Senegal - Economic trends and prospects (Vol. 3 of 4) : The industrial sector

Report No. 1720a-SERe p The Economic: Trends and Prospects of Senegal (In Four Volumes) Volume 111. The Industrial Sector December 1979 Western Africa Regional Office FOR OFFICIAL USE ONILY Document of the World Bankc This document has a restricteci clistribution and may be used by recipients only in the performance of their ofl icial duties. Its contents miiy not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS US $1.00 = CFAF 220.37 CFAF 100 = US$0.454 WEIGHTS AND MEASURES 1 kilogram (kg) 2.20 lbs 1 millimeter (mm) = 0.04 inch 1 meter (m) = 3.28 feet 1 kilometer (km) = 0.62 miles 1 hectare (ha) 2.47 acres GLOSSARY OF ABBREVIATIONS BCEAO '-Banque Centrale des Etats de l'Afrique de l'Ouest CEAO Communaute Economique de l'Afrique de l'Ouest CPSP Caisse de Perequation et de Stabilisation des Prix ECOWAS Economic Community of West African States EEC European Economic Community ICS Industries Chimiques du Senegal ONCAD Office National de Cooperation et d'Assistance pour le Developpement SIES Soci&te Industrielle des Engrais du Senegal SMIG Salaire Minimum Interprofessionnel Garanti SOFISEDIT Societe Financiere Senegalaise pour le D6veloppement Industriel et Touristique SONACOS Societe Nationale de Commercialisation des Oleagineux du Senegal SONEPI Societe Nationale d'Etudes et de la Promotion Industrielle UDAO Union Douaniare de l'Afrique Occidentale UDEAO Union Douaniere des Etats de I'Afrique de l'Ouest UMOA Union Monetaire Ouest Africaine VAT Value Added Tax 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 Maia Report Statistical Annex Vlolume II The Agricultural Sector Vlolume III The Induistrial Sector 0olume 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 payments), H. tubell (human resources), J.D. Shilling (economic modeling), El.R. Steeds (agriculture), and M.P. van Dijk (informal sector), and com- plemented with findings of the Bank's regular sector work and more recent economic missions. This volume was written by Mr. Brendan Horton on the basis of the results of an IBRD regional research program on incentives and comparative costs in four West African countries, a program which started in 1973; and Mr. Horton's participation in the basic economic mission in November 1976 and his subsequent missions to Senegal. This volume was finalized in March 1978; new developments since that date have been added in footnotes. Working Papers 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 V'olume III. The Industrial Sector Table of Contents Page No. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS ....................... Market Orientation: Export or Local Market .......ii Weight to Accord to Each Export Market ............... iv Choice of Technique ................................... vi Obstacles to the Development of Exports .... ........ vi The Productivity of Labor, the Quality of the Labor Force, etc. .................. vii The Competitivity of Exports and the Structure of Incentives ....... viii Chapter I. BACKGROUND ....... ......................... 1 A. The Structure of the Industrial Sector .... ...... 1 Basic composition of the sector ................. 1 Tradiag relations with the European Commzon Market ...... ..................... 3 RelatLons with former French colonies in West Africa .............. . .................. 7 B. Recent Evolution of the Industrial Sector ....... 9 Receni: performance ...... ................... 9 Governlment Policy towards Industry-- Developments since 1972 .................... ... 10 C. Government's Plans for the Future .... ........... 13 Cayar Complex ................................... 13 Faleme Iron Ore Deposits ... ..................... 14 Dakar Marine .................................... 14 Indust:rial Free Trade Zone ................... ... 14 Decentralization of Industry .................. .. 15 Promotion of Senegalese Industry ............. ... 15 Agro-Industry ................................... 16 Develcopment of Other Industries .............. ... 17  Table of Contents (Continued) Page No. Chapter II. GOVERNMENT POLICY INSTRUMENTS ...................... 18 Principal Conclusions ............................ 18 A. Instruments ........... ........................... 19 The Protection of Local Industry ..... ............ 19 Quotas ...................................... 19 Import Duties ........... .................... 19 Indirect Tax Policy .......... .. ............. 20 The Investment Code .............. ................ 22 Direct Tax Policy of Profits .......... ........... 24 Credit Policy ...................... 24 Price and Wage Policy ............................ 24 Export Promotion Policy .......................... 24 Temporary Admission and Rebate of Import Taxes ............... .............. 25 Drawback Facilities ......................... 26 The Free Trade Zone ......................... 26 Export Taxation ............................. 26 Conclusion ............... .............. 27 B. Quantitative Evaluation of the Impact of these Policy Instruments ........................ 28 The Analysis .. ........................... 28 The Results ............... .............. 28 C. A Proposal for Reform .. .......................... 33 Export Subsidies (Step 1) ......... ............... 35 Reform of Import Duty Structure and Sales Taxes (Steps 2, 3, 4) .......... ........... 37 D. An Ad Hoc Scheme for Export Subsidies .... ........ 39 Design and Calculation of Rates of Subsidy ....... 40 Budgetary and Administrative Aspects of the Problem . .................................... 42 Chapter III. SECTOR ANALYSIS .................................. 44 Principal Conclusions ............................ 44 A. The Economic and Social Profitability of some Industrial Activities in Senegal .... ....... 45 The Results ........................................ 48 Analysis of the Results by Activity ................ 51 Conclusions of this Analysis by Activity .......................................... 53  Table of Contents (Continued) Page No. B. The Industrial Sector Component of the Fifth Plan .......... ................................. 54 Problems & Prospects of Phosphate Development ... 54 The Deposit ............... ................. 55 Operating Costs ............ .. .............. 55 The Phosphoric Acid Plant ....... .. ......... 58 Fishing ...... ................................... 59 Textiles .................... .................... 60 Grey Cloth ............................................... 62 The Production of Finished Cloth ........... 64 Spinning of Yarn, Knitwear and Finished Clothing ....... ........................... 65 Imports of Used Clothes ................... . 68 The Choice of Market ......... .. ............ 69 The Development of Marketing .. ............. 70 Future Strategy and new Projects for the Whole Sector ................. ............. 70 Groumdnut Oil Production ......................... 72 The Nature and Recent Development of the Industry ....... ........................... 72 Groundnut Production in Senegal, and its Relation to Capacity Utilization .......... 73 Markets and Marketing Problems ...... ....... 75 Incentives for Groundnut Processing Under the New Structure .......... .. ............. 77 The Food Industry excluding Groundnuts .... ...... 78 Sugar ............ ... . ........................... 78 Tomato Concentrate ......................... 78 Flour ........ .............................. 79 Cement & Other Products ......................... 80 Cement ................................................. 80 Packaging .................................. 80 The Industrial Free Trade Zone . . 80 Conditions of Entry ............... . ........ 80 Producer Margins . * ........... ............ 81 Links to Economy ................. . ....... 81 C. Estimation of the Growth Rate . .... 82 Export Projections ........ . ....... 82   Tables Table No. Title Page No. 1 Evolution of the percentage share of value added by sector from 1959-75. 2 2 Compositi:on of value added in the industrial sector by branch (1974) ....................................... 4 3 Distribut:ion of exports by country - 1972 ............... 5 4 Distribut:ion of Ivorian exports by country - 1959-74 .... 6 5 Distribut:ion of principal exports by country within Europe - 1972 .6 6 Distribution of principal exports by country within the CEA0 ............. ...................................... 6 7 Reform of import taxes ................................... 21 8 Reform of consumer taxes ................................. 21 9 Incentives to domestic and export sales .... .............. 30 10 Exports from Senegal 1975 . .............................. 42 11 Cost of export subsidies by principal class of export 42 12 Incentive and resource cost measures 1972-74 .... .......... 49 13 Sample wide estimates of the resource cost of foreign exchange ....................................... 48 14 Domestic resource cost of foreign exchange by subsector: exporting vs. import substitution .................... 50 15 Characteristics of phosphate rock deposits in Senegal .... 55 16 Taiba's cost structure ...................... 55 17 Forecasts of world prices for Moroccan phosphates .... .... 56 18 Variable ,0osts for selected principal producers .......... 56 19 Domestic resource cost of foreign exchange in fishing .... 59 :20 Resource cost of foreign exchange in textiles .... ........ 61 21 Resource cost of foreign exchange in the textile sector .. 62  Tables (continued) Table No. Title Page No. 22 Resource cost of foreign exchange in textiles by activity 65 23 Effective rates of protection by markets . .67 24 Groundnut oil milling capacity .72 25 Groundnut production and marketing in Senegal (1956-1975) 74 26 Capacity utilization in groundnut transformation (1971-1976) 75 27 Percentage refining margins (cif Europe) for groundnut products as a function of the rate of extraction .76 28 Resource cost of foreign exchange in tomato concentrate 79 29 Estimates of growth rates in the industrial sector excluding groundnuts and mining .83  SUMMARY, CONCLUSIONS AND RECOMMENDATIONS i. This report analyzes in depth the problems of industrial development in Senegal over the next decade or so. Its principal conclusions, described in this s;ummary, are: (a) Protection policy and tax policy should be revised to correct some major defects, particularly strong negative protection to some industries which make capital goods, and excessive positive protection to other Lndustries; 1/ (b) Similarly positive measures should be undertaken to promote exports, both inside and outside the Free Trade Zone. Particular attention should be given to the use of direct export subsidies, either on a global scale or for selective industries; 1/ (c) Resources should be allocated only to industries which are economically profitable; and (d) In terms of industries to promote, emphasis should be put on: (i) labor-intensive rather than capital-intensive industries; (ii) principally but not exclusively to the develop- ment of exports. In the medium term, the local market is less likely to generate growth than exports; and (iii) not neglecting existing industries--these are more likely to be able to grow rapidly than new industries. The report is organized as follows. Chapter I reviews recent progress, and gives a summary of the Government's plans for the next decade and a half. Chapter II reviews Government policy instruments and the structure of incen- tives, and Chapter III analyzes in detail a sample of activities, the Fifth Plan and the Free Trade Zone. This report draws upon and extends a number of preceding reports of the Bank 2/, as well as Government documents (notably the Fifth Plan) and a number of working papers prepared in the context of the present ongoing tax reform. 1/ -The 1979 tax reEorm did not remedy the underprotection of industries producing capital goods or producing for export markets. 2/ Current Status and Prospects of the Industrial Sector in Senegal (1974, unpublished).  - ii - ii. The principal finding is that Senegal, like any other country, has some industrial activities which are economically profitable and some which are not. The problem for future development is how to increase the overall economic profitability of the sector as its size increases. Such a strategy would have to take into account policies toward existing industry and those toward new industry both inside and outside the Free Trade Zone. The best kind of policy should: (a) limit investments in activities which are economi- cally unprofitable by not renewing the capital installations of the enterprises involved; (b) increase investments in existing profitable activi- ties so as to increase their share of total output; and (c) encourage new investments only in activities which are economically profitable. An immediate implication of this is that the Government should create a unit within the administration to continually evaluate the profitability of existing activities, and to properly analyze new projects. Such a unit would have to be more than a simple project evaluation unit; its members would not only have to deal with issues of short-run strategy but also with questions of the effects of price and tax incentives on industrial development. iii. The following question of strategy therefore arises: Should future growth of industry be oriented towards the local or export market? An ancillary question is also important: What kind of industry should be promoted: labor intensive or capital intensive? Market Orientation: Export or Local Market iv. The debate as to whether growth should be based upon the local mar- ket or on exports is a long standing and heated one. However, provided that a particular activity is economically profitable for a country, rather than just financially profitable for shareholders, it does not matter whether this activity is import-substituting or export-oriented. Market orientation of growth is important, nevertheless, because market size can be a significant contributing element to economic profitability. In many West African countries, the local market is small; thus, the construction of plants, small enough to serve just the local market profitably in economic terms is often very difficult, if not impossible. The realization of economies of scale (and the associated increase in economic profitability) is then only possible if firms have access to, and are able to compete in, export markets. The implantation of "large firms", unable and/or unwilling to export, thus condemns them to low rates of capacity utilization (and thus  - iii - high average fixed costs) which in turn drives up prices in the local market. Thus, the welfare of the local consumer is linked tco the ability of the firm to export. v. Not only can exporting increase the profitability of a particular activity, but an export-oriented strategy can also lead to a substantially higher growth rate than one based on the growth of a small local market, as is brought out by the very rapid economic growth of some Asian coun- tries. For example, South Korea (population 36 million) had a GNP per capita of $90/head In 1961, but in 1976, this had grown to $260 expressed in 1961 prices and exchange rates as a result of export-led growth: exports were $41 miLlion in 1961, and $7.8 billion in 1976 (a mere $3.5 billion in 1961 prices). Senegal's exports were $140 million in 1961, and $429 million in 1975 (only $193 million in constant 1961 prices); its GNP per capita was $190 in 1961, and in 1976, $180 in constant 1961 prices and exchange rates. In other words, the conventional argument according to which growth should be based primarily on the local market, and only marginally on exporl:s is inappropriate for Senegal. First, per capita income is low,' and the local market small and destined to grow slowly--since it depends on the rate of growth of the agricultural sector. Thus, in the next decade there is likely to be little scope for new investment in import substitution. The consequences of "forcing" import-substitution growth would be serious; the Government would be obliged to heavily protect new firms, resulting in a high cost structure, which would in itself reduce the chance that such firms would be able to export. On the other hand, Senegal does have access to two regiontal markets--the CEAO and the ECOWAS, 1/ which are both larger than the local market. But most important, Senegal's exports have guaranteed duty free! entry into the European Common Market (EEC), under the terms of the Lome Convention. Furthermore, exports from non-ACP 2/ devel- oping countries are subject to quota as well as tariff restrictions at EEC frontiers. Next, the United States has started to ease restrictions on imports from less developed countries (LDCs) with the introduction of generalized preferernces. Finally, Senegal is geographically well placed to exploit export markets in the United States, Europe, the whole West Coast of Africa and Latin America. In short, large markets are available. vi. This implies that export orientation can be justified not only per se, but also in terms of a phasing argument that further import substitution will only be economically profitable when there has been considerable growth in purchasing power in the local market. The best way to raise the growth rate and real income in the next ten to fifteen years will be to aim for an export-led growth, with the local market being accorded a less important, but nonetheless significant role. 1/ CEAO - Communaute Economique de l'Afrique de l'Ouest. Membership includes the Ivory Coast, Niger, Upper Volta, Mali and Senegal. Togo and Benin are observers. ECOWAS - Economic Community of West African States. Membership includes Benin, Cap Vert, The Gambia, Ghana, Guinea, Guinea Bissau, Ivory Coast, Liberia, Mali, Maurit:ania, Niger, Nigeria, Senegal, Sierra Leone, Togo, and Upper Volta. 2/ ACP countries are the cosignatory developing countries of the Lome Convention (African, Caribbean and Asian and Pacific courntries).  - iv - Weight to Accord to Each Export Market vii. The success of export-oriented growth depends initially on unrestricted access to foreign markets, and subsequently on the non-intro- duction by these markets of both tariff and non-tariff barriers as the volume of Senegalese exports rises (Japan, for example). The Government should therefore evaluate carefully the existence and/or the likelihood of trade restrictions in each of the markets it proposes to penetrate, and seek to increase exports more rapidly to countries least likely to restrict trade. viii. Under the terms of the Lome Convention, the EEC agrees to import manufactured goods duty free from associated LDCs for the duration of the agreement and, moreover, the Community, in principle, cannot restrict trade by the use of quotas and/or non-tariff barriers to trade. Thus, it may be argued that Senegal faces an infinitely elastic demand curve within the EEC, at least during the life of the Lome Convention. Moreover, imports into the EEC from non-associated LDCs must pay duties, and are in addition sometimes subject to quota restrictions. The main source of competition for Senegal will be other LDCs, also signatory to the Lome Convention. Entry into the United States market is not as easy for Senegalese products which could be exported in the short run, particularly textiles, since they would be subject to import duties. If Senegalese prices are competitive, however, the US demand-elasticity for their products would probably be very high, for example in printed and dyed shirts (both knit and cotton). Senegal has in the past exported to the United States, only to lose the market because of the depreciation of the dollar. ix. Turning now to the CEAO, the treaty emphasizes the need to reduce trade restrictions, particularly the quotas and bans which were in evidence under the UDEAO 1/ arrangements. Available evidence indicates, however, that progress is slow on this point, primarily because the indus- trial structures of member states are so similar. This is particularly true, and also important, for Senegal and the Ivory Coast. The small size of the local market (in the importing country) means that the threat of economic damage to the local industry, resulting from imports, is much greater than in the case of European countries for example. In turn, given the magnitude of the unemployment problem in other CEAO countries, the threat of economic damage to local industry makes it very easy poli- tically to justify protective quantitative restrictions even against CEAO members. Since the inception of the CEAO, barriers in some countries have increased rather than decreased. For these reasons, it is argued that the elasticity of demand is lower, and the threat of restrictions greater, in the CEAO than in Europe or the USA. x. There is, however, a second problem within the CEAO: preferential trading arrangements with respect to intra-CEAO exports have changed radi- cally. Under the UDEAO arrangements, goods manufactured within the UDEAO were subject to a fixed duty rate of 50 percent of the total duty applied 1/ Union Douaniere des Etats de l'Afrique de l'Ouest.  by the importing country to similar goods imported from the EEC.. This was an automatic preference with the cost borne by the importing country. Under the CEAO arrangements, however, the preference is no longer automatic and the duty rate is subject to negotiations on a goods-by-goods bilateral basis between Senegal and each of its trading partners within the CEAO. 1/ Furthermore, the exporting country (Senegal) must compensate the importing country via payment through the CEAO development fund of the difference between the duty negotiated for Senegalese exports and the duty the im- porting country applies to the same goods imported from the EEC., This system could make exporting locally manufactured goods within the CEAO substantially more difficult than under the UDEAO. xi. The ECOWAS is potentially a much larger market than the CEAO since Nigeria is a member. Very little can be said about trading pros- pects, save that the possibility of protective measures is quite high as 'well within this organization. It seems unlikely that Senegal's exports will grow very fast in the near future within the ECOWAS, or within other African markets. Markets within Latin America are of different sizes, and almost totally unpenetrated by Senegal. In the short run, exports to this continent are unlikely to grow to significant levels, but they should not be neglected. Many of the protective tendencies likely to prevail within the CEAO are also likeLy to be in evidence in Latin America. xii. In summary, this evidence suggests an export-oriented strategy, based primarily on the exploitation of the EEC market under the terms of the Lome Convention. The Lome Convention is protective of Senegal's exports, which would not then be submitted fully to the rigours of unpro- t:ected competition in the world market. Furthermore, the EEC cannot restrict trade by raising either tariff or non-tariff barriers to trade for any length of time as far as ACP countries are concerned, and it is most unlikely that future agreements would drop this clause. Regional markets should have an important second place. Senegal should insist that its partners within the CEAO conform to the letter of the Treaty, namely that quantitative restrictions of intra-regional trade of local production should be abandoned. 2/ Consideration should be given to trading agree- ments with major African trading partners--Nigeria for example, from which Senegal imports some of its crude oil. Other markets in developed or less d'eveloped countries should have tertiary importance, but should not be neglected, particularly if they are large. Finally, the local market should be allowed to grow naturally, reflecting the growth of agriculture, population and industrial exports. xiii. Available evidence indicates that the Government realizes the importance of export development. Nevertheless, within certain circles in 1/ In the absence of the negotiation of the preferential rate, the importing country applies automatically the EEC rate. 2/ In particular, the Ivory Coast should be asked to lift its restric- tions on imports of Senegalese knitwear.  - vi - Senegal, the argument is still made that development must be auto-centered and dependent on the local markets. Prolonged import substitution would lead to the creation of firms which must be excessively subsidized or protected by high tariffs, which either reduces public savings or drives up the cost of production; aspects representing basic constraints on the growth of the economy. While the local market should not be neglected, it is clear that in the medium term, export markets offer a much greater scope for product diversification (which is risk-reducing) and growth than the local market. Moreover, export-led growth will itself raise the rate of growth of local demand faster than import-substitution growth, particu- larly for presently imported semifinished goods. Choice of Technique xiv. Since labor is abundant, it is obviously preferable that industry should be as labor-intensive as possible, provided that this does not erode economic profitability. However, at the margin, relatively capital- intensive projects should not be excluded, provided they are economically profitable. All in all, one would expect that resources would be allocated principally to light rather than heavy industry if proper economic criteria for their selection were to be used. Moreover, the choice of technique should not be distorted by tax and interest rate policies, (in particular credit rationing by size of firm), which discriminate against small, relatively labor-intensive industries. Obstacles to the Development of Exports xv. The principal problems are the following: (a) supply side problems, such as the price of other locally produced inputs (electricity, packaging, services), and transport problems; (b) the productivity of labor (c) the competitivity of Senegalese exports--i.e., prices; and (d) a structure of incentives which gives a larger producer margin in the local market than in export markets. xvi. Electricity pricing deserves more study (given its cost) since there appears to be areas in which it could be rationalized. First, there is the issue of whether or not the cost of generation reflects the economic cost of production, since there seems to be a substantial burden on production in the form of import duties, special property taxes, etc. Second, there is the issue of whether or not consumption prices are set properly between different classes of users. In January 1977, prices were raised to industrial users outside the Free Trade Zone as part of a policy designed to lower prices to consumers (without causing a revenue loss to the electricity firm). This does not seem wise in a country which has to generate electricity from high cost imported fuel, and which hopes to develop industrial exports. Proper pricing of electricity is very  - vii - important, and subsidies should be used only as an absolute last resort, Since underpricing can easily result in wasteful and very costly consump- tion. In particular, wasteful over-consumption by households should be avoided by using a sharply escalated rate structure for such luxuries as air-conditioners and refrigerators. xvii. Packaging is also a very important input into the development of industrial exports. The local production of both cans and cartons is currently undertaken by a single firm for each, which sells to exporters in "admission temporaire" (producers do not pay import duties on tin plate and cardboard used in production sold to exporters). Despite this, however, the prices of cans and cartons are often above the border price of similar goods imported from Europe. Because both firms are able to control imports (via "autorisations prealables"), users (i.e., exporters) are often obliged to pay more than they need to, so export becomes less competitive. The firm's control over competing imports should be reduced by the elimination of "autorisations prealables" certainly for export orders. This could certainly be done over a trial period, to determine whether protection against imports is neeced. xviii. Banking is ar,other area in which reform is required. There is little effective competition between banks either in rates or in the prices cr quality of services. Several objections may be made to existing policy. First, there is discrimination by the Central Bank between borrowers by nationality and by size of firm. Private banks have reinforced this dis- crimination by size of firm in their interbank rate-setting agreement. This has resulted in medium-term capital (5-7 years) being offered at 15.25 percent to small economically profitable export-oriented firms. Second, the spread between lending and borrowing rates is too large, and maintained by the interbank protocol. Third, the lower risk associated with import-substitu- tion projects (because competing imports can be controlled) implies that existing credit policy effectively discriminates against exports0 Fourth, more progress is needed on improving banking relations between Anglophone and Francophone Africa. A thorough study is needed of the banking system and of bank policy in the framework of export promotion. xix. International Transport is another area which requires thorough study. Of particular importance are the high conference rates along the West Africa coast, the infrequence of boats plying between Dakar and the Americas, and the cost of air freight, which is crucial for the develop- ment of agro-industry. The Productivity of Labor, the Quality of the Labor Force, etc. xx. This subject Ls dealt with at greater length in Volume IV on Human Resources; only brief points will be made here. The minimum wage is now 107 CFAF/hour excluding fringe benefits 1/. The real cost of labor/hour 1/ This comment should not be interpreted as suggesting that economically productive high cost personnel, in and of themselves, are a hindrance to exports.  - viii - is substantially above this for two reasons: first, very few personnel are in the lowest category (there are seven categories in all), so even unskilled personnel are paid above the minimum wage; second, wage differentials in higher categories are large and, it may be argued, they do not always reflect higher productivity. This means that wage increases must be kept in line with productivity increases; for otherwise the price of exports will rise, making them less competitive, and thus worsening the country's long-run growth prospects. The Government should shift from the policy to maintain urban purchasing power to a productivity-oriented strategy. Failure in this respect will make it very difficult to sustain, in the long run, the real level of consumption at its existing level. Increases in productivity should not, of course, be limited to blue collar personnel; white collar personnel of all nationalities need also to increase their productivity in order to justify their present high earnings. Related to this issue, one of the most interesting findings of the present report was that many of the economically profitable firms (particularly the smaller ones) were export-oriented, and that they employed a significantly smaller percentage of high wage personnel than larger, more privileged firms, oriented principally towards the local market. In the context of export development, productivity, wages and wage differentials will be a very important issue. The Competitivity of Exports and the Structure of Incentives xxi. The price competitivity of exports depends on several factors. First, it has already been pointed out that certain markets outside the franc zone (among them the USA) have been lost because of the appre- ciation of the French franc over the period 1971-1975 (in 1971 the dollar was worth 277 CFAF, in 1975 only 225). Exports to the USA are marginal, but for example, only about a third of phosphate exports goes to franc zone countries, so the appreciation of the French franc was quite costly (at constant dollar prices). Since Senegal does not and cannot control its exchange rate with respect to currencies outside the franc zone, exporters sometimes have to accept large revenue losses as a result of exchange rate fluctuations. The second influence on the competitivity of exports is obviously the local cost of factors and other locally produced inputs (as already described) which compound the problems of exporting when the franc appreciates. Third, within the franc zone, the exchange rate of the CFAF to the French franc is fixed and thus linked to other EEC currencies via the "snake". Thus the competitivity of exports depends on the relative rate of cost inflation in Senegal and in Europe. In 1975 and 1976, exporters reported that even within the franc zone, Senegalese exports were less competitive than before. In recent years, this situation has again improved somewhat by stiff wage and price controls. However, the local market has been strongly protected by quotas and tariffs with the result that gross producer margins in local markets are greater than in export markets, so that regardless of whether or not exports are profit- able (price greater than cost), they are less profitable than selling to the local market.  - ix - xxii. Chapter II of this report analyzes the structure of incentives, giving particular attenation to the problem of export development both in and outside the Industrial Free Trade Zone. The following are its principal findings. First, existing incentives substantially favor production for the local market over exports. Moreover, on the local market there is also a bias in favor of production of finished consumer goods, and against semi- finished goods and machinery. In other words, there is discrimination both against exports, and also against "vertically integrated development" of the local market, where this is to be interpreted as simultaneous clevelopment of raw materials, capital goods and finished goods--as stated by the Government in its Fifth Plan. There are many unexploited possi- bilities, and the development of many local industries is impeded by current incentives--particularLy in shipbuilding and the machinery industries. The analysis goes on to point out the need for a complete reform of the structure cf incentives 1/ which would consist of two principal elements: (a) the relative equalization of import duties on raw materials and finished goods at rates lower than those currently applied to finished goods--all to eliminate the bias in the local market against production of raw materials and semifinished goods; (b) the use of export premiums to offset the bias against exports. Arguments for and against export premiums are considered, and it is con- cluded that their use by Senegal would not meet with objections from the international community. 2/ Other third world countries use them and there seems no reason why Senegal should not. The budgetary question is, moreover, not likely to pose a serious problem, if compensatory fiscal measures are undertaken. xKiii. The existing investment code has a number of defects of which the most glaring is the proposal to give a ten-year exemption from import duties on imported raw materials, regardless of whether the firm is installed inside or outside of the Cap Vert region. Under the old law, this exemption from duties was not total, and moreover it was given only for 5 years in the Cap Vert region and 8 years outside. This reform is both costly for the Government and directly contradictory to the Government's own desire to encourage decentralizatLon, and it also can sometimes promote unfair competi- tion between Common Law firms and priority firms. Second, for certain firms there is a strong case :or the use of direct subsidies to value added instead of duty exemptions as the special incentive of the Investment Code. At equal budgetary cost, such subsidies have the advantage of encouraging more job creation than duty exemptions. Finally, subsidies also can give a better incentive to buy raw mat:erials locally, so that there is a bigger incentive to create vertically integrated industries. 11' The 1979 fiscal reform was only partly successful in this respect. 2/ It is, however, important not to give pressure groups within the EEC easy arguments for protection from imports of Lome countries; special attention should therefore be given to the presentation of such measures.  x xxiv. Concerning export subsidies, two approaches are suggested: the first of which would be global and based on the implementation of a complete reform of the incentive system. The second approach, which is less ambitious, supposes that the reform of the incentive system does not proceed very quickly; and suggests the selective use of export subsidies, for sectors which have the most export potential. The net budgetary cost of such a scheme could be zero, by reducing subsidies to the local market by one billion CFAF, and transferring the subsidy to exports. xxxv. This brings us to the questions analyzed in Chapter III--on which industries should the Government concentrate its development effort. The chapter contains an analysis of the economic profitability of existing activi- ties, the principal ideas around which the Fifth Plan is built as well as its specific projects, and finally the Industrial Free Trade Zone. Leaving aside activities of secondary importance, economically profitable activities are phosphates, fishing and fish processing, cotton yarn, finished clothes, knitwear, cement, truck assembly and agricultural machinery. Further invest- ment in these activities which are already, for the most part, export-oriented, would make a positive contribution to the economy. Care should be taken before investing further in printing and dyeing of grey cloth, vegetables, fertilizer and groundnut processing. For groundnut processing, existing capacity is quite sufficient to process the local crop. For the other activi- ties, observed average capacity utilization was low so that new investment is unnecessary at present. Finally, extreme caution should be taken before investing in new capacity for the local manufacture of grey cloth, and in the spinning and weaving of finished cloth--and certainly not before understanding the reasons for the present economic unprofitability of these activities. xxxvi. The second part of the chapter analyzes in detail the various projects of the Fifth Plan, passing by a detailed analysis of the sectors in which they are incorporated. In the textile sector, the principal project is an integrated knitwear plant at Kaolack; despite some promising features, the project has not advanced very much, mainly because of uncer- tainty about possibilities for exports. Nevertheless, this is the kind of project which Senegal should aim to reproduce, because of the positive employment effects that can be expected from it. It is also shown that some existing firms in this sector have considerable export potential and should not be neglected. This is particularly true for knitwear and finished clothes, for which export development is an absolute necessity. The Government's decision to allow significant imports of used clothes is posing some of these firms serious problems in the form of loss of market share within Senegal. Exports are then essential to survival, and could be effectively stimulated by the use of direct export premiums. Fishing is an activity which is deeply under-exploited. Cement is another activity which could be profitably exploited if it can be shown that local demand is likely to rise by enough to justify a new plant rather than an extension of existing capacity. Turning to groundnut oil production, existing capacity is shown to be easily large enough for historical average crop sizes, and that investments in new additional capacity is likely to reduce the rate of capacity utilization to unnecessarily low levels. Any incidental surplus over the one million ton capacity should be exported as was done in 1976.  - xi - The recent Government decision to offset the planned new capacity in Diourbel by closing down an obsolete plant in Dakar should therefore be welcomed. Finally, the prospects for the non-groundnut food industry are reviewed, paying partizular attention to the transfer prices set by Govern- ment for agricultural 2ommodities sold to the industrial sector. Exces- sively high producer prices can too easily increase costs of industrial sector firms to the poLnt that they are totally uncompetitive with imports. The result is high protection and ultimately excessively high prices for the consumer. xxvii. Following this, attention is given to the development of the Endustrial Free Trade tone. The prospects for the Zone are clearly less irhan they would be at a lower exchange rate. The exis,ting high exchange rate also reduces subsitantially the backward linkage effects, because firms :Ln the Zone are allowed to buy all imports duty free, they prefer cheaper :Lmports to more expensive local production. This outcome is clearly unfortunate, but unavoidable since forcing firms to buy locally would reduce, perhaps even eLiminate, their financial profitability. xxviii. Finally, on the basis of the preceding analysis, an estimate of the industrial growth rate is made and is projected at 6.8 percent per year over the period until 1985.   Chapter I. BACKGROUND A. The Structure of the Industrial Sector Blasic composition of the sector ].01 Industrial investment in Senegal can be usefully divided into four time periods -- pre 1950, 1950 to 1960, 1960 to 1970 and post 1970. Before 1950, most investment had occurred in groundnut oil transformation, ship- repairing, and in one or two light consumer industries. In the 1950s there was substantial new investment with the construction of two flour mills (capacity 125,000 metric tons), three packaging industries, several textile mills (total capacity 2,600 tons of cotton), one calcium phosphate mine (reserves 70 million tons), and one aluminum phosphate mine. Whereas the phosphates were to be exported primarily to Europe, the other investments had been designed on a scale to serve the colonial French West African market (total population 20,0C'0,000). With independence in 1960 and the dissolution of the colonial customs union 1/, these industries lost a substantial part of their market, as some cther former colonies (notably the Ivory Coast) con- structed their own industrial plants. This led to low capacity utilization which lasted through the 1960s and persisted into the early 1970s. 1.02 Nonetheless, there was some new investment in the 1960s, though on a scale adapted to the local market, reflecting the reality that the new customs union was not yet sufficiently mature to permit rational planning of investments on a regional basis (with each nation wanting its owrn indus- trial plants). In the textile sector, investment was in finishecd clothes and underwear, 2/ whereas in the 1950s, it had been in the spinning, weaving and dyeing of grey cloth based on imported raw cotton. Following the example of many other less developed countries, Senegal constructed a small petroleum refinery (capacity 600,000 tons) and a small truck assembly plant. In the latter half of the decade, a fertilizer mixing plant was set up and local processing of fish was encouraged. A second integrated spinning and weaving pLant was commissioned (capacity 1,200 tons of cotton) which came on stream in 1970. 3/ 1/ The Ivory Coast, NLger, Upper Volta, Mali, Benin and Senegal. It was replaced by l'lJnion Douaniere des Etats de l'Afrique de l'Ouest (UDEAO), with exactly the same membership. 2/ These investments were undertaken in conjunction with French trading houses in order to benefit from their marketing expertise and contacts abroad. 3V In the latter part of the 1960s, the-Government also promoted the local growing of cotton for the local textile industry as well as for direct export.  1.03 The main new investments in the 1970s were those associated with the Government's decision to process locally all of the country's groundnut crop (total refining capacity was increased to one million tons), a sugar mill at Richard Toll, and an export-oriented agro-industry in the Cap Vert region. In the mid 1970s, there were attempts to build a tanker repair plant at Dakar which led to the purchase of a 60,000 ton floating dock in 1979. There are currently some five firms which are or will shortly be producing in the Free Trade Zone. Thus recent investment has been primarily oriented towards export. 1.04 The result of all these investments has been that between 1959 and 1975, industry grew at a compound rate of 5.7 percent compared to 2.7 percent for the whole economy and 3.9 percent for the agricultural sector. The performance of industry and agriculture have been quite different. The whole economy grew steadily from 1959 to 1968 (at an annual rate of 3.4 percent), but that from 1968 on, there have been fluctuations of increasing amplitude for which the agricultural sector is clearly responsible. Industry continued to grow steadily however, until 1973 when the groundnut processing industry in particular was severely afflicted by the drought which struck the whole of West Africa. However, in recent years industrial growth has come back to its original trend despite the severe increase in the cost of crude oil in 1974, and the problems which this created. 1.05 Table I shows the increasing importance of industry in the Senegalese economy. Industry now accounts for about 18.5 percent of value added compared to 11 percent in 1959. This increase has occurred not at the expense of the agricultural sector, which has maintained a share of roughly 23 percent, but rather at the expense of private services and public administration. Commerce--with about a 20 percent share in GDP--is still the second most important sector in the economy, but it is likely that it will soon lose its place to industry. Table 1: EVOLUTION OF THE PERCENTAGE SHARE OF VALUE ADDED BY SECTOR 1959 1964 1969 1973 1974 1975 Agriculture 23.7 24.0 23.4 22.6 23.8 30.2 Industry 11.3 13.6 17.0 16.1 19.8 18.4 Construction 5.5 4.5 3.5 3.6 4.0 4.4 Handicraft 1.2 1.3 1.3 1.3 1.2 1.2 Services 58.3 56.6 54.8 56.4 48.8 45.8 Total 100.00 100.00 100.00 100.00 100.00 100.00  - 3 - 1.06 Table 2 gives the decomposition of Senegal's industrial output by branch and it may be seen that 40 percent of industrial value added is explained by three products -- minerals, canned fish, and groundnut products. These products also account for 70 percent of all exports (see Table 3). Moreover, the country's exports are concentrated in a very few markets -- Table 3 shows that 67 percent of all exports go to the EEC (and of this 59 percent to France alone). Of the 21.5 percent going to Africa, 15.57 percent go to countries within the CEAO (and in value terms 7.5 billion of a total 8.47 billion CFAF go to three countries -- 2.75 to the Ivory Coast, 2.17 ;o Mali and 2.57 to Mauritania). 1.07 Thus, Senegal has four main trading partners -- France, Mali, Mauritania and the Ivory Coast. This situation is in marked contrast t:o the Ivory Coast which exports only 26 percent of its exports to France and 36 percent to other non EEC European countries (see Table 4). While no one would deny that Senegal has fewer natural resources than the Ivory Coast, it is evident that Senegal has not diversified its export markets. An additional striking feature of Senegal's exports is their geographical distribution by product. Tables 5 and 6 illustrate this well: Senegal exports its primary products (processed or unprocessed) to Europe and its manufactured products to its partners within the CEAO. Shoes and textiles are exported to both areas, but apart from these products, there is very little overlap between the kinds of products which go to the different markets. We shall return to this point in our discussion of industrial strategy. 1.08 This geographical concentration of exports is not of course t,he result purely of Senegal's concentration in production, but also of the tactical choice to maintain an extremely close, even special relationship with France (which has ]1ow emerged into a close economic relationship with the EEC). At the same time, within Africa, the Government has preserved, elven developed, its economic links with other former colonies notably the Ivory Coast. This tactical choice is well reflected by the nature of the country's trading agreements. Trading relations with the European Common Market 1.09 From 1960 to 1975, the Yaounde conventions governed trade rela- tions with the EEC. Under the terms of these conventions industrial goods from Senegal were admitted into EEC countries duty free, and the EEC exports enjoyed preferential entry into Senegal. In practice, this meant that goods from the EEC were exonerated from the customs duty, whereas non EEC goods were not. All imports were subjected to fiscal duties however, over and above the customs duty. In 1975, the Lome Convention replaced the Yaounde conventions. The principle of guaranteed reverse preferences was abandoned and the LDCs signatory to the convention now have the option, to re-establish duties on EEC exports 1/. On the other hand, exports from 1/ In 1979, with the tax reform, Senegal eliminated the EEC preference in import tariffs; 3ome protection still exists through existing quota on imports from non--EEC countries.  Table 2: COMPOSITION OF VALUE ADDED IN THE INDUSTRIAL SECTOR BY BRANCH (1974) (in millions of CFAF) value added - indirect value added % of total value added % of total market + subsidies taxes factor cost value added in constant value added prices 1971 CFAF Mining 20,991 - - 607 20,384 33.87 4,900 11.77 Food 18,815 5,775 6,083 i8,507 (30.76) 17 567 42.22 8% Canned Fish (2,676) - - ( 335) (2,341-0) (3.89) (2,220 (5.33) Ground Nut Oil (9,408) (3,450) (2,517) (10,3410) (17.18) (9,820) (23.60) 5% Grain & Flour (1,590) ( 325) ( 326) (1,590) (2.64) (1,510) (3.62) 12% Sugar ( 339) (2,000) ( 450) (1,889) (3-13) (1,790) (4.30) 5% Other ( 772) _ _ ( 130) ( 667) (1-10) ( 630) (1.51) 3% Beverages (1,649) - - ( 810) ( 839) (1-39) ( 79 0) (1.89) 3% Tobacco & Matches (2,381) - - (1,515) ( 866) (1.43) ( 82 0) (1.97) Textiles 5% 7,098 - _ 1,956 5,142 8.54 4,880 11.73 Wood 708 - - 130 578 0.96 550 1.32 5% Paper Products 1,533 - - 384 1,149 1.90 1,090 2.62 Chemicals 5,541 1,806 - - ,627 11.01 6,290 15.12 Non-metallic Minerals 1,418 - - 450 686 1.14 650 1.56 Fabricated Metal Objects 3,143 - - 732 2,411 4.0 2,290 5.50 Utilities 5,097 - - 714 4,383 7,78 3410 7.45 Total 64,344 6,861 11,056 60,150 100.00 41,600 100.00 Note 1: There may be small rounding errors Note 4: Col 6 = Col 4 deflated as in Table 2,11 of Note 2: Col 4 = Col 1 + Col 2 - Col 3 Appendix Statistical Note 3: Col 5 = share of each brance in Note 5: Col 7 = share of each branch in value added value added at factor costs. in 1971 FCFA  - 5 - Table 3: DISTRIBUTION OF EXPORTS BY COUNTRY - 1972 (in millions of current CFAF) Value _ EEC 36,464 67.010 France 31,665 58.190 Belgium 520 0.095 Netherlands 2,277 4.180 Italy 2,002 3.670 Europe Non Common Market 3,689 6.780 England 1,032 1.896 Norway 574 1.055 Spain 486 0.896 Greece 454 0.830 Switzerland 369 0.367 Rest 774 1.420 North America 281 0.52 U.S.A. 246 0.45 Canada 35 0.06 Africa 11,690 21.480 Lebanon 232 0.426 UDEAO 8,472 15.570 UDEAC 793 1.457 Togo 179 0.329 Madagascar 362 0.665 Morocco 134 0.246 Anglophone Africa 841 1.540 Other 677 1.244 Latin America 4 Eastern Europe 229 0.42 Australasia and Japan 771 1.41 Provisions de Board + (errors = 0,009) 1,284 2.38 Total 54,412 100.00 Note: We have chosen 1972 for the following reasons. First, the period 1973-75 was very unrepresentative of production because of the drought and the high phosphate rock prices of 1974-75. Second, dal:a for 1976 are not yet available.  