Senegal - Economic trends and prospects (Vol. 2 of 4) : The agricultural sector

Report No. 1720a-SE F IL E C The Economic Trends and Prospects of S)enegal (In Four Volumes) Volume II. The Agricultural Sector Decern ber 1979 Western Africa Regional Office FOR OFFICIAL USE ONLY Docume>nt of the World Bank This docugment has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not iotherwise be disclosed without WVorld Bank author-ization.  CURRENCY EQUIVALENTS Currency Unit - CFA Franc US$ 1 - CFAF 245 CFAF 100 - US$0.408 ABBREVIATIONS CPSP Caisse de Perequation et de Stabilisation des Prix ONCAD Office National de Cooperation et d'Assistance pour le Developpement SAED Societe d'Amrenagement et d'Exploitation des Terres du Delta SODEFITEX Societe pour le D'eveloppement des Fibres Textiles SODEVA Societe de Developpement et de Vulgarisation Agricole SOMIVAC Societe pour la Mise en Valeur Agricole de la Casamance STN Societe des Terres Neuves GOVERNMENT OF SENEGAL FISCAL YEAR July 1 - June 30  FOR OFFICIAL USE ONLY THE ECONOMIC TRENDS AND PROSPECTS OF SENEGAL Volume I Summary and Conclusions The Main Report Statistical Annex Volume II The Agricultural Sector Volume III The Industrial Sector Volume IV Human Resources This report is based on an economic mission which visited Senegal in November 1976 composed of Messrs. J. de Leede (Chief), J.C. Brown (para- public enterprises), B. Horton (industry), Ms. K.M. Larrecq (balance of pay- ments), H. Lubell (human resources), J.D. Shilling (economic modeling), D.R. Steeds (agriculture), and M.P. van Dijk (informal sector), and complemented with findings of the Bank's regular sector work and more recent economic missions. This volume was written by Mr. David R. Steeds on the basis of his participation in the basic economic mission in November 1976. It was finalized in March 1978 and updated in 1979 to include new developments. Working Papiers on the balance of payments, public finance, and the informal sector are available in West Africa Programs Department II, Division C. The Para-Public Sector Report No. 1619a-SE can be obtained from the Reports Desk. This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contenst may not otherwise be disclosed without World Bank authorization.   THE ECONOMIC TRENDS AND PROSPECTS OF SENEGAL Volume II. The Agricultural Sector Table of Contents Page No. SUMMARY AND CONCLUSIONS i - iv Chapter I. RECENT TRENDS 1 A. Background 1 B. Government's Sectoral Objectives 1 C. Recent Performance 2 Chapter II. SECTOR STRATEGY 6 Chapter III. SECTOR POLICIES 10 A. Organization of Agricultural Services 10 B. Pricing and Marketing 11 Crop Outputs 11 Crop Inputs 12 Consumer Prices 15 C. Production Systems 18 Rainfed Farming 18 Irrigation 18 Settlement 21 Chapter IV. SECTOR PLANNING AND PROJECTS 22 A. Planning 22 B. Erojects 22 C. Projections 23 List of Text Tables 1. Share of Groundnuts for Oil in Exports 2. Likely Effects of Different Strategies on Sectoral Objectives 3. Fertilizer Quantities, Prices, Subsidies and Private Benefit:Cost Ratios 4. Estimated Foodgrain Demand and Supply 5. Primary Sector Value Added and Rural Primary Sector Income, 1975-90 6. Composition of Primary Sector Value Added 1975 and 1990. Annex A Statistical Tables Annex B Supply Response of Groundnut Farmers in Senegal   SUMMARY AND CONCLUSIONS CObjectives and PerformniLnce i.. Changes in Gcvernment's agricultural sectoral objectives over the last 10-15 years have been more in emphasis than in direction. Performance wtith respect to these objectives has been mixed: some advances have been made toward the socialization of rural society but there has been little progress iLn improving the econoxaic efficiency of the sector; in export diversification, iish exports have expanded substantially and, in the last 10 years, cotton and confectionary groundnuts have been successfully launched; home-grown food production has fluctua:ced widely about a slowly rising trend and production has increasingly failed to keep pace with demand; rural incomes also fluctuated widely and showed no alppreciable increase until 1975; savings from the agricul- tural sector were, on the other hand, significant, being equivalent to an effective tax rate on marketed groundnut production of over 25% in the period 1971-75. Part of the disappointing performance of the rural sector through 1974 can be explained by the low groundnut prices paid to farmers in this period. Objectives and Strategy ii. It would be desirable to change the emphasis among sectoral objectives in favor of increasing rural incomes and promoting rural development. There are good economic grounds for the former, and the latter would help give precision to the stated objective of the socialization of rural society. In this volume, an approach is outlined by which the likely effect:s of possible strategies on sectoral objectives may be made explicit and compared. Using this approach, the most effective strategies to attain the given objectives would be greater empha,sis on domestic food production and on promotion of exports other than grcoundnuts for oil. Policies iii. In the organization of agricultural services, improvements have been made in the delivery of services to basic production units through the regional development agencies but no agreed criteria to assess perform- ance exist and, with one exception, no incentive structures have been devised. It is also important to begin, now, to build up a long-term self-sustaining financial base for these agencies. Nor are there any built-in checks to ensure that the services provided are responsive to local needs; clear definition of the reslponsibilities of the newly-created regional and district deVelopment councils and 'committees would help. The policy of delegating some of the marketing board (ONCAD)'s responsibilities to the regional agencies is sound, provided their performance is regularly assessed. iv. The policy of uniform farm-gate prices for crop outputs involves, contrary to its apparent purpose, a subsidy from producers located in the less favorable climatic zones to those located in the higher and more stable  rainfall areas since the latter are generally further from Dakar, the main market; it also denies any inducement for local delivery and storage, and distorts regional comparative advantage. A practicable improvement on current policy would be one of uniform prices payable at existing main depots. For cereals, fixed prices can even increase farmers' uncertainty and, worse, de-stabilize rural incomes. There is an overwhelming case for de-control of cereal prices and elimination of ONCAD's de jure marketing monopoly. v. Since growth in agricultural output will have to come from a steep increase in fertilizer use, the existing fertilizer subsidy policy has to be carefully reviewed. Once farmers are familiar with the use of fertilizer, a subsidy would be justifiable only if there were continuing discrepancies between national and private profitability of fertilizer use. Where such differences are due to taxes levied on outputs, it would be preferable to reduce the farmers' tax burden rather than subsidize one input. Differences between national and private profitability may also be due to risk which weighs more heavily on the individual than the state; in this case it would be preferable to offset the risk by insurance. It is also argued that fertilizer could prevent soil deterioration but this argument does not apply to the more costly nitrogen and potassium components and must therefore be rejected. The fertilizer subsidy policy is also advocated as a means of redistributing income to the less favorable climatic areas but, because the compounds used in the more favorable areas are costlier, the uniform subsidized price entails a higher subsidy to the latter areas than to the former. If no other means of redistributing income to the less favorable areas can be found, the undue subsidy to the more favorable areas should at least be curtailed; as a first step, a standard percentage subsidy might be adopted. vi. The overall effect on income distribution of the complex of producer and consumer price and subsidy policies was, through 1974, a transfer to urban consumers and the wealthier rural dwellers. Since then the thrust of the policy changes has generally been to reverse this situation but it is difficult to assess the overall effects since urban wage-earners have received wage-increases sufficient to cover their increased cost of living. The critical effect of subsidies on rice and wheat flour is to provide a cheap and convenient substitute for millet which facilitates the substitution, commonly held to be inexorable, of soft for coarse grains. The experience of 1975, on the other hand, illustrates the power of relative price policy and the scope for substitution of millet for rice. New techniques for producing an instant millet couscous and bread from a blend of millet and wheat flour have been developed but not exploited because of the contradictions in recent relative price policy. The potential increased millet production from the central and southern Groundnut Basin will be realized only if a strong pro-millet pricing and marketing policy is pursued. Moreover, in economic analysis, the valuation of millet in terms of US sorghum understates the relative value of millet and, given the new techniques for rendering millet a more convenient foodstuff, it is argued that millet should be valued at 60% of the reference price for 5% broken rice. It is also shown that on the domestic market there are three sharply differentiated rice products and it is argued that domestic paddy should be valued at 70% of the reference price for 5% broken rice.  - iii - vii. Concerning rainfed farming, increasing the productivity of the most limiting factor of production requires, in the outlying regions where land is nct scarce, promotion of labor-saving equipment, and in the Groundnut Basin where land is now scarce, promotion of more intensive land use. The scope for intensification in the northern and central Groundnut Basin is unfortunately limited and marginal returns may already be negative in the north. viii. Assessments of the economic returns to particular irrigation schemes have and will continue to show marginal returns at best; the first: attempt to apply the domestic resource cost method showed negative returns. Current methods of project analysis do not, however, capture the benefits of risk- avoidance achieved by full water control quite apart from the argument that the scale of the investments envisaged in the Senegal Valley is such that, for the region if not for the country, the overall program is non-marginal. The apparent question of rainfed versus irrigated production is, in effect, with the exception of parts of Casamance, a question of promotion of rainfed farming in area A or irrigated farming in area B or both. Despite the con- siderable unexploited potential of rainfed farming outside the northern and central Groundnut Basin, physical land limits will be encountered and diminish- ing returns to intensification will increasingly set in. Whereas development of rainfed farming will probably be relatively profitable, and should be undertaken as first priority, domestic production of foodgrains will fall increasingly short of demand unless major investments in irrigation are also undertaken. The apparent question of development of irrigation by large-scale operations or by family farms is not really relevant in the Senegal Valley because the initial experience there with large-scale operations was disappoint- ing, and the middle valley area that could be developed already has a densely- settled population. In addition, recent design work shows that the gradual development of large perimeters by and for the local people and at a pace determined by local population densities will entail considerable capital cost savings, perhaps as much as 35x. This is the direction in which invest- ments in irrigation should proceed. On the other hand, the pace of irrigation development envisaged by Government heavily overestimates the absorptive capacity of the executing agencies despite the strengthening of those agencies that is underway or planned. ix. Population densities vary from 80/km2 in Thies and Diourbel to less than 10 in Senegal Oriental and in much of Fleuve. Given the limited scope for further agricultural development in the northern and central Groundnut Basin, emigration from this area is already and will continue to be important. Fragmentary evidence points to spontaneous, unassisted rural-rural migration on an appreciable scale into eastern Sine Saloum and, increasingly, into Senegal Oriental. It is becoming apparent that: organized settlement schemes on a scale sufficient to have any impact on these movements will be both expensive and difficult to implement; a better approach would be to devise methods of assistance more appropriate to spontaneous migrants.  - iv - Planninz and Projects x. For most agricultural and particularly crop subsector projects being considered under the Fifth Plan, funds are already committed and, given management and staff constraints, it would be a commendable achievement if all such projects were executed as envisaged. By contrast there appears to be spare absorptive capacity in livestock and forestry. Crop intensifi- cation projects could usefully include a substantial cattle fattening component and actions to promote poultry, pigs and small ruminants should be intensified. Tropical timber could be improved and developed in Lower Casamance and char- coal production could be better organized and its productivity improved. Livestock and forestry are both fields where an improvement in project pre- paration should have a high payoff. In Lower Casamance, however, where funds are already committed to dam construction and design work on additional dams, the proposed solution to the problem of cultivation of mangrove soils has not really been locally tested; until it is, dam construction should be deferred. xi. Primary sector accounts, rural incomes and trends in foreign trade have been projected through 1990. The key result is that rural primary sector income is forecast to increase by 3.4%/year between 1975 and 1990, or only 1.2%/year per rural inhabitant. These rates are considerably higher than those observed over the past 15 years and require, for the crop subsector component, adoption of the policies advocated in Chapter III of this volume.  Chapter I. RECENT TRENDS A. Background 1..01 Despite poor soils and erratic rainfall, and although its con- t:ribution to GDP is considerably lower than in most countries of West Africa, agriculture in the broad sense of crops, livestock, fishing and forestry is a key sector in the SenLegalese economy. In the last five years, the sector accounted for about one-third of GDP and about half of merchandise exports; it employed 70% of the labor force and will continue to be the largest source of employment for the ioreseeable future. 1..02 The agricultural income per head of the rural population was about CFAF 30,700 ($125) in 1.975. 1/ This average conceals differences between regional income levels, between the variability of regional incomaes and be- t:ween the respective contributions of crops, livestock, fishing and forestry t:o regional incomes (AxLnex Table 1.8). 2/ The differences in variability and composition are qui.te striking although they are generally explicable by location with respect to the coast and to rainfall. l;. Government's Sectoral Objectives 1.03 Since independence, the Government's overall strategy has gradually put more emphasis on the agricultural sector. In the early sixties, while t:he sector was regardecd as the economic base required to provide the markets and savings needed to clevelop the other sectors of the economy, there was little public investment in agriculture. In the later sixties, the crucial importance of rural development for the rest of the economy was better recognized. The Government, with the help of foreign donors, started some large-scale rural development projects, but it continued price policies that were generally in favor of the urban sector and therefore success was meager. Only after 1974, when price policies were changed in favor of the farmers, did agricultural production increase substantially. ]..04 Over the past ten years, the Government's agricultural objectives have been to: (a) soci.alize rural society; (b) diversify exports; (c) in- crease domestic production of food crops; (d) generate public savings; and (e) increase rural income. These objectives have been expressed in various official statements except for (d), the role of agriculture in the forma- t:ion of public savings, which has, however, been clearly manifested in I/ Income from the cI'op year 1974/75; for the other subsectors, income per calendar year. 2/ See Annex A.  actual policy decisions. These sectoral objectives are somewhat contradic- tory and over the past decade there have been changes in emphasis in response to changing perceptions of the national interest. The following section dis- cusses recent performance with respect to these objectives, although perfor- mance also has to be assessed against the overall growth of production in the sector. C. Recent Performance 1.05 In the first decade of independence, highest priority was given to the socialization of rural society. Government launched a cooperative move- ment as a means of replacing foreign traders with national institutions and of encouraging participation of the rural people in the development process. These institutions have been remarkably successful in providing access to cheap agricultural credit and in ensuring respectable repayment rates. In many other ways, however, their efficiency has been poor and their limita- tions are now widely recognized. Private ownership of non-urban land has been abolished and this land is now vested in the state. There were never more than a few large private farms so the question of redistribution of substantial land-holdings was not an issue. What was at issue, however, was the control of vacant land and the desirability of changing the complex and numerous forms of small-scale share-cropping and leasing arrangements. It is unclear to what extent the new law has 'brought the land to those who work it'--there are continuing complaints that a major obstacle to de-stumping is the uncertain status of land ownership--but perhaps the new law has not yet been sufficiently put to the test. To promote greater participation in the development process, Government has created in a number of rural district local authorities (communautes rurales) of about 10,000 people, governed by elected councils. These councils will manage their own budgets partly financed from general tax revenues and partly from local taxes and will have an involvement in land allocation. The Government tries to align the dis- tricts of the communautes rurales with those of the farmers' cooperatives. The reform has been implemented in a small number of districts, and will gradually be extended over the entire country. It is too early to tell what effects this reform will have, but the initial indications are encouraging. 1.06 In export diversification, the continuing predominance of ground- nuts for oil conceals three important changes in agricultural exports. Although production of confectionary groundnuts and seed cotton did not begin until the mid-1960s (Annex Table 1.1), exports in 1975 1/ reached about 12,000 tons of confectionary nuts and 12,000 tons of cotton fibre. The initial development of both crops required important innovations in input supply, farming practices and primary marketing. These changes, notably the purchase of confectionary groundnuts at primary markets according to six grades, have now been introduced on a sufficiently wide scale that 1/ I.e., the exportable surplus from output in the 1974/75 crop season.  a sound basis has been ;laid for sustained expansion. Since the mid-1960s, irmpressive increases have also been recorded in fish exports, which reached over 50,000 tons of fresh fish equivalent in 1974, of which about half were frozen fish. The markel: for frozen sardines in the larger West African coastal countries has scarcely begun to be tapped and Senegal is now in a good position to exploit it. In summary, over the last 10 years two new export commodities, conlfectionary groundnuts and cotton, have been success- fully launched and there has been a substantial expansion in fish exports. These are three commodit:ies in which the country, or particular areas of the country, clearly have a comparative advantage and expansion of these exports represents successful exploitation of potentials other than groundnuts for oil; they are complements rather than substitutes for the latter. Nonetheless, groundnuts for oil continue to dominate agricultural exports although some reduction in the share is discernible: Table 1: SHARE OF GROUNDNUTS FOR OIL IN EXPORTS 1964-66 average 1970 1971 1972 1973 1974 -------------- (CFAF billion) ---------------- (1) Exports of groundnuts for oil and associated products 25.7 20.3 12.3 28.7 15.1 33.4 (2) Exports of primary sector products n.a. 25.5 18.2 35.2 22.7 45.6 (1) as % of (2) 80% 68% 82% 67% 73% (3) Total merchandise exports 32.9 42.2 34.7 54.4 43.8 94.0 (2) as % of (3) 60% 52% 65% 52% 49% (1) as % of (3) 78% 48% 35% 53% 34% 36% Source: External Trade Statistics 1.07 With regard tc, domestic production of food crops, there was a regular increase in cereals output from 1961-66 followed by almost a decade of widely fluctuating production about a slowly rising trend. That trend was appreciably lower than the growth of demand for cereals, arLd imports have increased accordingly. For the three-year average 1969-71, imports of cereals reached 420,000 tons gross while domestic production was about 700,000 tons. The situation in 1972-73 was aggra- vated by the drought while in 1974 domestic production and imports were similar to the 1969-71 Everage (Annex Table 2. 3). Data on production of roots and beans, and vegetables and fruits, are even more unreliable than for foodgrains but it seems that, following modest growth 1961-67, the trend has been falling (Annex Table 1.1). Data on livestock are yet more un- certain; the trend in production of meat and offals, and milk, was one  - 4 - of very slow growth (Annex Table 1.3). Despite the difficulty of dis- cerning trends in the last decade, it is clear that domestic production of food crops and livestock products has grown slowly if at all, and that it has increasingly failed to keep pace with demand. The only encouraging sign is in fishing where there has been sustained rapid growth in the landed catch from canoes, from about 85,000 tons in the early 1960s to about 250,000 tons in the mid-1970s (Annex Table 1.5). That the essentially small-scale, private fish marketing system was able to handle such an increase in pro- duction was also an impressive achievement. 