Agricultural Pricing in Togo SWP-467 World Bank Staff Working Paper No. 467 Julr 1981 SECTOlRAL LIBRARY INTERNATIONAL BANK FOR RECONSTRUCTIC.ON ANn I)EVELOPMENT NOV 5 1986 Prepared by: David Bovet and Laurian Unnevehr Country Programs Departrnent n West Africa Regional Office (i 1981 d Bank MD rt *** HD9017.T62 B68 c.2 9O 11F Dr Agricultural pricing in Togo / .TE,1 4 IlilililUlililil I BBllilililililillXlililil DHilililililill authors c.2f' di SLC029707 Dns,-or to any n cMaual aaing in tneir Denalt. The views and interpretations in this document are those of the authors and should not be attributed to the World Bank, to its affiliated organizations or to any individual acting in their behalf. WORLD BANK Staff Working Paper No. 467 July 1981 AGRICULTURAL PRICING IN TOGO This paper analyzes the pricing policies pursued by the Government of Togo for its export crops -- cocoa, coffee, and cotton -- and proposes an analytical framework which could be used in price-setting. Because these crops are handled through a marketing board system, the conflicting pricing policy objectives of government revenue maximization and of foreign exchange maximization determine the type of analysis used. Supply elasticities and measures of private and social profitability are computed, and the effects of competing foodcrop prices and cross-border trade are analyzed. The results suggest that in the case of cocoa, the policy of pursuing revenue maximization in the short-run means that special measures must be introduced to provide an incentive for cocoa replanting. For cotton, the effect of input subsidies is shown to be more than offset by the relatively low producer prices. Prepared by: David Bovet and Laurian Unnevehr Country Programs Department II West Africa Regional Office Copyright kc 1981 The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. PREFACE This paper analyzes the price response of Togo's major cash crops, discusses options for pricing policy, and recommends a frame- work for price analysis which could be applied by the Government. The data gathering,analytical work, report writing and discussions with Government were carried out by Mr. David Bovet and Ms. Laurian Unnevehr, both of the West Africa, Programs Department. Annex A was written in conjunction with Mr. Michael J. Hartley of the Development Economics Department. This review was carried out in consultation with the West Africa Projects Department, Agriculture Division. AGRICULTURAL PRICING IN TOGO Table of Contents Page No. Background ........................................1............ Pricing Policy Objectives ...................................... 6 Methodology .................................................... 9 Analytical Results - Cocoa ................ .. ................... 10 - Coffee .................................... 12 - Cotton .................................... 13 Suggestions for Future Study ............... .. .................. 16 Conclusions and Recommendations ................................ 17 ANNEX A: Cocoa ........................ ........................ 23 Supply Response ...... ................... ............. 23 Replanting .. .......--. 29 Policy Objectives and Strategies ........ .. ........... 35 ANNEX B: Coffee ....................... ........................ 39 ANNEX C: Cotton ....................... ........................ 46 Supply Response ...................................... 46 Structure of Incentives ..... ......................... 51 Policy Implications ............... .. ................. 54 STATISTICAL APPENDIX ....... .............. ...................... 57 BIBLIOGRAPHY ......................... .......................... 76 MAP OF TOGO .................................................... 77 List of Text Tables and Figures Table of Contents Tables Page No. 1. Composition of Exports - % of Value ................... 1 2. Comparison of Producer Prices in CFA Zone ............. 3 3. Range of Discount Rates for which Replanting Shows a Positive Return .............. 11 Figures 1. Price Movements of Cash Crops and Foodcrops ........... 4 2. Cash Crop Marketed Production ...................... 5 3. Outcome of Alternative Pricing Policies for Export Crops ................................................. 6 4. Government Revenue - Maximizing Producer Price as a Function of Supply Elasticity ......................... 8 5. Cocoa Prices ................. 20 6. Coffee Prices ................. 21 7. Cotton Prices ................. 22 The purpose of this study is to investigate the effects of past agricultural price policies in Togo on production incentives and to evaluate alternative pricing strategies against macroeconomic objectives. In addi- tion, it is hoped that the analytical methods presented can be used by the authorities in carrying out their own analyses of agricultural prices in future years. The present study was prompted by the Government's and the Bank's desire to analyze cash crop pricing issues more closely in light of the emphasis now placed on agricultural development. This paper is speci- fically directed toward three major export crops, all of which have been the focus of Bank projects in Togo: cocoa, coffee and cotton. Incentives for growing these crops have been analyzed in relation to competing food- crop prices. Attention has been focussed on the export crops, since their prices are set annually by the Government marketing agency. Background Agriculture is an important sector in Togo in terms of both economic product (30%) and employment (75%). It is a less important component of export earnings because of Togo's mineral resources (see Table 1). TABLE I COMPOSITION OF EXPORTS - % OF VALUE 1970 1971 1972 1973 1974 1975 Phosphates 25 35 38 46 76 65 Cocoa 42 31 29 26 12 18 Coffee 18 18 21 13 4 6 Cotton 1 2 2 3 2 1 Agricultural institutions in Togo are divided between cash crops and foodcrops 1/. OPAT, a parastatal agency, serves as a marketing board in setting prices for cash crops (cocoa, coffee, cotton, groundnuts, palm kernels and others), providing domestic marketing and processing, and selling for export. TOGOGRAIN, a relatively new agency, is supposed to stabilize food grain prices by operation of a buffer stock. So far, its activities and impact have been small. There are also extension agencies: the ORPV's (ex-SORADS) are responsible for regional agricultural development but have been hampered by a lack of clearly defined goals. Agencies for individual cash crops have been more active: SRCC (cocoa and coffee), SOTOCO (cotton), and SONAPH (oil palm). Currently, farmgate prices for cash crops are fixed by OPAT while prices of foodcrops are determined by market forces. Producer prices of cash crops have been about half of world market prices (based on F.O.B. Lome): 1/ "Cash" crops and "export" crops are used interchangeably in this paper to refer to those crops marketed by OPAT. Foodcrops are the foodgrains and starchy staples which may be produced both for subsistence consump- tion as well as for cash sale. This distinction is not absolute since groundnuts and palm oil are marketed both domestically and for export. - 2 - Nominal Protection Coefficient 1/ Cocoa (Avg. 1967-76) .47 Coffee (Avg. 1967-76) .45 Cotton (Avg. 1972-77) .51 Palm Kernels (1977) .57 Togo producer prices have tended to be lower than producer prices in other CFA zone countries (see Table 2), especially the Ivory Coast 2/. For cocoa, this is a reversal of the earlier situation (up to 1973-74) in which Togo-s prices were above those of the Ivory Coast and Cameroon. Foodcrop prices have exhibited a rising trend over the last 10 years, though with considerable cyclical fluctuations related to rainfall conditions. The drought in 1976-77 caused prices of staple food to triple, though they subsequently declined somewhat in response to heavy 1977 grain imports. A much better harvest in 1977-78 has brought foodcrop prices to very low levels. Increases in cash crop prices have not always kept up with trends in foodcrop prices (see Figure 1). Cash crop production performance has been mixed (see Figure 2). Cocoa exports (Togo½s major export crop) have declined sharply since 1971, due partly to aging trees and partly to lower smuggled volume from Ghana. Coffee exports have declined slightly since 1971, but recent efforts to expand the planted area have been quite successful. Cotton production fell off sharply in 1975 with the beginning of substantial foodcrop price increases, but there are indications that cotton outputs is up in 1978. Consistent times-series data are not available for foodcrop production, which seems to be trending upward at slightly less than the population growth rate. Foodcrop production suffered in recent drought years but has rebounded strongly with better rainfall in 1978. The Bank Group has been involved in three projects in the rural sector so far. These are a Cocoa/Coffee Planting Project (approved in 1975), a Cotton Project that also promoted foodcrop production (1978), and a Rural Development Project in the Maritime Region (1976). A Second Cocoa/Coffee Project was appraised in September 1978 and future plans call for several more rural development projects. The Togolese government's stated goals with respect to agriculture are (1) to increase the productivity of cash crops, (2) to promote self- sufficiency in foodcrop production, and (3) to promote regional development to mitigate regional disparities in income. There is presently considerable interest in water management as a necessary step in countering the marginal rainfall conditions prevalent over much of the country. Recent agricultural projects have tended to combine efforts directed toward cash crop with intro- duction of improved techniques and varieties for foodcrop production. 1/ NPC = Producer Price _t C F.O.B. Lome price - internal marketing cost. 2/ This is a superficial comparison of returns to producers since costs of production may vary among countries. TABLE 2 COMPARISON OF PRODUCER PRICES IN CFA ZONE COCOA Togo Ivory Coast Cameroon 70-71 93 85 85 71-72 93 85 90-75 72-73 93 85 90-75 73-74 95 110 100-80 74-75 115 175 120-100 75-76 120 175 130-120 76-77 130 180 150 77-78 150 250 220 78-79 200 COFFEE Togo Ivory Coast Cameroon Benin 70-71 75 105 125 71-72 75 105 125 72-73 80 105 125 73-74 90/95 120 130 98 74-75 105 150 135 103 75-76 115 150 145 110 76-77 125 180 195 115 77-78 145 250 230 COTTON Togo Ivory Coast Benin 72-73 35 40 36 73-74 37 45 36 74-75 46 70 40 75-76 48 70 45 76-77 50 80 50 77-78 60 Source: BCEAO, UMOA: Conjoncture economique, fin 1977. No. 258, Feb., 1978 and Rapport d'Evaluation du Projet en Cours, SRCC. - 4 - FIGURE 1: AGRICULTURAL COMMODITY PRICE MOVEMENTS (PRICE INDEX, 1963 = 100) 400 350- 250t oo Q 200 \ / / , ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Co . li Oa 300 _ |/ r__/0 150 - ,"~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~00 250 t966i67 1 968/69 1970/71 1972/73 1974/75 1976 /77 1977/78 CROP YEAR W o0d Bank -- 20177 MARKETED PRODUCTION, METRIC TONS a -~~~~~~~~~~~~~~~~~~~~~~~~~~~- 41- -_ m 9 ... _ ,.,,. ''''' \ ~~~~~~~~~~.4. _0C 4. I -__ C)> M .......... m~~~- C) -4---- 4~~~~~~~~~~~~~~~~~~ -6- Pricing Policy Objectives In order to evaluate the effectiveness of pricing policies, it is useful first to define the objectives that the policies are intended to serve. Since this analysis is limited to export crops, the two primary objectives are likely to be the generation of foreign exchange earnings and the generation of government revenues. If the primary goal of the government is to maximize economic product (e.g., maximizing foreign exchange earned from export crops), then producer price should equal world price. 1/ However, for the government to maximize its revenues from the marketing of cash crops, the producer price should be considerably below the world price, the exact level depending on the characteristics of supply. There will be some trade off between maximizing economic efficiency and maximizing government revenue since a producer price below world price will reduce production. (graph below) FIGURE 3: OUTCOME OF ALTERNATIVE PRICING POLICIES FOR EXPORT CROPS - o~~~~~~~~s Government Gain to Price Revenue Society Pt - Output Qt~~Q Worla Bank - 20179 This argument is illustrated in Figure 3. SS is the cash crop supply curve, Pw is world price and Pt is producer price. Qt is produced, purchased by OPAT at Pt, and resold at Pw. OPAT revenues are the diagonally- shaded area, which represents a transfer from producers to OPAT. If producer price were raised to world price levels, Qt would be produced and the net 1/ World price refers to the economic or border price at the farmgate; i.e., F.O.B. Lome price less domestic transport, marketing and handling charges. - 7 - gain to society of the extra production would be the horizontally-shaded triangle, which represents additional foreign exchange earnings. The size of this triangle will depend on the magnitude of the price distortion, Pw - Pt, and the elasticity of the long-run supply curve which will determine Qt' - Qt. The cost of this foreign exchange-maximizing policy is the eliaination of the surplus collected by OPAT. In Togo, the marketing board collects revenues that the Government would be hard put to replace from other sources. For the case of a relatively inelastic supply curve, the loss of foreign exchange will be smnall relative to government revenue generated. One could argue that government revenues are preferable in this case if they can be invested more efficiently in an eco- nomic sense by OPAT than they could be by cash crop producers. This amounts to weighting government revenues more highly than private incomes, and may well correspond to present Togolese policy. Maximizing government revenue wsill depend on world prices and the elasticity of supply for the cash crop. T'he more elastic supply is, the greater the revenue.maximizing producer price EPT -lsiiyo will be, as given by this formula 1/: Pt/Pw = IEPT where EPT supply with respect to producer price. The government revenue.maximizing producer price is shown as a function of supply elaE-ticity in Figure 4 below. Thus, for a supply elasticity of 0.5, producer prices would be set at only 0.33 of the world market price in order to maximize government revenues under a marketing board situation. As suply elasticity increas;es, the revenue-maximizing producer price asymptoti- cally approaches the wo-rld market price. Any rational producer price will be between the world price and the revenue-maximizing