Table 4: DISTRIBUTION OF IVORIAN EXPORTS BY COUNTRY (1959-74) (in percent) 1960 1965 1970 1974 % change France 51 38 33 26 -24.4 Other EEC 15 23 30 36 21.6 USA 14 16 19 7 - 7.3 Other DME's 4 9 9 12 7.9 UDEAO 6 4 4 7 1.0 Other Fr. countries 5 5 2 5 - 0.4 Other countries 5 5 3 7 1.6 Total 100 100 100 100 Source: Customs statistics of the Ivory Coast Table 5: DISTRIBUTION OF PRINCIPAL EXPORTS BY COUNTRY WITHIN EUROPE 1972 (Millions of current CFA francs) Groundnuts Phosphates Fish Cotton Other Total France 23,200 1,500 2,350 259 4,256 31,665 Belgium 313 - - - 207 520 Netherlands 1,310 780 - - 187 2,277 Italy 1,267 - 16 242 477 2,002 England - 770 - - 262 1,032 Norway 569 - - - 5 574 Spain 117 - 45 189 135 486 Greece 417 - - 37 454 Switzerland - - - - 369 369 Other 319 277 - - 178 774 Total 27,195 3,744 2,411 690 6,113 40,153 Percentage 67.73 9.32 6.00 1.72 15.23 100% Table 6: - DISTRIBUTION OF PRINCIPAL EXPORTS BY COUNTRY WITHIN THE CEAO (Millions of current CFA francs) Groundnuts Textile Phosphates Petrol Cement Other Total Ivory Coast - 794 (170) - 19 177 2,753 Mali - 361 (407) 330 87 982 2,167 Mauritania 284 116 - 656 222 1,312 2,590 Others NOT AVAILABLE 962 962 Total 284 1,271 (577 986 328 5,026 8,472 Percentage 3.35 15.0 2.01 11.64 3.87 64.13 100%  - 7 - participating third world countries are still admitted duty free to the EEC. However, under Lome, Senegal's exports to the greatly enlarged EEC are subject to much stronger competition from other less developed countries because of the sharply increased number (46) which enjoy duty free entry into Europe. These countries are in Africa, the Caribbean and the Pacific. Asian countries are excluded and moreover, certain of their exports (notably textiles) are subject to quota restrictions in addition to tariffs. At the present point it seems that Senegal is not going to entirely eliminate preferential entry of Common Market goods. This point is discussed in greater detail in Chapter II. Relations with former French colonies in West Africa 1.10 From 1959 to 1973, Senegal was a member of customs unions which successively replaced the colonial unions. The first l'Union Douaniere de l'Afrique Occidentale (UDAO), was formed in 1959, to assure a fair distribution of customs revenues between the coastal and the interior countries: the former were supposed to rebate to the latter a certain proportion of import duties. However, this led to serious difficulties and in 1966, the UDAO was replaced by the l'Union Douaniere des Etats de l'Afrique de l'Ouest (UDEAO) which consisted of the same countries. 1/ The UDEAO was supposed to overcome the shortcomings of the UDAO. A common external tariff was to be established but goods produced by one member of the Union could be imported into another member country at 50 percent of the duties charged on imports from the EEC. Senegal has interpreted this to mean 50 percent of all duties plus sales taxes. Intra-Union re-exports of goods not produced within the Union continued to be taxed at the full EEC rate. l.11 Senegal also concluded a bilateral trading agreement with ihe Ivory Coast according to which some Ivorian goods are imported into Senegal duty free, although they are subject to the sales tax. In principle, l:his agreement covered all exports of domestically produced raw materials and goods manufactured exclusively with raw materials. 2/ In practice this agreement proved extremely difficult to implement. Thus in 1971, when it came up for renewal, the two countries took a more modest approach. A list of products was fixed for which the reduced level of taxation would apply. Eligible Ivorian exports include pineapple conserves, biscuits, noodles, fourteen textile artic:Les and five wood articles. Such agreements have also been concluded wil:h Mali and Benin but on a much smaller scale. L.12 In 1976, the UDEAO was replaced by the Communaute Economique de ]'Afrique de l'Ouest (CEAO). 3/ The principle of a common external tariff was maintained, but it has not yet been enacted. Intraregional trade was however accorded a radilcally different treatment. Under the new system, e!ach country may negotiate a trade agreement with each importer of its 1/ The Ivory Coast, bliger, Upper Volta, Mali, Benin and Senegal. 2/ Imported packaging material is however, permitted. 3/ Togo and Benin have observer status in this organization.  - 8 - goods, which sets the preferential import duty to be applied to each item, but if no trade agreement is negotiated, the goods are subject to the same import duty as goods imported from the EEC. Thus, for example, Senegal would negotiate on behalf of its local shoe companies, agreements with Niger and Upper Volta concerning the duty to be applied to shoes imported from Senegal. 1/ At the end of each quarter a regional develop- ment fund pays the importing country two-thirds of the difference between the negotiated duty and the full duty applied to EEC imports. The Develop- ment Fund mainly clears the different intracountry obligations, but the resulting overall deficits are funded through a levy on all inter-regional trade except for bilateral Senegal/Ivory Coast trade which is covered by a special agreement. 1.13 The main effect of this policy is that the burden of export develop- ment within the community is put on the budget of the exporting rather than the importing country -- in contrast to the situation under the terms of the UDEAO. Moreover, since the preferential duty rate is granted on a firm-by- firm basis, it is up to the individual producer to negotiate a preferential duty rate, whereas under the UDEAO the preferential rate (generally 50 per- cent) was guaranteed by the Treaty. 1.14 It follows from the analysis of the preceding paragraphs that practically all of Senegal's exports go to "protected markets", whether inside or outside of Africa. That is Senegal enjoys access to markets (notably the EEC) on preferential terms relative to some of its trading partners (whether developed or developing). Exporters can therefore be less efficient than they would have to be in the absence of this protection. 1.15 Some years ago, perhaps in response to this extreme concentration of exports, the Government has widened considerably its international contacts with countries of all economic and political persuasions. This policy has led to some diversification of imports. More importantly, the Government recently imposed normal import duties on EEC commodities thus abandoning the old system of trade preferences. Some diversification of exports is to be expected from the newly created Industrial Free Trade Zone, the purpose being to process imported or local raw materials and to export finished goods to Europe and/or other countries. At the time of writing, the Zone had not been in effective operation long enough to have had a significant impact on diversification of either markets or products. 1/ Nothing prevents Upper Volta and Niger from charging a different rate of duty, nor from concluding separate agreements with the Ivory Coast at, again, different rates.  - 9 - B. Recent Evolution of the Industrial Sector Recent Performance 1.16 The first part of this chapter has been concerned with the struc- ture of the industrial sector. In the second part, we describe rapidly the evolution of production, wages and prices on the one hand, and on the other, the main policy developments. Finally, we shall review briefly Government's plans for the future. 1.17 The level of industrial activity, defined to encompass activity in the mining, manufacturing and energy sectors as well as production of construction materials has tended to fluctuate with capacity utilization in l'he groundnut processinag industries. Owing to developments in this branch, lhe aggregate index of industrial production declined by 9 percent to 119.4 in 1973 (1969 = 100), rose by an average of 10.6 percent between 1973 and 1977, i-ollowed by a 25 percent fall in 1978. Excluding groundnut processing, the index showed major slowdowns in industrial growth about one year after major (iroughts; average growth of industry without groundnuts amounted to 7.4 percent a year between 1973 and 1977, but showed a decline of 6.8 percent in 1978. These serious declines in industrial production other than groundnuts clemonstrate the heavy dependence of the Senegalese economy on the groundnut crop. 1.18 One of the most important economic parameters for Senegalese industry is the cost of labor, since the exchange rate is fixed. The minimum wage (SMIG) 1/ is fixecd by law and changes only if inflationary pressure becomes too heavy. 2/ Price developments in Senegal are indicatesd by the changes in a consumer price index which is based on the spending patterns of moderate income Senegalese families in Dakar. The index, while reflecting changes in prices of domestically produced goods (mainly foodstuffs) as well as of imported goods, is strongly influenced by the Government's policy of fixing retail prices of major food articles, e.g., rice, groundnut oil and sugar. 1.19 After a period of relative price stability, the consumer price index began to rise in 1973 when prices for foodstuffs were allowed to increase substantially. In that year the index of food prices (including imported cereals) which has a weight of nearly 56 in the general index, rose by nearly 19 percent, pushing the overall index up by 12 percent. 1/ Salaire Minimum Interprofessionnel Garanti. 2,/ In 1977, the SMIG was CFAF 107 compared to 310 CFAF in France, and 110 CFAF in Korea and Taiwan. Moreover, in the latter countries, there are no expen,ive expatriates. This point serves to stress the need to keep wages and price inflation as low as possible, and certainly lower than the rate in Europe.  - 10 - Accordingly, the SMIG was raised in August 1973 from 50.60 to 58.19 CFAF/hr. Subsequently, with the severe increases in the price of food and petroleum imports, the consumer price index rose sharply again, so wages were increased again on March 5, 1974 (SMIG - 66.91). In November 1974, the Government was no longer able to afford the subsidy on rice, sugar and oils, so, they were substantially reduced, causing a severe increase in prices. To palliate the effect on the cost of living, wages and salaries were again increased (the SMIG was set at 107.06 CFAF/hr). 1.20 As a result of these developments, the overall consumer price index for 1974 was about 17 percent above the 1973 average. However, the index for December 1974 was significantly higher; in particular, food prices were some 65 percent above June 1974, and transportation prices were up by about 56 percent. Thus on an average basis, the full impact of the reduction in subsidies on essential commodities was felt in 1975 when the general index was about 31 percent above the 1974 average. During the first half of 1976, the cost of living index was stable, owing to a decline in the prices of certain imported goods. With world market prices for essential commodities declining, the Government reduced the price of rice by 20 percent, and for sugar by 50 percent. Government succeeded in keeping price movements under control; the average increase between 1975 and 1978 was 6.5 percent a year. Government Policy towards Industry -- Developments since 1972 1.21 Government policy towards industry over the last few years has been primarily directed towards the Senegalization of capital and jobs. In addi- tion, there has been a substantial revision of credit policy, in the context of the reform of the operating rules of the BCEAO 1/ of 1974-75. 1.22 Concerning the Senegalization of capital, the Government has been primarily concerned with taking a larger share of the ownership of the formerly entirely private foreign firms and the creation of new parastatal companies in the context of the development of para-public sector. By 1972, the Government had taken total ownership of the electricity and water companies. In 1975, it took a 50 percent share in three main mineral companies (la Compagnie Senegalaise des Phosphates de Taiba, la Societe Senegalaise des Phosphates de Thies, la Societe des Salins du Sine Saloum). Effective January 1, 1976 it created SONACOS 2/ which is a 65 percent Government owned company and has a quasi monopoly of the selling of groundnut products leaving only one independent firm with majority Senegalese ownership in competition. It has also taken 50 percent in one bank (la Banque de l'Industrie et du Commerce du Senegal). Its policy in the tourism sector has been to take most of the capital in hotels. In the agro-industrial sector, it now owns 68 percent of BUD-Senegal 3/. A number of smaller 1/ Banque Centrale des Etats de l'Afrique de l'Ouest. 2/ Societe Nationale pour la Commercialisation des Oleagineux du Senegal. 3/ Now under liquidation.  - 11 - candidates for major State participations: the Government now has a majority of SIV's capital (la Societe Industrielle du Vetement), but TMS (Tricotage Mechanique Senegalais) which was to have merged with SIV was bought out by SCOA, a French owned private trading company. 1.23 In 1977, wheni it became clear that most of the para-public esnterprises were not performing well and tended to become a drag on the budget, the Government reversed its policy. It re-emphasized the impor- tance of the private sector for economic development, made strong moves against further bureaucratization of mixed enterprises and liquidated some unprofitable enterprises, while considering to sell back to private parties :its shares in some others. L.24 The second aspect of this policy has been the Senegalization of jobs, which has made significant progress in recent years. Such a policy iLs bound to encounter some occasional problems, but one of the many benefits iLs that the nature of the relations between the Government and the industrial sector is beginning to change as their day-to-day contracts tend to be less between French businessmen and Senegalese functionaries. 1.25 A third major policy revision has been the change in monetary and credit policy enacted at the beginning of 1975. New statutes of the West African Monetary UJnion (UMOA) were approved and put into effect on October 11, 1974 by a conference of heads of state of the member countries, and a set of new intervention rules for central banking operations in the UrMOA was adopted by the Board of Directors of the West African Central Bank (BCEAO) on May 3, 1975. In the past, the BCEAO has relied on rediscount ceilings as the main irLstrument to control credit, but this instrument gradually proved to be insufficiently effective for pursuing an active monetary policy. Under the reformed system, the Central Bank obtained all modern instruments for monetary policy including the establishment of liquidity ratio's and sectoral lending objectives. A call money market was introduced for the Unicin and a system under which all bank credit in excess of CFAF 100 million has to be submitted to the Central Bank for preauthori- zation ("authorisation prealable"). It also abandoned, effective January 1976, the practice of establishing individual bank lending ceilings and the distinction between rediscountable and nonrediscountable credit. The banks' access to rediscount facilities is determined by annually established objectives for the Monetary Union and for each member country, particularly with respect to external reserves of the Central Bank. Furthermore, the banks are expected to rely first on their own resources and those obtained through the call money market, and to use Central Bank as the lender as a last resort. The BCEAO has also adopted a flexible interest rate policy and, effective July 1, 1975, tied the entire structure of lending and deposit rates to its basic discount rate ("taux d'escompte normal"); also, in principle, the Governor of the BCEAO (who is responsible to the council of heads of states) has the power to alter the basic discount rate as appropriate in the light of monetary developments within the lJnion and abroad.  - 12 - 1.26 The new market for daily call money operations between the deposit money banks and the Central Bank works as follows: deposit money banks indicate to the national branch of the BCEAO each morning their needs or availabilities. Each national branch in turn indicates to the BCEAO head office by mid-morning its net cash requirements resulting from these operations. The head office then decides by later afternoon the amount of funds that it can put at their disposal for the day's lending operations and allocates these funds among the various countries, taking into consid- eration targets regarding the overall credit expansion. All eligible commercial paper which can be accommodated within the amount indicated by the BCEAO head office is accepted at a rate fixed by the BCEAO. Interest rates on the call money market of the Union are set so as to be broadly in line with the movements of market rates in major financial centers (mainly France). Initially set at 7-1/8 percent, this rate has been changed several times. The basic rediscount ratio was set at 8 percent with a preferential rate ("taux d'escompte preferentiel") of 5.5 percent, both put into effect in July 1975. The preferential rate applies to Treasury overdrafts, to crop marketing credit extended to Government marketing agenciet, to small credits of up to 10 years extended to nationals or enterprises owned by nationals, and to credit of up to 10 years extended to the Government to finance development expenditures. 1.27 Thus, prior to 1975, it can be said that the industrial sector had easy access to credit and to rediscounted credit at the Central Bank. It was argued that this system led to an abusive outflow of cheap short- term capital for relending in Europe, and so, in 1975, the rediscount system was drastically curtailed. Foreign firms were required to borrow abroad when their local credit ceiling was insufficient. At the same time preferential credit was reserved for the agricultural sector (ONCAD) 1/ and small national firms. Thus the notion of sectoral favoritism had developed in favor of preferential credit based on nationality. Whilst the desire to promote national prices is highly desirable, it need not imply sectoral discrimination. In 1977, the monetary authorities tried to change this policy to more clear sectoral objectives, but had relatively little success. The monetary picture in recent years is entirely dominated by the large needs for credit by ONCAD and the Central Government. 1.28 Other aspects of industrial policy notably, indirect and direct tax policy, import duty policy, and the control of competing imports, have not changed at all in the last few years. A comprehensive reform of indirect tax and duty policy is currently under way, 2/ and the investment code is to be revised. A later chapter of this report gives an overall view of the various instruments. 1/ Office National de Cooperation et d'Assistance pour le Developpement. 2/ The main objectives of the reform which took place in June 1979 were the elimination of distortions and administrative simplification. The reform diminished the average level of import duties somewhat, while it increased taxation on consumption through the introduction of a taxation on value added.  - 13 - C. Government's Plans for the Future 1.29 Government's policy towards industry reflects its desire to promote more diversified industrial growth, oriented primarily toward vertical inte- gration of the different agricultural and industrial activities in Senegal, and to create as many jobs as possible. In 1975, at the beginning of the current Plan period, the Government's plans were: (a) to construct a large petrochemical complex at Cayar, 67 kms north of Dakar; (b) to develop the Faleme iron ore deposits;y (c) to construct a ship repairyard, Dakar Marine; (d) to develop the Industrial Free Trade Zone; (e) to continue with decentralization of industry over the entire territory of Senegal; (f) to pursue the creation of small-scale industry, entirely owned and run by Senegalese nationals; (g) to promote agro-industries; and (h) to opt for irtdustrialization which fits well over the overall industrialization plans of the CEAO and the OMVS 1/. Cayar Complex 1.30 The idea was to construct a new pole of industrial development 67 kilometers north of Dakar. The Governments of Iran and Senegal were partners in a project to construct a huge petrochemical complex consisting of a refinery, a phosphate mine, a fertilizer plant, and a power plant. The Government of Iran would buy one million tons of Senegalese phosphates (produced by the new mine of capacity 1 to 1.5 million tons) and in return it would sell Senegal crude oil for a new refinery (2-3 million ton capacity) supposedly oriented towards exports in Europe. A key element of the project was the construction of a port, designed initially just to handle the export of the phosphate rock and the import-export of crude oil and refined oil products. The fertilizer plant would produce ammonia and urea for the local and export markets (capacity 50-60,000 tons). This project would be jointly financed by the Governments of Senegal and Iran, although the latter expressed willingness to lend the former enough to be able to make the initial invest- ment. In 1976, when Iran found phosphate deposits within its own territory and world phosphate prices dropped to less than half their 1974 value (when tihe project was designed), the Iranians lost interest. The project was cancelled, but a new, more modest but also more promising project replaced it and is now in an advanced stage of preparation. The project aims at proces- sing unexportable low grade phosphate rock into phosphoric acid. The 1/ Organisation pour La Mise Valeur du Fleuve Senegal.  - 14 - company, Industries Chimiques du Senegal (ICS) has a majority of private share- holders including a French company which controls most of the West African fertilizer market. Faleme Iron Ore Deposits 1.31 In the longer run, the Government hopes that the iron ore deposits of western Senegal will prove to be of sufficiently good quality that their extraction would be worthwhile. In this event, 700 km of railway track would have to be built, as well as a seaport for the evacuation of the ore. Several points should be made about these iron ore deposits: (a) the quantities of ore so far identified are not yet sufficient for exploitation; (b) since the deposits have low iron content, considerable benefi- ciation is required to make them exportable; (c) hydropower generation for the beneficiation of the ore is possible, but would require substantial investment, or would be dependent on the construction of the Manantali Dam in Mali; and (d) the inland location of the deposits makes its exploitation relatively expensive compared to other known deposits in Africa. Although eventually the Faleme iron ore will most likely be exploited, the Bank does not foresee this to happen within the next ten years. Dakar Marine 1.32 The Government has decided to reduce the size of this project, and has ordered a floating dock of 60,000 mt. It will be extremely difficult for Dakar Marine to find sufficient work for the shipyard in this highly competitive market of repair work on ships up to 60,000 mt. 1.33 A far more interesting proposition than Dakar Marine is the creation of repair facilities of small boats (up to 30,000 tons) and of fishing boats, which fish in the area around Senegal: for example, Polish trawlers which would avoid the considerable downtime involved in an annual trip home or to other European repair facilities. Such arrangements could be made part of the international fishing agreements which Senegal is negotiating for the use of its national waters. Industrial Free Trade Zone 1.34 The Industrial Free Trade Zone, operational since late 1975, is entirely export oriented, and firms therein are accordingly allowed to import their raw materials duty free for re-export. Electricity is priced at the ex-factory price (no transport costs are included). It is hoped that the Zone will create at least 5,000 jobs by 1980. At the present point, there are five firms which have begun or are scheduled to begin production. The firms produce, or will produce, plastic sheets, trousers, bicycles, rattan furniture and will outfit specialized trucks.  - 15 - Decentralization of Industry 1.35 The Government attaches great importance to a judicious decentra- lization of industry, 1/ a process it has begun with the creation of a textile mill at Thies (1970), a sugar refinery near St. Louis (1974), a Eactory making tomato concentrate in the Fleuve region (1969) and a cotton ginning plant in Eastern Senegal. It wished to continue this process, partly by the expansiona of the extractive industry, which is inherently decentralized, but partly by the creation of industrial zones in the neighborhoods of Thies, Ziguinchor, St. Louis and Kaolack. Two types of irirms will be in these zones: (a) large units capable of competing on the international market; and (b) smaller units, which will be generally owned and run by Senegalese nationals, in the context of the Government's policy to promote the growth of purely national industry. 1.36 Concerning the larger firms, Ziguinchor and Kaolack seem the best suited in the short run, since they both already have some existing indus- t:rial infrastructure. An extremely interesting and pioneering project is t:he Government's plan to build an integrated textile plant at Kaolack which will transform 1,500 tons of local cotton into finished knitwear for export. This project will be jointly undertaken by the Chase Manhattan Bank and a Swiss textile group, and it in fact amounts to the creation of a one firm, free trade zone in Kaolack. However, this interesting project is n,ot advancing very much because of uncertainties about the marketing possibilities. Promotion of Senegalese Industry 1.37 The promotion of purely Senegalese business remains a principal objective of the Government. SONEPI (la Societe Nationale d'Etudes et de la Promotion Industrielle) was created in 1969 to give technical assistance to small- and medium-size firms owned by Senegalese. This organization has played a very important role in the growth of these firms. Spec:Lfic achievements include: (a) the creation of a special fund which loaned capital to these firms (now managed by SOFISEDIT 2/, a mixed investment bank); 1/ There has always been a certain decentralization of the groumdnut industry. There is a small groundnut crushing plant in Diourbel and in Ziguinchor, which also led to the creation of small autonomous power plants for these cities, owned by the groundnut mills. In Ziguinchor, there are a few small fishing firms, since the city is on the Casamance River. There is a large groundnut mill at Lyndiane, near Kaolack -- which is on the sea. 2/ Societe Financiere Senegalaise pour le Developpement Industriel et Touristique.  - 16 - (b) the creation of a Fonds de Garantie, which guaranteed money lent by the banks to these firms; (c) a special productivity center which provides technical assistance to these firms; (d) a marketing service; and (e) a service specialized in the preparation of dossiers, bids, etc. 1.38 Moreover, some attempt has been made to create small industrial estates where a number of firms could operate together. Such estates are planned for St. Louis, Kaolack, Ziguinchor and Thies, and they are also an integral part of the decentralization plan discussed above. Finally, SONEPI was responsible for the creation in 1972, of a special investment code specifically adapted to the needs of small- and medium-size firms. Thirty-one firms have been admitted to this regime. 1.39 The purpose of all this is to eventually create a group of national entrepreneurs sufficiently efficient and well equipped to be able to integrate themselves well into the rest of the industrial sector. To date, no comprehensive study has been undertaken of this group of national firms. Agro-Industry 1.40 With the exception of groundnuts, Senegal has a young, as yet limited agro-industrial sector, but given that the country is primarily agricultural, it would seem that there should be a good future in this kind of industry, and the Government intends to increase investment there quite substantially. At the present point, there is an industrial farm (BUD- Senegal) which exports fresh vegetables off-season to Europe, a sugar refinery (using some locally grown cane), and a tomato concentrate factory (there will soon be a second one). The problems these firms have faced may be classified as follows: (a) those related to water supply; (b) those related to the reliability of peasant production; and (c) those related to the institutions organizing the production. BUD-Senegal could never reach the stage of profitability partly since no water was found in the environment and irrigation with drinking water pumped from the northern part had to be terminated when the full piping capacity became needed for the population of Dakar; at the moment the company is in the state of liquidation. The sugar company suffered from insufficient locally produced sugar cane when the Senegal River was low, a problem related to the prevalence of salt in the soil. The tomato factory suffered from low capacity utilization in some years because farmers did not produce enough tomatoes and in 1978 because the organiza- tion responsible for the transportation of the crop could not provide the  - 17 - necessary trucking capacity on time. Most of the future growth of this sector is thus contingent upon the satisfactory resolution of these diffi- cult problems. Development of Other Industries 1.41 Other industries will be developed as the occasions present themselves. The Government is planning a second 500,000 ton cement plant, the construction of another groundnut mill in the Casamance, a plant to make paper sacks for cement, and other investments most of which are oriented towards the local market, including a project to enrich the aluminium phosphate presently manufactured by SSPT (la Societe Senegalaise des Phosphates de Thies), and a project on the part of ICOTAF to completely renew its capital equipment.  - 18 - Chapter II. GOVERNMENT POLICY INSTRUMENTS 2.01 Although the last chapter made some mention of particular aspects of economic policy, the purpose of the present chapter is to review all of the instruments which the Government uses directly or indirectly to in- fluence the development of the industrial sector. These are: (a) Protection policy (encompassing duty policy and indirect tax policy); and direct controls of competing imports via quotas; (b) Investment policy, direct tax and credit policy; (c) Price control policy; and (d) Export promotion policy. Principal Conclusions 2.02 First, existing incentives favor production destined to the local market over exports, to the point that there is positive discrimination against exports. Moreover, on the local market there is also a bias in favor of production of finished consumer goods, and against semi-finished goods and machinery. In other words, there is discrimination both against exports, and also against "vertically integrated development" of the local market, where this is to be interpreted as simultaneous development of raw materials, capital goods, semi-finished goods and finished goods - as stated by the Government in its Fifth Plan. There are many unexploited possibilities, and many local industries whose development is impeded by current incentives, particularly in shipbuilding and machinery industries. Second, the Government has done little so far to offset this bias in ways likely to have a significant impact on these problems. In particular, the terms and conditions of the Free Trade Zone do not seem likely to assure a very rapid growth of exports. Third, the most fundamental problem is whether the existing combination of the exchange rate and import duties is appropriate for development of the industrial sector. Fourth, the change from the UDEAO to the CEAO is likely to make intraregional exporting more difficult, as the CEAO provides a less auto- matic system of preferences than was the case under the UDEAO. Fifth, the Government should seriously consider a reform of the incentive system along the lines discussed in this chapter, to develop exports.  - 19 - A. Instruments 2.03 We saw in the last chapter that exports have been concentrated in a few commodities (mainly groundnuts, phosphates and fish). Recently, t:he Government has repeatedly stressed the need to increase nontraditional exports; this objective has not yet been well translated into concrete economic policy. Present economic policies consist of a mix of measures 1:ending on the whole to favor the development of the local rather than the external market. SpeciLfically, some exports are effectively taxed, others subsidized - wil:h no clear justification as to the economic benefit of either situation. On the other-hand, practically all production for the local market is subsidized through protective measures. IThe Protection of Local Industry Quotas 2.04 Senegal continues to discriminate against imports from non-EEC countries through a system of quantitative import restrictions. The principal means of this discrimirtation is the "programme general des importations", which fixes for all non-EEC, non-CEAO, non franc zone countries the annual ceiling of imports. 1/ The main justification for its continued existence being the need to control the balance of payments on a bilateral basis. 2/ It should, however, be clear that this kind of discrimination involves economic costs for consumers which are far greater than the benefits conferred on those who are lucky enough to dispose of the quotas. Since most quotas apply to goods which are not prcduced in Senegal (such as cars), their elimination should not pose any great problem. For textiles and shoes, however, care should be taken to address the problem taking into account the repercussions on local industry. In principle, the quotas should be removed and replaced by an appropriate level of import duties. This would have two advantages; first, the remaining imports would come from the,cheapest suppliers in the world, rather than from the sometimes more expensive bilateral trading partners; second, the rents gained on the quotas are now going to a few privileged traders, but with an appropriate system of import duties, they would go to the Treasury. Import Duties 2.05 The local market is also protected by import duties. The guiding principle for the administration of duty policy is that duties should be lower 1/ The most common mechanism is that of "autorisation prealable", which allows the local producer to block imports of competing goods, unless his price is at least 10% more than that of the imported good. For some industries, notably textiles, controls are much tighter, and for others (flour, matches) imports are banned. 2/ Some imports from the EEC are also subject to quantitative restric- tions, but only when they threaten (or when they compete with) local industries.  - 20 - on raw materials than on finished goods. For some industries (finished goods) the duty structure - so escalated - had created real protection (later called effective protection) much greater than nominal protection. 1/ Other indus- tries were untouched by duty policy, in the sense that the rates applied to their raw materials were the same as those applied to the finished goods they make. But, for capital goods, the duty on raw materials is often far greater than on finished capital goods. As a result, local producers of machinery and capital goods complained that they were unable to compete with imports. This problem is important in construction, metal working, shipbuilding, etc.; at the moment, the structure of duty discriminates against them. Thus, the Government's attempt to provide positive escalation led necessarily to negative escalation for some industries, at obvious cost to the economy. The 1979 tax reform improved the situation, but did not eliminate this problem. Indirect Tax Policy 2/ 2.06 The tax reform aimed at diminishing price distortions in the productive sector of the economy by distinguishing more clearly between taxation for revenue raising and taxation for the protection of domestic industries. Import duties do create price distortions even if applied across the board and are therefore less suitable for revenue raising; as will be seen in paragraphs 2.18 to 2.21, systems of drawback and temporary admission normally cover detaxation of imported inputs and equipment bought by the exporting firm, but not the import taxation element in the prices of domes- tically purchased inputs. Import taxes have therefore a price increasing impact on exports. The Value Added Tax system (VAT), introduced with the reform, is completely neutral in respect to the relative prices in the pro- ductive sector. Under the VAT system, firms can deduct the tax included in the suppliers invoices for all their inputs and equipment from their tax obligation on sales. Since no tax obligation exists on exports, but taxes on the inputs are nevertheless reimbursed, the VAT system assures--contrary to an ordinary sales tax system--that exports are completely detaxed. Since the tax paid on equipment goods can be deducted from the firm's tax obligations on its sales, investment goods are also completely detaxed under the system. In fact, the full tax burden is borne by domestic consumers, a feature which makes the VAT an ideal instrument to raise public revenues at the cost of private consumption. 2.07 Under the reform, all export taxes except for phosphate rock and groundnuts were abandoned and the overall price impact of import taxes was reduced, both measures resulting in lower cost prices for exported goods. Since the old system was so complicated, it is only possible to give an impressionistic view of the changes in tariffs brought about. 1/ Nominal protection is the ratio of the price of the local output in the Iocal market and the price CIF Dakar of an imported good of similar quality. Effective protection is measured by the ratio of value added in domestic prices to that in world prices. 2/ This section was updated after the 1979 tax reform.  - 21 - TabLe 7: REFORM OF IMPORT TAXES (in percentage) Before Reform After Reform Import Duty (droit de douane) 0.5 EEC 0.5 (all) 0-10.0 non-EEC Fiscal Duty (droit fiscale) 0-25.0 10.0 (taux reduit) 35.0 (taux ordinaire) 40.0 (taux majore) 70.0 (taux special) Statistical Tax (droit statistique) 4.0 T'ransaction Tax (taxe forfaitaire) 0-22.1 I'he new structure dimirLishes considerably the spread in taxation which went from 4 to 142 percent under the old system, to 10 to 78.5 percent in the new system. Of particular interest is the increase in taxation (10%) on equipment goods, which will give better protection to local production of equipment goods and eliminate part of the bias in favor of capital inten- sive investments or creation of overcapacity. Also the excessive escala- tion of protection through zero taxation on the imports of raw materials was considerably reduced. An important feature of the reform was the suppression of the numerous individual exonerations of the existing tariff structure, a phenomenon which had reduced the tax base to 65 percent of the annual imports. 2.08 The introduction of the Value Added Tax system implies a new step in a progressive modernization of the sales tax (TCA). In 1977, the Government had already removed the difference in sales tax for imported and non-imported goods. The TCA also gave some firms the right to deduct sales tax incorporated in bills of suppliers, but with the reform, the system will b,s extended over all commodity producing firms, including the construction sector. Only the tax on services (Taxe sur les Prestations des Services) is not yet incorporated in the VAT, and was maintained at slightly lower rates (5%, 12.5%, 15% and 45%). Table 8: REFORM OF CONSUMER TAXES (in percentage) Before Reform TCA After Reform VAT Reduced (taux reduit) 4.92.. 5.0 Normal (taux normal) 15.93 18.5 (taux majore) 39.33 45.0  - 22 - The reform of the sales tax implied a considerable higher revenue from this source to compensate for losses in revenue from the reform of import taxes. Moreover, rates had to be higher to compensate for the improved possibilities for deductions. The Investment Code 2.09 The Investment Code gives special treatment to firms which invest more than 100 million CFAF in three years, or which create at least 100 jobs in three years. Conditions are described in Law 72-46, but this is being revised. Under the proposed new law, profits are no longer exempted from taxes, although dividends are. The most important exemptions, however, are from import duties on raw materials for a period of five years in the Cap Vert Region (eight years outside) and from import duties on capital goods deemed essential to the completion'of the project. 1/ 2/ 3/ Whilst this kind of exemption from import duties is commonplace in many third world countries, it should not be forgotten that problems can result. If there are firms already operating within the country, granting a duty exemption to the new firm creates unfair competition for those older firms which, for the most part, do pay customs duties on their imported raw mate- rials, 4/ and this is an increasingly important problem in Senegal. It suggests the need to think in terms of sector promotion, rather than firm promotion. Third, duty exemptions of the kind described are unable to create an additional incentive to export. To summarize, the Investment Code increases the incentive to produce for the local market, while doing very little to further the development of exports. 1/ The "5 year, 8 year" disposition still has to be voted by Parliament. At the time of writing there was a uniform exemption of ten years, regardless of the location of the investment. This is a step backward from the 1972 law, and the "5 year, 8 year" option should be introduced. 2/ The Code also provides for exemptions from many other taxes, notably property taxes, company registration taxes, etc. 3/ If the firm invests more than one billion CFAF, it is eligible to request the status of "societe conventionnee"; and it has a special contract with the Government. This entitles it to additional rights, negotiated on a case by case basis, over and above those available to a priority firm. Of these, perhaps the most noticeable and the most controversial is a long-run fiscal stabilization clause which can protect firms from perfectly ordinary modifications of common taxes for periods up to 20 years. As far as possible, however, a limit of ten years should be used, since longer exemptions are of little value to firms, but very costly to the Government. 4/ It should be pointed out that there is also a danger of discordance between the conditions granted to different priority firms operating in the same sector.  - 23 - 2.10 However, these problems can be solved by changing the form in which concessions are granted. In this connection, it is essential to mnderstand that any tax or duty concession costs the Government revenue; therefore, it clearly pays the Government to 'spend" its lost receipts in ways most likely to maximize benefits to the economy. In this respect, one may argue that it is better to give incentives to Investment Code firms only in the form of temporary direct subsidies to value added (calculated ils a percentage of value added in world prices), than to grant them in the iform of exemptions from duties on raw materials. More precisely, one would require priority firms to make all their purchases at full cost prices, as if they operated as Common Law firms. 1/ I2.11 Such direct subsidies are preferable for several reasons. First, t:hey do not discriminat:e against the local production (real or potential) of raw materials and semi-processed goods, as do exemptions from duties on these goods. In other words,, direct subsidies stimulate local production, and also provide a greater incentive to buy locally. Second, value added is generated both by production for the local market and for exports - hence subsidies to value added reward both exports and local sales. Third, firms have an incen- tive to maximize value added (i.e., the sum of profits and wages) whereas duty exemptions constitute an automatic subsidy to profits alone. In a labor- abundant economy, it is clearly better that firms should have an incentive to earn increased profits by increasing employment. The use of subsidies to value added instead of to profits is a small, but important step, towards the elimination of the bias, in existing incentives, against labor and in favor of capital. The subsidies could be easily administered by allowing them as a credit against indirect taxes. This procedure would also faci]itate the problem of payment, since it would reduce (and sometimes eliminate) the need for direct cash payments by the Treasury. 2.12 Finally, it should be emphasized that these subsidies would be provided for only five years to Investment Code firms inside the Cap Vert Region and for eight years to those outside, as is the case for duty exemptions at present. This is an important point, because it serves to remind us that the Investment Code should provide limited additional incentives, whatever form they take, over and above those available to firms under Common Law. Neither direct subsidies nor duty exemptions, given for five years, can solve problems posed by insufficient protection of the local market and of exports. After this five or eight year period, these priority firms must operate within the confines of Common Law. It may well be that for the non-priority firms the structure of incentives is also inappropriate. In this case, it is the Common Law and not the Invest- ment Code which must be changed. We return to this very important problem later. 1/ A firm processing imported raw materials would then pay the duty on this raw material, if the final destination of the finished product is the local market. For exports, of course, the imported raw materials would continue to be exempt of duty, although the system of direct subsidies could be extended to include exports in such a way that exporting firms would pay duties on imported raw materials.  - 24 - Direct Tax Policy of Profits 2.13 Under Common Law profits are taxed at 33 percent, with an exemp- tion granted for the first five years within the Cap Vert Region and eight years outside. The Government is proposing to abandon this arrangement. Dividends are taxed at 16 percent. Credit Policy 2.14 The most important point to repeat is that credit is now accorded on preferential terms, on the basis of nationality rather than on the basis of the sector in which a firm operates. Price and Wage Policy 2.15 Wage and price controls are applied to most goods which are imported or produced in Senegal, as a means of controlling "excessive profits" of the commercial and industrial sectors. The Government is reconsidering ,the existing policy in order to diminish administrative controls which proved to be burdensome for the enterprises and less effec- tive in containing price increases. The Government traditionally followed a strategy tif maintaining the purchasing power of the urban workers, but under the force of the economic stagnation in recent years, the Government already de facto abandoned this strategy. It should be replaced by a more productivity-oriented strategy. Export Promotion Policy 2.16 Export promotion is again a mixture of policies with no central guiding principle. A number of points can be usefully made. First, as already pointed out in Chapter I, the framework of the West African economic union is becoming with each modification of the treaty a less effective instrument to promote inter-African trade: it appears clear that in the future, Senegal has to base its policies more on competitive prices than on institutional arrangements. Second, for exports outside the CEAO, the Government is placing great hope in the Industrial Free Trade Zone (tax holidays, duty free imports of raw materials, preferential elasticity charges, etc.), but this Zone is as yet young and so far, has not shown a strong ability to attract new industries. Third, for non-traditional exports (excluding phosphates and groundnuts), manufactured within the customs barrier, the Government does little to make exporting more profit- able. 1/ Fourth, the export of phosphates and groundnut products are effectively taxed at the frontier. 2.17 The following aspects of the export promotion policy need a more detailed discussion: (a) admission in suspension of duties for imported raw mate- rials incorporated in exports and/or a drawback of duties on imported inputs incorporated in exports; 1/ Apart from the usual measures such as Admission Temporaire, (see par. 2.18).  - 25 - (b) the export taxes on groundnuts and phosphates; ancl (c) the Industrial Free Trade Zone. Temporary Admission and Rebate of Import Taxes 2.18 Admission in suspension of duties is of course a standard customs procedures used by many countries. Nonetheless, it tends to encourage the use of imported, rather than local raw materials and semi-finished good for export because local goods are not completely tax free. Thus, temporary admission regimes do not encourage the creation in Senegal of vertically integrated export industries. Yet such is one of the targets of the Fifth Plan. 2.19 It may be added that there are policy instruments which can stimulate the development of such vertically integrated industries. In the Senegalese context, this problem can be most easily illustrated in the textile sector. This is one of the few industries in which a local raw material (cotton) is made into a semi-finished good (grey cloth), which is in turn made into finished products (cloth, and dyes and cloth). The printing and dyeing industry uses this locally produced grey cloth only for production destined for the local market. For exports, the printing and dyeing industries import (duty free) grey cloth from China; the price of local grey cloth is 50 percent more expensive than imported grey cloth. The result, of course, is that local grey cloth production is half of what it could be. Suppose, however, thal: imports of grey cloth for exports were subject to import duties; this would clearly stimulate local production, but the exports of printed cloth would be drastically reduced; in other words, there would be no market for the grey cloth. This problem can, however, be solved by sub- sidizing directly the exports of the printed cloth. The mechanism of direct export subsidies then permits not only the maintenance of existing exports of finished goods, but aiLso the creation of a new factory which would produce more semi-finished goods - themselves made from local raw materials. The use of these subsidies should not be automatic; their adoption should only be allowed for sectors where their economic benefits can be shown to be greater than their costs. It is also worth pointing out that the Government does in fact use them, albeit in a disguised form, within the CEAO - through the regional cooperation r-ax. (See Chapter I). 