1.08 Savings from the agricultural sector, more particularly the crop subsector, have been realized through the operation of the price stabilization funds. Separate commodity funds were consolidated in FY73 into a Caisse de Perequation et de Stabilisation des Prix (CPSP) which now manages the pro- duction accounts for groundnuts, cotton and agricultural inputs and the con- sumption accounts for rice, sugar, groundnut oil and wheat flour. At the time the separate commodity accounts were closed, they had reserves of CFAF 12 billion (over $50 million) of which only CFAF 0.6 billion was transferred to the CPSP. In the last three years of operation of the separate accounts, the production accounts together netted CFAF 3.3 bil- lion/year or about 10% of the gross value of crop production. The consump- tion accounts showed a loss of CFAF 2.1 billion/year. In FY74, the pro- duction accounts netted about CFAF 14.5 billion or about 30% of the gross value of crop production, but in FY75 these accounts showed a loss of about CFAF 1 billion, the small surplus on the groundnut account being outweighed by the loss on agricultural inputs. The fiscal outcome of FY76 is still unclear, but the loss on the production accounts may be as low as CFAF 5 billion, or less than half of what was anticipated in early 1976. Through FY75, the aggregate surplus on the production accounts for the five-year period from FY71 was CFAF 23 billion or about $100 million. By contrast, the gross value of marketed crop production, estimated from Annex Table 1.1 at 20% of food crops and 80% of industrial crops, was CFAF 117 billion in the same period; the effective tax rate on marketed output was about 20%, net of subsidies on agricultural inputs. After netting out the cancellation of farmers' debts after drought years, the net taxation averaged about 10%. Excluding those subsidies, the surplus on the production accounts was CFAF 33 billion or an effective tax rate of over 25%. 1.09 Rural incomes, as measured by the proxy of value added excluding modern sea fishing and incremental herd growth, increased by about 5% per year through 1966. During the next five years, incomes in current prices were lower than in 1966 and inter-year fluctuations were about 15%. In the following four years, 1972-75, incomes in all cases were greater than in 1966, but inter-year fluctuations reached 30%. These fluctuations were partly due to exceptional weather conditions, but also to the 1968 termina- tion of special groundnut marketing arrangements with France, which rendered the country more vulnerable to price changes in the world market; the Stabili- zation Fund failed to even out these fluctuations. In 1971 constant prices, rural incomes in 1966 were exceeded in only three years of the following nine--1968, 1972, and 1975, i.e., the fundamental reason for the disappoint- ing rural income performance through 1975 was the slow increase in physical  pcoduction. On the oth,er hand, rUral agricultural income per capita in- creased by over 40% between 1974 and 1975, from CFAF 21,400 (US$87) to CFAF 30,700 (US$125) (Aanex Table 1.7). This rise was mainly due to an increase in the groundnut price paid to the farmer, which was initially given to reflect higher world market prices, but which also resulted in the elimination of taxation on groundnuts. This change reversed the pre- viously widening gap between urban and rural incomes, but it did not make up for the ground lost since 1966. 1.10 From the above analysis of performance in respect of particular sectoral objectives, it is clear that slow growth of agricultural production threatens Government objectives and sharpens the contradictions between them. If production had been growing quickly, increases in local food production could have been complementary to developing diversified agricultural exports; new types of exports would not have excluded maximization of the most profit- able export crop, groundnuts; and finally, fast growing agricultural income would have allowed certain increases in taxation of the sector without destroy- ing the incentives for farmers to produce for the market. The slow rate of growth is closely linked to the prices paid to farmers for groundnuts. After 1968, when real groundrut producer prices were substantially reduced, annual groundnut production fell by about 40% over the following five years--after eliminating the weather factor from the series--until prices were increased again in 1973. Amazingly enough, there is no evidence that lower groundnut production led to higher production of other crops. This may be explained since the only other main crop in the Groundnut Basin, millet, is essentially a subsistence crop which, under present conditions, has a strict ceiling on i.ts useful production. (Typically, the price elasticity of supply of sub- sistence crops is very low.) The higher groundnut price paid to farmers since 1974 explains to a large extent the high groundnut production in recent years and confirms that most development efforts will not succeed umless supported by appropriate price policies. In sum, it appears that cgriculture cannot generate large contributions to public savings through taxation on production without destroying prospects for increased rural production and income and, consequently, growth of the domestic market for l-ocal industry.  - 6 - Chapter II. SECTOR STRATEGY 2.01 With partial success in the achievement of two objectives (sociali- zation of the rural sector and export diversification) and failures in the other three (diminished reliance on food imports, increased rural incomes and public savings) the Government will have to rethink the emphasis it wants to give to each of these aims. To arrive at a successful strategy, however, some additional objectives have to be taken into account. For political reasons the goal of increased rural income has to be combined with a policy of regional balance in the development effort. Higher rural income should also be combined with efforts to reduce the risk of starvation that threatens the Senegalese rural population during years of drought. On the economic side, the balance of payments prospects have deteriorated to an extent that export promotion should be given priority over diversification of exports. 2.02 The optimum strategy to realize all these sometimes contradictory goals cannot be determined objectively. One would need to know how much weight the Government attaches to each of these objectives and detailed data on the effects of different policy packages on realizing them, a knowledge that simply does not exist. It is nonetheless possible to make some systematic judgments on the effects of alternative strategies. In Text Table 2, the judgments of the mission are summarized. None of the strategies indicated entirely excludes the others. The table indicates only how different emphases in the total strategy mix would affect the respective objectives. It shows that the best results would be achieved through promotion of exports other than groundnuts and through domestic food production (columns (i) and (ii)). The only negative effect connected with the priority for developing (i) exports other than groundnuts is in the limitation in scope and geographical location of the new export acti- vities which have so far been developed. Promotion of (ii) rainfed food production could have a negative impact on the production of groundnuts because of the competition it creates for land. If the increased rainfed production has to be achieved by giving higher price incentives to the farmer; it would also diminish exports of urban industries; the higher food prices would increase the urban cost of living and most likely drive up Senegal's industrial production costs and reduce their internatinal competitiveness. However, appropriate policies can limit these negative effects. The strategies with the heaviest impact on the objectives are promotion of groundnuts and irrigation. Promotion of (iii) groundnut exports would certainly help to alleviate the difficult balance of payments situation and increase the average income of the majority of the farmers. Because Senegal has such strong comparative advantage for groundnut production this strategy may in the long run also be most beneficial for public savings. The most serious drawback of this strategy would be the increase in risks connected with droughts and fluctuating prices of groundnuts on the world market.  Table 2 LIKELY EFFECTS OF DIFFERENT STRATEGIFS ON SECTORAL OBJECTIVES Sectoral objectives Sectoral Strategies (i (ii) (iii) (iv) (v) promote promote promote promote promote alternative rainfed groundnut irrigated migration exports food p t f^od t_ .cgions where regions production where for oil production with comparative have a regions have a advantages comparative comparative advantage advantage Social Economic ObJectives 1. Socialize rural society = = = + + 2. Increase rural incomes + + + + + 3. Equalize incomes par region - + + + -4 4. Reduce risks + + -- ++ + Macro-Economic Objectives 5. Increase of exports + ++ = 6. Diversify exports + = - 7. Food import substitution ++ _ ++ 8. Generate public savings = = + Key: + positive effect on objective no effect on objective - negative effect on objective.  High incentives for groundnut production will probably also lead to some decline in the production of cereals and make the farmer and the country as a whole more dependent on food imports. This also implies a risk due to fluctuating world market prices for wheat and rice. The trade-off between the risks connected with higher groundnut production and the likely positive impact of this strategy on farmers income, balance of payments and public finance is one of the most difficult decisions to be made by the Government. Promotion of (iv) irrigated crops is also limited to certain areas and has to be part of a wider strategy covering the other regions. Since this strategy is very costly in terms of capital and marginal in pro- fitability, the impact on public savings (after debt service) is likely to be negative. On the other hand, this strategy would require a highly organized approach which would allow for a quick modernization of the rural communities. It would also have beneficial effects on the balance of payments. A final alternative strategy is based on (v) migration conceived in a highly organized way. This strategy is likely to have limited impact on the overall level of production and would hardly make a dent in regional income imbalances. Moreover, it could easily become a drain on Government savings because of the cost of incentives for the migrants. A migration strategy based on spontaneous movements would have a negative impact on socialization of the rural sector, but would be less costly for the budget. 2.03 The above analysis should be completed by a discussion of the relative weight the Government gives to its different objectives. The Govern- ment has been quite successful in socializing the agricultural sector. It has almost entirely eliminated foreign commercial interests and now reaches the individual farmer with its extension service and credit programs. The time may have come for a consolidation in this area and for more emphasis to be given to considerations of economic efficiency. Some special atten- tion could be focussed on the credit policy. The private savings capacity of the farmer is probably being underestimated, in particular since the farmer has been made aware of the advantages of modern inputs and invest- ments and since secure profitable outlets for his crops have been organized. There is a growing body of opinion which suggests from experience elsewhere that many credit projects are successful not because of the credit, but because of the technical assistance that goes with the credit. Extension services have now penetrated so deeply into the rural economy that it is also time to make farmers more critical of the services offered by making them share the costs. For instance, well-established farmers cooperatives should be able to execute more business themselves outside the centralized bureaucracies, particularly as far as supplies of inputs are concerned. 