Nigeria: Longman, 1966. Posada, Alvaro. "A Simulation Policy Analysis of the Northern Colombia Beef Cattle Industry." Unpublished Ph.D. disserta- tion, Michigan State University, 1973. 155 Prest, A. R., and Turvey, R. "Cost-Benefit Analysis: A Survey." Economic Journal (December, 1965). Reutlinger, Shlomo. Technique for Project Appraisal Under Uncertainty. Washington, D.C.: International Bank for Reconstruction and Development, 1970. Rogers, Everett, Diffusion of Innovations. New York: The Free Press, 1962. ~ Rossmiller, George e£_el: Korean Agricultural Sector Analysis and Recommended Development Strate iES, 1971—1985. East Lansing: Michigan State University, 197 . Sanders, Dean. "The Price Responsiveness of Nigerian Cocoa Farmers." Unpublished Ph.D. dissertation, University of Michigan, 1968. Stern, Robert. "The Price Responsiveness Of Primary Producers." The Review of Economics and Statistics,(May, 1962). Thodey, A. "Analysis Of Staple Food Behavior on Western Nigeria." Unpublished Ph.D. dissertation, University of Illinois, 1969. Von Bertalanffy, Ludwig. General System Theory: Foundations, Development and Applications. New York: George Braiiller, 1972. Yaghoobi, Rahmatabadi-Hossein. "The Most Efficient Production Techniques for Providing Nutrition and Income from the Agricultural Sector of Nigeria." Unpublished Ph.D. dissertation, Michigan State University, 1971. "71111111111111ii11111114111111'“ 3 1293 03046 3438