price. The higher govern- ment revenues are we'ght:ed in relation to private incomes, the closer producer prices should be totherevenue-maximizing price. This latter price can be regarded as a minimum rational producer price. The extent to .wh4ch ary policy objectives can be realized will be constrained by the level of world prices, over which Togo has very little control since it produces at most 1% of the world supply of any crop. Another constraint is the cost of domestic resources used in production of the export crop which will be determflned in part by prices of competing foodcrops. The level of world prices wilL affect the earning potential of export crops and the "shadow" or "economic" prices of land, labor and capital will determine whlch activities are the mnost efficient earners of foreign exchange. The spre4d be-qeen world prices and domestic costs will also affect the margin of profit that the government can capture through intervention in export markets. It is clear that the objectives of maximizing foreign exchange earned by the economy and maximizing Government revenues from export crop mnarketing provide inherently conflicting signals for agricultural pricing 1/ For a mathematical derivation of the revenue maximizing producer price Pee Annex A. - 8 - FIGURE 4. GOVERNMENT REVENUE - MAXIMIZING PRODUCER PRICE AS A FUNCTION OF SUPPLY ELASTICITY RANGE OF RATIONAL PRODUCER PRICES WORLD MARKET Pt/P°6E~~~~~~~~~~~~~~~~P IC 0.6 ptIPw REVENUE - MAXIMIZING PRODUCER 0.4 - PRICE AS FRACTION OF VWOR LD M AR KET PR I CE 0.2 - t ~I I 0 1 2 3 4 5 SUPPLY ELASTICITY,EPT World Ban k-22438 - 9 - policy. Policy choices are further complicated by uncertainties associated with world markets and climatic conditions. This paper does not attempt to choose between objectives, but rather to quantify, as far as possible, the economic choices involved. The approach developed should provide the Government with better tools to analyze the policy options available. Methodology The methodology employed allows an evaluation of the effectiveness of pricing policies in achieving the objectives outlined above. The estima- tion of supply response, based on past production and price data, is used to determine the optimal producer price from a revenue-maximizing viewpoint. An analysis of farm budget data provides two important types of information. First, economic profitability can be computed to determine whether the crop does in fact utilize domestic resources to generate foreign exchange in an efficient manner. Second, the structure of incentives faced by private pro- ducers can be analyzed t:o check whether prices are sufficient to encourage production of the crop in question. In order to determine supply response, supply functions were esti- mated for cocoa and cotton using historical data and standard statistical techniques. These functions give estimates of the supply elasticity, or change in production for a given price change, which quantifies the farmers response to price incent:ives. Lack of accurate acreage data precludes esti- mation of the price responsiveness of acreage planted in perennial crops. Therefore, for cocoa, only estimates of short-term production response could be made assuming acreage as given. The estimates of short-term price response for cocoa production were also complicated by the presence of smuggling (see Annex A). Coffee supply response was not computed in detail as the techniques involved are similar to those applied to cocoa. The estimated price elasticities are then used to stimulate revenue maximizing producer prices for cocoa and cotton. This allows an evaluation of past pricing strategies in terms of revenue maximization and provides an estimation procedure for future price-setting if this objective is of major importance. Using the farm budgets, economic profitability is evaluated first, using the world price of the cash crop, world prices of tradeable inputs and opportunity costs of non-tradeable inputs. For perennial crops, future values of inputs and outputs are discounted to the present. Economic profitability calculations indicate the contribution of export crop production to economic product; in other words, from the national economic point of view, whether production of the crop represents a sensible use of resources or not. Private profitability is then evaluated, also using farm budget data, except that the actual prices facing farmers (producer prices of outputs and market prices of inputs) are used in place of the economic (or efficiency) prices. Analysis of private profitability reveals whether current output prices and input subsidies encourage or discourage production by the farmer.. Additional measures of comparative advantage and incentives have been used in the analysis of cotton (see Annex C). - 10 - Analytical Results Cocoa Cocoa is the major foreign exchange earner among agricultural exports and hence the most important source of OPAT profits. Between 1966 and 1975, cocoa provided an average of 27% of recorded export earnings, and 64% of OPAT profits. A number of factors complicate an analysis of price response: (1) the nature of perennial crop production, (2) the existence of substantial smuggling from the Volta Region of Ghana, and (3) the lack of available land for new plantings. There is an eight year lag between planting and production for cocoa trees, and once planted a tree will continue to produce for forty to sixty years. This means that production in any given year will be primarily determined by the existing stock of trees. Therefore current production will reflect the planting response to prices eight or more years ago. Current production will also vary according to how intensively the existing trees are harvested. This will depend upon the opportunity cost of labor, which in turn will depend upon returns in foodcrop production. In Togo, observed quantities include an unknown amount of smuggling from the Volta Region of Ghana. This may explain the abnormally high levels of marketed production observed in 1968-70. The existence of smuggling means that price response in any given year will be more elastic than is usually the case for a perennial crop. The elasticity of short-term supply was estimated at .51. Using this elasticity optimal revenue maximizing prices were calculated for 1967-76. The results show that OPAT's policies have maximized government revenue taking acreage and age structure as given. In other words, the short-term revenue maximization prices maximize revenue treating acreage as exogenous, but do not necessarily provide incen- tives for planting. In the long run, supply will be more elastic as acreage responds to price. The more elastic the supply curve, the higher the optimal price for revenue maximization. Hence the optimal price calculated on the basis of supply elasticity = .51 is somewhat lower than optimum from a public revenue point of view. Forty-four percent of the cocoa trees in Togo are over 38 years old. This aged stock of trees implies declining production prospects. Suitable conditions for cocoa cultivation exist only in limited areas that have already been densely planted. This means that future increases in productive capacity will have to come about through replanting of very old trees. Since these trees still yield 150 kg/hectare, replanting entails opportunity cost of income foregone from cutting down the old trees. In economic prices, the internal rate of return on replanting a hectare of cocoa is 15.8%. This is greater than the economic opportunity cost of capital in Togo, which has been estimated at 8%. An increase in labor costs of 20% only reduces the rate of return to 14.5%. Replanting thus appears to be economically profitable. Two different types of budgets are used to evaluate private profit- ability. The first uses hired labor throughout and the second uses hired labor in the first six years and sharecroppers after new production begins. Table 3 shows discount rates for which replanting is privately profitable, under various assumptions of producer price. Because of the lost income in early years due to cutting down the old trees, very high private discount rates will make the investment decision appear unattractive. At lower discount rates, the higher value of output achieved in later years becomes more important. Depending upon producer price and labor cost assumptions, the investment will be privately profitable if discount rates below 15%-20% are assumed. Increases in producer price will make the replanting harvest more attractive unless very high discount rates apply. TABLE 3 RANGE OF DISCOUNT RATES FOR WHICH REPLANTING SHOWS A POSITIVE RETURN Producer Prices (present price = 200 CFAF/kg) 200 250 Positive Return for all Discount Rates less than: Hired Labor 18% 20% 21% Sharecropper 12% 14% 15% Source: Annex A, Tables 2 and 3. Perhaps the most important factor affecting private costs is the nature of labor markets in the cocoa producing region. Sharecroppers are employed on 85% of the plantations in the Litime area and 41% in the Kloto area. Of these, three-quarters are paid by taking a 1/3 share of the crop. 1/ As Table 3 shows, this type of arrangement does not make replanting very attractive to the plantation owner. 1/ "Sharecropping" takes two forms. The laborer may work on the harvest only and is paid a fixed rate (currently 600 CFAF) per bag. The more common arrangement is for the sharecropper to manage the farm and harvest the cocoa for a 1/3 share of the crop. - 12 - Sharecroppers are, of course, not interested in replanting since this deprives them of their income from the harvest of existing cocoa. Due to lack of land in the cocoa producing areas, many sharecroppers migrate to their homes in northern regions in order to plant foodcrops, and then return to harvest cocoa. Replanting would require year-round labor, which these sharecroppers could not provide even if economic incentives were right. In order to strengthen productive capacity new trees must be planted. The Government must focus attention on alleviating institutional constraints to labor availability and providing economic incentives to replant cocoa. Two options are available to government: (1) Raising the producer price substantially, thereby raising long run price expectations and encour- aging replanting. (2) Continuing current price setting policies but using some revenues to subsidize replanting. Current price setting policies maximize revenue in the short run, taking into account inflation and food- crop prices. The use of subsidies to encourage replanting would introduce a distortion to correct for already existing distortions and is therefore not desirable from a strictly economic point of view. However, it does have the advantage of being directed at a specific group of planters. The decision to raise the 1978/79 cocoa producer price to 200 CFAF per kg is in keeping with OPAT's price strategy which has maximized revenues in the short run. This increase should have some favorable effect on planting decisions. How- ever since world cocoa prices are currently high, the government has consider- able scope for implementing either of the above options, to further promote replanting. Coffee Coffee is Togo's second most important agricultural export, account- ing for 15% of total export earnings between 1965 and 1976. While the very high recent world market prices have declined, the long-term outlook for coffee prices is quite favorable. An analysis of farm budget data and pro- duction characteristics, which differ substantially from cocoa, result in rather different implications for price policy. Rainfall should be a more important factor in coffee production than cocoa production for two reasons. First, coffee yields are more susceptible to low rainfall combined with lack of maintenance. Second, coffee in Togo is planted in some areas that have marginal rainfall for tree crops. Average rainfall in the coffee producing areas is lower than average rainfall in cocoa areas. There are several factors that make coffee replanting more profit- able than cocoa replanting. First, the lag between planting and production is much shorter: 3-4 years instead of 7-8. Second, old coffee trees that could potentially be replanted have been abandoned and no longer produce. Bringing these areas back into production would not entail any opportunity cost of foregone production to either the producer or the economy. Finally, the Bank's projected world prices for coffee are much higher than those for cocoa. - 13 - Replanting coffee is highly profitable in both economic and private prices. At producer prices greater than 115 CFAF/kg, replanting has a return of greater than 20%. During the years when the tree is bearing, coffee is a more labor- intensive crop than cocoa. 