2.20 Although the "admission temporaire" system of duty suspension and the "entrepot industriel" system are similar in most respects, one important difference between the two systems should be mentioned. The admission temporaire system authorizes the duty free import of raw materials for re-export, for a period of up to six months at the end of which import duties and penalty interest must be paid. Exporters who over purchase are therefore penalized. On the other hand, the entrepot industriel system allows producers to keep imported raw materials in duty-free status indefinitely provided that the firm exports more than 60 percent of its production. Moreover, according to some firms, the Government wishes to eliminate the "admission temporaire", so that firms which do not export at least 60 percent of their production will be obliged to stop exporting, or to use the drawback  - 26 - system, which is so inefficient that firms try to avoid using it. As a result, small exporters could be put at a serious disadvantage, and one may reasonably expect a fall in their foreign earnings. Drawback Facilities 2.21 As stated above, the drawback system is less popular with entre- preneurs in Senegal because of the way the system is administered. In principle, agreed percentages of duty drawbacks to be deducted from other tax obligations is the most efficient way to eliminate import tax elements (including those in domestically purchased inputs) from the cost-price of exported goods. Some Senegalese authorities argue that duty drawbacks are equivalent to export subsidies, because the Government foregoes revenue by their use. Since this argument is technically correct, the proposed export premiums may be incorporated and justified by the establishment of an improved duty drawback system. The Free Trade Zone 2.22 At the moment, the Government is pinning most of its hopes on the Industrial Free Trade Zone. Firms therein are totally exonerated from all import duties, and from all other taxes both direct and indirect 1/, provided that they invest at least 200 million CFAF or create at least 150 jobs within two years of starting up. The Zone is, however, as yet young and beset by problems, the most serious of which are: (a) the indisputably generous conditions of the Zone do not appear to be sufficient to offset the high cost of local purchases; (b) administrative procedures are slow; and (c) the conditions of the Zone do not and cannot do anything to encourage the creation of links to the rest of the economy. For example, no firm in the Zone could afford to buy locally produced grey cloth. We shall come back to these problems in the next chapter. Export Taxation 2.23 Phosphate exports pay a fiscal duty of 2.5 percent for the first 500,000 tons exported, and 5 percent thereafter. 2.24 Since 1979, groundnuts and groundnut oil are taxed at 20 percent of an annually determined administrative value; for groundnut cakes the tax is 10 percent of the posted value. The use of administrative values can easily lead to undertaxation in years when world market prices are high and overtaxation when world prices are low. 1/ It can also purchase electricity at a special price, and it can buy from local firms in suspension of sales taxes; local firms can also consider their sales to the Free Trade Zone as exports and so import their raw materials duty free.  - 27 - Conclusion 2.25 Although the Fiscal Reform of 1979 brought considerable improve- ment, the price structure in the industrial sector is still biased towards the local market and away from exports. The principal causes of this are the escalated tariff structure (coupled with high rates on final goods) and the dispositions of the Investment Code which have left producer margins significantly larger in the local than in export markets. Firms have naturally tended to concentrate their production on the local market and will continue to do so in the absence of adequate incentives to export. The arrangements of the Industrial Free Trade Zone, which allow duty-free entry of raw materials for re-export and a long tax-holiday, do not satisfac- torily solve the problen of the smallness of producer margins in exporting activities: other more direct measures, such as direct export subsidies are also needed in this case. Finally, the co-existence of two groups of privileged firms -- those in the Free Trade Zone and those which benefit from the Investment Code -- with an unprivileged group (Common Law firms) is tending to create a highly distorted industrial sector with the fol- lowing attributes: - erosion of the tax base - unfair compet.Ltion - the use of ad hoc solutions to solve complicated problems - excess capacit:y all of which diminish the contribution of the industrial sector to the economy's growth. The following section of this chapter provides evidence wbich corroborates these conclusions.  - 28 - B. Quantitative Evaluation of the Impact of these Policy Instruments The Analysis 2.26 The purpose of this section is to report the results of an in-depth study being carried out by the World Bank, 1/ on the structure of incen- tives. The results which follow remain preliminary and have been overtaken by the 1979 tax reform which was partly based on the findings of this study; nevertheless, they do provide some very interesting insights into the impact of the whole gamit of measures (described in Section A) on private and social profitability. The study uses the concept of effective protection to measure the incentive given to firms to invest in different activities by the struc- ture of tariffs, indirect taxes and trade controls. For each firm, incentives as existing during the 1972/74 period are measured separately for the local and all export markets, both in and out of Africa, 2/ (in other words the CEAO, the EEC, and other markets). From the results, certain conclusions about the social profitability of different kinds of industrial activity in Senegal can be derived. 3/ The Results 2.27 In brief, the principal results are as follows: the structure of incentives 2/ (as measured by the effective rate of protection) have large positive incentives to produce for the local market and regional markets, and in general negative incentives to export to the EEC and elsewhere, at the existing exchange rate. The export of phosphates was taxed, reducing the incentive to export. The evidence presented thus corroborates the hypotheses advanced in the preceding section concerning the 1/ This study is being prepared under the direction of Mr. Bela Balassa, with country studies being prepared by his associates. For convenience, it is referred to as the Balassa study. 2/ The period covered was 1972-74, with most analysis being undertaken for 1972. 3/ In the next chapter, the concept of domestic resources cost of foreign exchange is used to examine whether the different branches of the industrial sector are economically and socially profitable, rather than just financially profitable for the shareholders of the firms (either national or foreign). The sample used covers the whole industrial sector, with the exception of metal working, packaging and cigarettes.  - 29 - the distortions created by the escalation of the tariff structure and the investment code in the period observed. I/ 2.28 Effective protection measures the extent to which the tariff struc- ture allows, at the observed exchange rate, value added in market prices to exceed value added in world prices, measured as EPC - value added in market prices value added in world prices where EPC is called the effective protection coefficient. 2/ Thus the EPC takes account not only of protection to sales, but also of taxes on inputs, and of protection to locally produced tradeables - by calculating the ratio of value added in domestic market prices (i.e., the sum of actual factor payments before income taxes) to value added in world prices. The EPC measures the extent to which the structure of protection permits domestic factor payments to exceed the value they would have had with zero protection at the existing exchange rate. If the EPC is greater than unity, the pro- tection system is said to give positive effective protection to the activity and, by implication, a positive incentive to the firm to "operate" the activ- ity in question. Correspondingly, if the EPC is less than 1, the activity is negatively protected. Analysis of Table 9 shows that: (a) in the local market, the EPC varies between 0.64 and 3.48; and most of the time the EPC is greater than unity; (b) in regional export markets, the EPCs are between 0.92 and 2.16. Except for minerals, chemicals and construction goods, these EPCs are greater than one; 1/ The exact meaning of negative (positive) incentives should be clearly understood. At the existing exchange rate, producer margins are estimated in two ways: in market prices, taking account of duties and taxes (VAd); and in world prices, assuming zero duties and taxes (VAw). In the present report a negative incentive means only that VAd<VAw; and not necessarily that VAw<0. 2/ Nominal protection (or the nominal protection coefficient - NPC) ex- presses the relation between the observed price in a given market, and the price of this same good in world prices (at the official exchange rate). For example, the nominal protection coefficient of 1.18 for grain means that the price of flour in the local market was observed to be 18% above the world price. The NPC is a useful measure because it enables us to determine at a glance whether the price is above or below the world price; and because it can serve as a benchmark of the "ap- parent" protection given to a particular industry resulting from the interplay of duties and quotas on competing imports, on the one hand and price control on the other.  Table 9: INCENTIVES TO DOMESTIC AND EXPORT SALES 1972-74 PRODUCT Export Shares Nominal Protection Effective Protection OR Local Export Sales Local ACTIVITY CEAO EEC Others Total Sales CEAO EEC, etc Together Total Sales CEAO EEC, etc Together Total Extractive Mining 0.00 o.67 0.28 0.96 0.94 1 00 0.98 0 98 0.97 0,79 0.92 0 86 0.86 0 86 Flour Mills 0.01 0.01 0.07 0.10 1.19 1.18 1 00 1 02 1.18 1.18 1.14 0 87 0 92 1.61 Other Food Processing 0.10 0.80 0.06 0.96 1.32 1.19 1 00 1.02 1.03 1 64 1 81 0 65 0 76 0.80 Other Food - Tobacco 0.05 0.00 0.02 0.08 1.23 1.13 1.00 1.09 1 22 1.29 1 06 0.64 0 94 1.27 Textiles + Clothing 0.20 0.04 0 07 0.30 1.46 1 26 1.00 1 17 1.38 1 88 2.16 0.79 1 66 1 80 Petroleum Refining 0.15 0.00 0.22 0.37 1 22 1 09 1.00 1 04 1.15 3.48 1 50 0 20 0.74 2 46 Chemical Industries 0.16 0.11 0 03 0 29 1.32 1.07 1 00 1.04 1.24 1 92 0 98 0 90 0.94 1 55 Construction Mat. - Glass 0.07 0.00 0.11 0 18 1.01 1.10 1.00 1.04 1.02 0.86 1.02 0 84 0.92 0 87 Transport Equipment 0.33 0.00 0.04 0.37 1.12 1 06 1 00 1.04 1.02 0.84 0.89 0 78 0.96 0.89 Other Mecan. - Electr. 0.58 0.00 0 00 0.58 1.19 1 05 1.00 1.04 1.11 0.64 1.04 0 93 1 04 0 67 Other Industries 0.00 0.00 0.00 0.00 1 39 - - - 1.39 1.47 - - - 1.47 Priority Firms 0.13 0.32 0.14 0.69 1.33 1.17 1.00 1 04 z 1.55 1.78 0 90 0.98 1 14 0 Co-nuon Law Firms 0.04 0.50 0.03 0.57 1.07 1.14 1.00 1.01 . 1.47 1.13 3 21 2 73 2 02 Years Old or Less 0.13 o.32 0.00 0 45 1 47 1.21 1.00 1.06 > 2 26 1 76 0.95 1 19 1 86 Between 2 - 5 years old 0.07 0.45 0.06 0.58 1 14 1 16 1.00 1 02 1.50 1 39 1 67 1 63 1.57 Between 5 - 10 years old 0.19 0.22 0.07 0.48 1 17 1.07 1 00 1.03 1.14 0 98 0.91 0.93 1 03 More than 10 years old 0.05 0.47 0.06 0.59 1 12 1 20 1 00 1 02 1.51 1.55 1.87 1 83 1 69 Foreign Capital - 10070 0.03 0.52 0.03 0.58 1 08 1.18 1.00 1 01 1 62 1.24 2.90 2 64 2 10 Foreign Capital - 50-1007. 0.14 0.26 0.15 0 55 1.27 1 15 1.00 1 04 1.33 1.47 0.89 0 96 1 09 Foreign Capital - 0-50% 0.26 0.74 0.00 1.00 0.00 1 20 1 00 1 05 0.00 1 87 0.95 1.19 1 19 Foreign Capital - 0% 0.00 0.00 0.00 0,00 0 00 0 00 0.00 0.00 0 00 0 00 0.00 0.00 0 00 Sample Total 0.07 0.45 0.06 0.58 1.27 1.16 1.00 1.04 1 17 1.54 1.37 0 087 0.095 1 12  - 31 - (c) in the EEC export markets, the EPCs are less than unity; and (d) with the exception of phosphates (and groundnuts), we find: (i) EPCs in the local market greater than in the EEC. EPCs in regional exports greater than in the EEC. (ii) EPCs in the local market sometimes greater than and sometimes less than in regional markets. Thus, protection policy causes wide dispersion of EPCs 1/ both between markets for a given activity, and between activities themselves. 2.29 Since these research findings date from 1972/74 when the UDEAO rules were still in force, it is obviously important to know how these conclusions are modifiel given that the UDEAO has been replaced by the CEAO. Since this replacement affected only trading arrangements within the UDEAO, the findings for the local market and for export markets outside the CEAO are unaltered. Coincerning the UDEAO markets, it has already been pointed out that the nelw preferential trading arrangements will probably result in a structure oE incentives just as distorted as under the UDEAO arrangements. However, the general conclusions of the study are essentially unaltered: it may safely be said that the system of incentives generally provides positive protection to the local market, negative protection to exports outside the CEAO, and tends to give positive protection to CEAO markets. 2.30 It is useful 1:o express these results in terms of the protection which could be obtained by a modification of the exchange rate - which wculd be a better way, c:eteris paribus, of protecting the local market. The results of paragraph 2.28 would say the following: with the exception of groundnuts, the structure of protection at the existing exchange rate is equivalent to a devalued exchange rate (say 75 CFAF to 1 FF) for the local and regional market; and to a revalued exchange rate for exports to the EEC (i.e., the existing tax policy reduces local value added/unit to below its value at zero duties and tax). But quite clearly, in local currency, a devaluation would increase value added (in both local and export markets). 1/ It is useful to point out the difference between the NPC and the EPC. The first gives the relation between the ex-factory price and the world price, whereas the latter calculates the ratio of value added in market prices to that in world prices. Thus, consider the case of an export industry for which the NPC is less than the unity because of an export tax, but that the EPC is greater than the unity. In this case, we should conclude that, despite a policy which aims to tax exports, the Government is in fact subsidizing them.  - 32 - Given that the Government wishes to promote exports (particularly from the Industrial Free Trade Zone), it is clear that existing policy requires substantial revision in the sense of, at least, an equalization of the structure of protection so that EPCs and NPCs are the same in local, regional and EEC markets. 1/ 2.31 One might rejoin that industry needs protection on the domestic market in order to create jobs, the more so since unemployment is one of Senegal's most pressing problems; and that all industrial policy must be seen in this light. However, the creation of jobs depends principally on the growth of output 2/; and so, in the Senegalese context, this means encouraging the growth of exports since the local market is not likely to grow fast. Protecting only the local market, while doing nothing to promote the growth of exports, can of course, protect the loss of jobs (if it is a successful policy), but it is unlikely to lead to a high rate of growth of jobs. Moreover, a proliferation of job creating but economically inefficient firms imposes a deadweight loss on the economy and restrains its growth. Furthermore, beca!use of the high cost of these firms, they are unlikely to be able to export. Finally, these firms are very difficult to close down, because of the political impossibility of eliminating jobs. For all of these reasons, it is argued that the best way to increase employment is to alter incentives so as to make exporting more profitable. 2.32 The results of this analysis therefore corroborates statistically our conclusions of the last section: the structure of protection needs a complete overhaul. Specifically: (a) How much protection does the local market need at the existing exchange rate? (b) Which activities should be protected more than the norm, and for how long? (c) How should incentives be equalized between different markets? Such questions would take us beyond the scope of this study, but are obviously very important. 1/ The 1979 tax reform did reduce the differences in EPCs, but did little to improve the differences between EPCs in local and EEC markets. 2/ It is, of course, true that labor-intensive technology can create more jobs than capital-intensive technology; this is a different point, the importance of which we do not deny. Our concern in the above argument is with the impact of the system of protection on the rate of growth of jobs.  - 33 - C. A Proposal for Reform 2.33 The evidence accumulated from the preceding analysis suggests that the tariff structure provides for a protection of the local market of aLbout 50 percent. Now protection as high as this may not be needed, in which case, one would think that the way to eliminate the discrimination against exports would be to progressively reduce duties to oblige firms to be more competitive. 1/ However, there is universal concensus, at least within Senegal that industry would still need to be protected by about 25 to 30 percent, in which case significant discrimination against exports would still exist. In theory, a devaluation of the CFA franc, accompanied by appropriate tax and duty measures, would be a good way to restore the balance between producer margins in the local and export markets. The study assembled available evidence which might point to whether the offi- cial parity of 50 CFAF = 1 FF is appropriate - and it was concluded that a more realistic rate would be about 62.5 CFAF to 1 FF. The principal benefit of the lower rate would be to permit exports to make a far greater contribution than at present to improving the balance of payments, thus easing the share of the adjustment burden which will otherwise have to fall exclusively on general import reduction and import substitution in the agricultural sector. Also, the growth rate and employment would be raised. 2.34 Such a devaluation package would consist of three measures: (a) a pure devaluation with respect to the French franc; (b) modifications of import duties and exports to correct existing distortions; and (c) adjustments of the incidence of the sales tax and other taxes to get the right amount of revenue for the Government. The devaluation could enable Senegalese producers to export with greater ease than at present, because they should be able to reduce their prices in foreign currency, thus increasing foreign demand (assumed elastic) for Senegalese output. Duty and tax modifications are necessary to ensure that these potentially beneficial price reductions are not eroded by increases in the prices of locally produced goods and manpower used in exportables. This will be difficult in view of rising prices for imported goods. 2.35 The modification of duties and taxes must first attempt to keep Government revenues as close as possible to their existing level. Second, it must avoid an increase in protection to the local market, the existing level being ger,erally sufficient. Specifically, duties would be progressively harmonized to the same rate (lower than at present) and the level of the sales tax set so that budgetary revenues remain at the same level. The harmonization of duties would lead over the longer run to the reduction of producer margins in the local market, thereby reducing the 1/ The 1979 tax reform made some progress in this direction.  - 34 - discrepancy in the relative size of margins between the local and the foreign market; and so there will be a relatively greater incentive to export. If these complementary measures are successful, the reduction of export prices in foreign currency will lead to an increase in producer profits per unit of output in CFAF and to an increase in the volume of exports. 2.36 A devaluation is, however, formally ruled out by Senegal's membership in the Monetary Union. Although the country could devalue, such a move would require it to leave the Union, a move clearly undesirable since it would probably adversely affect investor confidence and credit- worthiness. 2.37 At the present exchange rate, however, prospects for export growth are not good. Existing policy aims at increasing exports, via the use of tax holidays and some indirect subsidies inside, and via the Invest- ment Code outside the Free Trade Zone. These measures are generally defective and have so far borne little fruit. It is the contention of this report that a viable economic alternative does exist, namely the use of uniform import duties and direct export subsidies, set at the same rate (with some exceptions) so as to approximate a lower exchange rate for industry,achievable by a devaluation package. The main components would be: Step 1. All exports will be subsidized, 1/ for example, 25 percent of the FOB value, with the exception of groundnuts, their derivatives, and possibly phosphates, for which the subsidies will be separately determined. The subsidies will not be given to goods exported to the CEAO. When locally produced exportables (such as cotton) are sold on the local market, they will be sold at 25 percent above the world price, regardless of whether the end product is for final consumption in the local market or for export. This is an important component of the package. Step 2. This step involves duty reductions at the existing exchange rate, with duties being levied at 25 percent on imports. Step 3. The level of the value added to tax will be determined so as to recover any revenue shortfall of steps 1 and 2, and in a way that will minimize the impact of the cost of living on the workers. 1/ If the export subsidy system is introduced, the system of temporary admission can be eliminated. Under the A.T. system, an imported intermediate good is admitted duty free. Under the FOB export subsidy scheme, an exporter pays the duty at the point of import, but receives the subsidy on the FOB value of the export of the finished good, at the point of export. Since the FOB value of the finished good is greater than the CAF value of the raw material, and since the export-subsidy and import-subsidy rates are equal in per- centage terms, the firm gets its money back at the point of export. Moreover, the fact that the duty is applied to the raw material gives an additional incentive to produce this commodity.  - 35 - Step 4. Some easing of the administrative control of imports is under- taken. The following analysis assumes protection of 25 percent, in. line with the estimate of the shadow exchange rate. Step 5. For some industries, additional temporary protection might be justified by dumping or by the inefficiency of local industry. In this case, the duty on competing finished goods would be set, after justifica- tion of the need for additional protection, at a higher than standard level for a fixed period, at the expiration of which the duty would be reduced to the standard level. (In Chapter I we suggested to use a subsidy on value added for this purpose.) Steps 1-4 constitute the essense of an export subsidy package which is equivalent to a devaluatLon of 25 percent, with duties set at zero. The "protection" to the locaL market is the same under both policies and therefore there is no discrimination against exports. Export Subsidies (Step 1) 2.38 Subsidies would be given at a flat uniform rate of the FOB value to all exports going outside of the CEAO (for which the TCR regime is equivalent to an export subsidy). For groundnuts and phosphates, special regimes would be devised, including the possibility that export taxes would be applied. There are several reasons for preferring uniform direct subsidies to indirect subsidies. First, direct export subsidies give a uniform premium to all earners of foreign exchange regardless of the sector they operate in. Second,, direct export subsidies increase margins only in export markets and not simultaneously in the local market, whereas many indirect subsidies make the local market needlessly more profitable; this can happen very easily when indirect subsidies are awarded via rebates of property taxes and of other taxes which are levied on total production rather than just on exports. Third, concerning the local production of exportable raw materials that are also used domestically, direct export subsidies can avoid the n,isallocative effect of low domestic transfer prices for these raw materials which discourages their production and stimulates direct exports of these goods. This is particularly important for cotton, groundnuts and phosphate rock which are both exported unprocessed and also sold to local producers for transformation into final products - also to be sold at home and abroad. 