2.04 On the other hand, because of the precarious balance of payments prospects, more weight should be given to all objectives which will lead to higher exports or import substitution. In most cases this can be done by providing the right price incentives. The objective of balance of payments improvement, therefore, parallels to a large extent the Government's objective to increase rural income. There are three arguments why more emphasis on increasing rural incomes may be beneficial to the economy as a whole. The argument that agriculture is the dominant sector and that it should, therefore,  - 9- be taxed to provide revenues for capital and recurrent expenditures in other sectors no longer applies, since agriculture now constitutes only 25-30% of GDP. On the contrary, the Government should have an explicit policy that the urban sector, where incomes are much higher, is not sub- Esidized by the rural sector and that rural savings are not used for urban investments. Second, the elasticity of crop supply is probably greater than is often assumed ]./ so that high producer prices could entail in- creased rural incomes without a proportionate loss in Government revenues. There is clear evidence in Senegal that farmers react to small differences in the relative prices for cash crops, but it is not known to what extent higher groundnut prices will reduce millet production. Because of the vast possibilities that exist for simultaneous production increases in millet and groundnuts through use of fertilizers, a fall in millet production could be prevented with appropriate policies. Third, the multiplier effect of rural incomes is higher than for urban incomes since the import component in consumption and investment in rural areas is less than that in urban areas. A high income multiplier implies that one unit of additional income in the rural areas will. generate more demand for locally-produced goods than in urban areas and consequently will generate more local income. There are, however, no family budget data in Senegal to quantify this difference. 2.05 Until recently, there has been little discussion in Senegal of rural development as a sectoral objective. Today, however, it is being increasingly invoked as an agricultural or primary sector objective. While increased rural production and income is a necessary condition for rural cdevelopment, it is not sufficient. Rural development includes both production and social services and. its advocacy as an objective implies giving some weight to welfare considerations in addition to production. Senegal is not, however, among the more fortunate countries where resources are sufficiently abundant that a considerable sacrifice of either production or welfare can be avoided. The discrepancy between urban and rural areas in the provision of social services and amenities is striking (see Volume IV on Human Resources). But since it is not possible to make substantial improvements in both areas simultanetously, the critical choices would appear to be in the allocation of such services between urban and rural areas. This inter-sectoral problem is pursued in Volume IV. Within the rural sector, hLowever, rural developrment is now being advanced as an objective in the s;ense of the creation cf opportunities for participation of the rural people both in government and in the provision of amenities and services. T'his might help give a new meaning to the stated objective of the socializa- tion of rural society. '/ In Annex B the supply elasticity of groundnuts to prices paid to the farmer is estimated at 0.5.  - 10 - Chapter III. SECTOR POLICIES 3.01 Despite the ambiguity concerning sector objectives and the cor- responding difficulty of defining sector strategy, progress can be made in clarifying policy questions since in many cases they are common to most if not all possible strategies. A. Organization of Agricultural Services 3.02 Given that the basic production unit will remain the small family farm, herd or boat, the key questions are: how to ensure effective delivery of services to those units, and how to ensure that the services provided are responsible to local needs. After Independence, the Government opted for the cooperative approach and created a structure with trading monopolies for inputs and products that made membership in a cooperative almost compulsory. The advantage of this approach was that the cooperative movement could be set up as a large-scale operation with deep penetration into the economy. The system had the disadvantage of being considered by farmers as an imposed structure. About 2,000 cooperatives were formed, often however far below the size required for operational efficiency. Recently the Government has started to consolidate smaller cooperatives into larger units, which will further increase the gap between the farmers and the central administration of the cooperatives. 3.03 The cooperative marketing board (ONCAD) has a de jure monopoly on trade in inputs and products and provides individual farmer cooperatives with services (like transport) and technical assistance. For several years its performance has been criticized as inefficient, even allowing for the severe constraints under which it operates. These constraints include lack of financial autonomy, unsuitable administrative procedures and controls over day-to-day financial management, unsatisfactory hiring methods, and staffing problems arising from the Government's insistence on civil service employment regulations. Efforts to streamline this huge bureaucracy in the early seven- ties failed, and the Government has now decided to decentralize the organi- zation. The plan is to decentralize it into regional bureaus, but it is doubtful whether this will bring about the necessary change. More important is the test being made to delegate ONCAD's responsibilities for extension services and supply of inputs and implements to regional development agencies and to allow cooperatives with a sizeable groundnut turnover to provide for their own transport arrangements. ONCAD should be encouraged to further reduce its marketing monopoly; this would put its own efficiency to the test on a permanent basis. 3.04 Regional development agencies have been created by the Central Government to execute crop and livestock development programs. SAED and SODEVA were created in the mid-1960s and, more recently, SODEFITEX, STN and SOMIVAC. The Government intends to have BUD-Senegal take on this role for the coastal area. In this case all of agriculture would be  - 11 - covered by the regional organizations. These agencies have varying degrees of financial autonomy, administrative flexibility, freedom in recruitment, scope to devise inducenents for their staff, and are reputedly fairly effective in delivery of services. But in the absence of agreed criteria to assess their perfornance and of incentive structures to promote better iperformance, their reputation for effective delivery should not be assumed permanent. Like government departments, the regional agencies are centralized decision-making bodies and there are no built-in checks to ensure that the services they provide are responsive to local needs. Nor will the recent reform in local government necessarily provide such checks. This reform includes the creation of regional and district development committees (tech- aicians) and councils (political officials and elected representatives), 'but the responsibilities of these authorities, and especially the power of the political bodies over the technicians, have not yet been clearly defined. Some power of authority by the elected bodies over the local work programs of the regional agencies will be needed to make these agencies responsive to local needs. 3.05 Criteria for the assessment of performance of the regional agencies are difficult to devise but it is now important to make that effort and to develop incentive schemes. One such scheme has been instituted at SODEFITEX whereby part of the agency's remuneration is based on the volume of cotton production in excess of an agreed minimum. 3.06 Despite the lack and difficulty of performance assessment, the regional agencies have been fairly successful in attracting foreign, soft- term finance. They have not, however, with the possible exception of SODEFITEX, even begun to build up a long-term self-sustaining financial base. Three main possibilities are open to them to do so and these should be explored with some urgency: the Government could, on the basis of a fee, incorporate an incentive element; the 2,000 cooperatives could be given some participation in the supervision of the regional development agencies; and costs could be recovered by a levy on cash crops through ONCAD. 3.07 The regional agencies have also been fairly successful in under- taking applied agricultural research, designed to test themes to be adopted by the extension services. Such a decentralized approach to applied research is advisable in that the users are fully involved in the tests and a convenient division of labor is achieved whereby the agricultural research institute concentrates on basic research. B. Pricing and Marketing Crop Outputs 3.08 Prices of most crop inputs and outputs are controlled by the Govern- ment. Output prices are reviewed and announced annually towards the end of the calendar year but several months after planting. Estimates of production cost and supply response are considered only if undesired shifts in production  - 12 - occur. The main factors taken into account in price reviews seem to be, for groundnuts and cotton, likely export prices, Government revenue needs and, almost as a residual, rural incomes. For millet and sorghum, maize and paddy, where ONCAD has a de iure marketing monopoly, official prices are not of great relevance since the marketing board purchases are small as compared to total production and do not cover the full production surplus above subsistence consumption needs. Prices in the existing parallel markets often differ substantially from the official prices. 3.09 For all crops for which official prices are fixed ONCAD, and SODEFITEX for cotton, are required to buy at a uniform price regardless of location and of seasonal variations in supply. It is difficult to understand the apparent appeal of a uniform national price. It hardly results in equal pay for equal work since the implied subsidies on transport costs go in fact to regions with higher and more stable rainfall. Moreover, a uniform price encourages production in areas lacking regional comparative advantage as, for instance, the present substitution of groundnuts for unirrigated paddy in Casamance and cotton in Senegal-Oriental. An acceptable alternative to the uniform national price might be a decrease in the number of groundnut depots in regions with comparative advantages for other crops. Since the farmer has to pay for the transport between farm and depot, the transport subsidy would be accordingly diminished. 3.10 A policy of fixed prices can increase farmers' uncertainty and adversely affect production. This applies in particular to cereal crops and to an increasing extent if the ONCAD monopoly becomes more effective. When prices are announced late, they are the unpredictable result of largely political factors and are set without due attention to the effect on alternative crops. Worse, unless subject to fine tuning, fixed prices de-stabilize rural incomes by their slow or lack of response to yield fluctuations. For cereals, there is an overwhelming case for de-control of prices and for elimination of the de jure marketing monopoly. There may be a role for a guaranteed minimum price in association with a grain storage program but it has not yet been clearly shown that this would be in the national interest. Moreover, very difficult managerial and financial problems would be entailed. In general, prospects for grain storage programs would improve if prices were de-controlled, since purchasing in years of abundance and selling in years of shortage should normally be a profitable activity. Crop Inputs 3.11 The main input subsidy is on compound fertilizer, supplied as credit- in-kind by ONCAD to cooperatives. The quantity distributed grew steadily throughout the 1960s, from 14,000 tons in 1961/62 to 63,000 tons in 1967/68. It then declined in each of the next three years to 15,000 tons in 1970/71 and has subsequently increased steadily to about 100,000 tons in 1976/77. Compared to the total acreage under cultivation, this is a relatively small quantity. If one assumes that on the average 150 kg/ha of fertilizer is applied yielding 0.2 additional tons/ha of millet or groundnuts, it appears that fertilizer use presently explains less than 10% of total crop production. This situation could be improved considerably. Fertilizer use for ground- nuts has fluctuated even more widely than aggregate use (Text Table 3).  