'Cocoa requires an average of 125 mandays/ha of labor while coffee requires 193 mandays/ha. As a result, the returns per manday for coffee are lower than for cocoa (590 CFAF vs 844 CFAF) at 1977/78 producer prices. For maintenance activities alone, the difference is even more striking. Cocoa requires 48 mandays/ha, while coffee requires 103 mandays/ha. This is due to higher labor requirements for weeding and pruning. Perhaps due to higher labor requirements associated with coffee, more family labor is employed in coffee production. Fewer farms can rely on sharecroppers in the coffee producing areas. The sharecroppers who do work in these areas usually raise foodcrops also, in contrast to cocoa sharecroppers. The factors outlined above lead to the conclusion that economic incentives will be more effective in promoting coffee production than they would be for cocoa production. Private producers do not face as long a lag between planting and production. There is no opportunity cost of income fore- gone from cutting down old coffee trees. Institutional constraints on labor availability are not as acute for coffee.' Maintenance of coffee, especially when rainfall is deficient, will affect production. Higher prices will raise not only incentives to plant but also incentives to maintain existing trees, which will be more important for coffee production than for cocoa. Cotton In terms of government revenue or foreign exchange earnings, cotton is not as important as the perennial crops. However, cotton could be produced in all regions of Togo, and so the potential for significant expansion of cotton production is much greater. Cotton could provide cash income to a large segment of the agricultural population. Until 1965, only the low-yielding mono variety of cotton was produced in Togo. This variety is intercropped with yams. Starting in 1965, single- stand production of allen variety cotton was promoted and mono production was discouraged. Lower prices were offered for mono cotton and marketing arrange- ments were less favorable. Fertilizer and insecticide inputs for allen cotton production are subsidized. Production of mono declined rapidly while allen production increased slowly. Then, rising foodcrop prices starting in 1975 and drought in 1976/77 caused allen cotton production to decline. Data was available on allen cotton production by region for 1968/69 to 1977/78. Since production systems differ between regions, supply functions were estimated separately for the Plateaux and Centrale regions. These two regions account for more than 95% of all cotton production during this period. Each function was estimated using rainfall and foodcrop price data specific to the region. Rainfall was only significant for the Centrale Region. - 14 - It was difficult to separate effects of foodcrop prices and cotton prices since the two series moved closely together. The reported supply elasticities are for cotton prices deflated by an index of foodcrop price movements. Elasticity of cotton supply with respect to producer price * foodcrop prices Plateaux Region 2.2 Centrale Region 3.4 These high elasticities demonstrate the importance of prices of competing activities in producer response. While these elasticities over-estimate price response due to the use of production rather than acreage data, it seems safe to conclude that supply response of cotton production in Togo is elastic with respect to relative cotton/foodcrop prices. 1/ Price responsiveness to competing crops is further illustrated by an analysis of farm budget data. First, it has been computed that returns per manday for cotton production in the Plateaux Region were higher than those in the Centrale Region, due to differences in the labor requirements. The greater profitability of cotton-growing in the Plateaux Region may partially explain the more rapid adoption of allen cotton in the Plateaux Region. Second, an analysis of the different technological levels of cotton production indicates that the higher stages (using fertilizer, insecticides, etc.) yield considerably higher returns per manday than the more basic techniques. This may explain the fact that during recent drought years, the farmers who aban- doned cotton production were primarily those in Stage I, the least remunera- tive production technique. Cotton production is economically profitable in Togo, and more sophisticated techniques of production are relatively more profitable. This is to be expected since the more advanced techniques use less labor (Togo's scarcest resource) per unit of output. Production in the Plateaux Region is relatively more profitable than in the Centrale Region, because of the greater rainfall which allows double-cropping and hence fewer labor inputs. For cotton production in Togo, there are price distortions on both outputs and inputs. Fertilizer and insecticides are sold to farmers at prices below cost. Subsidized inputs do not offset the producer price distortion, however. The net effect of the price distortions is to discourage cotton production relative to the situation that would obtain with no price distortions. 1/ Estimates of cotton supply elasticities mentioned in "Price Prospects for Major Primary Commodities" are as follows: United States, 1.97; Greece, 1.00; aggregate for Mexico, Guatemala, El Salvador, Nicaragua, and Colombia, 1.55; Pakistan, 1.65; Turkey, 0.21; and Ivory Coast, 3.69. - 15 - It is argued that these subsidies are needed to overcome farmer resistance to modern techniques of production. Since the subsidized inputs do.require less cash outlay, the risk to the farmer of adopting the new techniques may be lessened. The subsidies do have the effect of making the more advanced techniques slightly more profitable to the private producer. This is desirable since the advanced techniques are more economically profitable. The subsidies are essentially costless to government since cotton output is heavily taxed. These subsidies may make sense for promotion of more modern techniques of production. However, this policy should be re- evaluated periodically as use of modern inputs becomes more common or extends to foodcrops. The analysis above indicates that cotton supply response is elastic. This result coupled with,a producer price that is on average only half of the world price will mean a substantial loss of earned foreign exchange. For example, taking only production in the Plateaux Region, and the estimated elasticity of 2.2, the loss of foreign exchange due to the price distortion in 1977 is 170,000 CFAF while government revenues are 197,000 CFAF. The loss to society is nearly as great as government revenues. This loss may be exaggerated if the estimated elasticity, 2.2, is higher than actual response would have been. However, this calculation demonstrates that when the supply curve is elastic, a large price distortion will cause large losses of foreign exchange. In terms of government revenue, an optimal producer price can be found by using the estimated elasticity for the Plateaux Region and a world price equal to the economic farmgate value minus the cost of subsidies per kg of output for Technique I. This "world price" includes the value of both seed and lint. Simulated Actual Producer Price Producer Price (designed to maximize Government revenue) 72 21-24 35 73 40-46 35 74 67-78 37 75 37-43 46 76 75-88 48 77 64-74 50 78 -- 60 Actual prices in the last few years have tended to be too low to maximize government revenue, if we take the Plateaux Region as representative. Estimated price responsiveness in the Centrale Region has been even higher than in the Plateaux Region, which should raise the aggregate elasticity for the country as a whole. The higher price responsiveness is, the higher the optimal price from the point of view of government revenue. - 16 - It seems that low cotton producer prices have caused substantial losses of foreign exchange and may even have been too low to maximize govern- ment revenues. This is because producer response is very sensitive to prices of competing activities. If cotton prices are raised, providing higher in- centives relative to foodcrop prices, cotton production should increase more than proportionally to the price increase. This means that government reve- nues will increase more than proportionately to the price increase, up to the revenue maximizing price. If the government provides greater incentives to cotton production this may pull resources out of foodcrop production, which may be undesirable if cotton producers are better off than foodcrop producers. The risk of drought will constrain revenue maximization. In years of disastrous drought, the perceived risk of obtaining food may be so high that no cotton price will maintain production levels. In such situations, foodcrop production may be the best earner of foreign exchange by reducing the need for food imports. This implies that, in setting competitive cotton prices, OPAT must try to distinguish between longer-term trends in foodcrop prices and those which reflect drought conditions in a single year. Suggestions for Future Study Among the cash crops, it is recommended that further analysis of coffee price response be carried out. This could include an analysis of farm budgets to study private incentives to maintain existing trees in order to determine the scope for short-run price response. As more data becomes available on acreage planted, a supply analysis of planting response could be made. Future study of price policy in Togo should also include analysis of foodcrops in terms of farm budget data, supply response and marketing. The specific types of analysis which could be carried out include: economic profitability, private profitability, and effective rates of protection. Returns per man-day are a useful measure of profitability if labor is the only scarce resource (which is the case in most of Togo). This type of analysis would provide information on which crops and which production techniques are economically profitable and what incentives private producers perceive. This information could provide guidance to the Government on which crops to encourage, through extension or integrated projects. Supply response could be analyzed once sufficient time series of production data have been built up. The price data collected for a group of 7 markets throughout Togo from 1966-1976 are quite useful, and can be tied into the more detailed pricing series available from 1977 on. However, pro- duction and acreage data have only been gathered in detailed form since 1972. Supply analysis could not therefore begin until around 1982. Marketing issues will be of interest if the Government becomes seriously involved in marketing foodcrops. The present series of foodcrop prices could provide two types of information: (1) Seasonal indices of price movements could be calculated. These would give some idea of storage costs - 17 - from harvest to harvest. (2) Differences in prices between towns would give some notion of transportation differentials. If seasonal or spatial differ- entials seem too high this may indicate some constraint on the marketing system such as lack of storage facilities or transportation. The types of analysis outlined above will provide useful information on the effects of any given price policy. Choosing a price policy will depend on the objectives chosen and the constraints - political and social as well as economic - on achieving those objectives. Conclusions and Recommendations The results of the present study confirm, for the Togolese case, that agricultural production responds to economic incentives. Farmers will divert their attention from cash crops to foodcrops during periods of drought to assure a food supply and to benefit from higher foodcrop prices. Farmers will not make a heavy investment in perennial crops when positive returns are far away and substantial costs (whether opportunity or out-of-pocket costs) are foreseen over the medium-term. Depending upon relative prices in neighboring countries and the degree of border surveillance, agricultural commodities will move across boundaries in accordance with economic incentives. Since producers respond to price and private prices diverge from economic prices, there will be a loss of foreign exchange entailed in OPAT's intervention in export markets. This intervention generates important public revenues that represent a substantial tax on the agricultural sector. These revenues have essentially been used to support the industrialization of Togo. Given Togo's difficult medium-term financial outlook, they are likely to remain critical in the years ahead. In order for the agricultural crops to continue to generate revenues, production incentives must be adequate to ensure a growing supply. In the case of perennial crops, prices have not been sufficient to assure long run growth in supply. For cotton, prices have not been high enough to provide incentives to produce in the face of rising foodcrop prices. The outcome of any price policy will of course be subject to the uncertainties of world markets and rainfall. In setting prices, these very real constraints on perfect knowledge must be recognized. The analyses of individual crops show how price policies affect pro- duction incentives and how they serve the conflicting objectives of earning foreign exchange and of generating government revenue. The highlights of these analyses and the resulting recommendations are as follows: (1) COCOA The government has pursued a short-run revenue maximization policy, treating acreage as exogenous. Meanwhile, productive capacity has declined due to the aging stock of trees. Increased production can only come about - 18 - through replanting of existing acreage due to the lack of new land. Replant- ing was found to be economically profitable but may not be privately profit- able due to the need to rip out still-bearing trees, coupled with high private discount rates and institutional constraints on labor. Subsidies may be the most cost effective method of overcoming institutional constraints and providing incentive to replant. The recent producer price increase from 150 CFAF/kg to 200 CFAF/kg is in line with the short-term public revenue maximizing policy pursued to date (see Figure 5), and does improve the private profitability of replanting. However, to assure the success of replanting programs, further economic incentives need to be provided to overcome high private discount rates. Subsidies would be most effective and could be used to provide incentives to sharecroppers. Another option would be further investigation of technical alternatives to replanting such as rehabilitation of old trees. However, any increase in productive capacity must come about through more labor intensive techniques, which means that increased private incentives will be necessary. (2) COFFEE The analysis of coffee shows that production is quite profitable in economic terms. Labor inputs are important in maintaining yields of mature trees, which means that producers can respond quickly to price incentives. Incentives to promote coffee production should be more effective than for cocoa production due to the importance of labor inputs, the shorter lag between planting and production, and the availability of new land. Bank projected prices for coffee are high (see Figure 6), indicating considerable latitude for price increases. Given the greater importance of labor inputs for coffee, the coffee producer price should be at least as high as the cocoa producer price. Increases in producer price could begin next year, allowing time for reevaluation if world market trends reverse. (3) COTTON Cotton production is economically profitable, at least in the two regions analyzed (Plateaux and Centrale). Since the supply analysis shows that cotton production response to price is very elastic, there will be large losses of foreign exchange for a producer price below world price. This highly elastic price response also indicates that producer prices may have been too low to maximize government revenues. World Bank cotton price projections indicate high world prices in the future (see Figure 7). Since present producer prices are low even from a revenue generating standpoint, an increase to at least 80 CFAF/kg is rec- ommended. This increase could be accomplished in stages in order to allow for evaluation of market trends. The techniques illustrated here--supply response analysis and farm budget evaluations of private and social profitability--could be applied by the Togolese Government on an annual basis as part of the price-setting exer- cise. These analyses would indicate explicitly the trade-offs involved and provide a rational basis for a pricing policy. OPAT would be the agency - 19 - primarily concerned with the results of the analysis. Branches of Government having the appropriate technical skills and data - such as the D)irectorate of Agricultural Statistics and the Rural Development Division of the Plan - should carry out the necessary calculations. The Bank can assist in familiarizing Togolese personnel with the analytical techniques and can provide projections of world market prices. - 20 - FIGURE 5: COCOA PRICES 900 Soo 700 SODg | l { i/ i |E:sp01World Market Price!