2.39 Consider the case of cotton. At the present point, the Government sells it to local spinners at the world price (at the existing exchange rate) if the finished product is to be exported, and at less than the world price for the local market. This policy clearly subsidizes the local production of finished teKtiles for the local market and does nothing to increase the profitability of exporting these goods. Moreover, it dis- courages the local production of cotton because of the lower prices paid to the farmer for domestically used cotton. With a uniform export subsidy, the price of cotton would rise in the local market by the amount of the subsidy regardless of whether the cotton were to be directly exported or to be incorporated into cloth destined for the local market or exports.  - 36 - Peasants would have an incentive to increase their production because of the higher price of cotton. Exports of, for example, grey cloth would be directly subsidized at the same rate as cotton which would more than offset the higher costs of cotton used for its production. However, grey cloth sold locally would become more expensive and would better reflect the real costs of this product. Thus, export subsidies would eliminate a relative disincentive to produce cotton, would put an end to an unjustified income transfer from the cloth, and would stimulate exports. 2.40 Some concern might be felt here about the effects of an increase in the price of cotton sold to spinners, to be transformed into cloth for the local market. The spinners' tendency to rely more on imported cotton which may be created by this price increase, can be dealt with, if necessary, by the adjustment of import duties. The disadvantages of the increase in the price of local grey cloth and finished cloth which may reduce local sales would have to be balanced against the benefits of increased exports of finished cloth and of the increase in the production of cotton. 2.41 It is prioposed that these export subsidies be given on the total volume of exports, rather than on, say, the marginal increase in a given year. This would avoid the problems that firms which begin to export (either existing firms, or new firms) would receive the incentive on all their exports since they start from a zero base, whereas firms which already export would only get it on the increase in exports. Now it could be argued that giving subsidies on existing (intra-marginal) exports is equivalent to giving money away needlessly, but this is not so. First, the proposed reductions in protection of the domestic market will reduce the profitability of the local market. Thus, the whole package will redistri- bute profits from the local market to exports and not increase overall profitability. Second, export subsidies given for all exports presents the equivalence of a devaluation package, which would involve exactly the same kind of intra-marginal redistribution of profits. 2.42 Past World Bank reports have suggested two apparently different methods of subsidizing exports. The 1974 Industrial Sector report sug- gested the use of direct subsidies, calculated as a percentage of the FOB value of receipts, one of the principal arguments advanced in its favor being administrative simplicity. A year later, an unofficial working paper of the Balassa research project suggested that it would be preferable to give subsidies calculated as a percentage of economic value added calcu- lated in world prices. These are in fact, practically identical proposi- tions, if FOB subsidies are implemented in a way that they increase value added of the production for exports by approximately the same percentage. This report advocates the latter idea: implementation of FOB subsidies, together with the complementary measures (to be discussed later) designed to insure that they do not create more distortions than they eliminate. 2.43 Finally, the financing of these export subsidies is proposed as follows. Assuming for the moment that the rate is set at 25 percent, exporters would receive for their exports endorseable export certificates for 25 percent of the FOB value of their exports, as soon as payment in foreign currency has been received; these certificates would be accepted  - 37 - for payment of import duties and value added tax. Such a mechanism should also indicate clearly a. second problem for the Government - the need to recuperate the loss in revenue. In the short run, however, revenue short- falls will be relatively small since imports are much larger than eligible exports. However, in the long run, compensating increases in other taxes will have to be implemented. The advantage of the above system is that it could be introduced on a branch by branch basis, starting with those branches which will show the best export response to this facility. Reform of Import Duty Structure and Sales Taxes (Steps 2, 3 and 4) 2.44 The present system of protection causes high prices for consumers because it permits relatively high prices for locally produced goods; on the other hand, it creates revenue losses for the Government, in the form of low tariffs on capital goods and raw materials. An important part of the export subsidy package is to correct these distortions, since their exis- tence discriminates against exports. It is proposed to use Steps 2, 3 and 4 to achieve these ends. Under Step 2, the structure of protection will be modified by gradually reducing the level of duties on finished goods to 25 percent, and increasing the rates of duties on raw materials and capital goods up to 25 percent. Under Steps 3 and 4, the level of the sales tax will be increased, and the control of competing imports eased so as to minimize the increases of consumer prices of locally produced goods. The latter is the key to this stability of consumer prices, as it implies increasing competition for the local production by imports (competition is a more reliable means oF keeping prices down than price control). The increase of import duties on raw materials and investment goods, and the increase in the sales tax insure that the Government recovers the revenue loss resulting from the reduction of duties on imported finished goods and export subsidies. 2.45 Higher import duties on investment goods, intermediate goods and raw materials may create problems in an economy which is exporting finished goods, since the tax on these inputs apparently amounts to tax on the exports. As stated before, an extended duty drawback system (including tax restitution rights for used equipment) would take care of this problem. If export subsidies are available (which comes in practice very close to an extended drawback system), this problem would not exist. Again, the relative prices would be very similar to those after a devaluation. 2.46 Thus, if the e!ntire export subsidy package is adopted, [t would be possible to levy duties at the same level regardless of the good imported. If the entire export package is not adopted, import duty harmonization should go together with a revision of the drawback system. 2.47 The harmonization should not be done instantaneously, but gradually to permit both industry and the Government to learn by experience. A gra- dualist approach would reduce uncertainty, enable targets to be adjusted in the light of experience and provide the entrepreneur with the time needed to adjust to the new set of incentives. The 1979 tax reform can be considered as a first step in the direction of a more fundamental reform, however, without substantial action on the export side.  - 38 - 2.48 This brings us to the last question concerning the use of export subsidies - would the international community accept them? Would Senegal be inviting retaliation by importing countries? International acceptability must be discussed in the context of the General Agreement on Tariffs and Trade rules (GATT), which govern international trade. Article XVI, Section A of the GATT states the obligation of any country which provides subsidies that directly or indirectly increase exports to notify the contracting parties (i.e., the GATT Secretariat) as to the extent and nature of the subsidization of the estimated effect of the subsidization on the quantity of the affected product or products...exported from its territory and of the circumstances making the subsidization necessary. Furthermore, "In any case in which it is determined that serious prejudice to the interests of any contracting party is caused or threatened by any such subsidization, the contracting party granting the subsidy shall discuss upon request, with the other contracting party or parties concerned, or with the contracting parties, the possibility of limiting the subsidization." 2.49 There is thus an obligation to notify and to discuss, but not to eliminate an export subsidy. Under Article VI, however, importing countries may offset the export subsidy by countervailing duties, set equal to but not exceeding the estimated subsidy. Nevertheless, GATT requires that export subsidies cause or threaten material injury to a domestic industry in the importing country and it exhorts developed industrial nations to refrain from imposing countervailing duties on exports from developing nations. 1/ In practice the U.S. imposes countervailing duties, but the EEC does not. 2.50 The relevant questions for Senegal are then whether a good case can be made for the use of export subsidies; and whether importing coun- tries (particularly those of the EEC) are likely to impose countervailing duties. Senegal, and the other members of the Monetary Union, in fact, have an excellent justification for the use of subsidies, in that they belong to a currency union, and so they do not have a national money, nor by implication, control over their "own exchange rate." Changing the CFAF exchange rate with respect to the French franc is extremely difficult, requiring the accord of all the Union's members. Finally, although it would be possible to get a national currency by seeking a new parity with respect to the French franc, it would clearly be economically very unsound to take such a measure if its costs were expulsion from, or rupture of the Monetary Union. Thus, membership in the Union has costs, in that Senegal loses control of a very important variable - the exchange rate. Export subsidies may be seen as an attempt to use import and export tax-subsidy policy as a surrogate, to affect balance of payments, production incentives, etc. This argument would become much stronger if the Government of Senegal were to make clear that export subsidies are but part of a wider policy 1/ The discussion of this paragraph and the preceding one is drawn virtually unmodified from Bela Balassa, Reforming the System of Incentives in Developing Countries, World Bank Reprint Series No. 22 and also published Vol. 2 June 1975.  - 39 - aimed at approximating the exchange rate which would prevail with an inde- pendent currency; and t.hat, it would be prepared to impose upon consumers in Senegal (and on itself via control of its expenditure policy) the same costs as would have to be borne at this independent exchange rate. The argument suggests strongly that GATT would not be easily able to condemn direct sub- sidies. 2.51 The likelihood of retaliation will depend on the country con- cerned and on the quantity of exports involved since retaliation in prac- tice requires according to the GATT that material damage or threat of damage be caused to a domestic industry. Senegal's main export markets in 1975 were the EEC (68%) and the CEAO (13%) of the remaining 20 percent, about half goes to European countries outside the EEC, and the remainder to the rest of the world. Since exports to the CEAO would be exclucled from the subsidy scheme, only exports to the EEC are significant. They totalled about 64.5 billion CFAF. Adding the other exports to the rest of the world, this gives total subsidizable exports of about 75.7 billion CFAF. Compared to total imports into the EEC or into the USA, for example, this is insignificant, and expanding exports would clearly not threaten any domestic industry. On the other hand, they could provoke a non-marginal increase in Senegal's exports. If the export subsidy scheme is organized as an extended duty drawbatk system, pressure groups within the EEC are less likely to make an issue out of it. The United States law does require that countervailing duties be imposed when the use of export subsidies has been established, but this market is, at the moment, of negligible importance in the geographical distribution of Senegalese exports. D. An Ad-Hoc Scheme for Export Subsidies 1/ 2.52 Such a scheme would still be based on the principle of increasing value-added/unit of output in export markets, but subsidies could be applied either to all or to selected industries. In this case, subsidies would be levied at different rates for different industries, and because of this, the administrative cost of the scheme increases with the number of activities included. However, if only two or three branches are subsidized to begin with, the administrative cost would not be prohibitive. A further advantage of second-besl: subsidies is that they do not require the simul- taneous enactment of import duty reforms 2/, nor do they require the abolition of "admission temporaire". However, if adopted on a wide scale, the administrative cost does increase considerably, rendering the scheme progressively less flexible and manageable. Consequently, to begin with, these subsidies should be used selectively. 1/ A similar scheme iEs also discussed in the 1975 working paper of the Bank's West Africa Regional Integration report. 2/ It should be added that the level of the subsidy would depend expli- citly on the level and profile of import duties for each industry; consequently, with each revision of import duties, it would be neces- sary to adjust the level of subsidies accordingly.  - 40 - Design and Calculation of Rates of Subsidy 2.53 Notwithstanding the use of subsidies in this second-best way, the guiding principle is the same as in the first best case: namely, subsidies should be chosen so as to equalize effective protection coeffi- cients in export markets for all activities at a rate about equal to the difference between the shadow exchange rate and the official exchange rate (in the present case, about 25 percent). It follows that the subsidy rates (expressed as a percentage of FOB value) will vary from sector, because of the uneven structure of protection to tradeable inputs. 2.54 The desire for equity between different activities (which is expressed by the equalization of the effective protection coefficients) means that one has to have a common reference point for all activities. We use value added expressed in world prices because this is the value added in an activity in the absence of all protection. We then multiply this by the shadow exchange rate; this gives us the desired value added in domestic prices. We then subtract from this value added in market prices to get the subsidy, which is then expressed as a percentage of the FOB value. Thus we proceed in three steps: (a) calculate value added/unit of output in domestic market prices; (b) calculate the desired value added/unit of output in domestic market prices; (c) express the difference, which will be the subsidy, as a percentage of the FOB value of export. 2.55 Example No. 1 Consider a firm using imported raw materials worth 2FF, CAF Dakar, to make a finished good worth 5FF, FOB Dakar. It also uses locally produced traded goods of 1FF (50 CFAF), say cans and cartons. We calculate the subsidy as follows: Step 1 Calculation of value added in domestic market prices Value of output 250 CFAF less raw materials 100 CFAF less local materials (traded) 50 CFAF equals gross "value added" 1/ 100 CFAF 1/ The concept of value added used needs clarification, since it does not correspond exactly to the sum of profits and wages. Rather, it corresponds to wages and profits and purchases of nontraded inputs (electricity, water, banking services, insurance, etc.). Thus one is in fact subsidizing something closely akin to a gross transformation margin.  - 41 - Step 2 & 3 Calculation of desired value added in market prices; and calculation of subsidy rate expressed as a % of FOB value This involves two steps: (a) calculation of value added in world prices; (b) multiplication by the shadow exchange rate to get the desired value added. The first step involves calculating the duty rates for imported goods and the nominal protection coefficients for locally produced tradeable inputs. For example, suppose that local cans are 25 percent more expensive than imported cans. We proceed as follows: value in value ini market prices NPC 1/ world prices Output 250 1.00 250 less raw materials 100 1.00 100 less local purchases 50 1.25 40 equals "value added" 100 110 The information in this table may be interpreted as follows: the firm actually had only 100 CFAF to cover additional costs of transformation such as: (a) nontraded goods, such as electricity, services, banking; (b) international transport, if the company is selling CAF; (c) local wages and salaries. However, if the company had imported all of its raw materials, the firm would have had 110 CFAF to cover these costs. Thus, the high costs of local inputs effectively reduces value added in exporting activities. If the Government wishes to subsidize value-added in world prices at 25 percent, in line with the excess of the shadow exchange rate over the actual exchange rate, desired value added would be 137.5 CFAF. But actual value added is onLy 100 CFAF, so a subsidy of 37.5 CFAF would be necessary. Expressed as a percentage of the FOB value, this would give a subsidy rate of 15 percent, and the total cost to the Government would be 37.5 CFAF/unit. 1/ NPC = nominal protection coefficient, or the extent to which the price of the local Input exceeds its price in the world market.  - 42 - Budgetary and Administrative Aspects of the Problem 2.56 It is not easy to provide accurate forecasts of the budgetary cost, but one may obtain a rough idea in the following manner (based on exports for 1975). Assuming that this subsidy regime will apply neither to groundnut products, nor to phosphates, nor to exports within the CEAO, total subsidizable exports would be 23.6 billion CFAF calculated as follows: Table 10: EXPORTS FROM SENEGAL - 1975 (in billions of current CFAF) Total 99.1 less Groundnuts 40.3 less Phosphates 22.2 less CEAO exports 12.94 equals total subsidizable exports 23.66 However, it is the value added contained in, rather than the total of exports of 23.6 billion CFAF which is being subsidized. Now, available evidence-suggests that on an industry-wide basis, value added (as we define it here, i.e. wages and profits and purchases of services and nontraded goods) is about 30% of gross value. 1/ Total value added is therefore roughly 7 billion CFAF (23.66 x 0.30), and a subsidy of 25 percent would cost 1.75 billion CFAF on an annual basis, if applied to all activities. 2.57 Since we have already argued that global application would be administratively impractical, it is interesting and useful to calculate the cost by principal category of export. Using the same methodology, we obtain the results as shown in the table below, which can be used to estimate the cost of a scheme judged administratively feasible. Table 11: COST OF EXPORT SUBSIDIES BY PRINCIPAL CLASS OF EXPORT Exports Value added share Cost of subsidy (CFAF billions) Fish 7.3 30% 0.548 Textiles 3.48 35% 0.305 Transport 0.500 0.17% 0.021 Metal products 3.300 35% 0.289 Other (a) Agriculture 2.19) (b) Industry 4.42) estimated as a Miscellaneous 2.47) residual 0.587 Total 23.66 1.750 1/ Obtained from national income accounts and World Bank studies of the incentive system.  - 43 - Suppose, then, that rapid implementation is decided for fishing, textiles, transport equipment and metal products, total cost would be roughly one billion CFAF per year, and even this is an overstatement for the following reasons 1/: (a) the creation of new output means certainly more profits in the short run, and the creation of new capacity and so more jobs in the long run. Both contribute to an increase in direct tax receipts; and (b) there is clearly sufficient elasticity in the tax revenue base (particularly by reducing import duty exemptions) to come up with some additional receipts which would eliminate it. 2/ In short by selective use of subsidies, it would easily be possible to devise a program of which the net cost does not exceed one billion C:FAF. 1/ To this, we may adcL that there is almost certainly a double counting problem in the estimates, because some of the exports of textiles, for example, in fact go to CEAO countries. But the data did not permit more refined calculations in the time available. 2/ The total cost of these exemptions for 1976 was at least 20 billion CFAF. It would certainly be possible, and indeed desirable, to reduce them, since for the most part it is the local market which is over- subsidized. A reduction in this of 1 billion CFAF would be quite practical, and if the proceeds were used to subsidize exports, the net budgetary cost of the export subsidy scheme would be zero.  - 44 - CHAPTER III. SECTOR ANALYSIS 3.01 The analysis of this chapter is based partly on the preliminary results of the World Bank study on Incentives and Regional Integration in West Africa, and on more recent information collected in the Basic Economic Mission, and other missions effected by the author of this report. The methodology used is the domestic resource cost of foreign exchange variant of cost benefit analysis, in which an activity is said to be economically profitable if the cost of earning a unit of foreign exchange is greater than the shadow exchange rate, that is, the official exchange rate adjusted for any over or undervaluation. Using this methodology, the economic profitability of a sample of firms is assessed. Principal Conclusions 3.02 Groundnuts - The Government should be careful not to overinvest in new capacity 1/, since all available evidence indicates that installed capacity is largely sufficient to process average crops (as estimated from production figures of the last twenty-five years). Moreover, the Govern- ment should make sure that the protocol between SONACOS 2/ and the mills has a clause which requires the mills to make sealed bids for the ground- nuts they are to process. 3.03 Phosphates - All evidence points to the fact that Senegal is a marginal producer of phosphate. Because of this the Government's decision not to divide Taiba into two separate concessions is clearly justified. Sequential operation of Taiba into Tobene is preferable, not only on cost grounds, but also because it will prolong the economic life of the mine, a very important consideration for Senegal. The phosphoric acid plant seems to be a project worth pursuing, which is confirmed by recent analysis. 3.04 There are excellent prospects in fishing--this is one of the sectors the Government should aim to develop rapidly. 3.05 Textiles - Emphasis should be based on developing exports of knitwear and finished clothes, via the use of export subsidies. Their use would resolve the problem currently posed by imports of used clothes, since the subsidies would allow firms to divert their production from the local market to exports. As far as grey cloth and finished cloth production are concerned, the reasons for the current high costs of these industries should be investigated and the problems resolved before new investments in these activities are undertaken. One of the problems is almost certainly 1/ Nonetheless replacement investment of existing capacity (the new 200,000 ton factory at Diourbel) is desirable given that it will redistribute jobs away from Dakar following the closing of Petersen (an obsolete crushing factory in Dakar). 2/ Societe Nationale de Commercialisation des Oleagineux du Senegal.  - 45 - overmanning, which could be substantially reduced by encouraging management and unions of these firms to reduce their work-force, the released labor to be employed by SOTEXKA, when this integrated spinning and knitwear plant hopefully comes on stream in the beginning of the 1980s. Another problem for existing spinning and weaving firms is excessive protection of the local market; it seems clear that some increase in efficiency could be obtained by lowering production to the local market and increasing incen- tives to export. Nonetheless, in developing textiles, the Government should be aware of protectionist sentiment prevailing in the EEC about :Lmports from LDCs, even when these are signatories to Lome. Although the EEC officially continues to subscribe to the Multilateral Fiber agreement, :Lt has begun to engage in bilateral negotiations with textile exporting nations to effectively put a ceiling on imports from these nations. Senegal should therefore attempt, if the need arises, to negotiate a ceiling which allows for considerable growth above its presently small level of exports. This is, however, not likely to be a problem in the short run. 3.06 Foodstuffs and agrobusiness in general - The most important point is probably that the Government should ensure that agricultural producer prices are not set so high that industrial processors are unable to compete either in the local or international markets. The Government should also be aware of the potential costs of granting de jure monopoly power to firms in this sector (as has been done in the case of the local sugar company (CSS), for example). Either the Government should allow competition of imports or it should establish a tight ex post control over the operation. 