Table 3- FERTILIZER QUANTITIES, PRICES, SUBSIDIES AND PRIVATE BENEFIT:COST RATIOS Crop Volume Distributed 1/ Groundnut 3/ Season Groundnuts Millet Other Total Price to Price Apparent Price in 2/ Benefit:Cost ratio to groundnut Farmers ex-factory Subsidy preceding year- producers at incremental yields (kg/ha) 300 200 100 --------------('000 tons)----------------- -----(CFAF/kg)--------- (CFAF million) ---(CFAF/kg)---- 1961/62 11.7 1.6 0.4 13.7 10 25 205 21.5 4.30 2.87 1.43 1962/63 20.7- 2.3 1.3 24.3 10 25 365 21.5 4.30 2.8/ ;.43 1963/64 23.1 2.9 0.8 26.8 12 25 348 21.5 3.54 2.36 1.18 1964/65 32.2 4.7 0.9 37.8 12 25 491 21.25 3.54 2,36 1.18 1965/66 26.1 4.7 1.1 31.9 12 25 415 21.25 3.63 2.42 1.21 1966/67 38.4 9.1 1.7 49.2 13 25 590 21.75 3.15 2.10 1.05 1967/68 48.2 12.7 2.4 63.3 14 25 696 20.5 2.44 1.63 0.81 1968/69 25.9 9.6 2.3 37.8 16 25 340 i7.1 2.14 1.43 0.71 1969/70 12.8 8.4 3.7 24.9 12 25 324 17.1 3.08 2.06 1.00 1970/71 6.5 6.2 2.1 14.8 12 25 192 18.5 3.08 2.06 1.00 1971/72 12.2 10.4 3.4 26.0 12 25.9 361 18.5 3.85 2.57 1.28 1972/73 23.1 19.5 4.4 47.0 12 25.2 620 23.1 3.85 2.57 1.28 1973/74 19.9 13.1 11.2 44.2 12 26.3 632 23.1 4.25 2.83 1.42 1974/75 28.6 24.9 9.3 62.8 12 50 2,386 25.5 4/ 6.92 4.61 2.31 1975/76 46.6 28.3 7.3 82.2 16 60 3,617 41.5 5.19 3.46 1.73 1976/77 64.7 31.8 8.2 104.7 20 55 3,665 41.5 4.15 2.77 1.38 Sources: Volumes, through 1973/74: MPC preliminary documents for Vth Plan; 1974/75: MDRH preliminary document for Vth Plan; 1975/76 and 1976/77: inferred from SIES data plus an assumed 5,000 tons special compounds for cotton. Prices, through 1973/74 MPC (see above); 1974/75-1976/77 mission estimates. 1/ Paddy, cotton, vegetables and miscellaneous. 2/ Because orders for fertilizer must be placed for crop season (Y) only a few months after announcement of the producer price for groundnuts in the preceding crop season (Y-1) and 6-7 months before the announcement for (Y). 3/ At standard fertilizer application of 150 kg/ha; fertilizer and groundnuts valued at producer prices in the same crop season. 4/ A subsequent payment or bonus of CFAF 4/kg was also paid, but too late to have any effect on demand for fertilizer for the 1974/75 crop season.  - 14 - The subsidy has generally amounted to about 50% of ex-factory price; it reached about 75% in 1974 and was reduced to about 60% in 1976 and about 50% in 1977. In aggregate, the subsidy reached about CFAF 3.6 billion ($15 mil- lion) in 1975 and 1976. These amounts, and the net loss on the production accounts of the Stabilization Fund in 1976, have provoked serious questioning of the purpose of the fertilizer subsidy. 3.12 Variations of the infant industry argument are sometimes advanced in support of fertilizer subsidies but, after more than a decade of fertilizer promotion, this argument no longer applies. Discussion of the two more substantial arguments in support of subsidies is complicated by major dif- ferences of opinion on the validity and the interpretation of data on yield response. First, where in the average year fertilizer use is socially pro- fitable as measured by opportunity costs, it may not necessarily be privately profitable. 1/ Such a difference arises for crops that are heavily taxed; but in these cases it would be preferable to reduce the tax burden on the product rather than to subsidize one input. The difference could also arise from farmers' reluctance to take risks in an environment of highly variable rainfall. Risk could however be tackled more directly by crop insurance, organized and implemented through the existing cooperatives. One can con- clude that there is a case for increasing private profitability and reducing private risks, but fertilizer subsidies are the wrong way to achieve these aims. 3.13 A second argument in support of fertilizer subsidies--promoted especially in Senegal--is that the long-term benefits of fertilizer are greater than the short-term because, over the long-term, fertilizer prevents soil deterioration. The justification for fertilizer subsidies follows then from the reasonable assumption that individuals place less value on future benefit than does society in general. The proposition that fertilizer use prevents soil deterioration is, however, keenly debated. The argument may apply to the phosphate component of fertilizer, but for the more expen- sive nitrogen and potassium components, all scientific evidence shows that their effects on soil fertility will last one or two years at most, and that fertility of exhausted soil can be improved instantaneously by new fertilizer application. This argument should, therefore, be rejected. 3.14 It is also argued that a fertilizer subsidy serves an income redistribution function in favor of the northern and central Groundnut Basin where there are few income opportunities other than groundnuts and millet. In these drought-prone regions where agriculture is marginal, fertilizer subsidies may be essentialV to overcome the inherent risk of fertilizer use. With decreasing rainfall the profitability of fertilizer declines rapidly and in the extreme, the entire expenditure on this input 1/ Social profitability is calculated by pricing inputs and production factors at their real value for the economy and by eliminating subsidies. For example, if a venture has a private return of 10% but receives sub- sidies at 20%, the venture may be privately profitable but not socially profitable.  - 15 - is lost. On the average, fertilizer use may still be socially profitable in some of the more temperate of these areas despite the disincentives to private use. The present uniform national price for compound fertilizer entails, however, subsidies for the more favorable southern regions too, where the private benefit:cost ratio for fertilizer use on groundnuts typically exceeds 3:1 (Text Table 3). Since such a ratio already gives sufficient incentive to farmers in these areas, any subsidy that would further improve this ratio would imply unnecessary benefits for farmers who do not need them. In the absence of crop insurance, fertilizer subsidies may be a second best solution to maintaining agricultural production and income levels in the northern regions. If this second best approach is followed the undue subsidy going to the more favorable areas should be curtailed. Paradoxically the more favorable areas receive the highest subsidy per unit of nutrient. The recommended compound fertilizers in favorable areas are chemically more concentrated and therefore more costly than those recom- mended in the northern-and central Groundnut Basin. Since all formulas are sold at the same price in Senegal the former are subsidized more. But this also provides the clue to a practicable way out of the present unsatis- factory situation, namely the application of different prices for different ccmpound fertilizers. As a first step, a standard percentage subsidy might be adopted, which would equalize the subsidy rates in the less favorable areas with those in the favorable areas. Consumer Prices 3.15 Maximum consumer prices are fixed for a fairly narrow range of agricultural commoditieEs, including rice, wheat flour, sugar and groundnut oil. These are the four main commodities in the consumption accounts of the Stabilization Fund. Price fixing is an attempt both to stabilize and to restrain the growth in the cost of living. Subsidy payments became increas- ingly important in the early 1970s and reached intolerable levels by late 1974 when substantial price increases were authorized in November. World prices subsequently declined so that the Stabilization Fund realized a surplus in its consumpti-on accounts for 1975 and through early 1976 before consumer prices were redluced in May. It is too early to ascertain the effect of these price changes on the Stabilization Fund but rice at least will pro- bably show a surplus in FY77. 3.16 The overall ei'fect on income distribution of the complex of consumer and producer pricing and subsidy policies through end-1974 is clear-cut: urban consumers at all income levels and the wealthier rural dwellers benefitted from retail price controls whilst the rural masses gained very little, mainly from sugar. Rural producers of groundnuts and cotton made substantial net payments to the Government. Since end-1974 the overall effect of Govern- ment pricing and subsidy policies is more difficult to assess since, although the thrust of the price changes was generally to reverse the existing bias' in favor of the urban population, urban wage-earners have received wage in- creases sufficient to cover their increased cost of living. The concern here is not to argue the merits of the price policies from the viewpoint of income distribution; a general case in favor of increased rural incomes  - 16 - has been made in Chapter II. Rather, the concern is to examine the effects of consumer price policies on incentives for domestic production, particularly of cereals. 3.17 The critical effect of subsidies on rice and wheat flour is to provide a cheap and convenient substitute for millet. On the other hand to increase demand for the latter, by reducing rice and wheat consumption, simultaneous action is required on relative prices and on millet promotion. There is a pervasive conventional wisdom to the effect that millet is not a substitute for rice for most urban consumers. In 1975, however, and sub- sequent to the increase in the consumer price of broken rice from CFAF 60 to CFAF 100/kg, imports of rice which had previously been at about 200,000 tons/year fell to 100,000 tons. The late 1974 millet harvest of 700,000 tons was by far a new record; yet the fragmentary evidence on producer prices suggests that millet was traded in the illegal market at CFAF 50/kg or more in contrast to the official price of CFAF 35/kg. All available indications point to a considerable switch in demand from rice to millet, sufficient to challenge the conventional wisdom of very imperfect substitutability between the two commodities and to illustrate the power of relative price policy. 