- 600 I t\ - // .' *'. , 4 Export Price, .ob. m / <500I 1. 00__-, __. ____"- ____. _ ___ _ _ _ ___ I __________ ._ _ _ _ _ _ _ '___A_t__at__rod_e__P__c_4 400 300 A t a_* Prolt Rever ux M haximizing Sour :Producer Price 11 ~ ICODivPieIvrg,fo 200 / ,. .. . . .. . . .. . . .. . . . Actual Prod cer Price- Act ual-a- Ct Pp Pro jected 1967 1969 1972 1975 1978 1981 1984 1987 1990 YE AR Sources: 31 ICCO Daily Price Average, from Price Prospects for Major Pr, mary Commodit ies, IBRD Report No. 814/78, June 1978. 2, OPAT. Projected Pricet are astu med to WrdBn 08 be a Constant Proportion of Projected World Prices. 3/ Calculated using a Supply, Elasticity of 0.51 and f.o.b. Lome Prices. For dletails, see Annex A. 4 OPAT. - 21 - FIGURE 6:CoFFEE PRICES 120C- ______ _ lo o c -- _ _ _ _ _ _ _ _ _ _ _ * ~~~~~~~World Market Pricel/~' 08a .............. ~Price, fob. LormX 200 O__ __ Au Producer _ ...........AtuAclual Pr ducer PriceS 100 Actu1l 4P-PSojected Majo Priar Iommodites. IBR I ± 4.L L Ii L I L L± 1967 1971 1975 1979 1983 1987 1990 YEAR Sources: 1/ Guatemalan, Prime Weafed, Spot New York. from Price Prospects for World Bank - 20182 Major Primary Commodities, IBRD Report No. 814/78, June 19713. 2J OPAT. Projected Prices are Assumed to be a Constant Proportion of PNojected World Prices. 3/ OPAT. - 22 - FIGURE 7: COTTON PRICES 360 300- 00, World Market PricePo c 240 _ _Expo rtl_ EP Price, f.o.b. Lm LL~~~~~~~~~~~~~~~~4 tL~~~~~~~~~~~~~~~4 197 194 97 190 R98 wnue99 Actual Actual f Projected. Producer Pricee4 s 1972 1974 1977 1980 1983 1986 1990 YEAR 1JEconomic Farmgate Value of Seed Cotton, Computed uting Lint and Seed Prices from Price Prospects for World Bank - 20181 Major Primary Commodities, IBRD Report No. 814/78, June 1978. V? OPAT. Projected Prices are Assumed to be a Constant Proportion of Projected World Prices. ~JCalculated using a Supply Elasticity of 2.2 and f.o.b. Lomd Prices. For details, see Annex C. 4JOPAT. - 23 - ANNEX A Page 1 TOGO COCOA Cocoa marketings are the most important source of foreign exchange among agricultural exports and hence the most important source of OPAT profits. Between 1966 and 1975, cocoa provided an average of 27% of total export earn- ings, and 64% of OPAT profits. Cocoa is produced in the Kloto and Litime areas of the Plateaux Region. Current plantings are very old, with some 44% of plantLngs over 38 years old. The Togolese Government is naturally concerned with the declining production that this aging stock of trees implies. However, little area is available for new plantings. The current cocoa/coffee project has met with only limited success in replanting old cocoa, due partly to farmer reluctance to uproot old trees. A large component of observed cocoa production in Togo has been smuggled from the'Volt'a region of Ghana. A cursory examination of quantities purchased by OPAT between 1959 and the present reveals swings in production of up to 15,000 tons. Such variance in domestic production is not possible in so short a period due to the fixed nature of tree crops. Acreage planted in cocoa in the Volta region is three times that planted in Togo, indicating that smuggled quantities could be large relative to domestic production. This paper will examine the supply response of cocoa harvestings, since not enough data is available to estimate acreage response. These esti- mates will take into account the effects of smuggling. Then, the private and economic profitability of replanting existing cocoa acreage will be examined along with the institutional factors that cause the two to diverge. Finally, an analysis of price policy in terms of government revenue and economic objec- tives will be made. Supply Response An investigation of determinants of cocoa supply'in Togo must take into account smuggling from Ghana, and therefore production in Ghana. Random elements such as weather variations or pests will affect production on both sides of the border. By grouping the supply equations for production in Togo and Volta region together, it would be possible to utilize the existence of smuggling and the cross-equation correlation of errors to produce more effi- cient estimates. However, this is not possible due to lack of information about the Volta Region. Instead, estimates are made for Togo only of the price response of harvested quantities as well as the response of smuggled quantities to the price difference between the countries. Since these esti- mates are for Togo only without taking into account factors affecting produc- tion in Volta, they may be biased. However the results demonstrated that both smuggled and harvested quantities will respond to price'changes. - 24 - ANNEX A Page 2 Data on production and marketing board-determined producer prices is available for Togo from 1959/60 to 1976/77. This series is too short for an analysis of planted acreage response to price. Instead, the price responsive- ness of harvested quantities from the existing trees was estimated, using data from 1959/60 to 1972/73. The most important characteristic of perennial crop production is the long life of the trees and the lag between planting and production. Amelando, the type of cocoa planted in Togo, does not begin to bear until 18 years old. Yields increase until 10-15 years and then decline slowly (see graph). The tree may continue to produce until 60 or more depending upon weather and soil conditions. The fixed nature of the stock of trees means that in the short run the largest determinant of production will be acreage and the age structure of existing trees. Nineteen hundred and seventy seven's hectarage by age of tree was available for Togo. Plantings were high after World War II, declining in the 60's to the present situation of virtually no new plantings. It was possible to estimate past hectarage and the age of the average hectare for the period under study. 1/. Average.age over the period studied was always greater than 18 years. Age should have a negative relationship to production since 18 is past the peak bearing age. The cost of harvesting cocoa will be determined by the opportunity cost of labor employed. An indirect method of evaluating the opportunity cost of labor is the price of fooderops. The principal cocoa harvest occurs between September and December, and the smaller secondary harvest occurs in April. The principal harvest coincides with the second food season in the cocoa producing area. The secondary harvest coincides with land preparation for the principal foodcrop season. In Togo the practice of sharecropping is very common. In the two principal cocoa producing areas of Litime and Kloto, respectively 64% and 31% of the cocoa farms are sharecropped. Sharecroppers maintain the farms and harvest cocoa in return for 1/3 of the harvest. The trade-off between time spent producing food and the returns from only 1/3 of the cocoa harvest may be significant. 1/ The data gives hectarage by age group. Total hectarage in an age group is assumed to be distributed equally across individual years. The number of hectares in age group one is subtracted from total hectares in 1977 to give total hectares in 1976, and so forth to 1959. It was assumed that no trees dropped out of production due to age. - 25 - Yo,Id Kg/ba FIGURE A-1: COCOA YIELD PATTERN ' 0 --I 4)0 iX 5 10 15 20 25 30 35 40 Years from Field Planting Source: IFCC Upper Amazon Hybrids. Assumptions proposed for Second I DA Cocoa/Coffee Project, Togo. World Bank --20184 - 26 - ANNEX A Page 4 Fooderop prices were averaged over the August to July cocoa season. An index of price movements was constructed for each crop. 1/ An average of these indices was used to deflate cocoa prices received by the farmer. The principal foodcrops in the region are maize,-yams, and manioc. Foodcrop prices are those reported in Palime, a major center in the cocoa producing region. The average index was constructed by weighting each crop index by the proportion of cultivated area in the Plateaux Region devoted to the crop. 2/ Due to lack of information, it was assumed that good crop prices were constant from 1959/60 to 1963/64. It is difficult to specify Togolese and Ghanaian cocoa prices in commensurate units. Both prices must be expressed in a common unit of foreign exchange. The value of the Ghanaian cedi has fluctuated.widely over the period studies, and black market rates have often diverged widely from offi- cial rates. 3/ Since historical data is not available on black market ex- change rates, some way of estimating the real relative value of Ghanaian and Togolese currencies is needed. The exchange rate is specified as the ratio of the Ghanaian CPI to the Togolese CPI multipled by a constant official exchange rate. 4/ The exchange rate then depends upon relative movements in the rates of inflation. This measures the currencies in terms of relative purchasing power. It is assumed that smugglers respond to the proportional difference in price, which is the ratio of the price difference to the mean expected price (both prices expressed in commensurate units). The mean expected price is the production weighted average of Togolese and Ghanaian prices. Rainfall data was available for Atakpame, a town in the Plateaux Region but outside the cocoa producing area. This rainfall variable did not produce significant results. Actual production in Togo will be a function of the variables discussed above: the producer price deflated by foodcrop prices, acreage, and age of trees. Observed production will also contain a smuggled component which will vary in response to the difference between Ghanaian prices and Togolese prices. 1/ An average index of price movements was used rather than average prices because of the problem of averaging across non-homogenous goods. 2/ 1972-73 census. 3/ For instance, current black market rates are 5 to 60 to the dollar whereas the official rate is 1.30/$. 4/ 242 CFAF/1. Ghanaian CPI obtained from DMF statistics, Togolese CPI from Togo government. -27 - ANNEX A Page 5 The estimated equation is: Q = -194695.8 + 39.6-p + 6.4 ACR - 1333.9-AGE + 6893.5 pT + bG F ~~ ~~~~~~T G t-statistic (2.2) (1.0) (1.9) (1.0) (1.4) R2 = .77 where: QT = observed production in Togo T = cocoa price in Togo deflated by index of food crop prices F ACR = area planted in cocoa AGE = median age of cocoa trees P T-PG = difference between Togolese and Ghanaian cocoa prices a = .33 b = .66 The signs of the coefficients are all as expected and the equation provides a good fit. Both production and smuggling respond in a positive manner to price. The response of harvested production to cocoa price deflated by foodcrop prices is very inelastic: .17. This would be expected since levels of production are very stable in the short run. The low t-statistics for PT P- is to be expected due to the lack of information on foodcrop prices in PFT the early '60s. The response of smuggled quantities to relative prices is also inelastic: .34. This may reflect barriers to movement of production across borders. The estimated response to the price difference wlll be lowered by costs by eluding law enforcement barriers to movement of quantities and exchange of currencies. The smuggling elasticity; .34, is much higher than the price elas- ticity for harvested production. Response of smuggled quantities to price should be more elastic than response of harvested production since the scope for smuggling response will be greater than the scope for production response. This means that price responsiveness of marketed quantities, Q , will be higher in Togo than would be the case without smuggling. The objective of the government marketing board may be to maximize net revenue, NR, which is equal to marketed production times the difference between world price, Pw, and producer price, Pt. NR = QT (PW - Pt) - 28 - ANNEX A Page 6 Since Togo's share of the world cocoa market is very small, Pw will not vary with Qt. For a given Pw there should be a Pt that will maximize net revenue. Taking the derivative of the net revenue function and setting it equal to zero, a formula is obtained for Pt that will maximize net revenue: Pt - Pw EPT 1+EPT where: EPT = elasticity of supply with respect to producer price Substituting the estimated elasticities (.34 and .17 = .51), it is possible to calculate an optimum Pt, for a given Pw. Average price, F.O.B. Lome, is available beginning in 1967. Prices actually received in Togo are usually lower than world prices, but in an un- systematic fashion. The following table gives actual producer prices and optimal producer prices calculated on the basis of F.O.B. Lome price. Revenue Maximizing Producer Price Based on F.O.B. Lome Actual Producer Price 1967 42 55 1968 50 70 1969 68 80 1970 76 88 1971 56 93 1972 51 93 1973 68 93 1974 121 95 1975 101 115 1976 136 130 Actual producer prices have followed the long run trend of simulated prices. In simulating these optimum producer prices, it has been assumed that acreage and age structure are given. In the long run, acreage, and hence age structure, will be some function of producer price. In other words, the revenue maximization strategy illustrated above maximizes revenue, treating acreage as exogenous, but does not necessarily provide incentives for replanting. - 29 - ANNEX A Page 7 In the above formula for the optimum Pt, the more elastic response of Qt to Pt is, the higher Pt should be. If acreage responds to Pt, then this increases the elasticity of the long-run total supply curve and increases the optimum level for Pt. It seems