3.07 Industrial Free Trade Zone - It is as yet too early to judge whether it is a success or not. There are, however, two problems which are slowing its rate of growth. First, the "conditions of entry" are too stringent because they foreclose all possibilities of attracting small- or medium-size firms, which are the most likely to wish to invest in Senegal. Serious thought should 'be given to finding a more satisfactory selection criterion. Second, the overvaluation of the exchange rate means that producer margins are probably too small, and therefore that the rate of growth of fLrms within the Zone is smaller than at a more realistic exchange rate. 3.08 Despite these difficulties, the industrial sector may be expected to grow at about 6.8 percent over the medium term. A. The Economic and Social Profitability of some Industrial Activities in Senegal 3.09 In the preceding chapter, we analyzed the effect of the incentive system on value added irn domestic market prices, VAd, in relation to economic value added measured in world prices, VAw. The purpose of this chapter is to assess the real social and economic costs of creating this economic value added by estimating the domestic resource cost of foreign exchange (DRC). Thus, in addition to the effective protection coefficient (EPC) which discusses the impact of the tariff system on the incentive to produce in a particular activity and how VAw is distributed among the  - 46 - factors of production employed in that producing activity and the rest of the community, the DRC addresses the question of whether a particular good should be produced with the factors of production at Senegal's disposal. 1/ 3.10 Two principal measures were calculated: the unadjusted domestic resource cost of foreign exchange, and then an adjusted version (of which two estimates were made reflecting different treatment of foreign profits); in total then, three measures were used. The unadjusted resource cost of foreign exchange is the real factor cost of earning a unit of foreign exchange (or a unit of value added, measured in world prices) on the assumption that all factors of production are Senegalese. It is estimated from the effective protection coefficient by replacing domestic value added in world prices (the numerator of the EPC) by domestic value added (DVA) at factor cost, measured in opportunity cost terms (or shadow prices 2/). Thus: DRCU = DVA at factor cost in shadow prices DVA in world prices (at the existing exchange rate) The adjusted resource cost of foreign exchange uses the same shadow prices, and it takes account of factor ownership and specifically of the fact that foreign profits and wages are not benefits to Senegal. Thus, it is the ratio of national value added (NVA) at factor cost (in shadow prices) to national value added in world prices, 3/ or DRCA = NVA at factor cost in shadow prices NVA in world prices (at the existing exchange rate) The difference between the two adjusted versions is as follows. In the first, the DRC was calculated including foreign profits at their observed 1/ In other words, we can say that effective protection deals with the impact of the incentive system and market prices on the actual allocation of resources. The next step is to ask whether these incentives have fostered the implementation in Senegal of economi- cally, rather than just privately profitable activities. This is exactly the purpose of cost-benefit analysis whatever its form (the DRC method or the effects method). 2/ Two shadow prices are needed at the present point. For capital, a central value of 13.5% was chosen. For labor, a shadow price of 60% was used for unskilled African labor, since for all other categories, it was found that wages and opportunity costs were roughly in line. 3/ National value added in world prices is calculated by subtracting from DVA in world prices the sum of foreign after-tax profits and the foreign exchange cost of expatriate salaries. National value added at factor cost is obtained by eliminating from the numerator of the unadjusted DRCU the cost, in shadow prices, of foreign owned capital, and foreign wages; and by adding to it, the national factor cost of expatriate consumption.  - 47 - value. In the second, foreign gross profits were recalculated and assumed to be equal to 19.5 percent of paid-up capital plus revenues and that firms expect at least this rate of return before agreeing to invest in a country. This procedure was used because it was felt that foreign firms aim to amortize their investment in five years. The corresponding rate of 19.5 percent may then be considered as a normalized rate of foreign profit. Observed rates of profit were of course sometimes greater and sometimes 'Less than this normalized rate. In the first case, great care must be taken ascribing any uninformed disapprobation to this finding; rather it is to be explained. There are several possible causes: (a) the firm may have been exceptionally efficient with the high profit rate reflecting this fact; (b) the firm may be exporting a natural resource in a year with except:Lonally favorable prices (1974 for phosphates, for example); (c) the high prDfits may be the result of the market structure and/or excessive protection, etc. By the same token, an observed rate of less than 19.5 percent also requires explanation. 3.11 These concepts were used to analyze the following questions: (a) Which activities are profitable for Senegal at observed rates of capacity utilization? The question is important since there is in fact a lot of spare capacity in some branches. (b) Does passage to full capacity make a difference to these results? (c) For activities which are unprofitable for Senegal, what should the Government do in the short run? Thus, we first measure the resource costs of foreign exchange for each of the activities in the sample, and this provides useful information in itself. Second, in order to make the normative judgments required to answer the questions above, we must compare the DRCs to a benchmark, namely the true value of foreign exchange to the Senegalese economy. This is calculated as the shadow price of foreign exchange, and we have provisionally chosen it at 62.5 CFAF, i.e. 25 percent above the existing parity. 1/ An 1/ The research on this point has not yet been concluded, but most likely the final results will indicate an overvaluation somewhere between 15 and 35 percent.  - 48 - activity is economically profitable for the country if the DRC is less than 1.25 and unprofitable if more than that level. The Results 3.12 Table 12 gives the results, by activity for all versions of the DRC, both at observed and full capacity. The first point to be noted is the wide dispersion of the results under all measures around the benchmark shadow price of foreign exchange. The second point is that some of the adjusted measures are negative and others positive. A negative DRCA means that this activity cost rather than earned foreign exchange for Senegal after taking into account the distribution of benefits between the foreign shareholders and the country. 1/ 3.13 The next point is to note the relation between the observed DRCs and the shadow exchange rate (shown below for the whole sample). Table 13: SAMPLE WIDE ESTIM1ATES OF THE RESOURCE COST OF FOREIGN EXCHANGE DRCU 11 DRCFA /2 DRCA /3 actual capacity 1.05 0.92 0.89 full capacity 0.93 0.72 0.95 /1 Domestic Resource Costs Unadjusted. /2 Domestic Resource Costs Fully Adjusted (normalized foreign factor costs). /3 Domestic Pesource Cost Adjusted for foreign factor costs. We see from Table 13 that all measures of the DRC are less than the shadow exchange rate; this would appear to demonstrate that the Senegalese industry is efficient in economic terms. However, it should be noted that this kind of analysis does not lend itself well to aggregation, since aggregate results depend upon the coverage of the sample. Mloreover, it is dangerous to infer macroeconomic conclusions from sample-means, without making some reference to the dispersion of the results about the means. In the present case, in fact, the dispersion is rather large reflecting the fact that some activities are very profitable and others unprofitable. Thus, for particular activities we do find that excess profits and underutilization of capacity are costly in economic terms; and in order to highlight the dispersion of the results, most 1/ Adjusting foreign profits to their long-run supply price sometimes improve the results. As pointed out in the preceding paragraph, one must be very careful to avoid erroneously maligning foreign firms.  - 49 - Table 12: INCENTIVE AND RESOURCE COST MEASURES 1972-74 Resource Cost of Foreign Exchange _ FuL Capacitv Utilization DRC (_) PRODUCT Unadiusted Adjusted for Foreign Factors nad justed Adjust for Foreign Factors OR No Excess Excess No Excess Profit Excess Profit ACTIVITY Profit Profit Phosphates I 1.05 0.52 0.54 1.05 (1.05) 0.57 (0 57) 0.67 (0 67) Phosphates II 0.74 0.55 0.50 0.74 (0.74) 0.55 (0.55) 0.50 (0.50) ExtractLve Mining 1.01 0.52 0.54 1.01 (1.01) 0.57 (0.57) 0.64 (0 64) Flour Milling 1 0.99 1.00 -2.12 0.73 (1.22) 0.44 (1.27) -1.32 (-1.04) Biscuits 0.70 0.62 0.54 0.69 (0.70) 0.61 (0.62) 0.53 (0.54) Food 2.14 -10.74 2.82 1.47 (2.29) 2.55 (-12.32) 8.86 (12 82) Flour Milling II 1.04 1.11 2.58 0.80 (1.14) 0.61 (1.18) 3.69 (6.00) Fish 0.86 0.82 0.82 0.48 (0.62) 0.33 (0.47) 0.45 (0.56) Fish Products 0.91 0.94 0.76 0.78 (0.95) 0.66 (0.88) 0 72 (0.78) Tunafish 0.48 0.29 0.25 0.47 (0.49) 0.28 (0.30) 0 25 (0.25) Food Processing 0.64 0.50 0.45 0.50 (0.56) 0.32 (0.38) 0.33 (0.35) Milk 2.11 -0.77 1.22 0.87 (1.03) 0.62 (0.95) 1 06 (I 14) Vegetables 1.80 3.48 1.01 1.16 (1.32) 0.80 (1.01) 1 00 (1.04) Other Focd - Tobacco 1.90 -6.12 1.07 1.06 (1.22) 0.75 (1 00) 1.02 (1.06) Cotton Thread 0.88 0.92 0.95 0.88 (0.88) 0.92 (0.92) 0.95 (0.95) Cloth I 1.67 3.50 1.81 1.59 (1.69) 2 61 (3.20) 1.85 (1.88) Printed 1.24 13.71 20.45 1.15 (1.22) 3.09 (5.10) 52.11 (71.01) Cloth II 2.18 -3.17 -1.73 1.72 (1.86) 10.26 (-62.62) -1.32 (-1.32) Clothes Manufacturing 0.82 0.72 0.71 0.66 (0.76) 0.56 (0.67) 0.64 (0.71) Bags 1.31 1.66 1.20 1.18 (1.25) 1.30 (I 45) 1 18 (1.22) Sewing Th:ead 0.94 0.95 1.30 0.65 (0.90) 0.53 (0.87) 1 22 (I 74) Millinery 1.05 0.99 1.00 0.86 (0.98) 0.68 (0.86) 0.98 (1.04) Textiles q Clothing 1.33 2.97 2.56 1.20 (1.29) 1.77 (2.27) 2.66 (2.85) Lubricants 2.04 -20.43 -1.01 1.29 (3.78) 1.09 (-1.28) -0.60 (-0.60) Petroleum Refining 2.04 -20.43 -1.01 1.29 (3.78) 1.09 (-1.28) -0.60 (-0 60) Soap 0.96 1.36 1.19 0.83 (0.98) 0.85 (1.32) 1.13 (1.34) Paints I 1.31 6.76 6.45 1.04 (1.21) 0.90 (1.44) 10 15 (-24 44) Detergents 0.85 0.57 0.76 0.48 (0 56) 0.23 (0.28) 0 72 (0 76) Paints II 1.25 1.92 7.48 0 95 (1.09) 0.55 (0.77) 11.11 (-12.39) Plastics 2.10 -2.97 -1.93 1.48 (1.76) 3.75 (19.42) -2.11 (-1.84) Matches 2.75 10.61 -2.17 1.27 (1.54) 1.83 (2 96) -0.45 (-0.44) Fertilizer 1.30 1.68 0.81 0.89 (1.00) 0.58 (0.72) 0.80 (0.82) Chemical Irndustry 1.32 2.29 1.70 0.92 (1.06) 0.72 (0.97) 1.95 (2.26) Cement 1 1.50 3.24 1.12 1.50 (1.50) 3.24 (3.24) 1.12 (1.12) Cement II 0.62 0.43 0.38 0.62 (0.62) 0.43 (0.43) 0 38 (0 38) Construct. Mat.- Glass 0.70 0.53 0.44 0.70 (0.70) 0.53 (0.53) 0.44 (0.44) Trucks 0.74 0.64 0.47 0.52 (0.91) 0.35 (0.87) 0.39 (0 49) Agriculture Equipment 0.68 0.58 0.62 0.56 (0.67) 0.47 (0.59) 0.57 (0 66) Transport Equipment 0.70 0.60 0.57 0.55 (0.72) 0.44 (0 63) 0.52 (0.62) Enameling 0.75 0.56 0.46 0.75 (0.75) 0.56 (0.56) 0 46 (0.46) Other Mecan. - Electr. 0.75 0.56 0.46 0.75 (0 75) 0.56 (0.56) 0 46 (0.46) Paper 1.18 1.08 1.10 0.51 (0.59) 0 26 (0.32) 0.86 (1.05) Other Industries 1.18 1.08 1.10 0.51 (0.59) 0.26 (0.32) 0.86 (I 05) Sample Total 1.05 0.92 0.89 0.93 (1.01) 0.72 (0.89) 0.95 (0.99)  - 50 - of the following study is based on analysis of sub-samples (for which the aggregation problem remains, though it is of less importance) and of individual activities. 1/ 3.14 Before proceeding to this analysis activity by activity, we present the results of an experiment which examined the effects of classifying firms of the non-groundnut sector according to the amount of production they exported; firms that exported more than 40 percent of their production were considered export-oriented activities, and those that exported less than 40 percent, import-substitution activities. This experiment showed import-substitution activities to be less profitable than export-oriented activities. As can be seen from Table 14, the DRC is always higher in import-substitution activities than in export-oriented-activities. Particular attention should be paid to the results of the last column for this estimates the costs and benefits of the observed situation. The high value of DRCAs (1.51) for import-substitution activities indicates clearly that these activities are costly for the economy and that measures are needed to correct this situation, since the lack of economic profitability of this subsector reduces the overall profitability of the industry. Table 14: DOMESTIC RESOURCE COST OF FOREIGN EXCHANGE BY SUBSECTOR: EXPORTING vs. IMPORT SUBSTITUTION DRCU DRCFA DRCA Actual Capacity Export-oriented 1.01 0.68 0.65 Import-substitution 1.10 1.56 1.51 Total 1.05 0.93 0.90 Full Capacity Export-oriented 0.91 0.57 0.69 Import-substitution 0.96 0.99 1.55 Total 0.93 0.72 0.95 1/ It is important to reiterate the provisional nature of these conclu- sions, and mention should be made of an important analytical prob- lem. Much capital equipment is old in Senegal which means that the cost of capital may be, in economic terms, artificially low. This reduces the DRC, and may create the impression that it is economically profitable when in fact it is not. There is no easy way to deal with this problem, save to try to ensure that one's result makes good sense when assessed in the light of all available evidence. This is the approach taken in this report. Finally, note that the size of the sample is too small to permit easy generalizatation of conclusion at a macro- level - that's why most discussion concerns the individual activities.  - 51 - 3.15 The low profitability of the import-substitution sector can be primarily attributed tc the size of the local market, which is much smaller than Senegal's principal export market, the EEC. This means that the Govern- ment should be aware of encouraging import-substitution projects, unless it is clear that they are economically profitable. A preponderance of unprofitable import-substitution activities will make the whole industrial sector more economically unprofitable as the share of import-substitution activities in total production rises. Its unprofitability and high cost structure (implicit in its dependence on a highly protected local market) will be passed on, via higher prices to export-oriented activities, making exporting more difficult aS costs rise; thereby, the economic profitability of the whole sector is reduced. 3.16 There are at present few prospects for further large scale substitu- tLon of presently impori:ed consumer goods. When consumer purchasing power has rinsen enough to merit the creation of new economically profitable units oriented towards the local market. Such an increase in purchasing power can best be obtained in the short and medium run by export-oriented industry-- which will also create new import-substitution activities for semi-finished goods and raw materials. In other words, it will contribute to the creation of a more "dense" industrial sector. In short, our recommendation is one of judicious balance between import-substitution and export-oriented industries. Concentrating most new investment in the latter (provided of course that they are economically profitable) will raise the overall economic profitability of the sector, as the share of unprofitable import-substitution activities in total production falls. Moreover, this will raise local purchasing power (and thus demand for import substitutes) faster than a policy based on more import substitution, because there are so few profitable substitution possibilities left. Analysis of the Results by Activity 3.17 The next step is to analyze the results of the level of each activity. The objectives were to: (a) classify activities as economically profitable or not, (according to all criteria); (b) see which kind of firms fall into the export-oriented and import-substitution sectors; (c) relate profitability to the structure of incentives as analyzed by the effective protection concept; and (d) seek plausible real world economic explanations for our findings.  - 52 - This resulted in a subdivision of the whole subsample into three subsamples: (A) Economically profitable activities: phosphates, biscuits, fishing, cotton yarn, finished clothes, knitwear, detergents, cement, truck assembly, agricultural machinery and enamel kitchenware. Every firm in this group has the following properties: (a) at observed rates of capacity utilization, all versions of the DRC are less than the shadow exchange rate, both when foreign profits are normalized and when they are not; (b) at full capacity, all the DRCs are less than the shadow exchange rate; again, regardless of whether foreign profits are normalized or not. Furthermore, eleven out of fourteen firms were in the export-oriented subsector (the exceptions being truck assembly, cotton yarn and detergents). Note also that phosphates and fish were sold primarily in markets outside the CEAO and were subject to real (and/or implicit) taxation of exports in that the EPCs were less than one; local sales were insignificant for these firms because of lack of demand. The remaining activities (about 50% of value added) export principally within the CEAO, with positive incentives (generally with EPCs greater than 1), and they also sell a significant share of output in the local market (also EPCs greater than 1). In all cases, the EPC in the export markets for these firms was less than in the domestic market, so that there was a stronger incentive to sell at home than in any export market. (B) Economically unprofitable activities: spinning and weaving of finished cloth, grey cloth, matches and plastics. This group of activities has the following properties: (a) at observed capacity, DRCs are either negative or greater than the shadow exchange rate, both when foreign profits are normalized and when they are not; (b) at full capacity, the DRCs are either negative or greater than the shadow exchange rate, regardless of whether foreign profits are normalized or not. With the exception of matches, all of these firms are in the import-substitution subsector with a structure of incentives favoring the local market and regional exports (EPCs greater than one), with the local market being additionally protected by total bans or selective import controls ("autorisations prealables"). These firms had either the largest excess profits or financial losses. Moreover these findings occurred both at observed and at full capacity.  - 53 - (C) Activities which cannot be classi- flour milling, lubricants, asbestos, fied as definitely profitable or cement, milk, vegetables, soap, unprofitable: paints, fertilizers, sewing thread, jute sacks, printing and dyeing of grey cloth. Most of these are in the importing-substitution subsector, and closer analysis reveals the low level of capacity utilization, and the high level of profits cause the activities to be economically unprofitable. Only the simultaneous elimination of these two problems make the DRCs less than the shadow exchange rate (with the exception of jute sacks). This points to an important, often unrecognized, danger of import substitution: the creation of protected firms simultaneously tooa.large for the local market, and uncompetitive in export markets. They are thereby condemned to long periods of underutilization of capacity with its associated economic costs, even though at full capacity, they would be economically profitable activi- ties. 3.18 The analysis of the groundnut sector is still incomplete and so unreported. The principal factors which influence its economic profit- ability are: (a) the refining margin in world rather than domestic prices; (b) the level of ,apacity utilization; and (c) the level of Eoreign profits. Of most importance is the refining margin, defined as the difference (in world prices FOB Dakar) between the price of one kilo of refined produce and one kilo of shelled nuts. This margin determines whether positive or negative value added is created for Senegal. The evidence is that this refining margin fluctuates widely and is very thin. Small variations in the applied weights (for example, because of water added to the groundnut cakes) can easily lead 1:o negative margins in this calculation that do not correspond to the real profitability of refining activities. Second, as in any industry, the level of capacity utilization makes a considerable difference to profits, but in this case, capacity utilization is a function of the local crop (since raw groundnuts are de facto never imported into Senegal). Under these circumstances, Senegal's crushing capacity should stay well below the size of the crops in good years to assure full capacity utilization in the mediocre years, while exporting the incidental surplus over the reduced crushing capacity in unprocessed form. Conclusions of this Analysis by Activity 3.19 In paragraph 3.16, it was pointed out that future growth of the industrial sector should be based on a judicious balance between export pro- motion and import substitution, but with some preference for exports. The activity analysis just completed indicates in which activities future growth could or could not be profitably based, subject to various contingencies to be discussed in the next section. Leaving out activities of secondary impor- tance (judging by the size of the branch in which they operate), profitable  - 54 - activities are phosphates, fishing, cotton, yarn, finished clothes, knit- wear, cement and agricultural machinery. Care should be taken, however, before investing further in printing and dyeing of grey cloth, vegetables, fertilizers, groundnut processing - all of these activities have not yet established unequivocably their economic profitability. Finally, extreme caution should be taken before investing in new capacity in spinning of grey cloth, and in the spinning and weaving of finished cloth, since these activities do not appear to be economically profitable using existing techniques of production. B. The Industrial Sector Component of the Fifth Plan Problems and Prospects of Phosphate Development 3.20 At the present point in time this sector consists of three firms: la Compagnie Senegalaise des Phosphates de Taiba (CSPT) which produces calcium phosphate; la Societe Senegalaise des Phosphates de Thies (SSPT), which produces calcium aluminum phosphate; and la Societe Industrielle des Engrais du Senegal (SIES), which makes phosphatic fertilizers. CSPT exports most of its production directly, but sells some rock (up to 100,000 tons) to SIES. SSPT sells practically all of its production (under the brand name Phosphal) for export as direct application fertilizer, or for use in the production of animal feed. SIES keeps local rock from CSPT, which it mixes with imported inputs to make fertilizer for the local market and for export primarily to Mali. 3.21 As already explained in Chapter I, the Government originally intended to develop this group of industries within the framework of the Cayar Complex. This complex was originally designed to consist of the following elements: (a) The division of the phosphate deposit into two separate mining operations. (b) The construction of a phosphoric acid plant in the Free Trade Zone. (c) A new refinery at Cayar. (d) A new power plant. (e) A port. (f) A model city. 3.22 The decision to redesign the project's phosphate components is sensible and based on three factors: (a) the size and structure of the phosphate deposit; and (b) the cost of mining operations in Senegal relative to international prices likely to prevail over the next ten years, because of the probability of excess supply in world markets.  - 55 - The Deposit 3.23 The Senegal d'eposit is small: 70 million tons of marketable concentrate compared to 20 billion in Morocco alone. Moreover it is deeper to work than that in Morocco. The following table compares their principal characteristics: Table 15: CHARACTERISTICS OF PHOSPHATE ROCK DEPOSITS IN SENEGAL Senegal Morocco Taiba Tobene Reserves (million tons) 30 1! 70 20,000 Thickness of deposit 5-15M 6-7M Average 8M 6M Overburden /2 25M 35M /1 Original deposit to be exhausted in about 5 years. /2 The overburden is t:he layer of sediment covering the phosphate rock deposits. Operating Costs 3.24 Costs of production in Taiba have been estimated at about $25 per ton, much higher than in other parts of the world; for example, about $20 per ton in Morocco and Jordan. 