3.18 Furthermore, two recent technical breakthroughs have created the means to render millet a more convenient foodstuff. It is now possible to produce a dry, stabilized millet flour marketable as a plain flour or as an instant couscous; another technique has been developed to produce bread made of a blend of millet and wheat flour. Both techniques are now exploited on a very limited scale in Dakar but two problems must be solved before the scope for large-scale operations can be tested. First, there must be an assured supply of millet, which is more likely to be forthcoming if the ONCAD de Jure monopoly is eliminated (this has been argued, and on even more powerful grounds, in Chapter II). The introduction of millet as an element in the urban food supply should not however be postponed until all institutional changes in the rural sector have been made, but should start-- if necessary--on the basis of imported millet. Second, the relative prices have to be right. In late 1976, the subsidized price of wheat flour was lower than the cost of millet flour, the latter having been "pulled up" by the price of rice, which included a tax! The financial impossibility for bakers to incorporate millet flour into bread is a fine example of how distorted prices can lead to negative effects on the economy as a whole. Despite the observation that millet is more readily substitutable for rice than popular belief would have it, some consumer resistance can be anticipated. A millet promotion campaign is required. In addition, bread-making with millet and wheat flours requires that the doughs be made separately and bakers would have to be trained accordingly. 3.19 Besides the balance of payments gains that can be expected from a millet program, there are some important agronomic aspects to this matter. Agricultural development in the central and southern Groundnut Basin calls for intensification of farming systems; such activity is incon- ceivable without a substantial component of millet in the cropping patterns. The increased millet production would lead to overproduction if no strong pro-millet pricing and marketing policy is pursued.  - 17 - 3.20 There are also good grounds for reviewing the basis used for evaluation of millet and rice in project analysis and in analysis of regional comparative advantage. Hitherto, the economic price for millet has been based oni the "nearest equivalent" world price, namely U.S. No. 2 Milo yellow sorghum. This commodity is, however, scarcely a perfect substitute for millet; after the 1972-73 drought, pri:ce differences of the order of 30% were observed in Sahelian countries between local millet and late arrivals of U.S. sorghum provided under drought relief programs. Although conditions then prevailing were abnormal, consumers; evidently made a distinction between local coarse grains and imported feed grains. This suggests that a consumer-preference premium, perhaps of 20%, should be added to the US sorghum price in determin- ing a reference price for millet. On the other hand, as argued above, the technology has been developed to make millet a more convenient foodstuff and thereby more substit:utable for rice and wheat. Valuation of millet in terms of rice and wheat., and primarily the former in view of the far larger scope for substitution, is at least as defensible as valuation in terms of sorghum increased by a iiomewhat arbitrary preference premium. 3.21 The reference price conventionally adopted for rice is Thai, mLlled, 25-35% broken. Since regular quotations for this grade have not been available since 1973, the convention has been to apply the previously prevailing relation of 70% between 25-35% broken and 5%-broken to the continuing price series for the latter. This procedure overstates the value of rice imported to Senegal since the monopoly importer, ONCAD, boys 100%- broken whenever possible. The relation between ONCAD's annual average buying price and the price of 5%-broken was about 60% in 1974-76. This is the basis to be used for valuation of millet in terms of rice. This method entails a mLllet price for 1976 sLmilar to that obtained from US sorghum with a 20% premium but an increasingly higher price through 1980 since rice prices are forecast to increase and sorghum prices to decline. 3.22 Whereas the value of broken rice has conventionally been overstated, it does not.necessarily follow that the value of paddy has been set too high. The local rice market involves trade in three sharply differentiated products, namely broken rice, whoLe-grain rice or 5%-broken, and processed rice, for which the retail prices in late 1976 were respectively CFAF 80/kg, CFAF 160/kg and CFAF 225/kg or more. For processed, packaged rice, there is an annual quota for import licenses which, in the last few years at least, has been inadequate to satisfy detmand. For whole-grain rice, ONCAD is the monopoly imnporter but does not assure adequate regular supplies. Its major concern, rLghtly, is to provide regular supplies to satisfy demand for broken rice. Wiile imports of processed and whole-grain rice probably do not exceed 5,000 tons/year, this understates demand. In addition, whole-grain rice under a brand name other than OIICAD's and a rather higher quality is also retailed ina Dakar along side the ONCAD product, but sales are unknown. These points suggest that demand for unbroken rice, although very small in relation to broken rice, is not negligible and that it may be restricted at present by the rationing implic:it in inadequate and irregular supplies. Valuation oE domestic production of paddy should, then, be at a higher level than that implied by adoption of broken rice as the reference price. How much higher is difficult to estimate, precisely because of current marketing  - 18 - imperfections. The share of unbroken rice in total rice demand would, how- ever, have to reach 25% before the weighted average valuation reached 70% of the price of 5%-broken rice. This seems unlikely but it is conceivable if the substitution of millet for broken rice becomes important. In sum, the valuation of domestic production of paddy should be based on 70% of the price of 5%-broken rice, it being understood that this is a maximum. C. Production Systems Rainfed Farming 3.23 The potential for rainfed farming in Senegal is considerable, but actual production growth will be determined by the pricing and marketing policies adopted and by how quickly and efficiently improved techniques, which have already been tried out in some regions, reach other parts of the country. The improvement of farm incomes necessarily requires that the productivity of the most limiting factor of production be increased. In the outlying regions of the country, land is not scarce and present farming practices leave substantial areas available for further development. In these areas, labor is the constraint; labor-substituting farm machinery should therefore be promoted and the credit and input supply systems geared up accordingly. In the absence of animal-drawn equipment, the main labor shortages are at the times of planting and early weeding. Once these bottle- necks are removed, the labor constraint shifts to harvesting, and appropriate machinery should then be made available. This sequence has largely run its course in the Groundnut Basin where land rather than labor is now the limiting factor and more intensive land use must be promoted. Levels of technology corresponding to these different circumstances have been devised and now constitute the themes of all extension work, where they are being monitored, re-tested and further refined. The scope for intensification of land use in the central and, particularly, in the northern Groundnut Basin is unfortunately very limited. There is little scope for irrigation and no major crop possibilities other than groundnuts and millet. In the northern Basin, fertilizer use becomes increasingly risky and the average yield response declines with decreases in rainfall. Increased land productivity in this area will depend primarily on the further development of livestock within the farm- ing system. Incomes could, however, also be improved by paying a premium price for the recently-promoted short-cycle groundnut varieties that fortunately have the characteristics of edible groundnuts (known commercially as HPS) and which were exported at favorable prices in 1976. In the southern Groundnut Basin, there is still much unexploited potential for increasing land productivity. Irrigation 3.24 The present pattern and future prospects from crop production are determined above all by rainfall. The more northern the region, the lower the average yearly rainfall is and the larger the risk becomes that rainfall will be insufficient in the critical periods of the growing cycle. Even in the outlying areas to the south and east, where there is much unexploited  - 19 - potential in rainfed farning, risk is the main feature of rural life. Uncertainty is also brought about by fluctuations of groundnut prices on the world market since the Groundnut Stabilization Fund has not been able to protect the farmer against low world market prices for longer than a year. In fact, the uncertainty inherent in the agricultural sector in Senegal is extended over the entire economy. Government revenues, volumes of agricultural prcduce to be processed, the need for food imports, and general consumer demand all are heavily influenced by the size of the crop. With full water control, irrigation permits the near elimination of these rainfall-associated risks and is considered by many to be highly desirable if not a panacea. Full water control is expensive, however, even if storage structures are excluded, and it requires skilled management. The main crops that could be produced and readily marketed under this type of scheme are rice and wheat in the Senegal Valley and rice in Casamance. These crops are not high-value and the returns to irrigated production would not there- fore be high even if sufficiently skilled management can be assured. Assess- ments of economic returns to particular irrigation schemes have and will continue to show marginal returns at best; the first attempt to apply the domestic resource cost method showed negative returns. 3.25 Current methods of project analysis fail to capture three important effects of irrigation on the scale that may be eivisaged for the Senegal Valley. First, risk and uncertainty will be reduced. No matter how great the scope and how high the returns to investment in rainfed farming, risks and uncertainties are inherent in rainfed cropping. Their elimination via full water control would have a value in terms of security and well-being that exceeds the difference in value added in the "average year" with and without irrigation. That additional value is hard to assess, but those who have experienced the famine that often accompanies a bad rainfall year would assess it higher than those who have not. In developed countries risks that involve human life are not submitted to cost:benefit calculation because of the difficulty of attaching numeric value to human life. Moreover, risks are reduced to negligible proportion by legally enforced standards of project design. Based on these minimum legal standards a "least-cost solution" is sought. Similarly, the Senegalese Government should look for t:he cheapest way to reduce the heavy risk of periodic famine for the rural population. Irrigation should be evaluated, together with other possible solul:ions like lccal grain storage and periodic massive international food relief actions. A second effect of irrigation not taken into consideration by current methods of project analysis is the impact that such a large irrigation scheme in the Senegal Valley would have on the rest of the economy. That is, the effects of^ the overall irrigation program are non-marginal. Project analysis, as per- formed by the Bank, could theoretically take all these ramifications into account, but in practice this is not possible. Third, the cost:benefit method is not suited to analysis of inter-generational distribution of income such. as wculd result from the irrigation scheme proposed for the Senegal Valley. Such a scheme would require investments over 30 to 40 years and more than double the agricultural production in a country with very few investment choices, hence strongly affecting future distribution of income.  - 20 - 3.26 In addition to the above country-wide aspects of irrigation, the regional side should also be considered. Since the Government cannot neglect the economic problem of large, densely-populated regions it sometimes has to opt for less profitable investments to maintain income levels and to arrive at a reasonable regional balance in its development strategy. The Senegal Valley is an area where no improvements can be envisaged in present farming practices, neither in rainfed cropping nor in flood recession cropping. Risks are very high and the inputs that may wisely be ventured are accordingly small. By contrast, in much of the area of the country where there is unexploited potential in rainfed farming, the scope for irrigation is very limited. The apparent issue of rainfed versus irrigated production is in effect a question of promotion of rainfed farming in area A or promotion of irrigation in area B, or both; the rainfed/irrigation choice in the same area arises only in parts of Casamance where regional arguments for irrigation do not apply. 3.27 Whereas development of rainfed agriculture will probably be relatively profitable, and should be undertaken as a first priority, domestic production of foodgrains will fall increasingly short of demand unless major investments in irrigation are also undertaken. Despite the considerable unexploited potential of rainfed farming, an absolute limit will arise in that all unculti- vated arable land could be occupied by the mid-1990s. With demand for food- grains increasing at about 3%/year, total demand will increase from 1.1 to 2.0 million tons in the period 1970-1990. Domestic production of rainfed cereals would increase from 0.6 to 1.0 million tons in this period under the assumptions stated in Annex Table 2.1 which take into account the measures advocated in this chapter. If there were no increase in irrigated cereal production through 1990, the cereals import gap would be 0.9 million tons. From other sections of this economic report, it becomes evident that it will be impossible to design a realistic program of export promotion that would provide foreign earnings to pay for such a massive increase in basic food imports. The likely outcome would be economic stagnation because of a severe balance of payments constraint. Moreover, even assuming a reasonable pace of development of irri- gation facilities, the cereals import gap in 1990 would be of the order of 0.5 million tons. Table 4: ESTIMATED FOODGRAIN DEMAND AND SUPPLY 1970 1990 (million tons, untreated grain) Demand 1.1 2.0 Rainfed production 0.6 1.0 Difference 0.5 1.0 Irrigated production 0.1 0.5 Import Gap 0.4 0.5 3.28 Within the irrigation subsector, there is an apparent conflict between large-scale mechanized operations, which might be expected to be quickly implemented and rapidly attain full production, and development based  - 21 - on family farms. The iniltial developments in the Delta were, however, large- scale and the results were disappointing. Now that the Delta is or will soon be wholly developed, future irrigation investments will be concentrated in the Valley where, unlike the Delta, there is already a densely-settled popu- lation. Even if the initial experiences with large-scale operations had produced the benefits expected, application of such methods in the Valley is virtually impossible. Furthermore, recent design work by SAED shows that the gradual development oE large perimeters by and for the local people will entail considerable capit-al cost savings, perhaps as much as 35%. This is the direction in which investments in irrigation should proceed. Settlement 3.29 The density of rural population varies considerably between the regions, from over 80/km2 in the central and northern Groundnut Basin to less than 10 in Fleuve and Senegal Oriental. The average figure for Fleuve is particularly deceptive in that most of the people are concentrated along the Senegal Valley, where the density is also over 80/km2, while a larger part of the region is desert. But while the Senegal Valley has great potential for irrigation, the northern and central Groundnut Basin has little scope for further agricultural development; outward migration is already and will continue to be important from this area. Although information on this migration is scarce, there are both rural-urban and rural-rural movements; migration into the Kaffrine District in eastern Sine Saloum was estimated at 1,500/year in the 1950s and at 5,000/year in the 1960s, or about: 0.5% of the rural population of Diourbel and Thies Regions at that time. This amounts to rural-rural migration on an appreciable scale. Recently, the destination of the migrants has shifted further east into Senegal Oriental. Between 1960 and 1970 the population of this region increased by 7.8 percent due to migration, but this modest rate has probably increased during the 1970s. It is claimed tha.t this uncontrolled (or, from a different point of view, spontaneous and unassisted) migration is tending to create the very prcblems found in the Groundnut Basin--lack of tenure rights, uncontrolled cropping, fragmentation of holdings and over-concentration on groundnut production. Proponents of this view argue that the development of new land should be controllee! through settlement schemes in order to avoid the problems cited. But this is not the only possible conclusion since the ostensible problems are aLll susceptible to direct action and actions which, taken together, would constitute worthwhile assistance to spontaneous mi- grants. It is becoming apparent that settlement schemes on a scale suffi- cient to have any impact on existing spontaneous movements will be both expensive and difficult to implement.  - 22 - Chapter IV. SECTOR PLANNING AND PROJECTS A. Planning 4.01 Projects have rightly been called the "cutting edge" of develop- ment since they involve a concentration of resources on fairly specific objectives. It is precisely because projects may be readily distinguished and understood and partly, too, because staff can have a fairly free hand in their design and implementation that they receive so much attention. Projects are, however, the last link in a chain comprised of objectives, strategy, policies, and projects. In particular, it is at the level of policy formulation--the critical point in the translation of the general into the specific--that key decisions are made. These decisions will have direct effects on the utilization of existing investments and they will set the framework within which new investments will operate. 4.02 Senegalese planning is based on a 25-year perspective that pro- vides an opportunity to examine questions such as population growth, training and intersectoral links, which cannot be captured in a four-year plan horizon. In practice, however, there is undue focus on the short- term, and problems and opportunities arising from population growth, in the northern Groundnut Basin for instance, have received little attention. The planning function should, in addition, relate to both policies and projects. The problem of recurrent costs that are being or will be generated by new agricultural projects tends to be ignored or underesti- mated by staff working on particular projects, who also tend to give inadequate attention to possible cost recovery from project beneficiaries. The planners are in the best position to propose a consistent approach to recurrent cost financing and cost recovery both within and between sectors. Furthermore, the ability of departments of ministries and of other agencies to prepare projects and to interest financiers varies widely. These relative abilities do not necessarily reflect their capacities for implementation nor even the Government's sectoral objectives. The planners are in a good position to correct these distortions and would, thereby, have some effect on increasing the currently low absorptive capacity of the agricultural sector compared to the foreign finance available. B. Projects 4.03 For most agricultural projects being considered under the Fifth Plan, funds are already committed or are likely to be secured, with the exception of part of the cost of large irrigation schemes in Lower Casamance and much of the cost of livestock and forestry projects. Considering the absorptive capacity of the crop subsector, which is limited by managerial and staff constraints, it would be a commendable achievement if the Govern- ment should succeed in carrying out those projects for which firm commit- ments have already been obtained.  - 23 - 4.1)4 In Lower Casamance, the draft Plan envisages construction of one small and two large dams and engineering studies of one large and several small dams. In chis mangrove area, the soils suffer from exces- sive salinity and have to be supplied with fresh water to be cultivable. In addition, mangrove so:Lls become acid as they dry up and could become ir]reversibly unsuitable i-or agriculture. For several years prior to 1974, a consultant firm experimented with rice cultivation in Lower Casamance, using fresh water storage dams, without sluices (floodgates). In view of limited success, the firm recommended sluice dams for future development, whereby the sluice would be used to admit saline water to prevent acidification when the fresh water level becomes too low. This method is attractive in principle but it has not really been tested locally: there is as yel: no firm evidence that the method would solve the problem of cropping on mangrove soils. In view of the magnitude of the investments envisagect and the uncertainty of the proposed method, it would be preferable to test the method on a small-scale at the Djibelor research station, and def'er dam construction pending the research results. 4.05 In the livestock subsector, and although there are several pro- jects for which finance has yet to be found, it would be preferable to def:er new projects based primarily on cattle development until the ongoing "Zone Sylvo-Pastorale" project and the new project in Eastern Senegal can be assessed. On the other hand, the beef fattening component of the ongoing Sirne Saloum project, which is primarily a crop intensification project, is developing much better than anticipated. Similar agricultural projects in the future could therefore include a substantial livestock component. In addition, actions to promote poultry, pigs and small ruminants should be intensified. 4.016 Forestry is a relatively neglected subsector despite considerable sccope for development. Although the timber potential of the country is low, tropical timber could be improved and developed in Lower Casamance. Even more important, the growing scarcity of charcoal and inadequately controlled processing are causing increasing deforestation. Recent improvements in kiln technology, with increased conversion rates, have not yet been intro- duc:ed. In the draft Plarn, the proposed allocation from the national budget to the forestry subsector is higher than in any other subsector except irrigated agriculture, but commitments from foreign financiers are appreciably lower than in other subsectors. This is a field where an improvement in the method of project presentation should have a high payoff. C. Projections 4.07 Annex Tables 2.1-2.9 indicate primary sector accounts, rural income and trends in foreign trade for 1975-1990. For the crop subsector, the projections have been derived from specific assumptions concerning physical outputs to be expected from projects now being executed or being planned and from adoption of the policies, particularly those on relative  - 24 - prices, advocated in Chapter III. The choice of a base year proves an awkward problem in that the good harvests of 1974/75 and 1975/76 were due in part to good weather. The general principle adopted is to take the Government's own view of "normal" output for 1974/75: long-term average yields and short-term average planted areas, on which base the effects of specific physical changes are traced. The resulting projections, summarized in Text Tables 5 and 6, imply much higher growth rates than were realized in the past decade, although lower than those realized in the period 1961-1965. 