that OPAT has successfully maximized revenues, taking the productive capacity as given. However, the current age structure of trees reflects both a shortage of area available for new planting and a lack of incentives for replanting. In order to maintain productive capacity, appropriate incentives to replant cocoa must be maintained. The next section will examine private and economic 'profitability of replanting cocoa. Replanting The above analysis has identified the determinants of production in a given year. In order to maintain or increase cocoa production over the long term, new trees must be planted. In Togo, little area is available for expansion of cocoa plantings. Increases in production must come from re- planting areas currently occupied by old trees. Since old cocoa trees con- tinue to bear indefinitely, there is an opportunity cost of earnings foregone from old cocoa when it is ripped out to provide room for new cocoa trees. This increases the costs associated with the six year period before the new cocoa trees begin to bear. This section will evaluate the economic and financial profitability of replanting a hectare of old cocoa at the margin. As such, these calcula- tions will differ from those performed in project appraisal, where a project will have a greater than marginal impact on the sector. The costs and benefits associated with replanting are presented in Table 5-A of the Statistical Appendix. The alternative of inter-planting young trees among old, which decreases the opportunity cost of foregone cocoa production, has not been considered here. Evaluating alternative forms of project design is not the main focus of this analysis, but should be an important priority for further work. Several comments are in order regarding the assumptions underlying the economic costs and benefits. Information on material inputs and required mandays is drawn from farm budgets used in appraisal of the IDA second cocoa/ coffee project. Material costs have been valued at market prices. Labor inputs have been valued at 300 CFAF/day. This is the agricultural wage in cocoa producing areas and also accords with current returns per manday in foodcrop production. - 30 - ANNEX A Page 8 In evaluating economic profitability, the cost of extension services per hectare was estimated. Using information from the first cocoa-coffee project, the cost of. planting material, cocoa capsid control and extension workers has been estimated on a per hectare basis. The opportunity cost of land used in replanting is equal to the value of the cocoa produced by the old trees at world prices minus the costs of inputs. It is assumed that the old trees continue to produce 150 kg per hectare for 20 years. 1/ The cocoa produced is valued at world prices. Two different assump- tion of future world prices have been tried: (1) world price is constant over the life of the project and equal to the average 1975/76 farmgate economic price, and (2) World Bank commodity price projections have been used. In raising cocoa it is necessary to provide shade for the young trees in the first four years. Banana plantain plants are commonly used. If the plantains produced have market value, then their additional benefit should be included. The question of market value arises since the demand for plantain may be limited. Plantains were alternatively valued at zero or 15 CFAF/kg. The results of the four cases are presented in Table 1. The internal economic rate of return in all four cases is greater than or equal to the social opportunity cost of capital in Togo, which has been estimated at 8%. The use of the World Bank projected prices lowers the returns. This is because prices are currently high and projected to fall, thereby increasing opportunity costs in the first six years and reducing returns from the new cocoa. The returns are also sensitive to the assumptions regarding the value of banana plantains. Sensitivity to variations in labor costs does not appear to be great. Two different types of budgets have been constructed to evaluate private profitability. The first uses hired labor an~d the second uses hired labor in the first six years and sharecroppers after new production begins. These budgets differ from the economic budgets in three respects: (1) the price of cocoa (which affects both costs and benefits); (2) the cost of labor; and (3) the exclusion of extension costs. Banana revenues are included in the private budgets. Only budgets employing wage labor were used due to the shortage of family labor in the cocoa producing region. The first private budget values hired labor for replanting and maintenance at 300 CFAF/day. Harvest labor is paid 60 CFAF for each 30 kilograms of cocoa harvested. This gives an implicit wage of CFAF 240/day. Hence, harvest labor costs are lower than in the economic analysis. 1/ Observations of the appraisal mission for the current cocoa/coffee project indicate that this may underestimate yields of aged trees. - 31 - Annex A TABLE 1 Internal Rate of Return at Economic Prices Plantains No Plantains World Price = 20.5 16.5 334 CFA/kg. World Price at Bank 15.8 13.1 Projections With a 20% Increase in Labor Costs: Plantains No Plantains World Price = 18.7 15.2 334 CFA/kg. World Price at 14.5 12.0 Bank Projections - 32 - ANNEX A Page 9 A consistent internal rate of return cannot be obtained for the private budgets since the stream of net benefits changes sign more than once. However, the present value at various discount rates was calculated for the first budget under varying assumptions of producer price. The results in Table 2 show that the profitability of replanting increases with increases in price. At 1978/79 price = 200 CFAF/kg, replanting is profitable at any discount rate below 20%. At price = 100 CFAF/kg, which is closer to historical levels, replanting is only profitable up to a discount rate of 12%. At higher private discount rates, the cost of foregone income in the near term outweighs the benefits of increased incomes in the long term. Present values were calculated for the second private budget. Costs are identical to the first private budget in the first six years. After that, a sharecropper is given 1/3 of the cocoa in return for mainte- nance of the plantation and harvest labor. Net returns to the owner are the value of 2/3 of the crop. Present values for this budget at the 1977/78 producer price of 150 CFAF/kg are in Table 3. Under this arrangement, re- planting of a hectare of old cocoa is not very appealing to the plantation owner since labor costs are effectively-higher. Present value is only posi- tive up to a discount rate of 10% as compared to 16% for the hired labor budget. If replanting of cocoa is economically profitable and appears to be privately profitable in some circumstances, what are the factors that have made private entrepreneurs reluctant to replant? First, private entrepreneurs may have both a higher opportunity cost of capital and a different rate of time preference than society as a whole. Initial costs of replanting are high. The first year costs are three times the revenues from old cocoa. It may be difficult for farmers to sustain. 7 or 8 years of losses no matter how high the potential gains. Farmers may also have a much higher rate of preference for present consumption than society or government. The advanced age of the majority of cocoa farmers could be a factor. In the Litime area, 47% of the plantation owners are older than 45. Presumably these farmers/ owners are not very interested in an investment that has high initial costs and no substantial returns for the first 10 years. Another factor affecting private interest in replanting is the price level and farmer expectations of future price movements. As prices increase, profitability of replanting increases. However, if prices are expected to fall in the future, then current losses oveshadow future gains. Perhaps the most important factor affecting private costs is the nature of labor markets in the cocoa producing region. Sharecroppers are employed on 85% of the plantations in the Litime area and 41% in the Kloto area. Of these, three-quarters are paid by taking a 1/3 share of the crop. As was seen above, this type of arrangement does not make replanting very attractive to the plantation owner. Sharecroppers are of course not in- terested in replanting since this deprives them of their income from the cocoa harvest. - 33 - Annex A TABLE 2 Present value of rep:Lanting 1 ha. of cocoa for varying producer prices using all hired labor (CFAF) Price = 100 Price = 150 Price = 200 Price = 250 Discount Rate 8 134209 332683 531157 729630 10 71365 212458 353549 494641 12 27661 128597 229534 330470 14 -3321 68911 141143 213375 16 -25657 25664 76986 128307 18 -41999 -6166 29666 65499 20 -54108 -29915 -5723 18469 22 -63176i -47843 -32509 -17176 24 -70029 -61510 -52991 -44472 26 -75245 -72016 -68786 -65556 - 34 - TABLE 3 Annex A Present value of replanting cocoa using hired labor and sharecropper Discount Rate Price = 150 8 173967 10 65889 12 -8312 14 -60104 16 -96740 18 -122909 20 -141717 22 -155260 24 -164983 26 -171899 - 35 - ANNEX A Page 12 Many sharecroppers migrate to their homes in northern regions in order to plant foodcrops and return later in the year to harvest cocoa. From March to July, 45% of the labor force has migrated from the cocoa pro- ducing regions. The lack of opportunities for permanent settlement by share- croppers constrains the amount of labor available for replanting. In promoting replanting there are two problems that must be addressed. First, the constraint on labor availability caused by share- cropping arrangements. Some alternative form of land tenure or some method for compensating sharecroppers must be found. This problem must be solved before the second problem of economic incentives can be addressed. Economic incentives can be provided either by higher producer prices or subsidies to replanting. Higher producer prices increase the attractiveness of the replanting investment. Subsidies may have an advantage in that they will overcome the cocoa producer's preference for current consumption and could be divided between owner and sharecropper. Policy Objectives and Strategies OPAT policies set producer prices below world prices. Between 1967 and 1976 the average nominal protection coefficient (NPC) 1/ for cocoa was .47. This implies an export tax rate of .53. An NPC less than once means that production of cocoa is being discouraged by domestic policies which distort world market incentives. Producers have received only about half the economic value of their cocoa. In discussing price policy alternatives for perennial cash crops such as cocoa it may be useful to step back and examine the overall objec- tives.* In terms of economic efficiency, cocoa production does contribute to foreign exchange earnings and is economically profitable. If the primary goal of the government is to maximize economic product, then producer price should equal world price. However, the government also has the goal of obtaining revenues from marketing cocoa. There will be some trade off between maximizing economic efficiency and maximizing government revenue to the extent that a producer price below world price will reduce production. 1/ NPC Producer Price - World Price (Economic Farmgate Price) - 36 - ANNEX A Page 13 FIGURE A-2: OUTCOME OF ALTERNATIVE PRICING POLICIES FOR COCOA Price Government Gain to S Revenue Societyr Pw Pt Output at at World Bank - 20183 This argument can be formalized by the above graph. SS is the cocoa supply curve, Pw is world price and Pt is producer price. Qt is produced, purchased by OPAT at Pt, and resold at Pw. OPAT revenues are the shaded area. If producer price equaled world price, Qt' would be produced. The area under the SS curve between Qt and Qt' is the cost of producing more cocoa. The net gain to society of the extra production is the black triangle, which represents additional foreign exchange earnings. The size of this triangle will depend on the size of Pw-Pt and the elasticity of the long run supply curve. What then are the arguments in favor of the implicit export tax that a marketing board imposes? One is that production is kept low in order to keep world price high and maximize foreign exchange earnings. This means that world demand is not infinitely elastic as in the above graph. This argument cannot reasonably be applied to Togo since Togo's share of the world cocoa market is only about 1%. Another possible argument is that the marketing board will "'stabilize" producer prices by losing money in years of high prices. This is true if stability of income is perceived to have some utility apart from the level of income. The fluctuations of world prices have not always been mitigated for the producer. Graph A shows producer price movements over time. Prices, even in current terms, fell sharply twice in the early 60's, before OPAT administration. A marketing board does collect revenues that the government would be hard put to replace from other sources. One could argue that government revenues will contribute to economic growth if they are invested more effi- ciently by OPAT than they would have been by cocoa producers. - 37 - FIGURE A-3: ACTUAL COCOA PRODUCER PRICES .00 190 1 180 170 - 1 so 1 io 1.10 _ 1 30 B 1:10 100 0 w 110 0 :0 189/60 1962163 1965/66 1968/69 1971/72 1974175 1978179 CROP YEAR W.rld Bardk 20185 - 38 - ANNEX A Page 15 In Togo, an income distribution argument could also be made. Since cocoa production is limited to a small area, monopoly rents accrue to a small number of farmers. If OPAT funds were reinvested in other parts of the rural sector, it could be argued that the cocoa price distortion served social ends. OPAT funds have been primarily invested in industry or tourism projects, how- ever, so this argument would not apply. It seems that OPAT's primary function has been collection of govern- ment revenues. If the current marketing board system is accepted as a polit- ical reality, then what course should OPAT and/or the government take to minimize the distorting effects on the economy? One important way would be to reinvest OPAT earnings from cocoa productively. Another way would be to minimize the effect of the price distortion on production. Since cocoa production is economically profitable the question is how to provide private incentives to produce. At the present time long- run response to price in Togo will be limited by the availability of land. In Bateman's analysis of Ghanaian