1/ On the other hand, Taiba's finished product is of very high quality, about 82.5 percent BPL, for which buyers have been in the past prepared to pay a premium of about 10 percent of the current world market prices. The structure of Taiba's costs is revealing: Table 16: TAIBA'S COST STRUCTURE US$/MT Mining 7.5 Beneficiation 9.0 Dyeing 1.5 Transporit 2.4 Overhead 2.0 Loading 1.0 Finance 2.0 Several points should be noted. First, direct mining costs are high due priLncipally to a large overburden. Second, beneficiation costs are high. 1/ This may well explain why the DRC of phosphates is high.  - 56 - This is one of the mission's most important findings. Taiba's high quality produce is due to repeated processing of the ore. This implies that the recovery rate is extremely low, 30 percent compared to 50-60 percent in Togo, 60-70 percent in Jordan and 90 percent in Morocco. At present there are about 20 million tons of slimes (schlams) with more than 50 percent P205 content. If the recovery rate were as high as 50 percent, one could get an additional 800,000 tons of phosphate (at 67% BPL) out of these slimes, usable for example, in a phosphoric acid plant. Third, transport and loading costs are high (on a per ton basis), due to inefficiencies of the railway system and underutilization of the port's storage capacity (the port of Dakar is thus not a constraint on more exports). Fourth, finance charges are low because Taiba's equipment is fairly well depreciated. All in all at a world price of $35/ton, the mine is clearly a profitable operation and Taiba can afford to pay 3.4 percent export taxes. 3.25 However, for a mine operated with undepreciated equipment, profit- able operation is not nearly so sure. Estimates put financial costs at $10 per ton, so that full costs of production would be closer to $35/ton on the assumption that operating costs are the same as those of Taiba. However, the fact that Tobene has a larger overburden, would imply a higher operating cost, and this would clearly drive costs above current forecasts of phosphate prices over the next decade which are shown below: Table 17: FORECASTS OF WORLD PRICES FOR MOROCCAN PHOSPHATES (US$/ton) Year 1975 1976 1977 1978 1979 1980 1981 1982 1985 1990 Current Prices 67.0 36.0 30.5 29.0 33.0 36.8 41.5 45.3 60.1 76.7 1978 Prices 86.0 45.5 35.1 29.0 31.0 32.0 34.0 35.0 39.0 39.0 It is also useful to add to this that many producers have much lower variable costs than in Senegal as indicated by the following table: Table 18: VARIABLE COSTS PER TON FOR SELECTED PRINCIPAL PRODUCERS (US$/ton) Florida 12-15 Morocco 11-12 Tunisia 22-23 Jordan 16-18 Togo 14-16 Senegal 22-25 Because of its high level of recurrent costs, in a glutted market, Senegal is likely to be significantly more affected than other cheaper producers. Furthermore, when demand for fertilizers is low and fertilizer factories are working below capacity, the advantage of high grade phosphate rock is much less than in the opposite situation, and thus the premium Senegal normally obtains will diminish substantially.  - 57 - .26 The chances of a glutted market should not be underestimated given the structure of the world market. Typically, demand builds up to full capacity utilization and causes then a gradual rise in prices. In 1973, this process was accelerated by a worldwide shortage of cereals, with steep price increases for this commodity, and subsequent increases in prices for fertilizers. The price rise led to a sharp increase in capacity and fall in demand. 1/ In 1975, the fading demand for cereals caused a drop in phosphate rock prices of 46 percent. The new price level is still profit- able for Senegal, but it will take a long time before the huge additional world capacity that was in the process of being constructed in 1975 will be fully utilized. Since Morocco is in a position to lead the world market, one can, nevertheless, assume that a major collapse of the market can be avoided. 3.27 All of the evidence therefore points to the fact that Senegal is a marginal producer of phosphate. Because of this, the Government's decision not to divide Taiba into two separate mines is clearly sensible. It is evidently prefera;ble to allow sequential operation of Taiba and Tobene by the same company. This may also be desirable for another so far unmentioned reason; simualtaneous exploitation of both deposits would cause Senegal to exhaust its reserves in only 25 years, leaving the country with a need to import phosphate and with some 3,000 miners unemployed. The short run gain in employment (1,200 jobs) does not seem to justify the longer run economic and social costs, particularly when the same number of new jobs could be created in other sectors at considerably lower financial costs. For example, the cost/job created for Tobene is $175,000; whereas it: is only $20,000 for the new textile plant in Kaolack, and only $4,000 for job for expansions :in capacity by existing textile firms. 1/ Phosphate fertilizing can easily be interrupted for a few years without major losses in yields.  - 58 - The Phosphoric Acid Plant 3.28 The second element of the phosphate sector we must look at is the phosphoric acid plant. After the Government had rightfully dropped the plans to start a new and independent phosphate mining operation in the framework of the Cayar Complex, the one good component in this scheme -- the plan for a phosphoric fertilizer plant -- had to be entirely redrafted. The ICS 1/ project was redesigned with EMC 2/, a French Government-owned enterprise which controls most of the West African fertilizer market outside Senegal, where 61 percent of the production would have to be sold. The proposed project would produce 198,000 tons of phosphoric acid a year, all of which would be transformed into solid fertilizer (346,000 tons/year of MAP, DAP and TPS). This project would give the Senegalese phosphate sector a much stronger link with the ultimate users of its products and make the sector, therefore, less vulnerable. Since the factory would use one-half phosphate rock and one-half phosphate pebbles (which is a less upgraded quality of phosphate ore), the project would not be burdened with Taiba's exceptionally high beneficiation costs. The use of pebbles also increases the recovery rate of the mining operation, one of the objectives mentioned in the earlier paragraphs. Moreover, it would free beneficiation capacity which is the bottleneck in the existing mining operation; for this reason, the existing exports of phosphate rock from the mine would not have to suffer from the roughly 400,000 tons a year sales to ICS; on the contrary, overall exports of phosphate products would increase substantially. The project would be an even stronger proposi- tion if the new solid fertilizer plant could be merged with the existing one (SIES) to reduce investment costs. 3.29 One of the first problems that had to be solved was the site of the new plant. Since the port of the Cayar Complex was not going to be built, it was decided to put a phosphoric acid plant at the mine site to reduce transport costs (in particular, the low grade phosphate pebbles) from the mine to the factory, and to transport the phosphoric acid in specialized railway tankers to a plant producing solid fertilizer at a site near the port of Dakar where the imported inputs for solid fertilizer production (mainly 1/ Industries Chimiques du Senegal. 2/ Entreprise Miniere et Chimique.  - 59 - ammonia and sulphur) would have to carry the lowest transportation costs. Since the whole operation would have a relatively small scale, the very specialized overseas transport of phosphoric acid would have become very expensive, a factor that brought about the decision to process the phosphate rock to a solid fertilizer which does not have this disadvantage. 3.30 The financing plan of the project is in an advanced stage of pre- paration. So far, the Government expressed willingness to participate for 25 percent whereas the other partners are private companies and foreign public institutions. However, if the Government were to guarantee the long-term loans needed for the venture, Government's risk would amount to $189 million ouit of the $263 million project costs (excluding the cost of about $35 million infrastructures). For a project depending on a long-term contract with a local mine, and the accessibility of an export market controlled by a technical partner, this seems to be an unreasonable risk-sharing arrangement. The debt service of these loans (including Government borrowing for its share capital) from 1984 to 1986 would average roughly CFAF 20 billion a year, or almost 5 percent of Senegal's i:otal export earnings (including ICS exports). Fishing 3.31 It is well known that the ocean off Senegal is one of the richest fishing areas in the world. There are several factories in Dakar which make canned tuna, and several which freeze shrimp and sole (there are also some in Ziguinchor). Practically all of the production is exported, both inside and outside of the Common Market. The results of the Balassa study show that this is very profitable from an economic point of view, since all of the DRCs are less than the shadow exchange rate. Table 19: DOMESI'IC RESOURCE COST OF FOREIGN EXCHANGE IN FISHING DRCU DRCFA DRCA actual capacity 0.64 0.50 0.45 full capacity 0.50 0.32 0.33 These results show clearly that Senegal can make an economic profit in the fishing industry - thus, production should be expanded given that market prospects are good in the long run. 1/ Nonetheless the sector is not without its problems. In particular, the tuna canneries (notably Conserverie du Senegal) were plagued by extremely unreliable supplies from its principal supplier SOSAP. 3.32 SOSAP was created in 1962 and is principally owned by the Govern- ment (87%). With strong technical assistance from a consulting firm with long experience in the sector, SOSAP purchased well adapted tuna fishing 1/ At the present point, these firms do not benefit from any kind of subsidy from the Government.  - 60 - boats from French and German suppliers and had a relatively successful record in the 1960s. However by June 1973, it had accumulated losses of 450 million CFAF and was in serious trouble--the major reason being their decision to purchase boats entirely unsuited to Senegalese conditions. These boats, ordered in 1966, were reportedly obsolete at the time of construction (capital costs 227 million CFAF). Partly due to maintenance difficulties, operating losses amounted to 158 million CFAF a year, so SOSAP invested in its own repair facilities which further compounded the financial losses. The boats should have been sold. In addition to this, SOSAP suffered from bad management which also increased operating costs. Total catches fell from 10,800 tons in 1974 to 5,900 tons in 1975, with only 15 boats out of 25 in operation, and of these, only one made a profit. In 1977, Government decided to liquidate this State enterprise. The result, of course, was that some canneries were operating at low rates of capacity utilization, and by 1975, the rate had fallen to less than 50 percent. 3.33 Compared to the tuna situation, supply side problems in the sole and shrimp industries are practically nonexistent. Factories are supplied either by artisanal boats or by private French or Senegalese boats, owned by companies with no Government participation. Most of these companies have very few expatriates - relying on Franco-Senegalese management and technical personnel. This must surely throw doubt on the hypothesis that the failure of SOSAP could-be attributed to the poor quality of Senegalese fishermen. In the short run, these firms expect to experience some depression in demand; however, long-run prospects are good. 3.34 The priority for the Government in this sector should be to ensure that the supply side problems in tuna fishing are resolved. If possible, this problem should be solved without Government participation for two reasons: (a) the evidence from the rest of the sector is that private entrepre- neurs can supply fish without any special favors (and subsidies) from the Government; and (b) the Government cannot afford the commercial risks associated with this type of intervention. In addition to this problem, there is another - that of packaging. Both locally made cans and cartons are significantly more expensive than imports, and this obviously reduces the profitability of the industry to the extent that canneries buy the locally made rather than the imported product. Textiles 3.35 The textile sector is one of the most interesting in the whole industrial sector. It is fully vertically integrated from cotton to finished clothes and exports part of its production in all stages of production. Many of its problems are those likely to be encountered in other sectors when the Government makes progress in achieving the two principal objectives of the Fifth Plan:  - 61 - (a) to further vertically integrated industries; and (b) to promote exports. tMoreover, as is well krnown, low-cost import competition is very fierce in this industry, particularly from used clothes. 3.36 The Balassa study investigated this sector in particular detail, obtaining data for eight of the twelve firms operating in Senegal. Overall, the unadjusted DRC was 1.33 at the existing capacity, and 1.20 at full capacity. However, the adjusted versions of the DRC were both greater than the shadow exchange rate--both at existing and at full capacity. Table 20: RESOURCE COST OF FOREIGN EXCHANGE IN TEXTILES DRCU DRCFA DRCA actual capacity 1.33 3.97 2.56 full capacity 1.20 1.77 2.66 Source: op. cit. At: first glance, it might seem that the textile branch taken as a whole is unprofitable, but such a conclusion would be suspect as is clearly shown by disaggregating the results. Specifically, the study found: (a) underutiliza'tion of capacity, so that results are better at full capacity; (b) (i) the production of grey cloth to be economically unprofitable; (ii) the production of finished cloth to be economically unprofitable; (iii) printing and dyeing of cloth to be marginally profitable; and (iv) knitwear and finished clothes to be economically profitabLe; (c) a distortion or--incentives so-that the local market is encouraged oveir exports; and (d) small Common Law firms to be economically more profitable than large Investment Code firms. These small firms also export much more than large firms; indeed, it is interesting to report the results for these firms grouped according to whether they export more or less than 40 percent: of their production.  - 62 - Table 21: RESOURCE COST OF FOREIGN EXCHANGE IN THE TEXTILE SECTOR DRCU DRCFA DRCA Exporting subsample actual capacity 1.05 1.08 0.97 full capacity 0.89 0.82 0.91 Import substitution subsample actual capacity 1.43 7.06 5.59 full capacity 1.30 2.82 6.98 This table shows clearly that export-oriented firms are economically more profitable than firms selling primarily in the local market. 3.37 This sector typifies the problems for the Government concerning the development of the industrial sector, namely to design a policy which in- creases the economic profitability of presently unprofitable firms, so they constitute less of a burden on the rest of the branch, and at the same time increase the share in total output of the activities in which Senegal does seem to have a comparative advantage (subject to the condition that such increases are undertaken in a way which is profitable to the country). Such policy requires treatment of the following problems: (a) the development of exports; (b) the development of the local market; (c) the optimal share of each market in total output; and (d) the growth of labor productivity. These problems are crucial if the textile sector is to increase its contri- bution to real economic growth. In the Fifth Plan, the most important project is the integrated plant at Kaolack, which will transform Senegalese cotton into finished knitwear. We shall review this project after discussing the remarks of the preceding paragraphs. Grey Cloth 3.38 The first point is that grey cloth production appears to be eco- nomically unprofitable as judged by the domestic resource cost of foreign exchange (DRCU = 2.14 and DRCFA, DRCA both negative). Consequently, further investment in grey cloth production does not seem advisable unless it can be shown to be economically profitable. This would be unfortunate since as explained in a moment, there is sufficient demand to warrant a new plant. The second point is that grey cloth is very expensive, generally more than 45 percent more expensive than on the world market, particularly that produced in the Far East. This has several effects: locally produced grey cloth is used only for finished cloth to be sold in the local market, so that imported grey cloth is used to make printed and dyed cloth for  - 63 - etxport, since grey cloth can be imported duty free when destined for re-export; but, if local grey cloth were competitive in world markets, production could double as it would substitute for some 9,000,000 meters currently imported. F'inally, the high cost of local grey cloth means that finished cLoth is also mtuch more expensive than imported finished cloth. This activity therefore requires high protection and it does not export much. 3.39 Cotton is grown and ginned in Senegal by SODEFITEX which diverts part of its exportable production to local spinners and weavers, in prin- ciple at the world prices. They therefore buy cotton in the same competi- tive conditions as Far East weavers who import cotton and export grey cloth to Senegal. The observed 50 percent difference in price, and perhaps the lack of economic profitability, must therefore be explained by a combina- tion of: (a) lower labor productivity; (b) higher costs in Senegal of other intermediate inputs - glue, dyes and electricity; and (c) high factor costs, including the presence of rents. All three seem to be contributing factors. Labor productivity is less than that in Europe and other less developed countries - partly because of conscious overmanning and sometimes because the capital stock is obsoles- cent. High costs of intermediate goods play a minor role, but the most striking thing about the production of grey cloth is the rate of return to investment, which was 29 percent of sales in 1972, and 17 percent in 1974. 1/ Such high profits would cause no problem if they represented the fruit oE competitive hard wor'k, but in fact they seem due to high protection against imports, a protection which is reinforced by the vertically integrated nature of the grey cloth industry: the main grey cloth factory sells its output to its owner, the printing and dyeing factory. There is a very strong case for the reduction of the protection accorded to this industry. 3.40 Since the grey~ cloth factory is partly owned by the printing and dyeing industry, it would be appropriate to accompany the reduction in duty rates on grey cloth by some reduction in the rate on printed and dyed cloth, in order to oblige this enterprise to reduce the mark up it takes on the grey c:Loth. 3.41 There is some chance that the imports of grey cloth by the printing and dyeing factory could be eliminated by the use of indirect or direct export subsidies. Indirect subsidies would have to take the form of a special price for cotton, which have t:o be sold to the spinners at a discount on the world price large enough to make the production of grey cloth competitive with imported grey cloth. However, this policy would be unsatisfactory from a resource allocation point because it would either depress the rate of return 1/ Based on Ministry of Industry estimates.  - 64 - to a profitable industry, namely cotton production and ginning, or impose a burden on the Caisse de Perequation et de Stabilisation des Prix (CPSP) which is technically the recipient of the profits from SODEFITEX. It would be preferable to sell cotton to the spinners at the world price; to protect all of grey cloth production by subjecting all imported grey cloth to import duties, even when intended for eventual export; then, to subsidize directly the production of printed cloth. The advantages of direct subsidies in this case are: (a) the profitability of cotton production is not reduced; (b) the reserves of the CPSP, and thus investment in the agricultural sector is not reduced; and (c) the cost of the subsidy on the budget has to be compensated by increases in consumer taxes so the burden will fall on consumption and not on savings. These subsidies should, of course, be put only into operation if it can be demonstrated that grey cloth weaving is an economically profitable activity. The Production of Finished Cloth 3.42 It is beset by many problems. The activity originally wvas con- ceived to serve the whole of the West African regional markets (before independence) and was badly affected by the break up of the colonial customs union. The low economic profitability is reflected by its high resource cost of foreign exchange (unadjusted DRC 1.67; adjusted 3.50 and 1.81, respec- tively). The industry claims to have suffered from five types of problems: (a) low productivity of labor; (b) high costs of dyes and inks, which are imported; (c) extremely strong competition from imports; (d) poor quality of the capital stock; and (e) the diversified nature of production. The low productivity of labor has already been discussed. Imported dyes and inks pose a problem because they are subject to duties; whereas in the Ivory Coast (which exports to Senegal), they can be imported duty free. This situation makes it difficult for the factory to compete with Ivorian products, since they are imported duty free into Senegal under the terms of the bilateral trading agreement between the two countries. Concerning the competition from imports, these are controlled by the Special Subcommission of the Committee on protection (see Chapter II). The effect of those controls and tariffs is to give an effective protection coefficient in the local market of 1.97; so that despite claims about the fierceness of import competition, the local market is in fact well protected except in the unquantifiable effects of smuggling. The poor quality of the capital stock is clearly one of the main reasons for the poor performance of this branch. This is effectively admitted by the owners of one firm who are proposing to renew it completely over a three-year period. The diversified nature of production (400 articles) is also a problem because of the reduced size of the local market. There is clearly a need to reduce the number of articles produced for the local market,  - 65 - allowing thereby greater satisfaction of the local market by imports and diverting resources int:o production for export. Such a policy would permit the industry to exploit: economies of scale and to increase capacity utiliza- tion thereby reducing overhead costs/unit of output. 3.43 The high effective protection that the Bank's research team found in the branch was derived from the protection the industry used to have in the framework of UDEAO agreement on exports to the members of that trade zone, coupled with the duty-free imports of grey cloth. In the local market protec- tion was, of course, also high via high tariffs and the earlier mentioned quantitative restricticns. With the replacement of UDEAO by the CEAO a sharp reduction in the protection on the West African market has followed (see para. x.). Specific measures which the Government could consider to encourage exports are the use of direct export subsidies. They would be preferable to indirect subsidies given through preferential prices for cotton and electricity. Total exemption of duties on dyes and inks as a means of increasing protection does not seem necessary. Spinning of Yarn. Knitwear and Finished Clothing 3.44 This collection of firms is quite different from those in spinning and weaving. They are economically profitable as judged by the domestic resource cost of foreign exchange criterion: Table 22: RESOURCE COST OF FOREIGN EXCHANGE IN TEXTILES BY ACTIVITY DRCU DRCF'A DRCA Spinning of yarn actual capacity 0.88 0.92 0.95 full capacity 0.88 0.92 0.95 Knitwear actual capacity 1.05 0.99 1.00 full capacity 0.86 0.68 0.98 Finished Clothes actual capacity 0.82 0.72 0.71 full capacity 0.66 0.56 0.64 3.45 Spinning of yvrn - This is a small Common Law firm which spins only (does not weave). It sells part of its output to a weaver, part to a printer and dyer, part to local knitwear plants, and part to local artisans. In terms of the analysis of the first part of this chaptet, the firm has a resource cost of foreign. exchange of less than one, and the effective  - 66 - protection coefficient for the local market is 1.52, 1.63 for the CEAO and less than one in the EEC. Yet despite the disincentive to export to Europe, the firm seems to export more there than to the CEAO, perhaps because of the trade restrictions imposed by the Ivory Coast on Senegalese exports. Nonetheless, the results are very interesting because they indicate clearly that spinning can be an activity which can be economically profitable for Senegal, and financially profitable for shareholders, even though the firm in question does not enjoy the privileged status of the Investment Code. It is also pertinent to note that this firm fell within the import-substituting subsample of the Balassa study, thus establishing that an import-substitution firm can be economically profitable (not that it is a proposition ever denied by Balassa and associates). Moreover, it is interesting to comp