4.08 The projections for the rainfed sector visualize the long-term needs for additional projects in which there would be heavy emphasis on technical assistance to farmers through extension services. The importance of the extension of improved technologies is reflected in the projected fertilizer use. This is expected to go up from 100,000 tons in 1976/77 to 150,000 in 1980/81 and to 270,000 tons in 1989/90. The 8 percent in- crease a year that these figures imply can only be achieved if the private benefit:cost ratio of the use of fertilizer is close to 3:1, which appears to be a threshold value for Senegalese farmers. This threshold can be reduced by good extension services and measures to minimize the risks of fertilizer use as mentioned in Chapter III. The fertilizer policy is crucial for the realization of increased growth rates for rainfed agri- culture. 4.09 For millet and sorghum the main specific assumptions include expansion in planted area in line with rural population (2.2%/year), expan- sion in fertilizer use to the maximum area to which it may be profitably applied, namely two-thirds of the total planted area by 1989/90, induced expansion in planted area due to the reduction in fallow permitted by increased fertilizer use, and an increased rate of intensification through de-stumping, base phosphate and ploughing. The effects of these assumptions are, through 1980/81, an incremental production of 110,000 tons and, through 1989/90, a further 260,000 tons, or a long-term growth rate of 3.5% per year. Planted area would increase from 1.047 million ha in 1974/75 to 1.164 million ha in 1980/81 and to 1.370 million ha in 1989/90; average yields would increase from 529 kg/ha to 567 kg/ha to 670 kg/ha. 4.10 For groundnuts for oil, the main specific assumptions include expansion in fertilizer use to the maximum area to which it may be profit- ably applied, namely three-quarters of the total planted area by 1989/90, an increased rate of intensification through de-stumping, base phosphate and ploughing, and a small increase in planted area. The effects of these assumptions are, through 1980/81, incremental production of 150,000 tons and, through 1989/90, a further 170,000 tons, or a long-term growth rate of 2% per year. Planted area would increase from 1.153 million ha in 1974/75 to 1.187 million ha in 1980/81 and to 1.255 million ha in 1989/90; average yields would increase from 852 kg/ha to 952 kg/ha to 1,036 kg/ha. 4.11 For irrigation, the main specific assumptions are as follows. In the Senegal River Basin, about 2,000 ha per year will be brought to  Table 5: PRIMARY SECTOR VALUE ADDED AND RURAL PRIMARY SECTOR INCOME, 1975-90 1975 1981 1990 Annual Growth Rates -1975-81 1981-90 1975-90 ----(1975 CFAF billion)--- Primary Sector Value Added Crops 66.9 77.9 107.5 2.6 3.6. 3.2 Livestock 21.1 26.8 37.6 4.1 3.8 3.9 Fishing 18.5 29.3 37.7 8.0 2.8 4.9 Forestry 8.4 9.6 11.7 2.2 2.2 2.2 Total Value Added 114.91' 143.6 194.5 3_8 3.4 3.61/ Rural Primary Sector Income 2/ N) Total income 107.3 129.6 177.0 3.2 3.5 3.4 Rural population (million) 3.493 3.982 4.846 2.2 2.2 2.2 Rural primary sector income/head (CFAF/head) 30,700 32.500 36,500 1.0 1.3 1.2 Source: Annex Tables 1.7 and 2.9 1/ The official national accounts for 1975, which were issued in September 1977 and which value outputs at official producer prices, show total primary sector value added as CFAF 122.8 billion rather than CFAF 114.9 billion. This difference is mainly due to prices: revaluation of the latter in terms of the former's prices yields CFAF 119.6 billion. Revaluing the 1990 estimate in the same terms yields a 15-year growth rate of 3.5% in those price terms. 2/ Value added, excluding industrial fishing and incremental herd growth.  - 26 - Table 6: COMPOSITION OF PRIMARY SECTOR VALUE ADDED 1975 and 1990 (Values in 1975 CFAF billion) 1975 1990 Value Z by % by Value % by % by sector sub- sector sub- sector sector Crops: Cereals 23.2 20.2 34.7 42.3 21.8 39.4 Other food 7.1 6.2 10.6 10.3 5.3 9.6 Groundnuts 34.1 29.7 51.0 48.3 24.8 44.9 Other industrial 2.5 2.2 3.7 6.6 3.4 6.1 Subtotal 66.9 58.2 100.0 107.5 55.3 100.0 Livestock: Cattle 11.0 9.6 52.1 17.6 9.1 46.8 Sheep and Goats 3.3 2.9 15.7 5.7 2.9 15.2 Poultry 4.7 4.1 22.3 10.8 5.5 28.7 Pigs 1.0 0.9 4.7 1.7 0.9 4.5 Other 1.1 1.0 5.2 1.8 0.9 4.8 Subtotal 21.1 18.4 100.0 37.6 19.3 100.0 Fishing: Small-scale sea fishing 13.3 11.6 71.9 23.4 12.0 62.1 Industrial sea fishing 4.4 3.8 23.8 13.4 *6.9 35.5 Inland fishing 0.8 0.7 4.3 0.9 0.5 2.4 Subtotal 18.5 16.1 100.0 37.7 19.4 100.0 Forestry: Subtotal 8.4 7.3 11.7 6.0 Total primary sector value added 114.9 100.0 194.5 100.0 SOURCE: Annex Tables 1.2, 1.4, 1.6, 1.7, 2.2, 2.5, 2.7 and 2.9.  - 27 - full production during the four years from 1976/77-1980/81, starting from an initial level of 17,400 ha of which 10,600 ha under full water control. By 1980/81, the existing 6,800 ha which are not yet equipped for full water control will be so equipped, and an additional 1,200 ha would be executed in the Valley. The total area under full production would then be 18,600 ha, of which 13,000 ha in the Delta and 5,600 ha in the Valley. In the following four years, 3,000 ha per year will be brought to full production and then 4,000 ha per year. By 1989/90, the total area under full production would be 19,600 ha or all the irrigable land in the Delta and 31,000 ha in the Valley. It is assumed that the Diama Dam will have been completed and that all the land in the Delta will be double-cropped. lields of paddy would be 3 tons/ha/crop in the Delta and 4 tons/ha in the Valley. In the CasamarLce area, project preparation is less well advanced and some technical problems have not yet been resolved (para. 4.04). It is therefore assumed that no new land will be brought to full production by 1980/81 but that development in the 1980s will proceed at the same rate as i.n the Senegal River Basin, and that 32,000 ha will be brought to full production by 1989/90; yields of paddy would be 3 tons/ha. Total pro- cLuction of paddy would increase from a 1974/75 base of 115,000 tons to 1.97,000 tons in 1980/81. and to about 450,000 tons in 1989/90, or a long- t:erm growth rate of 9.5%/year. 4.12 The sum of thiese effects (Annex Table 2.2) yields a gross value of crop output at 1975 prices of CFAF 92.7 billion in 1980/81, or a rate of growth of 3.0%/year from the observed value of CFAF 77.6 billion in 1974/75. In view of increased input use, the growth of value added is only 2.6%/year. The diraft plan does not provide directly comparable estimates but it does cite production objectives for 1980/81 in physical terms. Expression of t-hese objectives in comparable terms yields in 1980/81 gross value of CFAF 104.7 billion and a rate of growth of value added of 4.7%/year. In general, the Plan's production objectives are only slightly higher (5% for groundniits, 15% for millet and sorghum) than the estimates derived from the speciEic assumptions cited above. The notable exception is paddy for which the production objective is 300,000 tons in contrast to the above estimate of 200,000 tons. This difference derives exclusively from different assumptions concerning the pace of developments along the Senegal River. It derives from a) the notion of "area brought into full production", used in the above estimates but not in the Plan, and which takes account of the lead time required for construction and initial on- farm development, and b) the volume assumptions, the above estimates being based on 2,000 ha/year and the Plan on 5,000 - 6,000 ha/year. The latter pace is unlikely to be realized by 1980. 4.13 For the non-crop subsectors, the projections are derived from broader assumptions although specific physical changes have been estimated as far as possible; the assumptions are detailed in the annex tables. Again, the draft plan does not provide directly comparable estimates, but it does cite production objectives for 1980/81 in physical terms. For the livestock subsector, the two approaches yield almost identical results; although there are differences in detail they are minor. For the fishing subsector, the  - 28 - two approaches yield almost identical aggregate results in volume, but there are important differences of detail which find expression in value terms; in contrast to the 8.0% estimate shown in Text Table 5, the Plan shows 7.1%/year. For the forestry subsector, there are no significant differences between the two approaches. 4.14 The overall results are in Text Table 5, which shows primary sector value added by subsector and rural primary sector income. Text Table 6 shows the changes in the composition of primary sector value added between 1975 and 1990. It reveals little change between subsectors, with small increases in fishing and livestock and small decreases in crops and forestry, but considerable change within subsectors, notably between cereals and groundnuts, poultry and cattle, and industrial and small-scale sea fishing. The sectoral rate of growth of value added through 1981 (3.8%/year) is lower than that implied in the draft plan on account of the different approaches to the crop subsector. It is nonetheless considerably higher than rates of growth observed over the past decade and requires, for the crop sub- sector component (2.6%/year), adoption of the policies advocated in Chapter III. Rural primary sector income would grow at about 3.5%/year through 1990 or a little over 1%/year per rural inhabitant. This means that, even if real incomes in urban occupations were held at present levels, there would still be a massive gap between urban and rural average incomes in 1990. The only way to make a substantial reduction in that gap would be to reduce real incomes in urban areas.  _ 29 _ ANNEX A List of Tables 1/ 2/ 1. Primary Sector Acecounts and Rural Primary Sector Incomes 1961-75 1.1. Crops: Production and Value 1961 - 1975 1.2. Crops: Value Added at current prices 1974 and 1975 1.3. Livestock: Production and Value 1961 - 1975 1.4. Livestock: Value Added at current prices 1974 and 1975 1.5. Fishing: Production and Value 1961 - 1975 1.6. Fishing: Value Added at current prices 1974 and 1975 17. Primary Sector Value Added and Rural Primary Sector Income 1974 and 1975 1.8. Rural Primary Sector Income by Region 1974 and 1975 2.. Primary Sector Accounts. Rural Incomes and Trends in Foreign Trade 1975-90 (at 1975 market prices) 2.1. Crops: Production and Value 1975, 1976, 1981 and 1990 2.2. Crops: Value Added 1981 and 1990 2.3. Crops: Trends in Foreign Trade 2.4. Livestock: Production and Value 1975, 1981 and 1990 2.5. Livestock: Valute Added 1981 and 1990 2.6. Fishing: Production and Value 1975, 1981 and 1990 2.7. Fishing: Value Added 1981 and 1990 2.8. Fishing: Trends in Foreign Trade 2.9. Primary Sector Value Added and Rural Primary Sector Income 1981 and 1990 11 Tables are compiled from data and information available as of November 1976. Meantime, the Department of Statistics has issued, in May 1977, the "Comptes EcorLomiques 1974 et Retrospective sur les Comptes Eco- nomiques 1959-1973" and, in September 1977, the "Comptes Economiques 1975." Where conmparable to the presentation adopted in the attached tables, the results of these recent official publications are shown therein. Although the methods used in the two approaches are some- times quite different, the discrepancies in aggregate results are very small. 2/' All. data are expressed, for national accounts purposes, by calendar year; output from the crop year (e.g.) 1974/75 is shown under 1975.  C-p.o P-d-t-o, -od Uj-t, 1961 -1925 1961 I/ 19.62 i2