cocoa, he found that response to price varies inversely with the length of time that a region has been planted in cocoa. Two options are available to government: (1) Raising the producer price substantially, thereby raising long-run price expectations and encour- aging replanting. However, this may be expensive given the constraints to further supply. (2) Continue current price setting policies but use some revenues to subsidize replanting. Current price setting policies maximize revenue in the short run taking into account inflation and foodcrop prices. The use of subsidies to encourage replanting would introduce a distortion to correct for already existing distortions and is therefore not desirable from a strictly economic point of view. However, it does have the advantage of being directed at a specific group of planters. The effectiveness of economic incentives will be constrained by the nature of labor institutions in the cocoa producing area. The ability of the government to both collect revenue and encourage replanting will be constrained by future levels of world prices. In the past, OPAT has been successful in generating government rev- enue in the short run, ignoring implications of price for long-run productive capacity. But the objective of maximizing government revenue in the short run will not be served by a continuation of this policy. Neither will the objective of maximizing foreign exchange earnings for the economy as a whole. In order to strengthen productive capacity new trees must be planted. The Government must focus attention on alleviating institutional constraints to labor availability and providing economic incentives to replant cocoa. - 39 - ANNEX B Page 1 T O G O COFFEE Coffee is the second most important agricultural export, accounting for 15% of total export earnings between 1965 and 1976. Coffee is primarily cultivated in the Akposso and Dayes plateaus. However, coffee production is less geographically concentrated than cocoa production because coffee is more tolerant of soil and climate variation. Scattered coffee plantings exist throughout the eastern Plateaux Region and the northeastern Maritime Region. Coffee acreage is estimated at 50,000 ha, but it is estimated that some 12,000 hectares are abandoned due to the advanced age of the trees. Since the analysis of coffee is similar to cocoa, this paper will be on those production characteristics that differentiate coffee from cocoa, and hence have different implications for price policy. Various reports indicate the following historical trends in coffee production. Coffee plantings were high between 1940-55. In the 1960's prices were low and plantings declined. Unsanitary harvest practices caused the spread of disease. Smuggling should not affect coffee production to any great degree since there are only 10,000 ha of coffee in the Volta Region of Ghana. Rainfall should be a more important factor in coffee production than cocoa production for two reasons. First, coffee yields are more suscep- tible to low rainfall combined with lack of maintenance. Second, coffee in Togo is planted in some areas that have marginal rainfall for tree crops. Average rainfall in the coffee producing areas is lower than average rainfall in cocoa areas. No satisfactory results were obtained for a historical supply function due to lack of information about acreage planted and age of trees. However, rainfall explained 26% of the variation in coffee production from 1959 to 1976. If damage from drought causes a permanent downward shift in the supply curve, this could explain the decline in production since 1971. An analysis of the economic and financial profitability of coffee was made. There are several factors that make coffee more profitable than cocoa. First, the lag between planting and production is much shorter: 3-4 years instead of 7-8. Second, old coffee areas that could potentially be replanted have been abandoned and no longer produce. There will be no oppor- tunity cost of foregone production to either the producer or the economy. Finally, the projected world prices for coffee are much higher than those for cocoa. - 40 - ANNEX B Page 2 Economic profitability was calculated, valuing labor at 300 CFAF/ manday and materials at market prices. An estimate of extension costs per hectare was made using information in the first cocoa/coffee project appraisal. Two different assumptions regarding world price were made: (1) world price equals 213.2 CFAF/kg, an average of the 1974 and 1975 farmgate economic price; (2) world price equals Bank projections in real terms. The results are presented in Table 1. Since the stream of returns for the first assumption changes sign more than once, a consistent IRR is obtainable. Instead present values at various discount rates are presented. Coffee production has a very good return, especially for Bank projected prices. Coffee is a more labor-intensive crop than cocoa. During the years when the tree is bearing, cocoa requires an average of 125 mandays/ha of labor while coffee requires 193 mandays/ha. As a result, the returns per manday for coffee are lower than for cocoa (590 CFAF vs. 844 CFAF) at 1977/78 pro- ducer prices. For maintenance activities alone, the difference is even more striking. Cocoa requires 48 mandays/ha, while coffee requires 103 mandays/ha. This is due to higher requirements for weeding and pruning. The coffee budgets used in this analysis assume regular pruning which periodically decreases yields but increases lifetime productivity. However, these periodic decreases in yields combined with a low producer price mean that the private producer will have negative returns in some years. Table 2 shows costs, returns and net benefits for the farmer who replants at the 1977/78 producer price. There is some question as to whether farmers will follow recommended pruning practices when they result in periodic losses. Table 3 shows present values of the coffee replanting investment for various producer prices. Costs to the private producer are materials and labor inputs, priced at 300 CFAF/manday. To the extent that some labor may be provided by family labor, the calculations may overestimate labor costs. Replanting appears to be quite profitable for the 1977/78 producer price, 145 CFAF/kg, and a hypothetical producer price of 175 CFAF/kg. At 115 CFAF/kg, the 1975/76 producer price, replanting is only just profitable at 20%. Institutional constraints are not as serious for coffee as for cocoa. In the Dayes area, only 47% of the farms employ sharecroppers as opposed to 85% in the Litime area. Of these sharecroppers, 90% in the Dayes area are also responsible for fooderops where only 60% are in the Litime area. Similar figures are not available for the Akposso area, but the following figures on use of family labor are: Coffee areas: Akposso 65% Dayes 80% Cocoa areas: Litime 55% Kloto 75% - 41 - ANNEX B Page 3 These figures outline a broad pattern. In cocoa areas, labor is hired for harvest activities only. Constraints on available land mean that sharecroppers cannot settle and cultivate foodcrops. Cocoa requires less maintenance in non-harvest periods. In coffee areas owners are more often farmers and operators too. Coffee sharecroppers are not hired just for the harvest but maintain the plantation year-round and also cultivate foodcrops. This leads to conjecture that institutions have developed in response to pro- duction characteristics of the two crops and land availability. Several factors outlined above lead to the conclusion that economic incentives will be more effective in promoting coffee production than they would be for cocoa production. Private producers do not face as long a lag between planting and production. There is no opportunity cost of income foregone from cutting down old coffee trees. Problems of labor availability are not as acute for coffee. Maintenance of coffee, especially when rainfall is deficient, will affect production. Higher prices will raise not only incentives to plant but incentives to maintain existing trees, which will be more important for coffee production than for cocoa. The average nominal protection coefficient for coffee for the 1967-76 period was .45. This is slightly lower than the average NPC for cocoa, .47. Graph A shows coffee NPC's over time. Since the NPC is the ratio of producer price to world price, the lower the NPC, the greater the disincen- tive to production. Incentives to production have tended to decline over this period. The implicit export tax on coffee has increased from .46 to .81. If Bank price projections are correct, the real price of coffee will not fall below 497 CFAF/kg in the next 10 years. Despite the fact that Togo receives only about 80% of average world coffee prices, there should still be considerable latitude for raising producer prices. The above arguments about producer ability to respond to price and the high level of projected future prices argue for an increase in coffee producer prices. - 42 - ANNEX B FIGURE B-1: NOMINAL PROTECTION COEFFICIENTS FOR COFFEE 0.60 F- 2 0.50 A w U. jLL 0 0 I-0.40 0 0r -J 2 0 0.30 0.20 I, I I I I _II/ 1966/6 7 1969/70 1972/73 1975/76 CROP YEAR World Bank - 20186 - 43 - Annex B Table 1: ECONOMIC PROFITABILITY OF COFFEE REPLANTING (CFAF) Discount Price Price = Rate 213.2 CFA/KG World Bank Projections 8 935,514 3,228,083 10 *748,732 2,611,811 12 605,127 2,142,255 14 493,057 1,778,712 16 404,381 1,492,982 18 333,317 1,265,237 20 275,700 1,081,333 22 228,484 931,037 24 189,416 806,845 26 156,803 703,183 - 44 - Annex B Table 2: COST & BENEFIT STREAMS FOR COFFEE REPLANTING AT PRODUCER PRICE = 145 (CFAF) YEAR TOTAL COST TOTAL BENEFIT NET BENEFIT 1 23,100.00 0.00 -23,100.00 2 64,150.00 0.00 -64,150.00 3 34,950.00 0.00 -34,950.00 4 41,500.00 36,250.00 - 5,250.00 5 59,100.00 72,500.00 13,400.00 6 93,100.00 261,000.00 167,900.00 7 76,100.00 145,000.00 68,900.00 8 75,600.00 145,000.00 69,400.00 9 41,900.00 29,000.00 -12,900.00 10 73,500.00 108,750.00 35,250.00 11 86,600.00 217,500.00 130,900.00 12 82,700.00 174,000.00 91,300.00 13 76,100.00 145,000.00 68,900.00 14 41,400.00 29,000.00 -12,400.00 15 75,000.00 108,750.00 33,750.00 16 92,100.00 217,500.00 125,400.00 17 82,700.00 174,000.00 91,300.00 18 76,600.00 145,000.00 68,400.00 19 40,800.00 26,100.00 -14,700.00 20 71,100.00 97,875.00 26,775.00 21 88,500.00 195,750.00 107,250.00 22 78,300.00 156,600.00 78,300.00 23 72,800.00 130,500.00 57,700.00 24 40,700.00 23,200.00 -17,500.00 25 69,200.00 87,000.00 17,800.00 26 82,200.00 174,000.00 91,800.00 - 45 - Annex B Table 3: PRESENT VALUE OF COFFEE REPLANTING UNDER VARIOUS ASSUMPTIONS OF PRODUCER PRICE (CFAF) Discount Price = Price = Price = Rate 115 145 175 8 182,055 385,750 700,089 10 131,469 299,404 549,852 12 92,605 233,133 436,161 14 62,367 181,516 348,698 16 38,571 140,769 280,374 18 19,652 108,204 226,243 20 4,470 81,884 182,799 22 -7,811 60,395 147,519 24 -17,818 42,691 118,564 26 -26,021 27,984 94,570 - 46 - ANNEX C Page 1 T O G O COTTON In terms of government revenue or foreign exchange earnings, cotton is not as important as the perennial crops. However, cotton could be produced in all regions of Togo, and so the potential for significant expansion of cotton production is much greater. Cotton could provide cash income to a large segment of the agricultural population. Until 1965, only the low-yielding mono variety of cotton was pro- duced in Togo. This variety is intercropped with yams. Starting in 1965, single-stand production of allen variety cotton was promoted and mono produc- tion was discouraged. Lower prices were offered for mono cotton and marketing arrangements were less favorable. Fertilizer and insecticide inputs for allen cotton production were subsidized. Production of mono declined rapidly while allen production increased slowly. Then rising foodcrop prices start- ing in 1975 and drought in 1976/77 caused allen cotton production to decline. Supply functions were estimated for. allen cotton production in order to test the responsiveness of prdduction to producer prices and prices of competing foodcrops. Then, to evaluate the effects of price distortions on both inputs and outputs, effective protection coefficients were calculated for various techniques of cotton production in various regions. Supply Response Data was available on allen cotton production by region for 1968/69 to 1977/78. Since production systems differ between regions, supply functions were estimated separately for the Plateaux and Centrale Regions. These two regions account for more than 95% of all cotton production during this period. Each function was estimated using rainfall and foodcrop price data specific to the region. A time trend was used to capture the upward trend in production due to promotion and adoption of new varieties. Equations were estimated with the cotton price and foodcrop prices specified separately. The results were significant and the coefficients had the correct signs. However, there is a high degree of collinearity between cotton prices and foodcrop prices. The correlation coefficient between the two variables was .99 for the Plateaux Region and .93 for the Centrale Region. This makes it difficult to distinguish the separate effects of each price variable. The reported equations use foodcrop prices only or relative prices in order to avoid collinearity problems. Despite the limited number of observations, significant coefficients were obtained for relative prices in both regions and for rainfall in the Centrale Region. - 47 - ANNEX C Page 2 Plateaux Region Double cropping is possible in the Plateaux Region and cotton is usually planted in the second season. Production (crop season 68/69-77/78) is a function of time, cotton price (67/68-76/77), and a weighted average of indices of maize, yam, and manioc price movements (Atakpame, August to July 68/69-77/78)*. 1/ A log-linear functional form produces the best results, and gives the elasticities directly. Equation No. 1: Foodcrop Prices Q = 6.26 + 1.60 T - 1.47 F significance levels * 99% 99% Adjusted R2 = .88 Equation No. 2: Cotton Price Deflated by Foodcrop Prices Q = 1.29 + 1.58 T + 2.17C F significance levels * 99% 99% Adjusted R2= .89 Q = log of quantity (expected) F = log of food prices T = log of time C = log of cotton price F food Centrale Region There is a single crop season in the Centrale Region. Cotton is planted in rotation with cereal crops. Production is a function of time, cotton price (67/68-76/77), a weighted average of indices of sorghum and yarm price movements (Sokode 1968-1977) and rainfall (Sokode 1968-1977). A linear functional form produced the best results. Equation No. 1: Foodcrop prices Q = 2400.9 + 361.2T - 722.2F + 1.8R Significance levels 95% 99% 97.5% 99% adjusted R2 = .81 * Not significant. 1/ Rainfall (Atakpame) was not found to be a significant variable. - 48 - ANNEX C Page 3 Equation No. 2; Cotton price deflated by foodcrop prices C Q = 8119.9 + 508.OT + 110.0 F + 2.1R Significance levels 95% 99% 99% 99% 2 adjusted R .88 Q = quantity (expected) T = time F = foodcrop prices R = rainfall C = cotton prices deflated by foodcrop prices F Elasticity computed at mean: E =3.4 c F E = -1.1 EF The estimated price elasticities are high and may overestimate price responsiveness. Since the equations are estimated using production rather than acreage data, the foodcrop price variable captures some of the effects of rainfall on cotton yields. As longer time series on acreage planted becomes available, it will be possible to separate acreage response to price from variations in yields due to rainfall. While the estimated elasticities may be high, it seems safe to conclude that supply response of cotton production in Togo is elastic with respect to relative cotton/foodcrop prices. 1/ Producer price response is very sensitive to the price of competing activities. 1/ Estimates of cotton supply elasticities mentioned in "Price Prospects for Major Primary Commodities" are as follows: United States, 1.97; Greece, 1.00; aggregate for Mexico, Guatemala, El Salvador, Nicaragua, and Colombia, 1.55; Pakistan, 1.65; Turkey, 0.21; and Ivory Coast, 3.69. - 49 - ANNEX C Page 4 Price responsiveness to competing crops is further illustrated by an analysis of farm budget data. Differences between the two regions in returns per manday may explain differences in the rate of adoption. Less labor is required in the Plateaux Region than in the Centrale Region due to the differ- ent cropping patterns. Average Returns per Manday 1968-72 1/ Cotton (Stage 1) Maize Sorghum Plateaux 210 275 - Centrale 162 285 264 Returns to cotton production in the Plateaux Region were more competitive with returns from cereal production than those in the Centrale Region. Graph 1 shows that adoption of Allen cotton was slower in the Centrale Region than in the Plateaux Region. Cotton production in the Centrale Region began to in- crease significantly only after cotton price increases began in 1973. A priori, differences in supply response between farmers producing at different levels of technology would be expected since they face different cost curves. When food price increases occurred in 1975-1978 the farmers who abandoned cotton production were primarily those in Stage I, the least remunerative production technique. Financial Returns per Manday, Plateaux Region Price = 60 Stage I 280 Stage II 398 Stage III 458 Returns per manday for 'Stage III are competitive with the high returns to foodcrop production in recent years. 1/ Returns have been calculated assuming constant yields. In reality, as prices go up due to drought, yields may fall. Therefore an average was taken over this period to approximate long-run expected returns. - 50- ANNEX C FIGURE C-i: COTTON PRODUCTION BY REGION 6,500 5,500 4,500~~~~~~~~~~~~~~~~~~~~~~~~~~ z~~~~~~~~~~~~~~~~~~~~~ 0~~~~~~~~~~~~~~~~~~~~ Plateaux Region S LU ~ ~ ~ ~ ~ ~~~~~~~~* 4,500 : . Plateaux Regionentale Regio O 3500*S5.IS 196869 971/2 174/7 197/7 o 5~~~~~~~~~RPYA Worl Bak - 018 -51- ANNEX C Page 6 Structure of Incentives Cotton is marketed by OPAT. In evaluating the price distortions for cotton production, it is necessary to find economic values for both lint and seed. F.O.B. Lome prices for cotton lint and cotton seed were used to calcu- late an economic farmgate price for seed cotton. Nominal protection coeffi- cients were then calculated for 1972-77. The results are below, and a sample calculation of the economic farmgate price is in Table 1. F.O.B. Lome F.O.B. Lome Farmgate Producer Lint Prices Seed Prices Value Prices NPC 1/ 72 127 16 43.9 35 .80 73 203 16 74.3 35 .47 74 307 19 117.5 37 .31 75 181 24 69.9 46 .66 76 322 33 131.2 48 .37 77 310 24 112.9 50 .44 Average: .51 The NPC is always less than one, indicating that incentives to produce have been less than if free trade prices had prevailed. The world price of cotton and hence the rate of disincentive to production have varied greatly over this period. Producer Price - Farmgate Price - 52 - ANNEX C Page 7 TABLE 1 COTTON ECONOMIC PRICE CALCULATION SAMPLE -- 1977 Lint Seed F.O.B. Lome 310 24 Transport 6.6 6.6 Value at Factory 303.4 17.4 Ginning Costs 23.7 279.7 x .40 = 111.8 17.4 x .55 = 9.6 Lint Value 111.8 Seed Value 9.6 Seed Cotton Value Pre-ginning 121.4 Transport 8.5 Economic Farmgate Value 112.9 Transportation and ginning cost estimates taken from 1977/78 Bareme. For cotton production in Togo, there are price distortions for both outputs and inputs. Fertilizer and insecticides are sold to farmers at prices below cost. It is argued that these subsidies are needed to overcome farmer resistance to modern techniques of production. Since the subsidized inputs do require less cash outlay, the risk to the farmer of adopting the new tech- niques may be lessened. The question arises of whether subsidized inputs offset the producer price distortion. To evaluate the net effect of price distortions on incentives, effective rates of protection were calculated for each technique in the Plateaux and Centrale Regions. 1/ The EPC's vary across regions and techniques because yields vary and therefore the cost of inputs per unit of output varies. The results of the calculations, based on 1977 producer price and economic price, are presented below. Effective Protection Coefficients Technique Plateaux Region Centrale Region I .41 .41 II .45 .45 III .46 .46 The net effect of the price distortions is to discourage cotton production relative to the situation that would obtain with no price distortions. 1/ The effective rate of protection is the ratio of the difference between producer price and cost of subsidized inputs over the difference between economic farmgate value and economic costs of inputs. It measures the degree of protection on value added. - 53 - ANNEX C Page 8 It is of interest to compare EPC's for the hypothetical cases of no subsidies and/or increases in producer price. Plateaux Region Centrale Region Price = 50 Price = 80 Price = 50 Price = 80 Technique W/Subsidies WI/o sub. W/sub. W/o sub. W/sub. W/o sub. W/sub. W/sub. I .41 .29 .75 .63 .41 .29 .75 .63 II .45 .28 .79 .62 .45 .21 .83 .59 III .46 .29 .79 .63 .46 .26 .81 .61 These figures indicate that subsidies affect not only overall profitability of production but the relative profitability of different techniques. With subsidization, the relative disincentive is less with increasing sophistica- tion of production technique. Without subsidization, relative disincentive is about the same for alL techniques in the Plateaux Region. In the Centrale Region, relative disincentives increase with increasing sophistication. These figures support the current government rationale for input subsidies, which is to encourage adoption of more modern techniques. The question remains of whether the techniques being promoted are socially profitable. Domestic resource coefficients were calculated for all techniques in both regions. 1/ It is assumed that there are no capital inputs, the shadow price of land is zero, and the shadow price of labor is 300 CFA/manday. The results are below. 1977 Price Domestic Resource Coefficients Plateaux Region Centrale Region Economic Value = 113 CFAF/kg I .57 .74 II .41 .68 III .37 .50 The smaller the DRC, the greater the foreign exchange earned per unit of domestic resource. Cotton production is economically profitable overall. The more sophisticated techniques are relatively more profitable. This is to be expected since the more advanced techniques use less labor per unit of output. Production in the Plateaux Region is relatively more profitable than in the Centrale Region. This is because of the greater rainfall in the Plateaux Region, which allows double-cropping and hence fewer labor inputs. However, cotton production is economically profitable in both regions. 1/ DRC is the ratio of the costs of domestic inputs at shadow prices to value added in world prices, converted at the official exchange rate. If the DRC < 1, the activity is socially profitable. - 54 - ANNEX C Page 9 Since cotton prices have fluctuated so widely, it is of interest to test the sensitivity of economic profitability to a 20% decline in economic value. Domestic Resource Coefficients Plateaux Region Centrale Region Economic Value = 90 CFAF/kg I .77 1.00 II .67 .95 III .50 .68 Bank projections for both cotton seed and cotton lint show fairly stable prices at about 1977 levels. It seems that cotton production should remain profitable through the medium term. Sensitivity tests to A 20% increase in labor costs showed that cotton production remains profitable. Policy Implications It is useful to draw implications for policy from the above analysis in terms of the objectives of maximizing government revenue and maximizing foreign exchange earned. There will be some trade-off between these two objec- tives to the extent that a producer price below world price will reduce pro- duction. For further elaboration of this argument, see Annex A. The supply analysis above indicated that cotton supply response is elastic. This result coupled with a producer price that is on average only half of the world price will mean a substantial loss of earned foreign exchange. For example, taking only production in the Plateaux Region, and the estimated elasticity of 2.17, the loss of foreign exchange due to the price distortion can be calculated: Plateaux Region: 1977 World Price- Production at World Price Loss of Gov't. Producer Price -Actual Production Foreign Exchange Revenues (CFAF/kg) (tons) (thousand CFAF) 62.9 5413 170,239 196,814 The loss of society is nearly as great as government revenues. This loss may be exaggerated if the estimated elasticity, 2.17, is higher than actual response would have been. However, this calculation demonstrates that when the supply curve is elastic, a large price distortion will cause large losses of foreign exchange. - 55 - ANNEX C Page 10 In terms of the objective of maximizing government revenues, the existing producer price may be too low. The revenue maximizing producer price, Pt, will depend oni the world price, Pw, and the elasticity of supply with respect to producer price, Ept. Pt =P Ept l-Ept The higher the supply elasticity the greater the optimal producer price from a revenue maximization viewpoint. (For a mathematical derivation of the above formula, see Annex A.) In order to simulate optimal prices for the 1972-77 period, Pw is taken to be the economic farmgate value minus the cost of subsidies per unit of output for technique I. Revenue Maximizing Actual Producer Price Producer Price 72 23 35 73 43 35 74 73 37 75 40 46 76 82 48 77 70 50 78 - 60 Actual prices have tendecl to be too low to maximize government revenue, if we take the Plateaux Region as representative. Estimated price responsiveness in the Centrale Region has been even higher than in the Plateaux Region, which should raise the aggregate elasticity for the country as a whole. The higher price responsiveness is, the higher the optimal price from point of view of government revenue. World prices for lint and seed as forecast by the Bank should be steady at about the 1977 level in real terms. This should leave enough latitude for producer price increases. If cotton prices are increased, providing higher incentives relative to foodcrop prices,-then cotton production should increase more than propor- tionally to the price increase, up to the revenue maximizing price. However, the occurrence of drought will qualify this rule. In years of disastrous drought, the perceived risk of obtaining food may be too high that no cotton price will maintain production levels. In such situations, foodcrop production may be the best earner of foreign exchange by reducing the need for food im- ports. This implies that in setting competitive cotton prices, OPAT must try to distinguish between longer-term trends in foodcrop prices and those which reflect drought conditions in a single year. - 56 - ANNEX C Page 11 Cotton prices are set below world prices while foodcrop prices are determined by market forces. Thus it would seem that cotton production is discouraged relative to foodcrop production. However, this is partially offset by the provision of extension services and subsidized inputs to cotton producers. If incentives to cotton producers are increased, resources may be pulled away from foodcrop production. This might not be desirable from a social point of view if cotton producers are better off than foodcrop pro- ducers. However if the welfare of fooderop producers and growth of food supplies is to increase, it will be necessary to provide foodcrop producers with better techniques of production. The existence of different production systems in the different regions of the country means that costs and returns will differ by region even though production is economically profitable in all regions. If the promotion of regional equality is one of the government's goals, then prices should be high enough to be competitive with fooderop production in all regions. Otherwise only the most productive area, Plateaux Region, will benefit from cash crop production. Along with promotion of the new cotton variety, modern techniques of production are also being promoted. Subsidies are provided for fertilizer and insecticide. The above analysis of incentives showed that this did in- crease the relative profitability of the more advanced techniques. It has been argued in another Bank analysis that these subsidies reduce the risk to the farmer of adopting new techniques. The policy of providing subsidies seems reasonable at the present time, but should be reviewed periodically as the use of modern inputs becomes more common or extends to foodcrops. In terms of either economic efficiency, government revenue or pro- motion of the new cotton variety, higher prices than post historical levels seem to be called for. World price prospects should allow the government enough latitude to make increases in the producer price. - 58 - TABLE 1-A SUPPLY FUNCTION DATA COCOA Crop Median pT T G Year Acreage 1/ Age 2/ pG 3/ PFT 4/ PFG / QT 6/ QV 7/ 59/60 34,119 18.56 2.11 100 202 8,889 23,000 60/61 34,881 19.59 2.11 95 202 12,616 31,000 61/62 35,234 20.42 1.58 65 202 11,460 29,000 62/63 35,587 21.25 1.59 62.5 202- 10,903 21,000 63/64 35,940 22.07 2.00 70 202 13,834 28,000 64/65 36,293 22.90 2.55 90.9 202 17,587 27,000 65/66 36,646 23.73 2.11 30.5 149 14,807 20,000 66/67 36,999 24.56 2.68 63.2 179 16,317 19,000 67/68 37,352 25.38 2.30 95.9 243 18,337 24,000 68/69 37,705 26.21 2.58 68.4 229 19,979 14,000 69/70 38,058 27.04 2.41 73.3 311 23,188 21,000 70/71 38,411 27.87 2.54 92.1 305 27,878 15,000 71/72 38,511 28.82 2.84 58.1 210 29,361 10,000 72/73 38,611 29.77 2.23 71.5 236 18,604 22,000 1/ Source: SRCC data 2/ Source: SRCC data 3/ PT = Togo Producer Price, Source: OPAT; PG = Ghana Producer Price, Source: Gill & Duffus 4/ PT as above; PFT = index of food crop prices in Palime, Source: Togo Statistics 5/ PG as above; PFG = index of food crop prices in Hohoe, Source: Ghana Statistics 6/ QT = cocoa production in Togo, Source: OPAT 7/ QV = cocoa production in Volta Region, Source: Ghana Agricultural Sector Review - 59 - TABLE 1-B SUPPLY FUNCTION DATA COCOA: SUPPLEMENTARY DATA Crop Ghana Togo Year CPI 1/ CPI 2/ P 3/ P 4/ 59/60FT 1 FG 59/60 1.04 1.00 1.00 1.00 60/61 1.10 1.00 1.00 1.00 61/62 1.20 1.00 1.00 1.00 62/63 1.26 1.00 1.00 1.00 63/64 1.41 1.00 1.00 1.00 64/65 1.81 1.00 .77 1.00 65/66 1.90 1.00 1.31 1.00 66/67 1.78 1.035 .87 .84 67/68 1.96 1.011 .73 1.00 68/69 2.07 1.014 1.17 1.14 69170 2.14 1.075 1.20 .96 70/71 2.25 1.125 1.01 .98 71/72 2.70 1.196 1.60 1.43 72/73 2.82 1.289 1.30 1.59 I/ Source: IMF Statistics, 1958 1.00 2/ Source: Togo Statistics, African CPI Lom6, 1958 = 1.00 (Prices assumed constant 1958 to 1965.) 3/ Source: Togo Statistics, Palime food crop prices: maize, manioc, and yams. (Prices assumed constant 1959/60 to 1963/64.) 4/ Source: Ghana Statistics, Hohoe food crop prices: maize and yams. (Prices assumed constant 1959/60 to 1965/66.) - 60 - TABLE 2-A SUPPLY FUNCTION DATA COTTON: CENTRALE REGION Cotton Food Crop Crop Production Price Price Rainfall Year (tons) 1/ (CFA/kg) 2/ Index 3/ (mm/yr) 4/ 68/69 498 32 .70 1,559.8 69/70 685 35 1.01 1,874.2 70/71 492 35 .99 1,439.6 71/72 488 35 .97 1,169.5 72/73 583 35 1.15 1,262.4 73/74 840 35 1.21 1,333.6 74/75 1,788 37 1.33 1,457.4 75/76 1,907 46 1.66 1,389.8 76/77 1,797 48 2.23 1,302.7 77/78 808 50 3.11 1,122.4 1/ Source: Supervision Report for Cotton Project 2/ Source: OPAT 3/ Source: Togo Statistics; food crop prices in Sokode: millet/sorghum and yams 4/ Source: ASECNA,Sokode annual rainfall - 61 - TABLE 2-B SUPPLY FUNCTION DATA COTTON: PLATEAUX REGION Cotton Food Crop Crop Production Price Price Year (tons) 1/ (CFA/kg) 2/ Index 3/ 68/69 408 32 1.00 69/70 1,353 35 1.22 70/71 3,125 35 1.05 71/72 4,967 35 1.35 72/73 3,580 35 1.37 73/74 5,400 35 1.24 74/75 6,150 37 1.43 75/76 5,000 46 2.74 76/77 3,129 48 3.06 77/78 2,291 50 3.73 1/ Source: Supervision Report for Cotton Project 2/ Source: OPAT 3/ Source: Togo Statistics; food crop prices in Atakpame: maize, yams and manioc. - 62 - TABLE 3-A NOMINAL PROTECTION COEFFICIENTS (NPC) COFFEE (CFA/kg) Marketing and F.O.B. Producer Distribution Lome Year Price Costs Price NPC 1/ 1967 70 20.1 148.7 .54 1968 75 20.4 151.1 .57 1969 75 19.7 156.7 .55 1970 75 19.8 178.6 .47 1971 75 20.2 211.0 .39 1972 75 20.2 216.9 .38 1973 80 20.5 203.5 .44 1974 90 22.7 241.2 .41 1975 105 23.9 231.8 .51 1976 115 25.0 616.2 .19 Average: .45 Source: OPAT Producer Price 1/ Nominal Protection Coefficient = F.O.B. Marketing Lome - and Price Distribution Costs - 63 - TABLE 3-B NOMINAL PROTECTION COEFFICIENTS (NPC) COCOA (CFA/kg) Marketing and F.O.B. 2ND Producer Distribution Lome Year Price Costs Price NPC 1/ 1967 55 13.6 123.94 .50 1968 70 14.5 147.11 .53 1969 80 14.6 202.78 .43 1970 88 15.1 225.40 .42 1971 93 15.6 167.16 .61 1972 93 15.6 151.06 .69 1973 93 15.2 202.73 .50 1974 95 16.4 357.78 .28 1975 115 17.1 299.28 .41 1976 120 17.7 403.32 .31 Average: .47 Source: OPAT Producer Price 1/ Nominal Protection Coefficient = F.O.B. Marketing Lome and Price Distribution Costs - 64 - TABLE 4-A EFFECTIVE PROTECTION COEFFICIENTS AND DOMESTIC RESOURCE COSTS 1/ Cotton: Plateaux Region Technique I Technique II Technique III Inputs/ha. 2/ Producer Costs 8,000 10,250 11,000 Actual Costs 14,500 23,500 26,500 Yields 600 900 1,100 Inputs/kg. Producer Costs 13.33 11.39 10.00 Actual Costs 24.17 26.11 24.09 Producer Price 50 50 50 Economic Farmgate Price 112.9 112.9 112.9 EPC 3/ .41 .44 .45 Labor/ha. 100 110 120 Labor/kg. .17 .12 .11 Labor Costs (300 CFAF/manday) 51 36 33 Economic Price -- Actual Input Costs 88.7 86.8 88.8 DRC 4/ .57 .41 .37 All prices in CFAF. 1/ Above data based on farm budgets in Appraisal of Rural Development Project in Cotton Areas, Togo. 2/ Technique I = 20 liter insecticides. Technique II = as Technique I 4 150 kg. fertilizer. Technique III = as Technique II + 50 kg. urea. Input prices: insecticide = 725 CFAF/liter or 400 CFAF/liter with subsidy; fertilizer and urea = 60 CFAF/kg. or 15 CFAF/kg. with subsidy. 3/ Producer Price - Producer Costs EPC = Economic Price - Actual Input Costs DRC = Labor Costs Economic Price - Actual Input Costs - 65 - TABLE 4-B EFFECTIVE PROTECTION COEFFICIENTS AND DOMESTIC RESOURCE COSTS 1/ Cotton: Centrale Region Technique I Technique II Technique III Inputs/ha. 2/ Producer Costs 8,000 11,000 11,750 Actual Costs 14,500 26,500 29,500 Yields 600 800 1,050 Inputs/kg. Producer Costs 13.33 13.75 11.19 Actual Costs 24.17 33.13 28.10 Producer Price 50 50 50 Economic Farmgate Price 112.9 112.9 112.9 EPC 3/ .41 .45 .46 Labor/ha. 130 140 150 Labor/kg. .22 .18 .14 Labor Costs (300 CFAF/manday) 66 54 42 Economic Price -- Actual Input Costs 88.7 79.8 84.8 DRC 4/ .74 .68 .50 All prices in CFAF. 1/ Above data based on farm budgets in Appraisal of Rural Development Project in Cotton Areas, Togo. 2/ Technique I = 20 liter insecticides. Technique II = as Technique I + 200 kg. fertilizer. Technique III = as Technique II + 50 kg. urea. Inputs prices same as for Plateaux Region. 3/ EPC Producer Price Producer Costs Economic Price - Actual Input Costs 4/ DRC Labor Costs = Economic Price - Actual Input Costs - 66 - TABLE 5-A ECONOMIC PROFITABILITY OF COCOA REPLANTING 1/ (CFAF) Opportunity Extension Material Labor Costs Cost of Land 3/ Year Costs 2/ Costs (300 CFA/manday) No. 1 No. 2 0 9,448 26,000 45,000 45,100 89,950 1 5,178 1,250 56,400 45,100 80.950 2 5,178 2,250 44,400 45,100 72,700 3 5,178 750 36,900 45,100 68,350 4 5,178 1,000 21,600 45,100 63,550 5 5,178 250 14,400 45,100 58,600 6 5,178 2,000 22,500 45,100 53,800 7 5,178 500 27,900 45,100 50,800 8 5,178 1,250 38,700 45,100 47,950 9 5,178 1,000 46,800 45,100 44,950 10 5,178 1,500 46,800 45,100 42,100 11 5,178 1,500 46,800 45,100 39,100 12 5,178 1,500 46,800 45,100 39,100 13 5,178 1,500 46,800 45,100 39,100 14 5,178 1,500 46,800 45,100 39,100 15 5,178 1,500 46,800 45,100 39,100 16 5,178 1,500 46,800 45,100 39,100 17 5,178 1,500 46,800 45,100 39,100 18 5,178 1,500 46,800 45,100 39,100 19 5,178 1,500 46,800 45,100 39,100 20 5,178 1,500 46,800 45,100 39,100 21-25 5,178 1,500 34,800 0 0 26-30 5,178 1,500 32,100 0 0 31-35 5,178 1,500 27,900 0 0 1/ Information taken from SRCC budgets for second cocoa/coffee project. ,/ Planting costs, costs of capsid control and extension worker estimated on a per hectare basis from information in the Appraisal of a Cocoa-Coffee Development Project, Togo. 3/ Opportunity cost of land equal to value of 150 kg. of cocoa minus 5,000 CFAF harvest costs. Value of cocoa is alternatively No. 1: World price = 334 CFA/kg or No. 2: World price = World Bank commodity price projections. - 67 - TABLE 5-A (cont.) Banana Value of New Cocoa 4/ Plantains Year No. 1 No. 2 (15 CFA/kg) 0 0 0 0 1 0 0 45,000 2 0 0 90,000 3 0 0 45,000 4 0 0 0 5 0 0 0 6 66,800 78,400 0 7 167,000 186,000 0 8 233,800 247,100 0 9 283,900 283,050 0 10 300,600 282,600 0 11 300,600 264,600 0 12 300,600 264,600 0 13 300,600 264,600 0 14 300,600 264,600 0 15 300,600 264,600 0 16 283,900 249,900 0 17 283,900 249,900 0 18 283,900 249,900 0 19 283,900 249,900 0 20 283,900 249,900 0 21-25 250,500 220,500 0 26-30 217,100 191,100 0 31-35 167,000 147,000 0 4/ Value of cocoa under two different assuntions of world prices as above. - 68 - TABLE 5-B ECONOMIC PROFITABILITY OF COFFEE REPLANTING 1/ Extension Material Labor Costs Value of Coffee 3/ Year Costs 2/ Costs (300 CFA/day) No. 1 No. 2 0 5,998 12,600 10,500 0 0 1 1,700 3,850 60,300 0 0 2 1,700 4,050 30,900 0 0 3 1,700 3,700 37,800 53,300 114,835 4 1,700 13,500 45,600 106,600 238,995 5 1,700 14,500 78,600 383,760 894,042 6 1,700 14,000 62,100 213,200 515,340 7 1,700 13,500 62,100 213,200 524,660 8 1,700 3,800 38,100 42,640 106,798 9 1,700 13,500 60,000 159,900 407,445 10 1,700 14,000 72,600 383,760 414,435 11 1,700 14,000 68,700 255,840 674,292 12 1,700 14,000 62,100 213,200 561,910 13 1,700 3,300 38,100 42,640 112,382 14 1,700 15,000 60,000 159,900 421,433 15 1,700 13,500 78,600 383,760 1,011,438 16 1,700 14,000 68,700 255,840 674,292 17 1,700 14,500 62,100 213,200 561,910 18 1,700 3,300 37,500 38,376 101,144 19 1,700 13,500 57,600 143,910 379,289 20 1,700 15,000 73,500 287,820 758,579 21 1,700 13,500 64,800 230,256 606,863 22 1,700 14,000 58,800 191,880 505,719 23 1,700 3,800 36,900 34,112 89,906 24 1,700 14,000 55,200 127,920 337,146 25 1,700 13,500 68,700 255,840 674,292 1/ Information taken from SRCC budgets for second cocoa/coffee project. 2/ Planting costs and costs of extension worker estimated on a per hectare basis from information in the Appraisal of a Cocoa-Coffee Development Project, Togo. 3/ Value of coffee calculated under different assumptions of world price: No. 1 World price-= 213.2 CFA/kg; No. 2 World price = World Bank commodity price projections. - 69 - FIGURE SA-1: MAIZE RETAIL MARKET PRICES IN TSEVIE EL 0 .0 4 !-, :*o 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 YEARS So,urce: Direction des Statistiques Agricoles, April 1978 World Bank - 20188 - 70 - FIGURE SA-2: MAIZE MONTHLY RETAIL MARKET PRICES IN ATAKPAME 80 70 60 50 W 40 30 20~~Y 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 YEARS Source: Direction des Statistiques Agricoles, April 1978 World Bank - 20189 - 71 - FIGURE SA-3: MILLET-SORGHUM RETAIL MARKET PRICES IN LAMA-KARA 70 80 , .iWl 60 .u 50 WL 40 0. 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 YEARS Source: Direction des Statistiques Agricoles, April 1978 World Bank -20190 - 72 - FIGURE SA-4: MILLET-SORGHUM MONTHLY RETAIL MARKET PRICES IN DAPANGO cr~~~~~~~~~~~~~ y 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 YEARS Source: Direction des Statistiques Agricoles, April 1978 World Bank -20191 - 73 - FIGURE SA-5: YAMS RETAIL MARKET PRICES IN TSEVIE 83 , , I 70 70 EO- U~40 C-) 10 0 0956 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 YEARS Source: Direction des Statistiques Agricoles, April 1978 World Bank - 20192 - 74 - FIGURE SA-6: YAMS RETAIL PRICES IN ATAKPAME 6C I I 70 60 - 50 LLi 40 U 10 20 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 YEARS Source: Direction des Statistiques Agricoles, April 1978 World Bank - 20193 - 75 - FIGURE SA-7: MANIOC RETAIL PRICES IN TSEVIE 8q) , , I 70) 6) 5D - t- 40 CL) -0- 966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 YEARS Source: Direction des Statistiques Agricoles, April 1978 World Bank - 2b194 - 76 - BIBLIOGRAPHY General: Bateman, Merrill. "Aggregate and Regional Supply functions for Ghanaian Cocoa, 1946-1962'. Journal of Farm Economics, May 1965, Volume 47, No. 2. Purseglove J.W. Tropical Crops: Dicotyledons. Wiley & Sons, N.Y., 1968. Askari, Hossein and J.T. Cummings, Agricultural Supply Response: A Survey of the Econometric Evidence. Praeger Publishers, 1976. Pearson, S.R. and J. Cownie. Commodity Exports and African Economic Development. Lexington Books, 1974. Theil, Henri. Principles of Econometrics. Wiley & Sons, N.Y. Country Sources: Interim Report on Intensive Cocoa Survey, Republic of Ghana Ministry of Cocoa Affairs. Twentieth Annual Report and Accounts, Ghana Cocoa Marketing Board. Togo: Prices Enquete Agricole Annuaire Statistique IBRD 14196 .1*> I J IP P; E . L T _ I I I ( . X | *C i d I . S , , . ( e . b ,~~~~ - - &~~~~~~O \ YVAli w } -^-,-. - - nI SA,,,'NJ NA -. - A 10 0 ¢ \' 1 Kandi 1- \K .'~o, 13 I ~~~~ ~ ~ ~~~~~~ J ) rnrog.u Lonlo Kamo I -.-.' ~~~~~~~~~~~~~Bafilo * Bassori C ENT R ALlhm:\ .N.JR / X~~~~~~~~ou,ssountouL TOGO FoZO O Sotoubouo TRANSPORTATION AND POPULATION DENSITY B) Btt L S Pglo 1975 POPULATION DENSITY BY REGION Po" 8. ~PERSONS PER K,,, '" k8- 0 -20.5 20.5- 40.5P \A A - * 11EMAIN ROAD \ 11 ''Z' t 1fik' 'Sko n - ~ RIVERS - REGIONAL BOUNDARIES INTERNATIONAL BOUNDARIE 24D6,0 Bp 100 Ah KILOMETERS 0. 10 RICENT PAPERS IN THIS SERIES Nc. TITLE OF PAPER AUTHOR 450 Education and Basic Human Needs A. Noor 4'1i Economics of Supplemental Feeding of Malnourished 0. Knudsen Children: Leakages, Costs and Benefits 4'12 Land Tenure Systems and Social Implications M. Cernea of Forestry Development Programs 453 Industrial Prospects and Policies in the B. Balassa Developed Countries 4 54 Labor Migration from Bangladesh to the S. Ali, A. Arif Middle East A.K. Habibullah, R. Islam A. Hossain, W. Mahmud S. Osmani, Q. Rahman A. Siddiqui (consultants) 4,5 Managerial Structures and Practices in M. Schrenk Manufacturing Enterprises: A Yugoslav Case Study 166 Nutritional Consequences of Agricultural Per Pinstrup-Andersen Projects: Conceptual Relationships and (consultant) Assessment Approaches 4:37 Industrial Strategy for Late Starters: R. Gulhati The Experience of Kenya, Tanzania and Zambia U. Sekhar (consultant) 4;8 Comparative Study of the Management and A. Bottrall (consultant) Organization of Irrigation Projects 4;9 Size of Land Holding, Living Standards and P. Visaria Employment in Rural Western India, 1972-73 4130 Some Aspects of Relative Poverty in Sri Lanka, P. Visaria 1969-70 461 Incidence of Poverty and the Characteristics P. Visaria of the Poor in Peninsular Malaysia, 1973 4132 Regional Aspects of Family Planning and 0. Meesook Fertility Behaviour in Indonesia D. Chernichovsky 463 What is a SAM? A Layman's Guide to Social B.B. King Accounting Matrices 454 Structural Adjustmient Policies in Developing B. Balassa Countries 455 Cost-Benefit Evaluation of LDC Industrial G. Pursell Sectors which have Foreign Ow-nership 466 Energy Prices, Substitution, and Optimal R. Martin Borrowing in the Short Run: An Analysis of M. Selowsky Adjustment in Oil Importing Developing Countries *** HD9017 .T62 B68 c.2 Bovet, David. Agricultural pricing in Togo I I HIGHSMITH 424312 PRINTEDIN U.S.A.